Surg Endosc (1997) 11: 729733

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopically guided bipolar radiofrequency ablation of areas of porcine liver

S. A. Curley,1 B. S. Davidson,1 R. Y. Fleming,1 F. Izzo,1 L. C. Stephens,2 P. Tinkey,2 D. Cromeens2
1

Department of Surgical Oncology, Box 106, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX, 77030, USA 2 Department of Veterinary Medicine, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030, USA Received: 19 June 1996/Accepted: 30 October 1996

Abstract Background: Bipolar radiofrequency ablation (BRFA) is a promising technique with which to treat unresectable primary and metastatic liver tumors. Its effects on normal liver tissue and postoperative liver function, however, are unknown. We performed this study to determine (1) the feasibility of using laparoscopic ultrasound to guide placement of BRFA needle electrodes in the liver and (2) the histopathologic, hepatic biochemical, and systemic hemodynamic responses to BRFA. Methods: Two BRFA lesions were created in the liver of adult domestic pigs to ablate 810% of the normal liver volume. Laparoscopic ultrasound was used to guide creation of one peripheral liver lesion and one central liver lesion (with a major hepatic or portal venous vein branch in the center of the BRFA lesions) in each animal. BRFA of liver tissue was performed by passing 12 W of RF power for 16 min across two 16-gauge active-needle electrodes placed 3 cm apart. Results: All animals survived the procedure without significant hemodynamic alterations during or after BRFA. All animals had a transient elevation in serum transaminase levels that returned to normal within 1 week of the BRFA of liver tissue. Gross and microscopic histopathology of the BRFA lesions revealed 2.02.5-cm zones of complete coagulative necrosis around and between the BRFA needle tracks without destruction of major blood vessel walls. Conclusions: This study demonstrates (1) that laparoscopic ultrasound can be used to guide placement of BRFA needles in the liver and (2) that BRFA produces focal destruction of liver without significant systemic hemodynamic responses or alterations in liver function. Further studies of this technique to ablate malignant liver tumors are ongoing. Key words: Bipolar radiofrequency ablation Liver Laparoscopic ultrasound

Correspondence to: S. A. Curley

Hepatocellular carcinoma is one of the most common human malignancies; approximately 1 million cases are diagnosed annually worldwide [9]. Further, the liver is second only to lymph nodes as the most common site for metastatic cancer [21]. While surgical resection offers the best chance for long-term survival for patients with hepatocellular cancer or colorectal cancer liver metastases, the majority of patients are not candidates for resection because the tumor is too large, there are multiple tumor foci, key vascular structures are involved by tumor, or hepatic reserve is inadequate because of cirrhosis. In such cases, alternative treatments are required. Cryosurgery to freeze primary and metastatic liver tumors in situ is being used to treat unresectable liver tumors. However, technical difficulties with this approach remain problematic. The cryoprobes range in size from 8 to 30 mm in diameter, so hemorrhage from the probe track can be a significant problem in some patients [16, 18]. Renal and hepatic insufficiency have also occurred following hepatic cryosurgery [16, 18]. Finally, the number of lesions that can be effectively frozen is limited by tumor proximity to major blood vessels and the length of each freeze-thaw cycle. Still other ablative procedures have been used to treat unresectable tumors, including percutaneous ethanol injection (PEI), but these, too, have serious drawbacks or limitations on their effectiveness. Systemic therapies, such as chemotherapy, immunotherapy, and biological therapy, are also used, but they have low response rates, significant toxicities, and usually dont provide a major improvement in overall patient survival rates. One local tumor ablative technique with considerable promise is bipolar radiofrequency ablation (BRFA). A radiofrequency (RF) current applied across needle-electrodes placed around a liver tumor can ablate the tumor in situ. The RF current produces coagulative necrosis of the tumor tissue [19]. Although the technique does not appear to have the technical drawbacks of cryosurgery or PEI, the effect of BRFA on normal liver tissue, including major hepatic veins, intrahepatic arteries, and portal veins, and on post-treatment renal and liver function, is not known. To clarify some of

730 Table 1. Systemic hemodynamic response to biopolar radiofrequency ablation (BRFA) of liver under laparoscopic guidancea Baseline Systolic BP Diastolic BP HR CO PAWP
a

After 1st BRFA 89.5 4.5 48.6 3.6 96.2 4.4 4.01 0.51 13.0 1.2

After 2nd BRFA 84.6 3.6 45.0 2.8 93.9 3.8 3.82 0.55 11.9 1.5

88.3 4.2 50.1 3.9 93.5 4.0 3.77 0.28 11.8 0.9

Baselinereadings before peritoneal insufflation with CO2 to a pressure of 1012 mmHg and before placement of BRFA needles into the liver. After 1st BRFAreadings at the conclusion of the first 16-min BRFA. After 2nd BRFAreadings at the conclusion of the second 16-min BRFA; systolic BP, systolic blood pressure in mmHg; diastolic BP, diastolic blood pressure in mmHg; HR, heart rate in beats/min; CO, cardiac output in liters/min.; PAWP, pulmonary artery wedge pressure in mmHg

Fig. 1. Photograph of the cut edge of a porcine liver treated 7 days earlier with bipolar radiofrequency ablation (BRFA) for 16 min at 12 W of power. The needle-electrodes for BRFA were placed into the liver under laparoscopic ultrasound guidance in parallel tracks 3 cm apart. The coagulative necrosis around and between the two needle tracks is evident (solid arrows). A large portal vein branch is adjacent to the coagulated area (open arrow).

8puthese issues, we undertook a study of BRFA of normal liver tissue in pigs in which the electrodes were placed under laparoscopic ultrasonographic guidance. Methods
Eighteen healthy, adult domestic pigs (3540 kg) were anesthetized with ketamine (50 mg/kg), acepromazine (0.22 mg/kg), and atropine (1.2 mg), all given by intramuscular injection. General anesthesia was maintained with isoflurane (1.52.5%) and 100% oxygen following endotracheal intubation of the animals. A 16-French Silastic catheter was placed into the left femoral artery for continuous blood pressure monitoring. An 8.5French Cordis introducer sheath was placed into an external jugular vein, and a Swan-Ganz catheter (Baxter Healthcare Corp., Irvine, CA) was introduced through the introducer sheath to provide continuous hemodynamic monitoring. A second 16-French Silastic catheter was placed into the opposite external jugular vein for intravenous fluid administration. The peritoneal cavity was insufflated for laparoscopy by placing a Veress needle in a supraumbilical position and then filling the abdominal cavity with CO2 to a maximum pressure of 1012 mmHg. A 10-mm laparoscopy port was then placed in the supraumbilical position and a laparoscopic video telescope was inserted. Twelve-millimeter ports were placed in the right upper and left upper quadrants under direct camera visualization. A complete ultrasound evaluation of the liver was performed through the two 12-mm ports using a sterile pivoting-head laparoscopic ultrasound transducer at both 5.0 and 7.5 MHz (B & K Medical Systems, Inc., Cambridge, MA). Each animal was connected to the RF generator (Zomed International Corp., Mountainview, CA) through a dispersive electrode safety grounding pad. The generator produces a 480-kHz frequency with a maximum power output of 26 RF W and a peak voltage of 44 V. Two BRFA lesions were created in the normal liver of each animal. The first lesion was produced in the periphery of the right or left lobe of the liver by placing two 16gauge active-needle electrodes (ZoMed International Corp., Mountainview, CA) percutaneously into the liver under direct laparoscopic guidance. The active-needle electrodes have an insulated stainless-steel shaft 24 cm long with an exposed tip of variable length. For these experiments, the exposed tip length was set at a 2.5 cm. The needles were placed 3 cm apart and were confirmed by laparoscopic ultrasound to be lying in parallel tracks. An RF power of 12 W was passed across the needles for 16 min to create a lesion. The tissue temperatures adjacent to the BRFA needles were monitored continuously using the thermistors built into each active-needle electrode. At the conclusion of the first BRFA of normal liver, the needles were withdrawn. A second BRFA lesions was then created in the central liver. The two needle-electrodes were placed 3 cm apart so that either the middle hepatic vein or an anterior branch of the left portal vein was mid-

way between the two needles, as confirmed by laparoscopic ultrasound. Again, the second lesions was created by passing 12 W of power across needle tips for 16 min. The needles were then removed. BRFA was performed in 16 animals, while two control animals underwent laparoscopy and placement of the BRFA needles without RF current being passed across the needles. Baseline serum liver function test values (total bilirubin, alanine aminotransferase [ALT], aspartate aminotransferase [AST], alkaline phosphatase, and albumin), creatine phosphokinase (CPK), lactate dehydrogenase (LDH), complete blood count, prothrombin time, partial thromboplastin time, blood urea nitrogen, and serum creatine levels were determined in blood samples obtained prior to hepatic BRFA in each animal. The identical blood tests were performed at the conclusion of the BRFA laparoscopic operation in all animals and then again on postoperative days 1, 3, 7, and 14 in the survival group of animals. Half of the animals were sacrificed immediately after the BRFA procedure, and their livers were subjected to gross and microscopic histopathologic evaluation. The surviving animals were sacrificed 7 or 14 days following the BRFA procedure; likewise, their livers were evaluated for gross and microscopic histopathologic changes.

Results All 18 animals used for this study survived the laparoscopic operation and the BRFA of two areas within the liver. In all 32 of the hepatic BRFA procedures, there was complete hemostasis in the needle tracks into the liver and at the liver capsule puncture sites upon withdrawal of the BRFA needles. Dynamic ultrasound during the 16-min BRFA procedure demonstrated a progressive increase in echogenicity of the liver tissue for 2 cm around and between the two needles. The size of this hyperchoic area visualized by ultrasound corresponded to the area of gross coagulative necrosis observed during pathologic evaluation (Fig. 1). With two BRFA lesions in each animals liver, the amount of normal liver volume ablated by the procedures was estimated by pathologic evaluation to be 810% of the normal liver volume. During the BRFA procedures, there were no significant changes in systemic hemodynamic parameters (Table 1). This was true whether the lesions were created in the liver periphery or in the central liver around a major intrahepatic blood vessel. The focal destruction of areas within the liver did not reduce systemic vascular resistance as evidenced by the stable blood pressure, heart rate, cardiac output, and filling pressures of the animals (Table 1). Serum CPK, LDH, and ALT levels were elevated significantly 1 day after BRFA in the eight animals in the

731 Table 2. Values of serum laboratory tests before (baseline) and for several days after bipolar radiofrequency ablation (BRFA) of normal liver in adult domestic pigsa Baseline CPK LDH ALT AST Alk Phos T Bili Creat WBC Hgb
a

Day 1 4676.7 1052.9* 806.6 181.4* 76.6 16.3* 32.4 3.4 103.0 5.7 0.2 0 1.3 0.1 26.8 2.5 11.4 0.7**

Day 3 4694.4 590.9* 807.7 125.8* 48.4 5.3* 35.3 3.8** 79.3 3.0 0.1 0 1.3 0.1 25.1 3.3 10.6 0.6

Day 7 1150.0 128.0 547.3 14.0 32.3 3.2** 37.3 3.4** 100.3 7.5 0.3 0 1.3 0.1 20.0 1.5 11.5 0.7**

Day 14 633.7 95. 515.0 16. 30.0 2.6 30.4 3.9 107.8 10. 0.3 0 1.3 0.1 23.3 2.6 11.3 0.6

1304.1 327.4 519.1 67.2 24.7 3.8 25.6 1.8 97.0 3.0 0.2 0 1.2 0.1 22.6 1.9 9.9 0.4

CPK, creatine phosphokinase (IU/l); LDH, lactic dehydrogenase (IU/l); ALT, alanine aminotransferase (IU/l); AST, aspartate aminotransferase (IU/l); Alk Phos, alkaline phosphatase (IU/l); T Bili, total bilirubin (mg/dl); Creat, creatinine (mg/dl); WBC, white blood cell count (103); Hgb, hemoglobin (g/dl) * p < 0.01 Students paired t-test, vs baseline value, n 8 ** p < 0.05 Students paired t-test, vs baseline value, n 8

survival group (p < 0.01, Table 2). Seven days after BRFA of areas of normal liver, the serum CPK and LDH levels had returned to baseline values, while serum ALT and AST levels remained slightly elevated (p < 0.05). Serum ALT and AST levels also returned to baseline values 14 days after the BRFA procedures. In contrast, serum levels of alkaline phosphatase, total bilirubin, creatinine, and the white blood count were not significantly different compared to baseline values at any time after the BRFA procedures (Table 2). Similarly, serum blood urea nitrogen levels and the prothrombin and partial thromboplastin times were not elevated following BRFA of normal liver (data not shown). Serum hemoglobin levels were slightly elevated after the BRFA procedures; no animal had a decrease in hemoglobin levels compared to baseline values. The gross pathologic findings showed that the laparoscopic ultrasound probe was accurate in guiding the percutaneous placement of the BRFA needles 3 cm apart in the liver. Gross examination revealed well-demarcated 2.02.5cm-diameter hemorrhagic zones around each needle track, overlapping between the needles (Fig. 1). This pattern was observed immediately, 7 days, and 14 days after BRFA. There was no evidence of hemorrhage into the needle tracks. Microscopically, the sharply defined gross lesions correlated with 2.02.5-cm zones of complete coagulative necrosis of hepatic parenchyma abutted by normalappearing hepatocytes and liver architecture (Fig. 2). There was no gross or microscopic evidence of necrosis of the walls of major intrahepatic portal or hepatic veins in any of the animals despite the necrosis of hepatic parenchyma circumferentially around the vessel. The temperatures measured at each of the two needles during the BRFA procedures are depicted in Fig. 3. In two control animals, BRFA needles were placed into the liver but no RF current was passed across the needles. There was no evidence of any gross or microscopic liver destruction in these two animals 7 days after the operation.

Discussion Primary and metastatic liver malignancies, particularly from colorectal cancer, are a major cause of morbidity and mor-

tality in the United States and worldwide. Hepatocellular cancer is associated with chronic hepatitis B or C virus infection and cirrhosis from any cause and has a mortality index of 0.94 [20]. A subset of patients with metastatic colorectal cancer have liver-only disease, but less than 25% of these patients are candidates for resection based on number or site of metastases [24]. Even in patients with colorectal cancer metastases in sites in addition to the liver, control of the liver metastases could provide significant palliative benefit. Several local tumor ablative techniques have been used in the treatment of liver malignancies. Percutaneous ethanol injection (PEI) with real-time ultrasound guidance has been used to treat hepatocellular cancer [11, 23]. PEI is generally an outpatient treatment that requires multiple injections of absolute ethanol over a several-week period, although a one-shot injection technique under general anesthesia has been described [12]. PEI for hepatocellular cancer can be curative in some patients with solitary tumors less than 4 cm in diameter, but PEI alone rarely cures patients with larger or multiple hepatocellular tumors and has not been effective against metastatic tumors [10]. With BRFA, it should be possible to completely ablate larger liver tumors that cannot be treated successfully with PEI because the BRFA needle-electrodes can be moved to several different positions in the primary or metastatic live cancer. Percutaneous and surgical placement of laser fibers under ultrasonographic guidance has been used to produce interstitial thermal ablation of liver tissue [1, 2, 4, 15]. However, laser fiber ablation produces smaller zones of necrosis, generally less than 2 cm in diameter, compared to BRFA lesions. The rapid vaporization of tissue near the laser fiber produces a dense coagulum that inhibits propagation of cytotoxic heat more than 1015 mm away from the fiber [4]. Cryoablation using circulating liquid nitrogen probes has been used to treat hepatocellular cancer and colorectal cancer liver metastases [16, 18, 28]. The probes are much larger than the 16-gauge BRFA needles, however, and life-threatening hemorrhage has been reported after removal of the cryoprobes [16, 18]. Compared with cryoblation, bleeding from the liver following BRFA is less likely because of the small diameter of the needles and the cauterization of the liver tissue along the

732

Fig. 2. A photomicrograph (magnification 65) of porcine liver treated 7 days earlier with bipolar radiofrequency ablation (BRFA) for 16 min at 12 W of power. There is a well-defined margin between the zone of coagulative necrosis produced by BRFA (open arrow) and the normal adjacent hepatic parenchyma (solid arrow). Fig. 3. Temperature readings from the thermistors built into the two active needle-electrodes used for bipolar radiofrequency ablation (BRFA) of normal porcine liver. Time 0 represents the beginning of the 16-min-long BRFA and 18 min is the time point 2 min after completion of BRFA. There is no significant difference between the liver tissue temperatures measured at the two active needle-electrodes.

needle-electrode tracks. Furthermore, the time needed to produce complete necrosis of tumor tissue is five to 10 times shorter for BRFA than for cryoblation [5, 7, 22, 27]. Coagulative necrosis of tissue occurs if tissue temperature is maintained above 60C for 2 min. In our experiments with BRFA of normal liver, we maintained the tissue temperatures between 60 and 70C for 16 min to determine the hemodynamic and hepatic biochemical effects of prolonged BRFA current application. In actual BRFA treatments of liver tumors, current would only need to be applied for 35 min in each area treated. A single 5-min application of BRFA can effectively kill a liver tumor up to 4.0 cm in diameter. For larger tumors, the BRFA needles can be placed at various points in and around the cancer to increase the probability of complete tumor necrosis. A further disadvantage of cryosurgery is that it is usually a one-time treatment, as repeated treatments are technically difficult and increase the risks of morbidity and mortality. In contrast, since BRFA of superficial liver lesions can be performed percutaneously with transcutaneous or laparoscopic ultrasound guidance, tumors deep within the liver parenchyma, near major intrahepatic blood vessels, or recurrent and new lesions can be treated repeatedly if necessary. In our study, the combination of BRFA of areas within the liver and laparoscopy was well tolerated in all of the animals. Excellent visualization of the liver for inspection and laparoscopic ultrasonography was obtained with peritoneal CO2 insufflation pressures of 1012 mmHg. The animals were hemodynamically stable at these insufflation pressures, unlike the decreased cardiac output that can occur at higher insufflation pressures [6, 26]. In our study, the BRFA procedures had no effect on the hemodynamic stability of the animals. BRFA of up to 10% of the normal liver volume in experimental animals produced transient but completely reversible elevations in serum liver function tests. No animal had an immediate or delayed decrease in serum hemoglobin levels, indicating that no significant hemorrhage from the needle tracks or into the zone of thermal injury occurred. This is particularly significant because half of the lesions

were created around at least a 2.5-cm length of a major intrahepatic portal or hepatic venous segment. It is important to assess the effects and possible life-threatening complications of BRFA in normal liver before proceeding with full-scale treatment in patients with liver cancers. Tissue temperature of 6070C for 16 min did not cause necrosis, hemorrhage, or thrombosis in large intrahepatic blood vessels. If it had damaged large vessel walls, BRFA for tumors near or surrounding these vessels would be precluded. It remains to be determined if BRFA will produce complete killing of liver tumors near large intrahepatic blood vessels or the inferior vena cava because of the heat sink effect produced by the high flow rates in these vessels. To ensure complete killing of a liver tumor by BRFA, it is necessary to produce an approximately 1-cm-wide zone of necrosis of normal liver tissue surrounding the tumor. The BRFA parameters we used12 W of power for 16 min with 2.5-cm-long exposed electrode tips placed 3 cm apart produced 2.02.5-cm-diameter zones of coagulative necrosis. Laparoscopic ultrasound was useful in guiding the placement of the BRFA needles and in monitoring the procedure while current was passed across the needles. The hyperechoic area around and between the needles correlated with the area of coagulative necrosis on pathologic evaluation. This confirms in vivo the prior ex vivo observation that the size of the ultrasonographic hyperchoic region around the needles corresponds to the area of thermal destruction in pieces of porcine liver subjected to RF currents [13]. Laparoscopic ultrasound guidance will be particularly important to ensure accurate placement of the needle electrodes and to monitor BRFA treatment of central or deep liver malignancies. The thermal injury produced by BRFA is not the result of direct heating from the needle electrodes. Rather, the RF

733

current passed across the needle electrodes produces ionic fluxes in the tissue surrounding the electrodes that follow the changes in the direction of the alternating current. The ionic fluxes produce frictional heating so that the tissues, not the electrodes, are the source of the heat [14, 17]. Thus, the thermistors in the active needle-electrodes are measuring tissue temperature. To produce thermal necrosis of liver and tumor tissue, it is necessary to heat the tissue to temperatures greater than 4345C [8, 25]. However, to produce complete, uniform, larger zones of necrosis, temperatures between 60 and 70C are necessary [5, 14, 27]. While it is possible to produce tissue temperatures that exceed 90C around the electrodes by increasing the wattage and current amplitude, this actually decreases the efficiency of BRFA because of rapid tissue desiccation near the needles and a resultant decrease in current flow. RF currents have been used to treat a small number of patients with malignant and benign liver tumors [8, 19]. These studies have not been sufficient to establish the optimal treatment approaches and long-term survival benefits of BRFA treatment for liver cancers. Clearly, this is a local tumor ablative technique which should be studied further because of the potential advantages compared to other local therapies. From our study, we conclude that BRFA (1) produces a predictable volume of tissue necrosis that can be accurately monitored by ultrasound, (2) can be performed percutaneously with minimally invasive laparoscopic techniques, and (3) does not produce significant hemodynamic or hepatic toxicities. References
1. Amin Z, Donald JJ, Masters A, et al. (1993) Hepatic metastases: interstitial laser photocoagulation with real-time US monitoring and dynamic CT evaluation of treatment. Radiology 187: 339347 2. Bosman S, Phoa SS, Bosma A, Van Gemert MJC (1991) Effect of percutaneous interstitial thermal laser or normal liver of pigs: sonographic and histopathological correlations. Br J Surg 78; 572575 3. Buscarini L, Rossi S, Fornari F, et al. (1995) Laparoscopic ablation of liver adenoma by radiofrequency electrocautery. Gastrointest Endosc 41: 6870 4. Dachman AH, McGehee JA, Beam TE, et al. (1990) Ultrasoundguided percutaneous laser ablation of liver tissue in a chronic pig model. Radiology 176: 129133 5. Daikuzono N, Joffe SN, Tajiri H, et al. (1987) Laserthermia: a computer-controlled contact Nd: YAG system for interstitial local hyperthermia. Med Instrum 21: 275277 6. Davidson BS, Cromeens DM, Feig BW (1996) Alternative methods of exposure minimize cardiopulmonary risk in experimental animals during minimally invasive surgery. Surg Endosc 10: 301304 7. Gage AA, Guest K, Montes M, et al. (1985) Effect of varying freezing

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

26.

27.

28.

and thawing rates in experimental cryosurgery. Cryobiology 22: 175 186 Haines DE (1993) The biophysics of radiofrequency catheter ablation in the heart: the importance of temperature monitoring. PACE Pacing Clin Electrophysiol 16: 586591 Kew MC (1986) The development of hepatocellular carcinoma in humans. Cancer Surv 5: 719739 Livraghi T, Vettori C, Lazzaroni S (1991) Liver metastases: results of percutaneous ethanol injection in 14 patients. Radiology 179: 709712 Livraghi T, Bolondi L, Lazzaroni S, et al. (1992) Percutaneous ethanol injection in the treatment of hepatocellular carcinoma in cirrhosis. Cancer 69: 925 Livraghi T, Lazzaroni S, Pellicano ` S, et al. (1993) Percutaneous ethanol injection of hepatic tumors: single-session therapy with general anesthesia. Am J Roentgenol 161: 10651069 McGahan JP, Browning PD, Brock JM, Tesluk H (1990) Hepatic ablation using radiofrequency electrocautery. Invest Radiol 25: 267 270 McGahan JP, Brock JM, Tasluk H, et al. (1992) Hepatic ablation with use of radio-frequency electrocautery in the animal model. J Vasc Interv Radiol 3: 291296 Nelsoe CP, Torp-Pederson S, Burcharth F, et al. (1993) Interstitial hyperthermia of colorectal liver metastases with a US-guided Nd-Yag laser with a diffuser tip: a pilot clinical study. Radiology 187: 333337 Onik G, Rubinsky B, Zemel R, et al. (1991) Ultrasound-guided hepatic cryosurgery in the treatment of metastatic colon carcinoma. Cancer 67: 901908 Organ LW (1976) Electrophysiologic principles of radiofrequency lesion making. Appl Neurophysiol 39: 6976 Ravikumar TS, Kane R, Cady B, et al. (1991) A 5-year study of cryosurgery in the treatment of liver tumors. Arch Surg 126: 1520 1526 Rossi S, Di Stasi M, Buscarini E, et al. (1995) Percutaneous radiofrequency interstitial thermal ablation in the treatment of small hepatocellular carcinoma. Cancer J 1: 7381 Rustgi V (1988) Epidemiology of hepatocellular cancer. Ann Intern Med 108: 390401 Schwartz SI (1990) Primary malignant tumors. In: Moody F (ed) Surgical treatment of digestive diseases. Year Book Medical, Chicago, pp 400407 Shier WT (1988) Studies on the mechanisms of mammalian cell killing by a freeze-thaw cycle: conditions that prevent cell killing using nucleated freezing. Cryobiology 25: 110119 Shiina S, Yasuda H, Muto H, et al. (1987) Percutaneous ethanol injection in the treatment of liver neoplasms. Am J Roentgenol 149: 949 Steele G, Ravikumar TS (1989) Resection of hepatic metastases from colorectal cancer. Ann Surg 210: 127138 Strohbehn JW (1983) Temperature distributions from interstitial RF electrode hyperthermia systems: theoretical predictions. J Radiat Oncol Biol Phys 9: 16551667 Westerband A, van de Water JM, Amzallag M, et al. (1992) Cardiovascular changes during laparoscopic cholecystectomy. Surg Gynecol Obstet 175: 535538 Zervas NT, Kuwayama A (1972) Pathological characteristics of experimental thermal lesions: comparison of induction heating and radiofrequency electrocoagulation. J Neurosurg 37: 418422 Zhou XD, Tang Zy, Yu YQ, et al. (1988) Clinical evaluation of cryosurgery in the treatment of primary liver cancer: report of 60 cases. Cancer 61: 18891894

Surg Endosc (1997) 11: 737740

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Abdominal fat tissue necrosis as a cause of acute abdominal pain

Laparoscopic diagnosis and therapy
D. Aronsky, K. Zgraggen, M. Banz, C. Klaiber
Department of Surgery, Spital Aarberg, CH-3250 Aarberg, Switzerland Received: 5 February 1996/Accepted: 25 June 1996

Abstract Background: Infarctions of the greater omentum and of the epiploic appendages are rare etiologies of acute abdominal pain. The aims of the study were to determine the incidence of abdominal fat tissue necroses and to discuss the clinical features and the role of laparoscopy in the treatment of these conditions. Methods: A retrospective study in 563 consecutive patients with acute abdominal pain was performed. In all patients diagnostic laparoscopy was indicated. Results: The incidence of abdominal fat tissue necroses in 563 patients with acute abdominal pain was 1.1%. Six patients had either infarctions of the omentum or of the epiploic appendages. Pain was the predominant clinical symptom and the preoperative diagnosis depended upon the location of the omental or epiploic necroses. Diagnosis and treatment were performed laparoscopically without morbidity. Conclusion: The incidence of abdominal fat tissue necroses in our patients was increased compared to the prelaparoscopic period. Omental and epiploic necroses are significant in the differential diagnosis of appendicitis, acute cholecystitis, and diverticulitis. Key words: Acute abdominal pain Omentum Appendix epiploica Fat necrosis Laparoscopy

verticulitis. Laparoscopy allowed us at the same time to identify and treat these rare pathologies. The incidence of abdominal fat tissue necrosis in our patient population is in contrast to their infrequent mention in the literature.

Methods
The technical details of diagnostic and therapeutic laparoscopies have previously been described by us [12]. Briefly, a transumbilical access by a Veress needle was gained and a CO2 pneumoperitoneum was established. Intraabdominal pressure was limited to 12/15 mmHg. An umbilical 10-mm trocar for a 25 telescope and one additional 5- or 10-mm trocar are inserted. The location of the second trocar depended upon the clinical diagnosis and the site of the expected pathology. A thorough inspection of the abdominal cavity was then carried out. This systematic revision allowed diagnosis of abdominal fat tissue necroses, i.e., omental necrosis or twisted epiploic appendages. At this stage another trocar is inserted to help in the resection of the pathology. Laparoscopic resection of the omental necroses and the necrotic epiploic appendages was performed by EndoGIA stapler (Autosuture) in five patients, using a Roeder loop in one patient. The resected specimens were removed directly from the abdominal cavity; no endobags were used to prevent wound infections. Prophylactic antibiotics were used in only one patient (No. 4) with the preoperative clinical diagnosis of acute cholecystitis. One patient (No. 5) was given antibiotic therapy starting the day of admission. Four patients did not receive antibiotics. Four patients (No. 13, 6) underwent concomitant appendectomy. Patient No. 4 had undergone appendectomy previously. Although it is our standard policy to perform routine concomitant appendectomy in patients under 50 years of age, patient No. 5 did not undergo this procedure. Resected specimens were sent for histopathological examination. Followup was complete and carried out by telephone.

Infarction of the greater omentum and of the epiploic appendages can cause acute abdominal pain. Although rare in occurrence, they are clinically relevant as they mimic more common diseases in patients with acute abdominal pain. We report six cases of abdominal fat tissue necrosis diagnosed by laparoscopy over a period of 5 years in patients showing clinical suspicion of appendicitis, acute cholecystitis, or diCorrespondence to: K. Zgraggen, Department of Visceral and Transplantation Surgery, Inselspital, University of Bern, CH-3010 Bern, Switzerland

Patients
Relevant clinical symptoms and findings are summarized in Table 1. The average age of our patients was 36 years (1661 years). Four of the six patients were male. Three patients were obese; the body mass indices of patients 35 were 26.9, 36.2, and 32.0, respectively. Clinical diagnosis was appendicitis in four patients, acute cholecystitis in one patient, and diverticulitis in one patient. The duration of symptoms prior to hospital admission was 1272 h. None of the patients experienced nausea, vomitus, or diarrhea. The axillar temperature averaged 37.7C (37.4C38.0C), the

738 Table 1. Summary of relevant clinical, laboratory, and pathological findings of the six patients with abdominal fat tissue necroses Patient 1 Age 29 Sex M WBC 11,300 Shift Y C-reactive protein 8 Localization of pain Right lower quadrant Umbilical/ right lower quadrant Right lower quadrant Right upper quadrant Suspected diagnosis Appendicitis Postoperative diagnosis Primary omental necrosis Omental necrosis with torsion Omental necrosis with torsion Omental necrosis with torsion Size (cm) 3.5 3 1.8 Histopathological findings Recent hemorrhagic necrosis with slight granular inflammation Hemorrhagic necrosis with fibrinogranular inflammation Hemorrhagic, purulent, focal granular inflammation Hemorrhagic, focal partly fibrino-purulent partly granular inflammation Granular partly fibrino-purulent inflammation near serosa, focal acute purulent inflammation in center Diffuse recent hemorrhage and venous congestion

17

10,500

<5

Appendicitis

511

49

6,000

<5

Appendicitis

7 1 1.5

61

7,900

29

Acute cholecystitis

8.5 4 0.5

43

8,500

<5

Left lower quadrant

PID, diverticulitis

Epiploic appendage of the sigmoid colon with torsion

2.2 1 1

16

7,800

43

Right lower quadrant

Appendicitis

Epiploic appendage of the sigmoid colon with torsion

4 2 0.5

mean axillar-rectal temperature difference was 0.9C. The white blood cell count (WBC) averaged 8,650/mm3 (6,00011,300/mm3). The WBC showed a shift in only one patient. The erythrocyte sedimentation rate (ESR) averaged 8 mm/h and the C-reactive protein 16 mg/l. In patients 13 and 6 laparoscopy was performed on the day of admission; in patient 5 it was done on the 4th day of hospitalization after an abdominal ultrasound and a water-soluble contrast enema. Patient 4 was operated on the 3rd day after a diagnostic workup with abdominal ultrasound, computed tomography (CT) scan, and upper-gastrointestinal (GI) endoscopy.

after an average of 10 months (715 months). None of the patients had remaining symptoms.

Discussion The literature on abdominal fat tissue necroses is scarce. Only a few references on omental infarctions [1, 4, 9, 18] and necrotic epiploic appendages [3, 14, 16] exist and laparoscopic diagnosis and treatment in particular are only described as case reports [6, 8, 11]. Pathogenetically the rare, primary segmental necrosis of the omentum is distinguished from the more common secondary necrosis, generally following torsion of a mobile omental segment [1, 1820]. There is no preferred age classchildren can also be affected [1, 9, 17, 19]. Obesity and cardiovascular diseases are considered predisposing conditions [1, 7, 18]. Obesity was present in 50% of our patients, and one patient had a preexisting cardiovascular disease. The infarctions tend to occur in the right side of the omentum [7, 9]. An anomaly with reduced vascularization of the affected part is discussed as the cause of this phenomenon [19]. A secondary impaired vascularization of the larger, right-sided part of the omentum due to systemic diseases, e.g., arteriosclerosis and diabetes mellitus, offers an alternative explanation. Necroses of epiploic appendages occur almost exclusively following torsion. Depending upon their localization, necroses of the omentum and the epiploic appendages present with a similar but atypical clinical pattern compared to the more common etiologies of acute abdominal pain. Nausea, diarrhea, and vomitus are only occasionally described and were absent in all of our patients. Abdominal pain is the predominant and

Results From 1990 to 1995 some 563 laparoscopies in patients with acute abdominal pain were performed. Laparoscopy was indicated in cases with suspected appendicitis, acute cholecystitis, and diverticulitis and in patients with unclear clinical presentation but suspected surgical pathology. In six patients the diagnosis of abdominal fat tissue necrosis was madean incidence of 1.1%. A primary segmental necrosis of the omentum occurred in one patient. A secondary segmental necrosis of the omentum with torsion was discovered in three patients during laparoscopy (Figs. 1 and 2). A necrosis of a twisted appendix epiploica was diagnosed once (Fig. 3). In one patient a torsion of the necrotic epiploic appendage could not be positively identified. In none of the six patients was free serous or sanguineous abdominal fluid detected. All procedures were performed without intraoperative complications. The mean operative time was 62 min (40120 min). The histologic examination confirmed the macroscopic finding of fat tissue necrosis in all patients. No complications occurred in the postoperative course and the duration of hospitalization averaged 5.8 days (47 days). Patients were followed up

739

only consistent symptom and, occurring for 15 days, was of short duration [1, 9, 18]. The discrepancy between the patients good general condition and the marked abdominal symptoms must be emphasized and is consistent with the literature and our experience. Laboratory results are usually nonspecific and show no or only a slight increase of WBC but hardly ever a shift to the left or toxic signs. Considering the above-mentioned possible pathogenetic factors, it is not surprising that omental necroses are usually located in the right lower abdomenaccording to the literature, 84% of the time [9]. The most commonly reported preoperative diagnosis was therefore appendicitis (6474%) followed by acute cholecystitis in 2126% [9, 10]. Necroses of epiploic appendages 40% of the time are situated in the left lower abdomen, corresponding to their frequent origin in the sigmoid colon [5], but they lead to the clinical diagnosis of appendicitis in 3860% [5, 13]. A correct preoperative diagnosis of abdominal fat tissue necrosis is rare [1, 5]. Even additional diagnostic procedures such as CT scan and ultrasound do not allow a conclusive diagnosis [17], and in the reviewed literature abdominal fat tissue necrosis was always diagnosed at laparotomy. In contrast to these descriptions our standard policy is to perform diagnostic laparoscopy in patients presenting with acute abdominal pain. Comparison of our reported incidence of 1.1% to the international literature is therefore difficult. The value of laparoscopy in the diagnosis and therapy of nonspecific acute abdominal pain is well known [11, 15]. Laparoscopy enables the surgeon to inspect the entire abdominal cavity, which is an advantage over a limited open access, e.g., a McBurney incision in suspected appendicitis. The result of the frequent use of diagnostic laparoscopy is that rare atypical findings such as abdominal fat tissue necroses can be detected consistently. The incidence of abdominal fat tissue necroses of 1.1% in patients with acute abdominal pain shows that these etiologies are rare but of definite clinical relevance. We assume that necroses of the omentum and the epiploic appendages are usually missed in open surgery, particularly in cases with suspected appendicitis. Therefore in patients presenting with acute abdominal pain, especially those with rare etiologies, laparoscopy represents a refined diagnostic procedure and closes a diagnostic gap. We are convinced that the use of laparoscopy will allow the diagnosis of abdominal fat tissue necroses more frequently in future. References
1. Adams JT (1973) Primary torsion of the omentum. Am J Surg 126: 102105 2. Angulo JM, Ruiz I, Villanueva A, San Vicente M, Tovar JA (1990) Patologia quirurgica primaria de epiplon. Cir Pediatr 3: 125129 3. Anton JI, Jennings JE, Spiegel MB (1945) Primary omental torsion. Am J Surg 68: 303 4. Bernatz PE (1956) Unusual conditions simulating acute appendicitis. Proc Staff Meet Mayo Clin 31: 5356 5. Carmichael DH, Organ CH (1985) Epiploic disorders: conditions of the epiploic appendages. Arch Surg 120: 11671172 6. Chung SCS, Ng KW, Li AKC (1992) Laparoscopic resection for primary omental torsion. Aust NZ J Surg 62: 400401 7. Crofoot DD (1980) Spontaneous segmental infarction of the greater omentum. Am J Surg 139: 262264 8. Diaco JF, Diaco DS, Brannan AN (1993) Endoscopic removal of an infarcted appendix epiploica. J Lap Surg 3: 149151

Fig. 1. Omental necrosis in the right upper quadrant in a patient with clinical signs of acute appendicitis (patient 2). Fig. 2. Omental necrosis in a patient with acute abdominal pain (patient 3). Histopathology subsequently found the appendix to be normal. Fig. 3. Torsion of an epiploic appendage of the sigmoid colon in a patient with acute abdominal pain (patient 5).

740 9. Epstein LI, Lempke RE (1968) Primary idiopathic segmental infarction of the greater omentum: case report and collective review of the literature. Ann Surg 167: 437443 10. Federmann G, Maeo-Guitierrez C, Schneider A (1991) Torsion von Netz oder Appendix epiploicaseltene Ursachen eines akuten Abdomens. Chirurg 62: 571 11. Giuly J, Franc ois GF, Gaujoux J, Reynaud B (1991) Infarctus idiopathique segmentaire du grand e piploon. J Chir (Paris) 128: 204206 12. Klaiber Ch, Metzger A, Petelin JB (1993) Manual of laparoscopic surgery. Hogrefe and Huber, Bern 13. Lynn TE, Docherty MB, Waugh JM (1956) A clinicopathological study of the epiploic appendages. Surg Gynecol Obstet 103: 423433 14. Mattarelli G, Roth B, Rossetti M (1985) Omentale und epiploische Torsion in der Differentialdiagnose des akuten Abdomens. Helv Chir Acta 52: 665669 15. Paterson-Brown S, Eckersley JRT, Sim AJW, Dudley HA (1986) Laparoscopy as an adjunct to decision making in the acute abdomen. Br J Surg 73: 10221024 16. Puppala AR, Mustafa SG, Moorman RH, et al. (1981) Small-bowel obstruction due to disease of epiploic appendage. Am J Gastroenterol 75: 382383 17. Puylaert JB (1992) Right-sided segmental infarction of the omentum: clinical, US and CT findings. Radiology 185: 169172 18. Schnur PL, McIlrath DC, Carney JA, Whittaker LD (1972) Segmental infarction of the greater omentum. Mayo Clin Proc 47: 751755 19. Tolenaar PL, Bast TJ (1987) Idiopathic segmental infarction of the greater omentum. Br J Surg 74: 1182 20. Vertuno LL, Dan JR, Wood W (1980) Segmental infarction of the omentum. Am J Gastroenterol 74: 443446

Surg Endosc (1997) 11: 707710

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Acidbase balance alterations in laparoscopic cholecystectomy

V. Ga ndara,1 D. S. de Vega,2 N. Escriu ,1 I. Garcia Zorrilla1
1 2

Service Anesthesiology, General Hospital of Mostoles, Madrid, Spain Complutense University of Madrid, Madrid, Spain

Received: 25 January 1996/Accepted: 29 May 1996

Abstract Background: The purpose of this study is to determine alterations of acidbase balance originated by pneumoperitoneum with CO2. Influence of other factors such as anesthetic technique, duration of procedure, and volume of CO2 insufflated has also been analyzed. Methods: Some 132 patients were divided in three groups according to anesthetic technique used. Arterial blood gases were determined before pneumoperitoneum, at 20 min after it, and every 30 min, until procedures end, and in postoperative period up to a total of four samples. Results: Pneumoperitoneum originated a fall of pH (p < 0.001), ion bicarbonate (p < 0.001), and base excess (p < 0.001) and an elevation of PaCO2 (p < 0.001). No correlation was found between these changes and duration of pneumoperitoneum or amount of CO2 insufflated. Changes were fundamentally of a metabolic type. There were no statistically significant differences among anesthetic techniques. Conclusions: In conclusion, pneumoperitoneum with CO2 originates alterations of the acidbase balance, mostly of a metabolic type. This could mean that besides CO2 absorption, there is a tissular hypoperfusion due to the increase of abdominal pressure. Key words: Acidbase balance Laparoscopic surgery Pneumoperitoneum

tions periods usually required in digestive laparoscopic procedures may imply some added risks [13, 19]. Abdominal insufflation with carbon dioxide (CO2) for laparoscopic techniques implies some degree of hypercapnia and respiratory acidosis. This, which in gynecological techniques is almost of no clinical relevance [9, 14] (because short insufflation periods are used in usually young and healthy women) may imply a greater risk in digestive laparoscopic procedures. The longer insufflations periods, in patients not always young and healthy, may lead to an important imbalance of the acidbase balance (ABB) due to an increase in abdominal pressure and CO2 absorption through the peritoneum serosa [11, 13]. Controversy exists, among different authors, over which factor in the ABB is basically disruptedthe respiratory or the metabolic one. In this situation, anesthetic drugs may play some sort of role, given their respiratory and hemodynamic effects. Our study analyzes the ABB alterations produced by laparoscopic cholecystectomy and its influence on the immediate postoperative period. The influence of other factors employed in each proceduresuch as the duration, total amount of CO2 and anesthetic techniqueis also analyzed. Materials and methods
A prospective study was carried out with 132 patients referred for laparoscopic cholecystectomy. Patients included were previously selected according to their pathologyonly uncomplicated biliary lithiasis, and ASA I and ASA II (anesthetic risk) patients were included. Those who needed of a laparotomy conversion for their cholecystectomy were also excluded. Patients were divided into three groups, in a randomized way, according to the anesthetic technique employed. In group I nitrous oxide/oxygen (O2) (60%/40%) and in group II isoflurane 1% in O2/air (FiO2 0.4) were administered. In group III propofol in continuous infusion (7 mg kg1h1 the first 30 min and then 5 mg kg1h1) with O2/air (FiO2 0.4) was employed. Other drugs employed were similar in all three groups. As preanesthetic medication, droperidol (30 g kg1), atropine (0.01 mg kg1), and fentanyl (2 g kg1) were intravenously administered. Anesthetic induction was achieved with thiopental in groups I and II (56 mg kg1) and with propofol in group III (2 mg kg1). For intubation, succinylcholine (1 mg kg1) was administered. A bolus of atracurium followed by a continuous infusion (5 mg kg1h1) was employed to obtain muscular relaxation. Fentanyl (35 ug kg1h1) was used as analgesic. For mechanical ventilation a volumetrical respirator (EXCEL 210 Ohmeda) was em-

Digestive laparoscopic surgery is one of the latest endoscopic techniques to be applied and offers certain advantages compared to the traditional surgical ones. Among them are a better aesthetic result and a faster postoperative recovery period that allows an earlier hospital discharge of the patient [6, 16]. However, compared to other laparoscopic techniques, and regarding anesthetic, the longer insuffla-

Correspondence to: V. Ga ndara, C/ Saliente n 6. Urb. Montealina, Pozuelo de Alarcon, 28223, Madrid, Spain

708 Table 1. Patient group characteristics Global Patients (number) Age (years) Sex (M/F) Weight (kg) ASA (I/II) Intraabdominal pressure (mmHg) CO2 volume (l) Surgery duration (min) 132 46 12 27/105 66 11 76/56 14 1.3 57 17 99 35 N2O 44 46 12 9/35 68 11 25/19 14 1.3 58 16 101 33 Isoflurane 44 47 13 9/35 67 10 23/21 14 1.1 57 16 99 34 Propofol 44 46 13 9/35 64 11 28/16 13.7 1.5 56 18 96 38

ployed. To better appreciate the peritoneum absorption of CO2, respiratory parameters were not modified during the procedures if not necessary, being previously established in a tidal volume of 10 ml kg1 and a respiratory rate 12 breaths/min. Intraoperative venous fluid therapy was supported with crystalloid solution at a infusion rate of 6 ml kg1h1, avoiding lactatecontaining ones to prevent altering the ABB. At the end of the procedure, all the patients were translated to the recovery room awake, exubated, and with spontaneous breathing. Pneumoperitoneum was established with CO2 insufflated through a Veress needle, with an intraabdominal pressure of between 13 and 15 mmHg, sustained through all the surgical procedure. Patients were positioned in dorsal decubitus with a slight reversed Trendelenburg and with the lower limbs in the French position. Arterial blood gases were obtained previously to the pneumoperitoneum, 20 min after it, and then every 30 min until the procedure was finished (PP, P20, P60, P90, P120). Additional measures were obtained at recovery-room arrival and then every 30 min until four samples were taken (R, R30, R60, R90). Regarding criteria of acidbase changes, a blood pH range of 7.357.45 was considered normalgreater and lower values were considered alkalotic and acidotic, respectively. Alkalosis was considered of respiratory origin when PaCO2 decreased under 35 mmHg with a normal serum bicarbonate and of metabolic nature when PaCO2 was normal and bicarbonate was above 24 mEq/l. Acidosis was considered respiratory in origin when PaCO2 was greater than 40 mmHg with a normal bicarbonate and as metabolic acidosis when there was a lowering of the pH due to a decreased level of bicarbonate and PaCO2 was within normal limits. Mixed acidosis and alkalosis where considered when both parameters were simultaneously altered. For statistical analysis a chi-square test was used for qualitative variables and the Students t test was used for the quantitative ones. ANOVA has been used to compare groups, applying a Newman-Kuhls test if the contrast was significant. Pearsons correlation coefficient was calculated to determine the correlation between two quantitative variables. Only the determinations at recovery-room arrival were employed to correlate the blood analysis changes with the duration of the procedures and the total amount of CO2 used. Values were expressed in terms of their mean and standard deviations. Significance level was established at 0.05, accepting the hypothesis for lower values and rejecting it for higher ones.

Fig. 1. Illustration of changes in pH from prepneumoperitoneum phase (PP), during pneumoperitoneum (P20P120), and in postoperative period in recovery room (RR90). Falls in a significant way (p < 0.001) with establishment of pneumoperitoneum (P20) and at arrival in the recovery room (R). At 90 min after the postoperative process (R90), values reach close to normal.

Fig. 2. Illustrates PaCO2. A statistically significant increase is observed (p < 0.001) with production of pneumoperitoneum (P20) and on its withdrawal (R). During this time, a stabilization is observed.

Results Age, sex, weight, anesthetic risk, intaabdominal pressure, total amount of CO2 insufflated, and duration of procedures (Table 1) were comparable between groups. Blood pH lowered significantly (p < 0.001) with pneumoperitoneum from its first determination, reaching its lowest level at the recovery-room arrival determination (p < 0.001). From this moment on, a significant gradual increase was observed, with nearly normal values 90 min later. No correlation was found between these parameters and duration of procedures (r 0.083) or total amount of CO2 used (r 0.012). No significant differences were found between the anesthetic groups (Fig. 1).

PaCO2 increased significantly with pneumoperitoneum (p < 0.001). There was a slight trend to increase thereafter, and it stabilized after 60 min; most patients kept CO2 within normal levels. At recovery-room arrival a maximum increase first (p < 0.001) and then a progressive decrease were observed (Fig. 2). No correlation was found between the PaCO2 and duration of surgery (r 0.1312) or total amount of CO2 used (r 0.1981). None of the patients presented hypercapnia before the pneumoperitoneum. Thirty-six patients (27%) presented it during pneumoperitoneum and 74 (56%) at recovery-room stage. Bicarbonate decreased significantly with pneumoperitoneum (p < 0.001) insufflation and continued to decrease all along it. At recovery-room arrival an increase (p < 0.01) was initially observed and then posterior progressive normalization (Fig. 3). No correlation was found between plasma bicarbonate at recovery-room arrival and duration of procedure (r 0.0244) or total amount of CO2 used (r 0.1988). Base excess also decreased significantly during pneu-

709 Table 2. Alterations of acidbase balance Acidosis Respiratory Metabolic Mixed Total PPN 0 3(2%) 0 3(2%) PN 10(8%) 58(44%) 27(21%) 88(67%) Recovery 21(16%) 47(36%) 55(42%) 118(89%)

Discussion Longer-duration procedures such as are usually needed for digestive laparoscopic surgery, could cause an important alteration of the acidbase balance due to increase in intraabdominal pressure and CO2 absorption through the peritoneum serosa. According to some authors, the consequences could be hypercapnia and respiratory acidosis [2, 9, 11, 13, 14, 17]. Our study results, however, differ significantly, not only quantitatively but qualitatively, from other authors. In our study, we have been able to determine that pneumoperitoneum production with CO2 at a limited and constant intraabdominal pressure, in healthy patients (ASA I and II), produces significant ABB alterations such as a decrease in pH, bicarbonate, and base excess, and an increase of PaCO2. Consequently, a high incidence of acidosis was observed67% of patients in our series. This means a statistically significant difference when compared with the acidosis rate in previous stages (2%). Metabolic acidosis was the predominant type observed, as has been reported in other series [3, 8, 10]. Thus, during the CO2 insufflation period, 44% of patients presented a pure metabolic acidosis, and a 21% a mixed type, with only 8% pure respiratory. In spite of the global increase of PaCO2 during this period, most of the patients (72%) kept it within normal values. With pneumoperitoneum deflation, these changes were exacerbated; blood pH reached the lowest values and PaCO2 the highest ones. Acidosis increased to 89%, but with important qualitative changes: The respiratory type rose to 16%, the mixed type rose to 42%, and the metabolic type dropped to 36%. Some authors have stated that changes of pH and PaCO2 produced during pneumoperitoneum follow a progressive tendency, decreasing or increasing respectively, in regard to the time and the total amount of CO2 used [2, 11, 13]. The fact, in our study, that the greater rise in PaCO2 is observed at pneumoperitoneum insufflation and deflation, with an intermediate stable period, suggests that CO2 absorption is mainly produced in those two moments. This is supported by the lack of correlation between ABB alterations and the total amount of CO2 used or the procedure duration, as has also been reported by other series [1, 12, 15]. This could be explained in two ways: first, because patients are kept in hyperventilation, and second, because intraabdominal hyperpressure may produce compressive mechanical phenomena over capillaries, limiting CO2 absorption [1, 7, 12, 15]. The metabolic component of acidosis observed during pneumoperitoneum in this study has higher significance than respiratory factor. Critchley et al. has explained this fact as a decrease of cardiac output and consequently as a decrease in peripheral perfusion. Other reports support this finding. Metabolic changes could thus be justified by ischemic phenomena induced by the increased intraabdominal

Fig. 3. Illustrates changes observed in bicarbonate ion. A statistically significant fall is observed with pneumoperitoneum (p < 0.001) continued throughout it. There is an increase (p < 0.01) with posterior progressive normalization in the immediate postoperative time (RR90).

Fig. 4. Illustrates the changes in base excess. With establishment of pneumoperitoneum and throughout it (P20P120), a statistically significant drop is observed (p < 0.001); a raise occurs after its withdrawal (p < 0.001) in recovery room (RR90).

moperitoneum (p < 0.001), but normalized (p < 0.001) during the recovery period (Fig. 4). No correlation was found between these values and duration of procedure (r 0.255) or total amount of CO2 used (r 0.1514). Hemodynamically, during the pneumoperitoneum period there was a significant increase in mean arterial pressure, basically due to an increase in the diastolic arterial pressure. Arterial blood gases obtained previous to pneumoperitoneum insufflation showed 28 patients (21%) with respiratory alkalosis and three patients (2%) with metabolic acidosis. During insufflation 88 patients (67%) developed some type of acidosis; ten (8%) respiratory, 58 (44%) metabolic, and 27 (21%) mixed; seven patients developed more than one type of acidosis. At recovery-room arrival the acidosis rate increased to 89% (118 patients): 16% respiratory, 42% mixed with a greater respiratory component, and 36% metabolic, the only type whose rate decreased. No significant differences were found between anesthetic groups at any period. Percentage data of each group are shown in Table 2.

710

pressure, as has been established in other studies. Thus Kotzampassi et al. [10] used helium vs CO2 in their experimental study to differentiate mechanical hyperpressure effects from the CO2 absorption ones. During peritoneal insufflation an increase in portal and inferior cava vein pressure (p < 0.0001), intestinal hypoperfusion with a decrease of 50% on jejunal blood flow (p < 0.0001), and tissular acidosis with a decrease in intramural jejunal pH from 7.4 to 6.8, were noticed in both groups, showing that tissular ischemia is inescapable. Furthermore, only in the CO2 group was an increase in PCO2 levels in portal and inferior cava vein blood and arterial and mixed venous blood noticed. Ishizaki et al. [7] have noticed increases in portal and inferior cava vein pressures and a decrease in splachnic blood flow due to the increase in abdominal pressure. Joris et al. [8] have shown an increase in lactate levels during laparoscopic periods, which supports the presence of an anaerobic metabolism, probably due to ischemia tissue phenomena. Once the pneumoperitoneum was deflated, acidosis showed marked modifications in respect to level, rate, and type. Parameters measured at recovery room showed the lowest levels of blood pH, with a significant increase in acidosis rate in up to 89% of patients, especially of respiratory mechanism. These findings could be explained by the absence of mechanical hyperventilation; by the postoperative ventilatory depression; by an increase in absorption of CO2 due to cessation of the mechanical factor; and by progressive draining of the deposits of CO2, up to basal values [4, 5, 18]. Regarding influence of anesthetic technique on ABB alterations, no significant differences have been found between the three groups in our study. For this reason, we think the changes observed are exclusively due to pneumoperitoneum. In conclusion, pneumoperitoneum with CO2 at limited and constant pressure causes significant ABB alterations, basically of a metabolic type. This means that beside the CO2 absorption, phenomena of tissular hypoperfusion due to an increase in the intraabdominal pressure also play a role. No correlations were found between ABB alterations and the duration of procedure or total amount of CO2 used. Not even the anesthetic technique influenced the results. In healthy patients, these ABB alterations did not have clinical repercussions. References
1. Blobner M, Felber AR, Gogler S, Weigl EM, Jelen-Esselborn S (1992) Carbon dioxide uptake from pneumoperitoneum during laparoscopic cholecystectomy (Abst). Anesthesiology 77: A37

2. Bongard FS, Pianim NA, Leighton TA, Dubecz S, Davis IP, Lippman M (1993) Helium insufflation for laparoscopic operation. Surg Gynecol Obstet 177: 140146 3. Critchley LAH, Critchley JAJH, Gin T (1993) Haemodynamic changes in patients undergoing laparoscopic cholecystectomy: measurement by transthoracic electrical bioimpedance. Br J Anaesth 70: 681683 4. Farhi LE, Rahn H (1955) Gas stores of the body and the unsteady state. J Appl Physiol 7: 472480 5. Farhi LE, Rahn H (1960) Dynamics of changes in carbon dioxide stores. Anesthesiology 21: 604614 6. Grace PA, Quereshi A, Coleman J, Keane R, McEntee G, Broe P, Osborne H, Bouchier-Hayes D (1991) Reduced postoperative hospitalization after laparoscopic cholecystectomy. Br J Surg 78: 160162 7. Ishizaki Y, Bandai Y, Shimomura K, Abe H, Ohtomo Y, Idezuki Y (1993) Changes in splachnic blood flow and cardiovascular effects following peritoneal insufflation of carbon dioxide. Surg Endosc 7: 420423 8. Joris J, Honore P, Lamy M (1992) Changes in oxygen transport and ventilation during laparoscopic cholecystectomy (Abst). Anesthesiology 77: A149 9. Kelman GR, Swapp GH, Smith I, Benzie RJ, Gordon NLM (1972) Cardiac output and arterial blood-gas tension during laparoscopy. Br J Anaesth 44: 11551161 10. Kotzampassi K, Kapanidis N, Kazamias P, Eleftheriadis E (1993) Hemodynamic events in the peritoneal environment during pneumoperitoneum in dogs. Surg Endosc 7: 494499 11. Leighton TA, Liu SY, Bongard FS (1993) Comparative cardiopulmonary effects of carbon dioxide versus helium pneumoperitoneum. Surgery 113: 527531 12. Lister DR, Rudson-Brown B, Warriner B, McEwen J, Chan M, Walley K (1994) Carbon dioxide absorption is not linearly related to intraperitoneal carbon dioxide insufflation pressure in pigs. Anesthesiology 80: 129136 13. Liu SY, Leighton T, Davis I, Klein S, Lippmann M, Bongard F (1991) Prospective analysis of cardiopulmonary responses to laparoscopy cholecystectomy. J Laparoendoscopic Surg 1:241246 14. Motew M, Ivanovich AD, Bieniarz J, Albrecht RF, Zabed B, Scommegna A (1973) Cardiovascular effects and acid base and blood gas during laparoscopy. Am J Obstet Gynecol 155: 10021012 15. Mullet CE, Viale JP, Sagnard PE, Miellet CC, Ruynat LG, Counioux HC, Motin JP, Boulex JP, Dargent DM, Annat GJ (1993) Pulmonary CO2 elimination during surgical procedures using intra or extraperitoneal CO2 insufflation. Anesth Analg 76: 622626 16. Schirmer BD, Edge SB, Dix J, Hyser MJ, Hanks JB, Jones RS (1991) Laparoscopic cholecystectomy: treatment of choice for symptomatic cholelithiasis. Ann Surg 213: 665677 17. Schoeffler P, Haberer JP, Manhes H, Henry C, Habouzit JL (1984) Re percussions circulatoires et ventilatoires de la coelioscopie chez lobese. Ann Fr Anesth Re anim 3: 1015 18. Seed RF, Shakespeare TF, Muldoon MJ (1970) Carbon dioxide homeostasis during anaesthesia for laparoscopy. Anaesthesia 25: 223 231 19. Wittgen CM, Andrus CH, Fitzgerald S, Baudendistel L, Dahms T, Kaminski D (1991) Analysis of the hemodynamic and ventilatory effects of laparoscopic cholecystectomy. Arch Surg 126: 9971001

Surg Endosc (1997) 11: 772773

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Ileocutaneous fistula formation following laparoscopic polypropylene mesh hernia repair

K. Miller, W. Junger
Second Surgical Department, Landeskrankenanstalten, Mu llner Hauptstr. 48, A-5020 Salzburg, Austria Received: 7 October 1996/Accepted: 14 October 1996

Abstract. A rare case of enterocutaneous fistula caused by chronic erosion of polypropylene mesh after laparoscopic repair of a recurrent inguinal hernia is described. Successful treatment was achieved by fistulectomy, total resection of the implanted mesh, and small-bowel segmental resection. The patient recovered well postoperatively, and at followup 18 months later, the herniorrhaphy has remained intact. This complication needs to be added to the differential diagnosis in patients who present inflammation, abscess formation, or cutaneous fistula following laparoscopic hernia repair. Key words: Enterocutaneous fistula Polypropylene mesh Laparoscopic hernia repair

Case report
O.B. is a 67-year-old male with a prior history of six inguinal hernia operations on the left side. The patient was admitted for a laparoscopic preperitoneal hernia repair. The first herniorrhaphy had been performed 30 years ago and was closed with Bassinis technique. The recurrent hernias were closed by primary repairs two times with and four times without resorbable sutures. A laparoscopic transabdominal preperitoneal herniorrhaphy was now performed for the seventh operation of the recurrent inguinal hernia. We used a 6-inch by 6-inch (15 15 cm) piece of polypropylene mesh (Prolene Mesh, Ethicon, Inc., Somerville, NJ). The peritoneum was properly closed by a running suture with an extracorporal knotting technique. The patient was doing well until 1 year prior to the present admission, when he developed a subcutaneous abscess formation in the left lateral inguinal scar. Incision and drainage of the abscess formation were done. The patient was doing well at 3 months; then he developed a draining suppurative tract and chills. There was no history of feculent drainage from the tract; cultures of this drainage grew mixed flora. Laboratory evaluation and chemistry profile except for white blood cell count (11.350 U/l) were normal. A fistulogram with Gastrografin injection into the sinus tract did reveal a fistula communicating with the ileum (Fig. 1). The operative treatment was achieved by fistulectomy, small-bowel segmental resection, and total resection of the implanted polypropylene mesh (Fig. 2). The laparotomy was closed by primary repair with Ethipond 0. A closed suction drain brought out inferior and lateral to the incision was applied. Perioperative antibiotic treatment was done with amoxicillin 2 g and clavulanic acid 200 mg (Augmentin 2,2 g) twice a day for 1 week. The patient recovered well postoperatively and at follow-up 18 months later, the herniorrhaphy has remained intact.

The use of polypropylene mesh (Marlex, C.R. Bard Inc., Billerica, MA) was initially reported by Usher as a means of repairing incisional and inguinal hernias [13]. The most common complication associated with polypropylene mesh is wound infection, and this is often managed conservatively with antibiotic therapy [5, 8, 13]. Fistula formation from a polypropylene mesh has been reported to occur particularly if the prosthetic mesh is placed adjacent to the stomach or in contact with the small or large intestine [24, 6, 10, 11]. We present the case of a patient who underwent laparoscopic transabdominal preperitoneal herniorrhaphy of a recurrent inguinal hernia, complicated 1 year later by an ileocutaneous fistula. Factors that may predispose to fistula formation and techniques to prevent this complication are discussed.

Discussion There are numerous synthetic meshes in current use. The polypropylene mesh is the most widely used artificial prosthesis commonly used for laparoscopic herniorrhaphy. To our knowledge, ours is the first reported case of an ileocutaneous fistula after laparoscopic herniorrhaphy. Fistulas can be caused by radiation, neoplasm, inflammation, trauma, or infection. The inflammatory reactions of subcutaneously implanted polypropylene (Marlex) and polytetrafluoroethylene (Gore-Tex) were compared by Stockeld et al. [12]. They found that the polypropylene mesh was associated with significantly greater chronic inflammatory reaction and fibrosis than was Gore-Tex. Kara-

Correspondence to: K. Miller

773

rials for repair of abdominal wall defects, that adhesion of the intestine to the biomaterial is the first stage of biomaterial-related intestinal fistula. Its prevention is logical for the elimination of this complication. The diagnostic management of fistula formation could be difficult and Gastrografin fistulogram could fail accurate diagnoses performed the first time [2]. Factors that may have contributed to the fistula formation in our case might include close contact between the mesh and the bowel, resulting in low-grade infection of the mesh. We believe that the mesh eroded through the visceral peritoneum, caused an inflammatory reaction and local infection of the ileum, and penetrated into the lumen. This resulted in sinus tract formation and development of an enterocutaneous fistula. This complication needs to be added to the differential diagnosis in patients who present inflammation, abscess formation, or cutaneous fistula following laparoscopic hernia repair. References
1. Amid PK, Shulman AG, Lichtenstein IL, et al. (1994) Experimental evaluation of a new composite mesh with selective property of incorporation to the abdominal wall without adhering to the intestines. J Biomed Mater Res 28: 373375 2. DeGuzman LJ, Nyhus LM, Yared G, Schlesinger PK (1995) Colocutaneous fistula formation following polypropylene mesh placement for repair of a ventral hernia. Endoscopy 27: 459461 3. Karakousis CP, Volpe C, Tanski J, et al. (1995) Use of a mesh for musculoaponeurotic defects of the abdominal wall in cancer surgery and the risk of bowel fistulas. J Am Coll Surg 181: 1116 4. Kaufmann Z, Engelberg M, Zager M (1981) Fecal fistula: a late complication of Marlex mesh repair. Dis Colon Rectum 24: 5354 5. Lewis RR (1984) Knitted polypropylene (Marlex) mesh in the repair of incisional hernias. Can J Surg 27: 155157 6. MacMillan JI, Freeman JB (1984) Healing of a gastrocutaneous fistula in the presence of Marlex. Can J Surg 27: 159160 7. Matapurkar GB, Gupta AK, Agarwal AK (1991) A new technique of Marlex peritoneal sandwich in the repair of large incisional hernias. World J Surg 15: 768770 8. Molloy RG, Moran KT, Waldron RP, et al. (1991) Massive incisional hernia: abdominal wall replacement with Marlex mesh. Br J Surg 78: 242244 9. Nagy SW, Marshall JB (1993) Aortoenteric fistulas: recognizing a potentially catastrophic cause of gastrointestinal bleeding. Postgrad Med 93: 211222 10. Schneider R, Harrington JL, Granda AM (1979) Marlex mesh repair of a diaphragmatic defect later eroding into the distal oesophagus and stomach. Ann Surg 45: 337339 11. Seelig MH, Kasperk R, Tietze L, Schumpelick V (1995) Enterocutane Fistel nach Marlex-Netz-Implantation. Chirurg 66: 739741 12. Stockeld DG, Granstrom L, Backman L, Dahlgren S (1992) Inflammatory response to subcutaneously implanted Marlex and Gore-Tex in massively obese patients. Biomaterials 13: 261263 13. Usher FC (1962) Hernia repair with Marlex mesh. Arch Surg 84: 325328

Fig. 1. Fistulogram with Gastrografin injection into the sinus tract (arrow) reveals a fistula communicating with the ileum. Fig. 2. Intraoperative situs. Fistula formation to the ileum.

kousis et al. [3] reported a 23% enterocutaneous fistula rate in patients with abdominal wall defects due to ablative surgery for carcinoma when no tissue interposed between the bowel loops and the mesh. Patients with tissue interposition between the bowel loops and the mesh developed no enterocutaneous fistula. Matapurkar et al. [7] described a new technique in which the mesh was placed between the two layers of peritoneum of the overstretched hernia sac to prevent wound infection. To minimize infection, the mesh was placed so as to not come into contact with the viscera or anterior abdominal wall. The mechanism of fistula formation is similar whether enterocutaneous, colocutaneous, or aortoenteric. After prolonged prosthetic graft contact with the bowel wall, erosion resulting in infection or abscess formation and subsequent fistula formation can occur [9]. In an experimental study, Amid et al. [1] could demonstrate, using different biomate-

Surg Endosc (1997) 11: 714717

Surgical Endoscopy

Common bile duct repair with titanium staples

Comparison with suture closure
A. Leppa niemi,1 D. Wherry,1 E. Pikoulis,1 H. Hufnagel,1 C. Waasdorp,1 N. Fishback,2 N. Rich1
1 2

Department of Surgery, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814-4799, USA Department of Pathology, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814-4799, USA

Received: 25 September 1996/Accepted: 19 November 1996

Abstract Background: Vascular Closure Staple (VCS) clips made of titanium were initially developed for microvascular anastomoses with little experience of their use in larger tubular structures. This study compares VCS clips and sutures in the closure of supraduodenal choledochotomy. Methods: In nine pigs, two longitudinal incisions of the common bile duct (CBD) were randomly assigned to closure with 4-0 interrupted polyglactin sutures or VCS clips. Results: Clip closure was significantly faster (116 28 vs 581 88 s). All nine CBDs were patent and without signs of calculus formation after 3 months. Clip closure resulted in slightly less narrowing of the duct lumen and thinner scar at the repair site. At histological examination, all 18 incisions had healed without signs of fistula formation or marked fibrosis. Conclusions: Choledochotomy closure with VCS clips results in as good or better wound healing than suture closure, with a comparable degree of narrowing. The time required for clip closure is only about one-fifth that of suture closure. Key words: Choledochotomy Biliary reconstruction Metal clips

19]. In addition to closure of the supraduodenal choledochotomy after exploration for calculus disease, reconstruction of the biliary passages is required in the management of biliary strictures [2], traumatic and iatrogenic biliary tract injuries [6], and after orthotopic liver transplantation [14]. The emergence of laparoscopic cholecystectomy has refocused the interest in iatrogenic bile duct injuries [3, 4] and the technical challenges of laparoscopic common duct exploration [5, 9, 13], emphasizing some of the problems with traditional suture reconstruction of biliary passages, whether performed during open or laparoscopic surgery. A new sutureless technique for microvascular anastomoses utilizing nonpenetrating, arcuate-legged titanium Vascular Closure Staples (VCS clips) applied with a special applier to evert the edges of the structures being connected was developed by Kirsch and Zhu [20]. In a preliminary study using a porcine model, we showed that the VCS clips are easily and quickly applied also to larger tubular structures and can be safely used to close linear incisions of major vascular structures, ureters, and the gallbladder [12]. The purpose of this study was to compare the VCS clips with sutures in the closure of choledochotomy with special reference to closure times, degree of narrowing, and stricture formation during a follow up period of 3 months.

Soon after open cholecystectomy was introduced by Langenbuch in 1882 to treat gallstone disease [11], surgeons in the United States and Europe developed and refined techniques for exploring and repairing the biliary passages [16,

Materials and methods

Nine Yorkshire pigs weighing 2738 kg were used. Animal care complied with the Principles of Laboratory Animals Care (formulated by the National Society for Medical Research) and the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86-23, revised 1985). The VSC clips are manufactured by the United States Surgical Corporation (Norwalk, CT) and are currently available in three sizes; large (clip dimension at the tip 2.0 mm), medium (1.4 mm), and small (0.9 mm). The animals were subjected to midline laparotomy under general isoflurane anesthesia and aseptic conditions. For infection prophylaxis, the animals received one dose of ceftriaxone sodium (1 g intravenously) at the beginning of operation. At operation, the common bile duct was exposed and two sequential 1-cm-long anterior longitudinal incisions were performed leaving a strip of intact duct of 0.51 cm between the incisions. The

Disclaimer: The opinions expressed herein are those of the authors and are not to be construed as reflecting the views of the Uniformed Services University of the Health Sciences, the Department of the Army, or the Department of Defense Correspondence to: A. Leppa niemi, Second Department of Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland

715 Table 1. Closure of a supraduodenal choledochotomy with VCS clips or interrupted 4-0 polyglactin sutures in a porcine model (n 9) Parameter studied Closure time (seconds) No. of clips/sutures required Inner diameter (mm) Inner diameter (% of intact duct) Smooth mucosal surface Wall thickness (m) VCS clip 116 28 10.2 1.0 5.2 0.4 104 4 4/9 628 85 Suture 581 88 7.2 0.9 4.6 0.6 92 5 9/9 883 172 p <0.001 0.036 0.423 NS 0.089 NS 0.015 0.164 NS

choledochotomy was closed with 614 (mean 10.2) medium-sized VCS clips or 513 (mean 7.2) interrupted 4-0 polyglactin sutures. The method of repair of the proximal incision was randomly assigned to closure with clips or sutures by flipping a coin, and the other method was then assigned for closure of the distal incision, respectively. Clips were used for closure of the proximal duct in five and sutures in four animals. The animals were followed for a mean of 86 (7597) days and reoperated under general anesthesia for assessment of the patency of the common bile duct and signs of bile leakage. The common bile duct was harvested, and the animals were euthanizedwhile still under anesthesia using saturated potassium chloride. The harvested bile ducts were opened longitudinally at the opposite site of the repairs and the repair sites were assessed for surface irregularities, bare clips or suture material, or calculus formation under an operating microscope. The inner circumference of the duct was measured and used to calculate the inner diameter. The inner diameter of the intact duct between the sequential incisions was used as a control value for each duct. The tissues were fixed in formalin and the clips were removed by microdissection. Multiple sections of the specimens were taken for histological examination using standard hematoxylin-eosin (HE) staining and were reviewed by a pathologist (N.F.) to assess the quality of healing at the repair sites and to measure the thickness of the wall at the thickest point of the repair site. Because of the distinct markings left by the clips to the specimens, the pathologist could not be blinded to the method of closure. The data are presented as mean SE. Unpaired t-test for continuous variables and Fischers Exact test for proportions were used for statistical analysis with significance at the 0.05 level. Fig. 1. Two longitudinal common bile duct incisions closed with sutures (on the left, arrow) and VCS clips (on the right).

Results The results are summarized in Table 1. The time required for a bile-tight closure of the choledochotomy (Fig. 1) was considerably shorter using clips in spite of a higher number of clipscompared with the number of interrupted suturesrequired for closure. After 3 months, all nine common bile ducts were patent and without macroscopic signs of bile leakage or fistulas (Fig. 2). Compared with sutured sites, the inner diameter of the duct was larger after clip closure, although the difference was not statistically significant (Table 1). There was no significant difference in the inner diameter between proximal and distal closure sites (4.9 1.5 mm vs 4.8 1.6 mm, p 0.888). The surface was smooth in all nine specimens closed with sutures but in 5/9 specimens closed with VCS-clips one or two clips were at least partially uncovered with mucosa. Of a total of 92 clips inserted, eight (8.7%) eventually remained uncovered after 3 months. No bile calculi were seen. The histological examination demonstrated intact mucosa (except for the bareclips in five specimens), healing with mild fibrosis in the submucosa and continuous serosa (Fig. 3). No evidence of microscopic fistula formation was seen. Mild or moderate mononuclear infiltrate was seen in

Fig. 2. Mucosal site of the common bile duct 3 months after repair of two linear incisions with VCS clips (middle part of the incision, marked with two needles) or sutures (marked with one needle).

2/9 specimens after suture closure and in 1/9 after clip closure. Prominent serosal vessels were seen in one specimen after clip closure. Although the duct wall was slightly thicker after suture repair (Table 1), none of the specimens showed extensive fibrosis or stricture formation.

Discussion In 1906, Robson closed the incised common bile duct with a continuous catgut suture being used for the margins of the duct proper, and a continuous fine green catgut or spun celluloid thread being employed to close the peritoneal edges of the duct [16]. Over the years, the closure methods have changed little [7, 18]. In this study, the standard suture closure of a longitudinal common bile duct incision resulted in reliable and uncomplicated closure with slight narrowing of the inner diameter of the duct, smooth mucosal surface, and no marked

716

Fig. 3. H&E photomicrograph of a common bile duct after incision and closure with VCS clips. The epithelium is continuous and the submucosa is intact. There is no inflammatory response and there is revascularization of the serosa.

stricture or calculus formation. Although the differences were not statistically significant, closure with VCS clips tended to produce less narrowing and a thinner scar, which could be attributed to the nonpenetrating nature of the VCS clip. The occurrence of bare clips in some of the specimens was thought to be caused by slipping of the edge of the duct at insertion rather than the clip penetrating the duct wall, as discussed below. The benefit of not perforating the entire thickness of the ductal wall with a needle has been previously emphasized [18]. In our preliminary study we demonstrated that some of the VCS clips used to close incisions of ureter and gallbladder were exposed to the lumen after a 1-month follow-up, and this is thought to be caused by technical problems (slipping of the edge) in applying the clips to thin-walled structures rather than pressure necrosis [12]. The occurrence of bare clips seems to be much rarer in blood vessels, where the accurate alignment of the edges is facilitated by the partially occluding vascular clamp. In this study, about 9% of the clips applied were devoid of mucosal cover at reoperation 3 months later. Although there was no calculus formation adjacent to the bare clip or elsewhere in the duct, the long-term effects and fate of the exposed clip remain unknown. However, the ability of the biliary epithelium to regenerate over various types of biological and synthetic materials including metal has been demonstrated [1, 8]. Quick and effortless closure is the main benefit achieved with the clips. Although in this study we used an interrupted suture technique, which is slower than continuous suturing, the clip closure retains its time-consuming advantage when compared to continuous suture by a 48-fold margin, as has been demonstrated in our laboratory using continuous sutures and VCS clips to close linear vascular incisions (data not shown). In addition to speed, another advantage of the clips in open surgery is associated with the ability of the surgeon to apply the clips in confined spaces, which could be useful,

for example, in the repair of traumatic injuries of the right or left hepatic duct [6]. To preserve the minimal invasiveness during laparoscopic cholecystectomy and to assure the complete treatment of cholecysto- and choledocholithiasis, laparoscopic exploration of the common bile duct has gained increasing popularity [5, 9]. The methods employed for laparoscopic common bile duct closure have been modified from open surgery and include the placement of interrupted 3-0 or 4-0 polyglycolic acid, polyglactin, or polydiaxone sutures [5, 9, 13]. Laparoscopic extracorporeal or intracorporeal knottying, however, is cumbersome and time-consuming [13]. In addition, several authors have pointed out the problems associated with laparoscopic suture closure of the common bile duct. Dion et al. [5] reported two patients with narrow ducts prior to choledochotomy who showed a stenosis of at least 60% on the postoperative cholangiogram caused by one of the stitches. They also pointed out that the part taking most of the total time, averaging 172 min, was the choledochoscopy and closure of the choledochotomy. In a series of 35 patients undergoing successful laparoscopic choledocholithotomy, seven complications (18%) including three major complications were recorded [9]. In one patient with late stricture, bleeding from the edge of the common bile duct was checked with excessive electrocautery and endosuturing. In addition, laparoscopically formed knots may be weaker than those tied by hand, although the results are controversial [10]. Although the VCS clips are not currently available for laparoscopic use, the quick and effortless application of the clips could have a significant influence in reducing the closure time of the choledochotomy after laparoscopic exploration. In addition, the magnifying effect of the laparoscope would facilitate the accurate alignment of the duct edges and placement of the clips, and, thus, avoid the slipping and exposure of the clip to the lumen. Among the more complex repairs of the biliary tract,

717

end-to-end anastomosis, usually performed with a single layer of interrupted 5-0 absorbable sutures, is associated with a 4050% long-term failure rate most likely related to technical problems and the relatively poor local blood supply [14, 15, 17]. Potentially, the VCS clips could have an advantage over conventional suture techniques, because the clips create an everted non-constricting closure line and are easy to apply, requiring little mobilization of the duct ends. This, however, requires further study.
Acknowledgment. This study was supported by a grant from the US Surgical Corp., Norwalk, CT.

8.

9.

10.

11.

12.

References
1. Belzer FO, Watts JM, Ross HB, Dunphy JE (1965) Autoreconstruction of the common bile duct after venous patch graft. Ann Surg 162: 346355 2. Blumgart LH (1988) Benign biliary strictures. In: Blumgart LH (ed) Surgery of the liver and biliary tract. Churchill Livingstone, Edinburgh, pp 721752 3. Cuschieri A, Dubois F, Mouiel J, Mouret P, Becker H, Buess G, Trede M, Troidl H (1991) The European experience with laparoscopic cholecystectomy. Am J Surg 161: 385387 4. Deziel DJ, Millikan KW, Economou SG, Doolas A, Ko S-T, Airan MC (1993) Complications of laparoscopic cholecystectomy: a national survey of 4,292 hospitals and an analysis of 77,604 cases. Am J Surg 165:914 5. Dion Y-M, Ratelle R, Morin J, Gravel D (1994) Common bile duct exploration: the place of laparoscopic choledochotomy. Surg Laparosc Endosc 4: 419424 6. Feliciano D (1994) Biliary injuries as a result of blunt and penetrating trauma. Surg Clin North Am 74: 897907 7. Gunn AA (1988) Supraduodenal choledochotomy. In: Blumgart LH 13.

14. 15.

16. 17. 18. 19. 20.

(ed) Surgery of the liver and biliary tract. Churchill Livingstone, Edinburgh, pp 647659 Hartung H, Kirchner R, Baba N, Waldmann D, Strecker E-P (1978) Histological, laboratory, and X-ray findings after repair of the common bile duct with a Teflon graft. World J Surg 2: 639644 Huang S-M, Wu C-W, Chau G-Y, Jwo S-C, Lui W-Y, Peng F-K (1996) An alternative approach of choledocholithotomy via laparoscopic choledochotomy. Arch Surg 131: 407411 Kadirkamanathan SS, Shelton JC, Heworth CC, Iaufer JG, Swain CP (1996) A comparison of the strength of knots tied by hand and at laparoscopy. J Am Coll Surg 182: 4654 Langenbuch CJA (1882) Ein Fall von Exstirpation der Gallenblase wegen chronischer Cholelithiasis: Heilung. Berl Klin Wochenschr 19: 725727 Leppa niemi AK, Wherry DC, Soltero RG, Pikoulis E, Hufnagel HV, Fishback N, Rich NM (1996) A quick and simple method to close vascular, biliary and urinary tract incisions using the new Vascular Closure Staples. A preliminary report. Surg Endosc 10: 771774 Lezoche E, Paganini A, Feliciotti F, Chan R (1993) Laparoscopic suture technique after common bile duct exploration. Surg Laparosc Endosc 3: 209212 Lewis WD, Jenkins RL (1994) Biliary strictures after liver transplantation. Surg Clin North Am 74: 967978 Northover JMA, Terblanche J (1979) A new look at the arterial supply of the bile duct in man and its surgical implications. Br J Surg 66: 379384 Robson AWM (1906) The technique of operations on the biliary passages. Surg Gynecol Obstet 2: 574582 Rossi RL, Tsao JI (1994) Biliary reconstruction. Surg Clin North Am 74: 825840 Sawyers JL, Herrington JL Jr, Edwards WH (1965) Primary closure of the common bile duct. Am J Surg 109: 107112 Walzel P (1923) VII Zur Therapie des Choledochussteines. Arch Klin Chir 126: 321328 Zhu YH, Kirsch WM, Cushman R, Becker K, McCabe W, Kornfeld M, Saland L, Cooper VR (1985) Comparison of suture and clip for microvascular anastomoses. Surg Forum 36: 492495

Surg Endosc (1997) 11: 718721

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic management of acute biliary pancreatitis

C. Ballestra-Lopez, X. Bastida-Vila, C. Bettonica-Larran aga, F. Zaraca, M. Catarci
Laparoscopic Surgery Unit, Department of General and Digestive Surgery, Hospital de Bellvitge Princeps dEspanya, Barcelona, Spain Received: 1 July 1996/Accepted: 10 November 1996

Abstract Background: The appropriate management of acute biliary pancreatitis has evolved considerably over the past decades. The advent of laparoscopic surgery made it necessary to reevaluate the traditional algorithms. Methods: This study assesses the outcome of 40 patients laparoscopically treated for gallstone pancreatitis. The severity of pancreatitis was scored by clinical and biochemical evaluation and CT findings. Laparoscopic cholecystectomy was performed during the same admission in all cases. In no case was a preoperative endoscopic retrograde cholangiopancreatography (ERCP) performed. In 32 patients (80%) with mild acute pancreatitis interval cholecystectomy was less than 1 week (group A) and in eight patients (20%) with severe disease it was more than 7 days (group B). All patients underwent intraoperative cholangiography. Results: The rate of common bile duct (CBD) stones was 5% (two cases), both occurring in the group A. There was one perioperative death (2.5%) in group B and one late CBD injury (2.5%) in group A, not requiring surgery. Complication rate was significantly higher in group B (50%) than in group A (9.4%). Conclusions: We consider that treatment of mild-tomoderate acute biliary pancreatitis can be satisfactorily accomplished by laparoscopy with routine intraoperative cholangiography and laparoscopic treatment of bile duct stones, showing no mortality and low morbidity rate. Laparoscopic treatment of patients with severe acute pancreatitis deserves further investigation. Key words: Laparoscopic cholecystectomy Acute biliary pancreatitis Gallstone pancreatitis

pain associated with biochemical abnormalities to a systemic syndrome (multiple organ failure), and several criteria are proposed to predict the prognosis [13]. Studies have stratified the severity of pancreatitis based on multiple clinical and laboratory factors [18, 19] and/or on CT scan findings [5]. Over the past two decades the role of surgical therapy for biliary pancreatitis has evolved considerably. However, the timing of surgery is still matter of debate [10, 21, 22]. Cholecystectomy during the initial admission for gallstone pancreatitis avoids recurring bouts of pancreatitis and can reduce hospital stay compared with a more conservative policy [15, 18]. On the other hand, many surgeons delay cholecystectomy until the acute phase resolves. The relatively recent popularization of endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic sphincterotomy (ES), and now laparoscopic cholecystectomy (LC) and laparoscopic common bile duct (LCBD) exploration, promises to dramatically change the traditional surgical approach. Since its introduction, LC has gained rapid acceptance as the preferred management of symptomatic cholelithiasis. Laparoscopic removal of CBD stones, however, requires greater technical skill and experience. Furthermore, experience with laparoscopic surgery for acute complications of gallstones is limited for both acute cholecystitis and pancreatitis. Nevertheless, data suggest that laparoscopic procedures, in these situations, are feasible and safe, although technically demanding. In order to assess the efficacy of this policy, we reviewed the records of all patients admitted in a Spanish emergency unit with a diagnosis of acute biliary pancreatitis who underwent a laparoscopic procedure during the same hospitalization.

Patients and methods Gallstones and biliary sludge are the most common cause of acute pancreatitis in the United Kingdom and in certain parts of USA [8, 19]. The symptoms vary from an epigastric
From January 1992, through June 1994, 44 consecutive patients were admitted at our surgical emergency unit for acute gallstone pancreatitis. Four patients with severe necrotizing pancreatitis underwent laparotomic treatment with laparostomy and were therefore excluded from the study. In the remaining 40 patients (15 men, 25 women) mean age was 60.7 years (range 3189) for men and 63.1 (range 3185) for women. Data of these patients

Correspondence to: M. Catarci, Piazza D. Gnoli, 6, 00162 Rome, Italy

719 Table 1. CT scan findings in acute pancreatitis according to Balthazars classification Balthazars score A B C D E CT findings Normal picture Focal or diffuse pancreatic enlargement Peripancreatic tissue involvement Single fluid collection Two or more fluid collections; presence of pancreatic or peripancreatic gas

Table 2. Ransons score of 40 patients with acute biliary pancreatitis Positive Ransons criteria Fig. 1. Algorithm for management of acute biliary pancreatitis. US: ultrasonography; CT: computed tomography; LC: laparoscopic cholecystectomy; IOC: intraoperative cholangiography; CBD: common bile duct; IO ERCP: intraoperative endoscopic retrograde cholangiopancreatography. were prospectively recorded and then reviewed. During the same period, 412 laparoscopic cholecystectomies (134 men, 278 women, mean age 57.4 years, range 2483) were carried out by the same surgical team for cholelithiasis. The diagnosis of biliary pancreatitis was based on history, findings of abdominal and back pain, and abnormal laboratory tests. Severity of disease was assessed by Ransons criteria [18, 19] and the operative risk was evaluated according to the ASA score. Patients were managed according to the algorithm shown in Fig. 1. After clinical and biochemical evaluation, all patients underwent ultrasonography (US) of liver, biliary tract, and pancreas in order to determine the biliary etiology. Patients found to have more than three Ransons criteria or ultrasonographic pancreatic abnormalities underwent abdominal computed tomography (CT). The CT scoring system proposed by Balthazar et al. [5] was followed in order to stage and manage the disease (Table 1). LC with intraoperative cholangiography (IOC) was undertaken at the next available operating session, usually within 7 days from admission, in patients with mild-to-moderate pancreatitis (group A) who had positive Ransons criteria of 1 to 3 and Balthazars score of A, B, or C. According to our algorithm, unsuccessful LCBD exploration or stone clearance leads to conversion to an open procedure. Alternatively, it is possible to attempt an intraoperative ERCP in the case of single CBD stone impacted in the ampulla of Vater. In case of failure, laparotomy is mandatory. In severe diseasepatients who showed Ransons criteria of more than 3 or Balthazars score of DELC with IOC was performed after medical therapy brought about improvement and stabilization of the acute phase, usually 830 days after admission. Unsuccessful medical treatment within 72 h, with onset of pancreatic necrosis, determines, in our algorithm, an open procedure, namely a laparostomy [4]. Seven out of 40 patients had severe pancreatitis according to Ransons criteria (R 4). All Ransons criteria were available for each patient, and the score of the whole series is listed in Table 2. Abdominal US showed gallbladder stones in 35 cases. It failed to identify the gallbladder and the pancreas, due to intestinal gas, in five cases (12.5%). In selected cases, we decided to perform further investigations. Eighteen patients (45%) underwent abdominal CT with vascular enhancement within 24 h from admission. The indications to abdominal CT are shown in Table 3. According to Balthazars classification, abdominal CT found a normal picture in one patient (A); in eight patients it confirmed a pancreatic focal or diffuse enlargement (B); in five patients a peripancreatic tissue involvement (C); in four patients a single fluid collection (D). In no patient was pancreatic or peripancreatic gas demonstrated by CT, indicating severe necrotizing pancreatitis with liquefaction (E). Preoperative evaluation of patients for surgical risk was performed according to ASA score (Table 4). In no patient was preoperative ERCP performed. All patients underwent LC during the same hospitalization. Thirty-two patients (group A) were operated on within 7 days from admission based on operating room availability (early operation), and eight patients (group B) were operated on 830 days from admission, following resolution of the acute manifestations of pancreatitis (late operation). Group B included all seven patients with Ranson score 4 and one patient with Ranson 3 and Balthazar score D. Ranson2 Ranson3 Ranson4 No. of patients 11 22 7 % 27.5 55.0 17.5

Table 3. Indications for abdominal CT in patients with acute biliary pancreatitis Indication Ransons score > 3 Failure of US imaging Pancreatic edema or enlargement at US scan No. of patients 7 5 6 % 38.9 27.8 33.3

Table 4. ASA classification of patients that underwent surgeery for acute biliary pancreatitis and for cholelithiasis Acute biliary pancreatitis ASA classification 1 2 3 4 No. of patients 8 15 14 3 % 20.0 37.5 35.0 7.5 Cholelithiasis No. of patients 164 138 98 12 % 39.8 33.5 23.8 2.9

The mean time until operation was 3.4 days in the early operation group and 15 days in the late operation group. All patients were examined 30, 60, and 90 days after hospital discharge, and then underwent abdominal CT control.

Results There was no conversion to open procedure. Intraoperative complications rate among the 40 patients was 5% (two cases). In group A, two patients operated on after 5 and 7 days from admission had gallbladder bed bleeding during cholecystectomy. It was controlled by local application of biologic glue and did not require conversion. Cystic duct cholangiography was performed successfully in all cases. In two cases of group A it revealed filling defects consistent with CBD stones. In one patient, the stones were extracted laparoscopically by placing a stone basket through the cystic duct into the CBD. In the other patient a laparoscopic choledochotomy and stone extraction was performed and a T-tube was left in place. Mean operative time was 86 min (range 45210 min).

720

The longest operations were those requiring laparoscopic extraction of CBD stones. There was one death (2.5%) in group B during the perioperative period. One patient, 85 years old, with operative risk of ASA 4, Ranson 4, under anticoagulant therapy, showed a cerebral thrombosis during the 3rd postoperative (p.o.) day and died within 24 h. Overall postoperative complications rate was 17.5% (7/ 40), significantly lower in group A (3/32; 9.4%) than in group B (4/8; 50%; p < 0.03, chi-square test). Major p.o. complications occurred in four patients (10%). One patient (group B) required a laparotomic reexploration. The patient developed intraabdominal bleeding from the site of trocar insertion and gallbladder bed detected by US scan. Relaparotomy at the 10th p.o. day showed a cul-de-sac abscess. Another patient (group A) showed a late CBD stenosis 2 months after surgery, probably due to monopolar electrocautery. The patient underwent ERCP and CBD pneumatic dilation that resolved the stenosis without subsequent complications nor sequelae. Two patients (group B) had p.o. minor bile leaks drained through abdominal drainage placed at the time of operation. Leakage persisted for longer than 7 days, but drainage was less than 100 ml per day, allowing discharge of both patients by the 15th p.o. day, and draining spontaneously ceased after 40 and 50 days, respectively. Minor complications occurred in three patients (7.5%). One patient (group A) experienced wound hematoma at the site of trocar insertion and another patient (group B) a selfresolving right basal pulmonary atelectasis. There were no bouts of pancreatitis, but one patient (group A) showed an asymptomatic 6-cm pancreatic pseudocyst at the abdominal CT control. The mean p.o. stay was 8.4 days. There was no significant difference in length of p.o. stay between those operated on earlier and those operated on later (7.9 vs 8.7). patients operated on earlier, therefore, had a significantly shorter total hospitalization (11.3 vs 23.7). There was no patient with retained CBD stones.

Discussion Management of acute biliary pancreatitis has evolved considerably over the past two decades. Delayed operative intervention was supplanted by early cholecystectomy. This policy avoids recurring bouts of pancreatitis and shortens the hospitalization. Stratifying patients into mild and severe categories is, in our opinion, helpful in directing appropriate management. Furthermore, this stratification allows direct comparison between groups studied at different institutions and comparison between different therapeutic modalities. For this purpose, in our series we adopted the Ranson score for the clinicobiochemical evaluation and, in selected cases, the Balthazar classification for the radiological evaluation. Abdominal US may be used to assist in the diagnosis of acute pancreatitis. The gallbladder can be assessed for stones and the CBD can be evaluated for size and the presence of stones. Unfortunately, the value of US is often limited by the presence of intestinal air obscuring biliary structures and the pancreas in almost one-third of patients with

acute pancreatitis [12]. Currently the best method to stage the acute pancreatitis is CT. Specific CT findings can be categorized into pancreatic and peripancreatic changes (Table 1). Pancreatic changes include diffuse or focal parenchymal enlargement, edema, or necrosis with liquefaction. Peripancreatic involvement includes blurring or thickening of the surrounding tissue planes. An approximate correlation exists between the degree of CT abnormalities and the clinical course and severity of acute pancreatitis. Biliary pancreatitis is probably due to the transient blockage of the ampulla of Vater by a migrating stone. Studies showed that the stone migrates spontaneously to the duodenum in the majority of cases within hours of the onset of pancreatitis [2]. Anyway, it is difficult to detect which stones are passed and which are going to cause persistent obstruction of the ampulla of Vater. These concepts have led some authors to suggest early systematic exploration of the biliary tract to determine if a stone is present and to remove it [23]. Several reports have proposed that ERCP could be safely performed in these situations, followed by LC during the same admission [3, 14, 23]. In our opinion, although ERCP with ES and stone extraction has been shown to be useful for early treatment of severe biliary pancreatitis [6], the incidence of bile duct stones at elective surgery is low and most of these ERCP are unnecessary. Accurate predictors of CBD stones are required; studies have shown that the sensitivity of preoperative abdominal US for predicting CBD stones is 42% and specificity is 86% [20]. Furthermore, an endoscopic approach is unable to fully resolve the patients biliary pathology with one procedure and one anesthesia. This adds substantial risk of morbidity and even mortality. Concern remains also regarding the potential long-term risks of ES. Although the immediate complications of ES are well documented, the long-term effects are less defined. Stricture formation and stone recurrence account for nearly all longterm complications. The results of endoscopic vs surgical treatment of CBD stones are still matter of debate [9]. Therefore we decided to perform LC without preoperative ERCP. Timing of laparoscopic surgery in acute biliary pancreatitis depends upon the severity of the disease. In the case of mild pancreatitis it doesnt matter when, within 1 week, laparoscopic cholecystectomy is performed. However, in patients with severe pancreatitis, laparoscopic cholecystectomy, when performed within the 1st week after the onset of symptoms, as other authors have observed [22], places patients at increased risk of operative morbidity and technical complications. In these patients, the management of complications of pancreatitis is strongly advisable before cholecystectomy. Delaying surgery for more than a week after hospitalization, in our experience, does not adversely affect technical difficulty. Patients of group B did not show greater technical problems than those of group A; in both groups the conversion rate was nil. The rate of intraoperative complications was only 5% (two cases), both occurring in group A. Delaying surgery for several weeks in severe acute pancreatitis allows acute inflammation to settle down and might allow stones in the CBD to clear spontaneously. However, studies showed that approximately one-quarter of patients have symptomatic recurrence within 6 weeks if gallstones are untreated, and it increases with time [16, 17]. According to our algorithm it is mandatory to obtain an

721

i.o. cholangiogram of good quality during LC, since the risk of CBD stones is 1420% [1]. This strategy minimizes the need for CBD exploration and still achieves the goal of a limited hospital stay and the prevention of recurrence of pancreatitis. If CBD stones are found at IOC they should be treated laparoscopically if at all possible. In most instances, it should be possible to retrieve the stones via the cystic duct, since acute pancreatitis is usually caused by the migration of small stones. If this is not feasible, one alternative is to perform a laparoscopic choledochotomy. These cases have a rather long hospital stay and delayed return to work, but their level of p.o. pain is diminished. Our current impression is that this procedure is possible though technically demanding. In case of failure, traditional CBD exploration is mandatory. The morbidity and mortality rates of 17.5% and 2.5% recorded in our series are higher than those reported for elective LCs, but there was a clear shift toward higher operative risk in this series when compared to our elective laparoscopic cholecystectomies performed during the same period (Table 4), and they are fully within the ranges reported for LCs in patients with acute biliary pancreatitis [3, 7, 14, 20, 22, 23]. A clear correlation existed between the degree of severity of the pancreatitis and the rates of postoperative deaths and complications, being significantly higher in group B. While the early laparoscopic approach to mild-to-moderate acute biliary pancreatitis proved to be effective in this series (no mortality, 9.4% morbidity, and no bouts of recurrent pancreatitis in group A), as it did in prospective randomized trials of open surgery [10, 21], 12.5% mortality and 50% morbidity rates in group B (severe pancreatitis) may raise a question about the safety of delayed LC during the same admissionnamely, would patients in this group have fared better if they had been allowed to go home for elective readmission at a later date? This question already existed before the advent of LC, with some surgeons stressing the fact that once biliary pancreatitis occurs there is more than 50% risk that a new episode of acute pancreatitis will occur within 6 months [16, 17], and a prospective randomized trial in open surgery failed to give a definitive answer [10, 21]. Nevertheless, it should be stressed that group B in the present study comprised only eight patients, a number too small to draw any conclusion about laparoscopic approach to severe acute pancreatitis. We had one late CBD injury (2.5%), which resolved with endoscopic dilation. A higher incidence of CBD injury after LC was demonstrated compared with open cholecystectomy. Its frequency may relate directly to the experience of surgeons and to the degree of gallbladder inflammation [11] and it should warrant continued concern. In conclusion, this study suggests that patients with mild-to-moderate acute pancreatitis can be safely treated by early LC and IOC, with LCBD stone extraction if indicated, whereas a delayed laparoscopic approach during the same

admission to patients with severe disease, after proper medical management, deserves further investigation.
Acknowledgments. Marco Catarci was supported by C.N.R. Italy grant #AI91.00134.04.

References
1. Acosta JM, Rossi R, Galli MR, Pellegrini CA, Skinner DB (1978) Early surgery for acute gallstone pancreatitis: evaluation of a systemic approach. Surgery 83: 367370 2. Acosta JM, Pellegrini CA, Skinner DB (1980) Etiology and pathogenesis of acute biliary pancreatitis. Surgery 88: 118125 3. Amaral JF, Mateo R, DiMase J, Califano N (1993) Laparoscopic cholecystectomy in gallstone hyperamylasemia/pancreatitis. Surg Endosc 7: 141 4. Ballesta C, Bastida X, Nieto B, Sanchez Cano JJ, Bettonica C (1989) La laparostomia nel trattamento della pancreatite necrotico-emorragica. Chirurgia 5: 226230 5. Balthazar EJ, Robinson DL, Megibow AJ, Ranson JH (1990) Acute pancreatitis. Value of CT in establishing prognosis. Radiology 174: 331336 6. Carr-Locke DL (1993) Role of endoscopy in gallstone pancreatitis. Am J Surg 165: 519521 7. De lorio AV, Vitale GC, Reynolds M, Larson GM (1995) The roles of laparoscopic cholecystectomy and endoscopic retrograde cholagiopancreatography. Surg Endosc 9: 392396 8. Goodman AJ, Neoptolemos JP, Carr-Locke DL et al. (1985) Detection of gallstones after acute pancreatitis. Gut 26: 125132 9. Graham SM, Flowers JL, Scott TR et al. (1993) Laparoscopic cholecystectomy and common bile duct stones. Ann Surg 218: 6167 10. Kelly TR, Wagner DS (1988) Gallstone pancreatitis: a prospective randomized trial of the timing of surgery. Surgery 104: 600603 11. McMahon AJ, Fullarton G, Baxter JN, ODwyer PJ (1995) Bile duct injury and bile leakage in laparoscopic cholecystectomy. Br J Surg 82: 307313 12. McKay AJ, Imrie CW, ONeill J, Duncan JC (1982) Is an early ultrasound scan of value in acute pancreatitis? Br J Surg 69: 369372 13. Millet B, Fingerhut A, Gayral F, Zazzo JF, Brivet F (1992) Predictability of clinicobiochemical scoring systems for early identification of severe gallstone-associated pancreatitis. Am J Surg 164: 3238 14. Molina E, Wise Unger S, Barkin J, Goldberg R (1994) Combined ERCP and laparoscopic cholecystectomy for biliary pancreatitis. Surg Endosc 8: 264 15. Osborne DH, Imrie CW, Carter DC (1981) Biliary surgery in the same admission for gallstone associated acute pancreatitis. Br J Surg 68: 758761 16. Patti MG, Pellegrini CA (1990) Gallstone pancreatitis. Surg Clin North Am 70: 12771295 17. Pellegrini CA (1993) Surgery for gallstone pancreatitis. Am J Surg 165: 515518 18. Ranson JHC (1979) The timing of biliary surgery in acute pancreatitis. Ann Surg 189: 654663 19. Ranson JHC (1982) Etiologic and prognostic factors in human acute pancreatitis: a review. Am J Gastroenterol 77: 633638 20. Soper NJ, Brunt ML, Callery MP, Edmundowicz SA, Aliperti G (1994) Role of laparoscopic cholecystectomy in the management of acute biliary pancreatitis. Am J Surg 167: 4251 21. Stone HH, Fabian TC, Dunlop WE (1981) Gallstone pancreatitis: biliary tract pathology in relation to time of operation. Ann Surg 194: 305309 22. Tang E, Stain SC, Tang G, Froes E, Berne TV (1995) Timing of laparoscopic surgery in gallstones pancreatitis. Arch Surg 130: 496 500 23. Tate JJ, Lau WY, Li AKC (1994) Laparoscopic cholecystectomy for biliary pancreatitis. Br J Surg 81: 720722

SAGES guidelines
Surg Endosc (1997) 11: 789792

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Guidelines for the surgical practice of telemedicine

Overview SAGES has established guidelines for post residency surgical education in its Framework for post-residency surgical education and training. [1] That document provides a format for obtaining appropriate training, attaining competence, and gaining privileges for advanced surgical procedures not learned during a formal residency program. It is upon the foundation laid by that document that this Guidelines for the surgical practice of telemedicine has been constructed. The two instruments are designed to be complementary and not exclusionary; both are intended to promote high quality patient care. Telemedicine has been practiced for over 30 years [2]. Recent technological advances, however, have expanded the scope of medical interaction that may be achieved. Whereas consultative services, examination of still documents (photos, x-rays, slides, or ECGs), and interactive voice sessions previously defined the state of the art, the telemedical event may now involve live manipulations of patients and/or tissues at a distance. In fact, there are now many levels of health care-related interaction that may take place in the telecommunications medium: physicianto-physician consultation, physician-to-student (physician, nurse, other care giver) teaching, physician-to-patient examination and consultation, and physician-to-patient treatment [310]. This document provides guidelines for establishing policies and procedures to promote safe, high quality application of telemedicine technology to the practice of surgery. Recognizing that the technology and the practice of telemedicine is in the process of rapid development, SAGES expects this document to be a vibrant resource for those interested in the field. As such, it is expected to evolve as the state of the art evolves [1113]. Both surgeons and telemedicine facilities involved in the practice of intrastate and/or interstate telemedicine are responsible for compliance with both appropriate state and federal licensing requirements [14, 15]. Existing definitions and concepts In order to provide continuity of concept relative to SAGES Framework for post-residency surgical education and training document, the following definitions have been extracted from that instrument [1]. Must or shall: Indicates a mandatory or indispensable recommendation. Should: Indicates a highly desirable recommendation. May or could: Indicates an optional recommendation; alternatives may be appropriate. Competence: Competence is defined as the minimum level of skill, knowledge and/or expertise, derived through training and experience, required to safely and proficiently perform a task or procedure. Credentials: Documents provided following successful completion of a period of education or training. Clinical privileges: Authorization by a local institution (usually a hospital) to perform a particular procedure. Course: A course is a limited period of instruction with defined objectives designed to educate participants in clinical skill, techniques or procedures. Course structure and duration may vary according to the course objectives. Pre-test: A quantifiable examination of a trainees level of clinical knowledge, manual skills or technical proficiency prior to commencing a training course. Post-test: A quantifiable examination of a trainees level of clinical knowledge, manual skills or technical proficiency upon completion of a training course. Preceptorship: An individual education program in which the physician acquires additional skills and/or judgment to improve his/her performance of specific medical or surgical techniques and/or procedures. The preceptorship should define eligibility for participation. Preceptor: An expert surgeon who undertakes to impart his/her clinical knowledge and skills in a defined setting to a preceptee. The preceptor must be appropriately privileged, skilled, and experienced in the procedures(s) and/or technique(s) in question. In order to serve as a preceptor in a specific procedure or technique, the surgeon (preceptor) must be a recognized authority (e.g. publications, presentations, extensive clinical experience) in the particular field of expertise. Preceptee/trainee: A surgeon with appropriate basic knowledge and experience seeking individual training in skills and/or procedures not leared in prior formal training. The trainee must have appropriate background knowledge, demonstrated basic skills, and clinical experience relevant

This document was reviewed and approved by the Board of Governors of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), March 1996. It was prepared by an adhoc task force on Telemedicine. Correspondence to: Society of American Gastrointestinal Endoscopic Surgeons (SAGES), 2716 Ocean Park Blvd., Suite 3000, Santa Monica, CA 90405, Tel: (310) 314-2404. Fax: (310) 314-2585. E-Mail: SAGESMail@aol.com. SAGES publications 21

790

to the proposed curriculum. The trainee should be pre-tested to demonstrate adequate basic training. The trainee should be board eligible or certified in the appropriate specialty or possess equivalent board certification outside the United States. Proctor: A person who supervises or monitors students. As defined here, a proctor differs from a consultant or a preceptor in that (s)he functions as an observer and evaluator, does not directly participate in patient care, and receives no fees from the patient. A proctor must be a physician/surgeon who has recognized proficiency or documented expertise in the specialty area being proctored. The proctor should be free of perceived or actual conflicts of interest, which might create a bias again, or in favor of, the applicant. A proctor may work at the same or at another institution. It is anticipated, that in rare cases, a proctor may be required to intervene in a procedure on an emergency basis, and assume responsibility for patient care in order to preserve the welfare of the patient. (This situation is described in detail in the Framework for post-residency surgical education and training document.) Skills: Practical ability and dexterity based on talent and knowledge usually derived from a period of education and training. Skills laboratory: A facility in which a practicing physician acquires, refines or improves his/her ability to perform specific medical/surgical tasks or procedures. Skills are the building blocks upon which procedures are constructed. A skills laboratory may teach one skill or the entire set of skills required to perform a procedure. A skills laboratory is usually a continuing resource that can be revisited. New definitions and concepts The following definitions and concepts regarding surgical practice and telemedicine have been formulated after a review of the literature on telemedicine, with an intent to provide continuity relative to the principles defined in SAGES Framework for post-residency surgical education and training document. Some of the items in this section are derived from the history of telemedicine; others are natural extensions of concepts presented in SAGES parent document. The purpose here is to develop a framework for understanding and categorizing basic telemedicine concepts and activities, and to establish guidelines for the safe surgical practice of telemedicine. Telemedicine The practice of medicine and/or teaching of the medical art, without direct physical physician-patient or physicianstudent interaction, via an interactive audio-video communication system employing tele-electronic devices. Appropriate Use: Some applications of consultation, diagnosis and teaching with the potential for treatment as defined below. Remote site (site of the primary activity) The site or location of the primary activity. This location may be the originating site of a conference, the laboratory

where a new technique, instrument, or technology is being demonstrated, the facility where a patient is being evaluated or treated, or the operating theater where a surgical procedure is being performed.

Central location (central site): The site of the teacher, demonstrator, evaluator student, or clinician which is not immediately adjacent or proximal to the primary site of the activity or procedure. The central or off-site location may be as little as 100 feet or as distant as several thousand miles from the primary site of a conference or patient interaction. The basic assumption here, is that the individual at the central site is not able to physically intervene immediately in the primary procedure without the telecommunications interface. Comments: The concept of interaction at a distance implies that some form of telecommunications medium is employed. The participants, facilities, and telecommunication service vendors involved in the event should coordinate their efforts so that the telecommunications interface is suitable for the planned activity.

Teleconferencing A real time and live interactive program in which one set of participants are at one or more locations and the other set of participants are at another location. The teleconference allows for interaction, including audio and/or video, and possibly other modalities, between at least two sites. Appropriate use: Teaching (e.g. didactic lectures, demonstration of surgical or other medical procedures, and demonstration of uses of equipment), consultation, diagnosis, or deliberations. (See: Teleconsulting) Comments: Teleconferencing may be a useful adjunct to hands-on experience in the instruction of new procedures but is not a substitute for on-site supervised hands-on training in the development of competency.

Telementor An expert surgeon, at a central site, who undertakes to impart his/her clinical knowledge and skills in a defined setting to a student. The telementor must be appropriately privileged, skilled, and experienced in the procedure(s) and/ or technique(s) in question. In order to serve as a telementor in a specific procedure or technique, the surgeon (telementor) must be a recognized authority (e.g. publications, presentations, extensive clinical experience) in the particular field of expertise. The telementor, by definition, does not have the ability to physically intervene on-site in the primary activity without the telecommunications interface.

Telementoring A real time and live interactive teaching of techniques or procedures by a telementor to a student. The telementor is in one location and the student is in another. The telementor

791

must have the ability to see the performance of the procedure or technique being executed by the student in real time. The telementor and the student must have the ability to verbally communicate during the session. Implicit in the definition of telementoring is that the telementor does not have the ability to physically intervene on-site and can therefore not assume primary patient care responsibility. Appropriate use: Demonstration and/or teaching technique or procedures using inanimate trainers. Demonstration and/or teaching techniques or procedures using animate ex vivo models. Demonstration and teaching techniques or procedures on patients as an adjunct teacher when a qualified preceptor is on-site with the student. Comments: Telementoring is not an acceptable substitute for an on-site preceptorship but may be a useful adjunct.

an activity that is included in the concept of telemanagement. Telemonitoring assumes that the ability of the telemonitor to physically intervene at the site of the primary procedure is not possible without the telecommunications interface.

Teleconsultant A physician at a central location who evaluates a patient, and/or patient data, and who presents an opinion of his or her findings and/or recommendations for further evaluation or treatment to the patient or other health care provider at the remote site, using a telecommunications interface.

Telemonitor A person who supervises or monitors students from a central location. As defined here, a telemonitor differs from a consultant or a preceptor in that (s)he functions as an observer and evaluator, does not directly participate in patient care, and receives no fees from the patient. The telemonitor acts as an agent of the privileging committee of the sponsoring hospital. The telemonitor, by definition, does not have the ability to physicially intervene on-site without the telecommunications interface and therefore cannot assume primary patient care responsibility. A telemonitor must be a physician/surgeon who has recognized proficiency or documented expertise in the specialty area being monitored. The telemonitor should be free of perceived or actual conflicts of interest, which might create a bias against, or in favor of, the applicant. A telemonitor may work at the same or at another institution.

Teleconsulting (remote patient evaluation and consultation) Evaluation of patient(s), and/or patient data, and consultation regarding patient management, from a distant site, using a telecommunications interface. The teleconsultant, by definition, does not have the ability to physically interact with the patient, except through the telecommunications interface. Appropriate use: [16] Initial urgent evaluation of patients, triage decisions, and pretransfer arrangements for patients in an urgent/ emergency situation. Supervision and consultation for primary care encounters in sites where an equivalently qualified physician/surgeon is not available. Routine consultations and second opinions based on history, physical findings, and available test data. Public health, preventive medicine, and patient education. Comments: Teleconsulting has heretofore represented the pinnacle of achievement in telemedicine applications. Its use in the fields of radiology and pathology has stimulated the development of specific guidelines regarding the minimum and suggested interface requirements for reliable interpretation of transmitted patient information [17, 18]. In other disciplines, minimum requirements for the telecommunications interface remain to be defined. Remote patient evaluation assumes that a remote health care provider, who is familiar with, and capable of using the telecommunications interface equipment, is present with the patient or that the patient has been instructed in the mechanics of, and is capable of applying the diagnostic and telecommunications instrumentation necessary to provide clinical information to the teleconsultant.

Telemonitoring A real time and live interactive monitoring (evaluation) of technique(s) or procedure(s) of an applicant seeking privileges, or a surgeon seeking to certify or document his competence in a specific technique or procedure(s). The telemonitor is in one location and the surgeon to be evaluated is in another. The telemonitor must have the ability to see the performance of the procedure or technique being executed by the student in real time. The telemonitor and the applicant must have the ability to verbally communicate during the session. Appropriate Use: Telemonitoring may be used as an adjunct to proctoring in the privileging process but should not alone be a substitute for proctoring to determine competency. Integration of telemonitoring into the proctoring process may reduce, but not eliminate, the number of on-site proctored cases required. Comments: The term telemonitoring is sometimes used to define remote patient surveillance. For the purposes of this document, telemonitoring is not used in that context, but rather as described above. Remote patient surveillance is

Telemanagement (remote patient management) Remote evaluation and non-operative treatment of a patient, using a telecommunications interface. Appropriate use: [16] Medical and surgical evaluation, follow-up, and medication checks

792

Management of chronic diseases and conditions requiring a specialist not available locally. Public health, preventive medicine, and patient education. Comments: Telemanagement of a patient assumes that the central physician has evaluated the patient, and/or patient data, concurrently with the management activity. Because it involves a level of physician-patient interaction comparable to, or more intense than teleconsulting, telemanagement requires that a remote health care provider, who is familiar with, and capable of using the telecommunications interface equipment, is present with the patient, or that the patient has been instructed in the mechanics of, and is capable of applying the diagnostic and telecommunications instrumentation necessary to provide clinical information to the remote physician.

in these events should coordinate their efforts so that the visual fideltiy and telecommunications interface is suitable for the planned activity. References
1. SAGES (1994) Framework for post-residency surgical education and training. Surg Endosc 8: 11371142. 2. DeBakey ME (1995) Telemedicine has now come of age. Telemedicine Journal, 1: 34. 3. Allen A, Cox R, Thomas C (1992) Telemedicine in Kansas. Kans Med, 93:323325. 4. Satava RM, Simon IB (1993) Teleoperation, telerobotics, and telepresence in surgery. Endosc Surg Allied Technol, 1:151153. 5. Eide TJ, Nordrum I (1994) Current status of telepathology. APMIS, 102: 881890. 6. Belmont JM, Mattioli LF, Goertz KK, et al. (1995) Evaluation of remote stethoscopy for pediatric telecardiology. Telemedicine Journal, 1(1): 133150. 7. Mexrich RS, DeMarco JK, Negin S, et al. (1995) Radiology on the information superhighway. Radiology, 195(1): 7381. 8. Sweet HA, Holaday BA, Leffell D, et al. (1995) Telemedicine, delivering medical expertise across the state and around the world. Connecticut Medicine, 59: 593602. 9. Nakamura K, Takano T, Akao C (1995) Assessment of the value of videophones in community model networks for developing a comprehensive home health care system employing multimedia. Paper presented at the Second International Conference on the Medical Aspects of Telemedicine and Second Mayo Telemedicine Symposium. Telemedicine Journal 1(2): 174. 10. Darkins A, Gough D, Opett L (1995) Inner city telemedicine management of patients with minor injuries by low cost videoconferencing. Paper presented at the Second International Conference on the Medical Aspects of Telemedicine and Second Mayo Telemedicine Symposium, Telemedicine Symposium, Telemedicine Journal : 177. 11. Bashshur RL (1995) On the definition and evaluation of telemedicine. Telemedicine Journal 1: 1930. 12. Houtchens BA, Allen A, Clemmer TP, et al. (1995) Telemedicine protocols and standards: development and implementation. J Med Syst, 19: 93119. 13. Sanders JH, Bashshur RL (1995) Challenges to the implementation of telemedicine. Telemedicine Journal, 1: 115123. 14. Buckler LB, et al. (1995) An act to regulate the practice of medicine across state lines. The Federation of State Medical Boards, Euless, Texas. 15. Buckler LB, et al. (1995) A model act to regulate the practice of telemedicine or medicine by other means across state lines: Executive summary. The Federation of State Medical Boards, Euless, Texas. 16. Gribsby J, Schlenker RE, Kaehny MM, et al. (1995) Analytic framework for evaluation of telemedicine. Telemedicine Journal, 1: 3139. 17. Black-Schaffer S, Flotte TJ (1995) Current issues in telepathology. Telemedicine Journal, 1: 95106. 18. Forsberg D (1995) Quality assurance in teleradiology. Telemedicine Journal, 1: 107114.

Telesurgery (remote surgery) Surgery, procedure or technique performed on an inanimate trainer, animate model, or patient, in which the surgeon or operator is not at the immediate site of the model or patient being operated upon. Visualization and manipulation of the tissues and equipment is performed using tele-electronic devices. Appropriate use: Demonstration and/or teaching technique or procedures using inanimate trainers as the objects of the procedure. Demonstration and/or teaching techniques or procedures using animate model for purposes of testing technology under supervision of IRB. Demonstration and teaching techniques or procedures on patients under strict guidance of an IRB and only when a qualified surgeon is present to intervene in a timely fashion if technical difficulties arise. Comments: Remote surgery, at this time, is highly investigational and should not be performed except under IRB approval and by persons thoroughly familiar with the technology. SAGES strongly urges surgeons and hospitals to defer clinical implementation of these modalities until the technology has been validated. It is our opinion that current clinical use of this technology should only be conducted under a protocol reviewed by an institutional committee for the protection of patients and should include the collection of quality assurance and outcomes data. The participants, facilities, and telecommunication service vendors involved

Surg Endosc (1997) 11: 741744

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Percutaneous endoscopic gastrostomy (PEG)

8 years of clinical experience in 232 patients
W. Amann, H. J. Mischinger, A. Berger, G. Rosanelli, W. Schweiger, G. Werkgartner, J. Fruhwirth, H. Hauser
Department of Surgery, University Hospital Graz, Karl Franzens Universita t, Auenbruggerplatz 15, A-8036 Graz, Austria Received: 31 May 1996/Accepted: 26 November 1996

Abstract Background: Percutaneous endoscopic gastrostomy (PEG) is now a standard method for providing long-term enteral nutrition in patients who are unable to swallow. The aim of our study was to document clinical data that would allow prediction of a possible complicated clinical course. Methods: The study was carried out retrospectively. Clinical data of patients having received a PEG tube by a single endoscopic technique were analyzed. Results: Some 5.17% of 232 patients showed complications requiring surgery including a mortality rate of 0.43%. Patients with complications had a significantly lower body mass index and there was a significantly higher complication rate in patients having obstructive malignancies compared with benign diseases. Conclusions: Low body mass index and advanced malignancies are predictors for complications after PEG application. Early installation should help prevent further nutritional deterioration and the related risk of complications. Key words: Percutaneous endoscopic gastrostomy Enteral nutrition Neurological disorders Obstructive tumor disease

patients in whom a complicated clinical course would be expected.

Method
The procedure we perform is a single endoscopic technique, as first described in 1980 by Gauderer et al. [3] under sedation and local anesthesia using ECG monitoring and pulse oximetry. If complications necessitate removal of the tube, reapplication with a special balloon replacement set (Kangaroo 1422 Charr., Sherwood Medical, St. Louis, MO, USA) is possible and feasible even without endoscopy. To evaluate nutritional status the body mass index (BMI weight2/height) was calculated in all patients and blood samples were taken to measure the levels of cholesterol, triglycerides, and proteins preoperatively. In the statistical evaluation of our data the chi-square test was used to compare complication rates between the main patient groups. The body mass index (BMI) and levels of cholesterol, triglycerides, and protein were compared between the patient groups by the Mann-Whitney U and rank-sum tests. Complications directly related to the procedure were excluded from the tests, because they cannot be influenced by the nutritional status or the underlying disease of the patient.

Patients From April 1986 to December 1994 a PEG tube was placed in 232 patients with inability to swallow that had persisted for more than 2 or 3 weeks. Contraindications for PEG are listed in Table 1. There were 97 (41.8%) female and 135 (58.2%) male patients. The median age was 59 (range 14 to 95) years. The patients were divided into groups with nonmalignant disease (201, 86.7%) and with advanced obstructive malignancies (31, 13.3%). In the first group the majority of our patients, 186 (80.2%), had benign lesions of the central nervous system. The remaining 15 (6.5%) patients in this group had other benign lesions (Table 2). The patients with malignant disease had obstructions of the upper gastrointestinal and/or respiratory tract (31, 13.3%, Table 3). PEG was reapplied with a balloon replacement set with-

One of the main problems in the care of severely ill patients with inability to swallow is providing adequate nutrition. Although methods for parenteral feeding have improved, the enteral route is more effective, safer, and less expensive. Percutaneous endoscopic gastrostomy was the first method to allow placement of a feeding tube in the stomach without requiring laparotomy and general anesthesia [3]. The poor general condition of patients requiring long-term enteral nutrition led to a relatively high complication rate [6, 10, 11]. The aim of our study was to use clinical data to identify

Correspondence to: W. Amann

742 Table 1. Contraindications for PEG Absolute Peritonitis Total gastrectomy Impaired translumination Total obstruction Partial gastric resection Extended abdominal operations Coagulopathies Portal hypertension Morbid obesity Table 3. Indications for PEG application Patients, n 232 Group I Central nervous disease Cerebrovascular Ischemic Hemorrhage Traumatic Others Other benign lesions Traumatic Esophagitis Esophagotracheal fistula Benign esophageal stenosis Sepsis Group II Malignant obstructive tumors 201 186 120 93 27 40 26 15 5 4 2 2 2 31 31

Relative

Table 2. PEG patients from 4/86 to 12/94 Total Male Female Median age Group I (benign) Group II (malignant) 232 135 97 59 (range 1495) 201 31

out endoscopy when the tube was damaged or obstructed (n 26). Results The procedure was successfully performed in 230 (99.1%) patients. In one patient, total obstruction of the esophagus rendered introduction of the gastroscope impossible; in another case the position of a small stomach in the hypochondric region impaired positioning of the tube in the abdominal wall. The duration of the endoscopic procedure ranged from 2.8 to 8.3 (average 5.1 1.9 min) min. The feeding tubes were retained for periods of from 1 week to 5 years 6 months (average 8 months). In 22 (9.5%) patients the tube could be removed because the underlying problem improved and they were able to eat normally again. Complications were observed in 23 (9.91%) patients. Severe complications with the necessity of surgical intervention occurred in 12 patients (5.17% of all). In three cases (1.3% of all) the complications were directly related to the endoscopic procedure. In two cases puncture of an arterial vessel (lienal artery, gastroepiploic artery) led to emergency laparotomy. One patient died 2 days after the operation due to multiple organ failure. In another patient the punction of the presituated transverse colon caused an acute abdomen with subsequent laparotomy. Severe septic complications (abscess and partial necrosis of the abdominal wall) required surgical attention in four cases (1.72%). The most frequent severe complication was perforation of the gastric wall with peritonitis; this was seen in five (2.15%) patients. In all cases, oversuture of the leakage was followed by intraoperative application of a Witzel fistula. Complications such as inflammation were seen in 11 (4.74%) cases. After removal of the tube these cases recovered under antibiotics and local application of povidone iodine ointment. Reapplication was performed after wound healing without any complications. The overall complication rate was not significantly different in either group, when complications directly related to the procedure which are not influenced by the underlying disease or nutritional state of the patient are

included, but there is a significant difference when comparing severe complications and all complications after exclusion of these complications (Table 4). The body mass index (BMI) was significantly lower in tumor patients compared with patients having benign disease (p < 0.001) and in patients with complications compared with those having an uncomplicated course (p 0.0066, Fig. 1). So low body mass index is not only correlated with tumor disease but also with a higher complication rate. There was no statistically significant difference in cholesterol, triglyceride, and protein levels either in patients with or without complications (Fig. 2) or in comparison between the two patient groups (Fig. 3). Discussion The advantages of enteral nutrition as the physiological means of feeding are obvious. The poor condition of patients, who are unable to swallow, makes them bad surgical risks. Mortality rates up to 9.7% following operations under general anesthesia are given for these patients [9]. The role of nutrition in reducing the rate of major surgical complications in malnourished tumor patients is described in the literature [2]. Rates for major complications following gastrostomy (method of Stamm and Witzel) range up to 28% [5]. Compared to these procedures, major complication rates for PEG range between 0% and 16% (median 2.8%) [4, 8, 10, 12]. There are no data in the recent literature on the probability of complicated clinical courses after PEG application. In our series, the overall rate of severe complications directly related to the procedure is low (three patients, 1.3%). The overall rate of major complications requiring surgical intervention is 5.17% (12 patients) and equals the reports in the literature. Arterial intraabdominal bleeding in two cases, including one lethal outcome and perforation of the colon transversum in another case, both occurred early in the series. When the tube is placed under exact endoscopic view, good air insufflation of the stomach and transillumination of the anterior abdominal wall, these complications will not occur. Difficult anatomical conditions including obesity, previous abdominal operations, and small stomach lying in the hypochondric region can aggravate the endoscopic procedure and so are risk factors. The

743 Table 4. Complications Group I (n 201) Dir. proc. related complication Arterial bleeding Colonic perforation Septic Severe (with reoperation) Minor (with cons. treatm.) Gastric perforation Complications requiring surgery Complications requiring surgery (Dir. proc. rel. compl. excluded) All complications All complications (Dir. proc. rel. compl. excluded) 3 (1.49%) 2 1 13 (6.46%) 2 (0.99%) 11 (5.47%) 1 (0.49%) 6 (2.98%) 3 (1.49%) 17 (8.46%) 14 (6.96%) Group II (n 31) 0 0 0 2 (6.45%) 2 (6.45%) 0 4 (12.9%) 6 (19.35%) 6 (19.35%) 6 (19.35%) 6 (19.35%) p n.s. n.s. p 0.0298 n.s. p < 0.00001 p 0.00012 p < 0.00001 n.s. p 0.022

Fig. 1 (a) Body mass index (BMI) in patients (a) with benign (group I) and malignant disease (group II); (b) with and without complications.

most frequent severe complication was peritonitis in patients in whom an adequate gastrocutaneous fistula failed to develop (five cases, 2.15%). In four of these five patients significant malnutrition compromising wound healing led to leakage of gastric contents. After we began to tighten the discs that hold the gastric wall to the anterior abdominal wall for at least 72 h we did not see this complication any more. Antibiotic prophylaxis is able to reduce septic complications [1, 7, 13] and in patients at risk because of malnutrition or malignant disease we recommend prophylaxis for at least 3 days. Pneumoperitoneum without clinical signs of peritonitis can be ignored. One report in the literature describes this phenomenon in 56% [14]. X-ray control of the abdomen was only done in cases with high risk of perforation. Our study identified two factors that contribute significantly to a complicated clinical course. The first ma-

Fig. 2 (a) Cholesterol, (b) triglyceride, and (c) protein levels in patients with and without complications.

jor factor is an advanced malignant disease and the second is a low body mass index, which is of course often correlated with malignant disease and indicating a poor nutritional state even when levels of proteins, cholesterol, and tryglycerides are normal. We draw the conclusion that

744

poorly nourished patients are highly susceptible to complications even when PEG is done by an experienced endoscopist. A feeding tube should be applied promptly via PEG to prevent further nutritional deterioration and the related risk of complications. References
1. Akkersdijk WL, van Bergeijk JD, van Egmond T et al. (1995) Percutaneous endoscopic gastrostomy (PEG): comparison of push and pull methods and evaluation of antibiotic prophylaxis. Endoscopy 27(4): 313316 2. De Vita V, Hellman S, Rosenberg SA (1989) Cancer principles and practice of oncology. 3rd ed. JB Lippincott, ISBN 0-397-50840-9, pp 20392041 3. Gauderer MWL, Ponsky JL, Izant RJ (1980) Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 15: 872875 4. Gibson SE, Wenig BL, Watkins JL (1992) Complications of percutaneous endoscopic gastrostomy in head and neck cancer patients. Ann Otol Rhinol Laryngol 101: 4650 5. Grant JP (1993) Percutaneous endoscopic gastrostomyinitial placement by single endoscopic technique and long term follow-up. Ann Surg 217(2): 168174 6. Haws EB, Sieber WK, Kiesewetter WB (1966) Complications of tube gastrostomy in infants and children: 15-year review of 240 cases. Ann Surg 164: 284290 7. Jain NK, Larson DE, Schroeder KW et al. (1987) Antibiotic prophylaxis for percutaneous endoscopic gastrostomy. A prospective randomized double blind clinical trial. Ann Intern Med 107(6): 824828 8. Miller RE, Castlemain B, Lacqua FJ, Kotler DP (1989) Percutaneous endoscopic gastrostomy: results in 316 patients and review of literature. Surg Endosc 3: 186190 9. Palmberg S, Hirsja rvi (1979) Mortality in geriatric surgery. Gerontology 25: 103112 10. Ponsky JL, Gauderer MW, Stellato TA (1983) Percutaneous endoscopic gastrostomy. Arch Surg 118: 913914 11. Samii AM, Suguitan EA (1990) Comparison of operative gastrostomy with percutaneous endoscopic gastrostomy. Mil Med 155: 534535 12. Shellito PC, Malt RA (1985) Tube gastrostomy: techniques and complications. Ann Surg 201: 180185 13. Strodel WE, Lemmer J, Eckhauser F, Dent T (1983) Early experience with percutaneous endoscopic gastrostomy. Arch Surg 118: 449453 14. Wojtowycz MM, Arata JA Jr, Micklos TJ, Miller FJ Jr (1988) CT findings after uncomplicated percutaneous gastrostomy. Am J Roentgenol 151: 307309

Fig. 3 (a) Cholesterol, (b) triglyceride, and (c) protein levels in patients with benign (group I) and malignant disease (group II).

Surg Endosc (1997) 11: 722728

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

A stratified intraoperative surgical strategy is mandatory during laparoscopic common bile duct exploration for common bile duct stones
Lessons and limits from an initial experience of 92 patients
J. F. Gigot,1 B. Navez,2 J. Etienne,1 E. Cambier,2 P. Jadoul,1 P. Guiot,2 P. J. Kestens1
1 2

Department of Digestive Surgery, St-Luc University Hospital, Louvain Medical School, Hippocrate Avenue, 10, 1200 Brussels, Belgium Department of Digestive Surgery, St-Joseph Hospital, Gilly, Belgium

Received: 7 May 1996/Accepted: 19 November 1996

Abstract Background: Open exploration and endoscopic sphincterotomy (ES) remain the preferred treatment of common bile duct stones (CBDS). The recent spread of laparoscopy has worsened the dilemna of choosing between surgical and endoscopic treatment of CBDS. The aim of this study was to critically evaluate the results of our preliminary experience with laparoscopic common bile duct exploration (CBDE) for CBDS. Methods: Ninety-two consecutive patients were prospectively submitted to laparoscopic CBDE. Surgical strategy included an initial transcystic approach or laparoscopic choledochotomy. Failure of stone clearance was managed by conversion to open CBDE or by postoperative ES. Electrohydraulic lithotripsy and papillary balloon dilatation were selectively used. Stone clearance was assessed by choledochoscopy and control cholangiography. Results: The overall laparoscopic stone clearance in this series was 84% (transcystic route 63% and choledochotomy 93%). Conversion to laparotomy was mandatory in 12% of the patients because of incomplete stone clearance and in 5% because of intraoperative complications. Postoperative ES was required in 4% of the patients, giving an overall surgical success rate of 96%. When indicated (small and limited number of stones located below the cysticocholedochal junction, with a dilated and patent cystic duct) the transcystic route had the lower success rate, the higher complication rate, and the shorter operative time and postoperative hospital stay. When indicated (accessible and dilated common bile duct over 7 mm), laparoscopic choledochotomy had the higher success rate, the lower complication rate, the longer operative time, and the longer postoperative hospital stay, which is related to associated external biliary
Correspondence to: J. F. Gigot

drainage. The hospital mortality included two high-risk patients (2%) and the complications rate was 15%. Conclusions: Laparoscopic CBDE is safe in selected patients. A stratified intraoperative surgical strategy is mandatory in deciding between a transcystic route and choledochotomy with specific indications for each approach. When feasible, laparoscopic choledochotomy is more efficient and safe than the transcystic route, but it is associated with a longer postoperative hospital stay, which is due to external biliary drainage. Key words: Biliary tract Common bile duct stones Laparoscopic Complications Choledochotomy

In the era of mini-invasive surgery, the best approach for common bile duct stones (CBDS) remains controversial [2, 14, 23, 28]. When open CBD exploration first began to be performed (CBDE), three successive randomized trials did not show the superiority of preoperative endoscopic sphincterotomy (ES) over open CBDE in one session [24, 31, 32], except in cases of acute, severe cholangitis [21] and severe pancreatitis [12, 25], where the advantage of ES has been clearly demonstrated. The development of laparoscopic surgery has worsened the confusion. However, the efficacy and safety of laparoscopic CBDE [4, 8, 11, 13, 16, 20, 22, 27, 29, 30] as well as the value of preoperative ES before laparoscopic cholecystectomy (LC) [3, 15, 1719, 26, 33, 34] have been increasingly reported. After a surgical experience with open CBDE involving about 500 patients, in 1990 we started a prospective consecutive evaluation of the laparoscopic approach in patients with CBDS. The present series is taken from a part of a total experience of 2,500 LCs performed by the two senior au-

723

thors. The purpose of this paper is to present the limitations, the technical algorithm used for the operative strategy, and the lessons learned with this approach during our initial experience. Materials and methods
During the past 5 years, the laparoscopic approach was attempted in 92 consecutive patients (62 females, 30 males) with CBDS. Patients with severe cholangitis or pancreatitis were not included in this series, and were treated by ES. The mean age was 62 years (range, 1786 years). Fifteen patients (16%) were classified as high-risk patients (ASA III) [1]. All patients underwent clinical and biochemical evaluation. Clinical presentation included right upper quadrant (RUQ) or epigastric pain in 35 patients, jaundice in 33, acute cholecystitis in nine, mild cholangitis in 12, and mild pancreatitis in two. One patient was free of symptoms. Preoperative radiographic evaluation included ultrasound in all patients. In order to exclude the presence of CBDS, it is our routine practice to perform intravenous (IV) cholangiography, which was performed in 68 patients (74%). Nine patients (9.8%) were referred to us who had had endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic sphincterotomy (ES) and three patients (3.3%) had had endoscopic ultrasonography. In 53 patients (58%), clinical presentation and biochemical tests were predictive of CBDS [5]. In 23 patients (25%), preoperative clinical indices for CBDS and biochemical tests were normal, and occult CBDS were detected by preoperative IV cholangiography. In 16 patients (17%), clinical indices, biochemical tests, and IV cholangiography were all negative, and occult CBDS were detected by intraoperative cholangiography (IOC), which was performed in 85 patients (92.4%). In seven patients, IOC was not completed because of the quality of preoperative biliary imaging. Intraoperative cholangiography failed in two patients and was negative in two others (one false-negative and one spontaneous migration) (5%). A single stone was demonstrated by preoperative cholangiography in 39 patients (46%) and multiple stones in 42 patients (49%) (mean number three, range two to 10). Our standard approach for the management of CBDS was an initial attempt at transcystic CBDE (TCD-CBDE) (76 patients). When the cystic duct was not dilated, a progressive dilatation was performed in 17 patients, using flexible urological dilators (five patients) or an angioplasty balloon catheter (Cook Surgical, Bjaeverskov, Denmark) pushed on a guidewire (12 patients). Transcystic stone clearance was achieved first by using a three-wire, 6-F Dormia basket introduced into the common bile duct. Stone capture and clearance were assessed under fluoroscopic guidance. A balloon catheter was not used to avoid migration of CBDS in the upper part of the CBD. A transcystic biliary drain was left in place (in 28% of the patients) in case of multiple stones, suspicion of residual debris, or transpapillary passage of the scope. When the transcystic approach proved impossible or unsuccessful, laparoscopic choledochotomy was used if CBD diameter was at least 7 mm. Longitudinal incision of the choledochus was performed with a laparoscopic knife after blowing up the CBD with saline through the transcystic cholangiographic catheter. Stone clearance was obtained by using a blind Dormia basket, a Forgati balloon catheter, or endoscopic extraction. Postoperative external biliary drainage was routinely employed in all patients either by transcystic drain (35%) or by T-tube (65%). The choledochotomy was sutured with interrupted Vicryl or PDS 4.0 stitches, with routine use of a methylene blue water-tightness test. Stone clearance was routinely assessed with a fine flexible choledochoscope (Olympus URF-P2, 3.5 mm of external diameter). When choledochoscopy was used, we inserted a fifth soft trocar in the epigastric area, just above the hepatoduodenal ligament. An attempt was made routinely to explore the inferior and the upper biliary tract, avoidingif possible transpapillary passage of the scope. In cases of residual CBDS, choledoschoscopic extraction was attempted via a Dormia basket (3 F) through the operative channel of the scope. At the end of the procedure, we routinely performed control intraoperative cholangiography to assess stone clearance, control correct positioning of the biliary drain, and exclude extravasation of contrast material related to biliary tract injury. Intraperitoneal drainage was used in all patients. When intraoperative stone clearance was incomplete, the procedure was converted to laparotomy or to postoperative ES, according to the preference of each senior surgeon. In converted patients, a transcystic drain or T-tube was inserted in 6% and 81%, respectively. Postoperatively, tran-

scystic or T-tube cholangiography was routinely performed at the 3rd postoperative day before hospital discharge. Patients with a biliary drain were readmitted 3 weeks later for cholangiographic assessment and biliary drain removal during a 1-day hospital stay. Long-term follow-up included clinical assessment and biochemical tests at a mean follow-up time of 14 months (range: 140 months). Twenty-seven patients agreed to undergo intravenous cholangiography, after a mean of 12 months (range 140 months), to exclude residual CBDS. The statistical analysis included appropriate use of chi-square, Students t-test, and Fishers Exact test. A p value <0.05 was considered to be significant.

Results

Stone clearance In one patient who had a preoperative clinical and biochemical history of stone migration and CBDS well seen by preoperative endoscopic ultrasonography, no stone was demonstrated 24 h later on intraoperative cholangiography during the TCD-CBDE because of probable spontaneous stone migration. In three patients (3%), the approach was rapidly converted to laparotomy because of the presence of severe acute cholecystitis, which made it impossible to achieve safe access to the hepatoduodenal ligament and the Calot triangle to perform laparoscopic CBDE. Seventy-six patients (83%) had an initial attempt at a TCD-CBDE. This failed in eight patients (11%) because of rupture of the cystic duct during instrumentation (two patients) or the presence of obstructive cystic valve (six patients) which precluded introduction of the instrument. Two of these patients were treated by conversion to laparotomy and the remaining six underwent secondary laparoscopic choledochotomy. Forty-eight patients (63%) had complete stone clearance during TCD-CBDE. In two of these patients (2.6% of the transcystic group), successful stone extraction was due to the use of flexible scope. The remaining 20 patients (26%) with incomplete stone clearance were treated by laparotomy (six patients), secondary laparoscopic choledochotomy (12 patients), or postoperative ES (two patients). Finally, at the end of TCD-CBDE in 76 patients, complete stone clearance was achieved in 63% of the cases, while secondary laparoscopic choledochotomy was done in 18 patients (24%), secondary conversion in eight patients (10%), and postoperative ES in two patients (3%). Thirty patients (33%) were treated by laparoscopic choledochotomy. In 12 patients this was an initial attemptin five because of a nondilated cystic duct, in two because of inadequate anatomy of the cystic duct (tortuous, obstructive valve), and in five because of large size or proximal location of the CBDS above the cystic duct entrance in the CBD. Successful laparoscopic stone clearance was achieved in all these patients. Of the 18 patients treated by secondary choledochotomy after transcystic failure, laparoscopic stone clearance was achieved in 16 patients, and two patients required postoperative ES because of impacted papillary stones. The overall stone clearance rate with laparoscopic choledochotomy was thus 93.3% (28 patients); there was a need for postoperative ES in 6.7% (two patients). The two groups of patients with choledochotomy are not comparable, since one contains patients who have already failed

724 Table 1. Reasons for conversion to laparotomy according to surgical experience and technical proceduresa Surgical technique (no. of cases) Transcystic Incomplete stone clearance Acute cholecystitis CBD stenosis (CDA) Technical complications Impacted Dormia basket Biliary injury Biliary drain malplacement 4 4 1 6 1 4 8% 1 15/76 (20%) Choledochotomy 1 1 3.3% 1/30 (3.3%) p < 0.03 Surgical experience (no. of cases) 10 cases 4 1 1 1 0 6 >10 cases 0 4 6 5 1 10

CBD, common bile duct; CDA, choledochoduodenal anastomosis

transcystic CBDE. Complete stone clearance was achieved in all patients undergoing conversion. At the end of the general anesthesia, complete stone clearance was thus achieved in 77 patients (84%) by laparoscopy and in 11 patients (12%) after conversion to laparotomy, giving an overall stone clearance rate of 96%. Postoperative ES was required in four patients (4%) for known residual CBDS. Electrohydraulic lithotripsy (EHL) was successfully used in three patients (3%) with impacted papillary stones. In two other similar cases during choledochotomy, endoscopic EHL was technically impossible because of small diameter of the lower CBD impaired access to the stone with the scope. Papillary balloon dilatation and flushing through the papilla were performed in six patients with multiple small CBDS, complicated in one case by transient cholestasis.

The conversion rate was higher during TCD-CBDE (20%) compared with choledochotomy (3%) (p < 0.003). Biliary tract injury due to instrumentation This unexpected complication was found in seven patients in our series, mainly with TCD-CBDE (six of 76 patients, 8%). Two patients presented a cystic duct perforation, but a small perforation in the posterior CBD wall at the level of cystic duct entrance occurred in five patients. During TCDCBDE the injury was related to difficult repeated and blind introduction of the Dormia basket in a cystic duct with obstructive valves (four patients) or during transcystic dilatation (two patients) when a semirigid angioplasty balloon catheter was used. The single injury that occurred during laparoscopic choledochotomy was related to the use of a semirigid cholangiographic catheter to perform control cholangiography after successful stone clearance. All these biliary tract injuries were detected during routine control intraoperative cholangiography at the end of the procedure (Fig. 1). Five patients were converted to laparotomy for optimal biliary repair. Two more recent patients were successfully treated laparoscopically. No postoperative complication occurred in these patients. Operative time The mean duration for combined laparoscopic cholecystectomy and CBDE was 170 76 min for TCD-CBDE compared with 266 105 min for laparoscopic choledochotomy and 260 109 min for conversion to laparotomy. When choledochotomy was performed as initial treatment, the mean operative time was 228 74 min, which is still significantly longer than the time for TCD-CBDE (p < 0.01). The mean operative time for initial conversion to laparotomy was 155 43 min compared with 285 105 min for conversion after initial attempt at TCD-CBDE or choledochotomy (p < 0.03). The operative time decreased with experience: 226 95 min, compared with 154 62 min for TCD-CBDE (p < 0.002), and 379 219 min, compared with

Choledochoscopic stone clearance assessment Choledochoscopy was performed in 75 patients (80.5%) to assess the effectiveness of stone clearance. During the transcystic approach, the procedure was possible in only 45 patients (59%), with an introduction failure rate of 18%. Complete exploration of the lower and upper biliary tract was achieved in only 15% of these patients. During choledochotomy, choledochoscopy was used in all patients, with no failure (p < 0.01) and complete visualization of the whole biliary tract in all patients (p < 0.001, compared with TCD-CBDE).

Conversion rate Conversion to laparotomy was required in 16 patients (17%) for various reasons (Table 1). But the reasons for conversion changed: with an increasing experience there was no need to convert for incomplete stone clearance, but conversions were necessary for inaccessibility to the CBD in the presence of acute cholecystitis or for technical complications.

725

leak, one with acute cholangitis following control cholangiography 3 weeks after surgery, and one with acute abdominal pain after Silastic transcystic drain extraction 3 weeks postoperatively. Six patients (5%) had major local complications: three with known residual CBDS treated by ES, one patient with wound abscess treated by reintervention, one 80-year-old confused patient who pulled out her transcystic drain and required reintervention, and one case of T-tube fracture during extraction which required endoscopic removal. The last-mentioned patient also had one residual CBDS, which was treated successfully by flushing through the T-tube. Problems or complications related to biliary drainage are listed in Table 2; the incidence was 8% during laparoscopic choledochotomy. In the transcystic group, the incidence was 43% with a biliary drain compared with 5.5% without a biliary drain (p < 0.004). Postoperative hospital stay Successful laparoscopic CBDE was associated with a mean postoperative hospital stay (POS) of 6.3 3.8 days compared with 14 9.9 days in patients undergoing conversion (p < 0.001). The transcystic group had a shorter POS compared with the laparoscopic choledochotomy group: 5.7 3.8 days vs 7.4 3.5 days (p < 0.03). With increasing confidence in the technique, a reduction in POS was observed in the transcystic group8.1 6 days with less than ten cases of experience compared with 5 2.6 days with more than ten (p < 0.006)but not in the group with choledochotomy8.9 1.6 days with less than ten cases compared with 7.1 3.7 days with more than ten (NS). In cases of biliary drain insertion (conversion patients excluded), the mean POS was 7.7 4.2 days compared with 4.7 2.5 days when no drain was inserted (p < 0.001). The POS in patients with a laparoscopically inserted biliary drain included a mean initial POS of 6.4 3.3 days and a mean stay during readmission for cholangiographic assessment of 1.3 days (range, zero to 15 days). Patients without biliary drain during TCD-CBDE had a shorter POS than those for whom a biliary drain was inserted: 5 2.5 days vs 8 5.4 days (p < 0.001). Long-term follow-up No residual CBDS were observed on the postoperative cholangiographic examination. Ninety patients (100% of the surviving patients) were assessed regularly during a mean follow-up of 18 months (median 14, ranges 365 months). Two patients (2.2%) presented residual CBDS and were successfully treated by ES (at 8 and 17 months, respectively). Twenty-seven patients had normal IV cholangiograms. All patients with biliary tract injury had normal IV cholangiograms, except in one case in which a slight narrowing of the CBD was observed at 1 year postoperatively. Discussion The main lesson from this series is the absence during laparoscopoic CBDE of a simple and uniform approach to treat

Fig. 1. Biliary tract injury (arrow) demonstrated on control intraoperative cholangiography at the end of the procedure, after blind transcystic introduction of instrument. Treatment in this case was by laparoscopic suture, which was successful.

245 59 min for laparoscopic choledochotomy (p < 0.008), depending on whether the surgeon had fewer or more than ten cases of experience. During TCD-CBDE approach, operative time was increased when a biliary drain was inserted: 243 77 min compared with 141 53 min without biliary drain placement (p < 0.001).

Postoperative complications The 2-month hospital mortality was 2% (two patients). One patient, a 75-year-old man, died from a myocardial infarction after successful transcystic exploration lasting 90 min. The second patienta 77-year-old womandied from respiratory complications following conversion to laparotomy after failure of an initial transcystic approach (overall operative time of 190 min). Both of these patients were classified as high risk (ASA III). Mortality was zero (zero of 76 patients) when age was less than 75 years, compared with 13% (two of 16 patients) over 75 years (p < 0.03). In ASA I and II, mortality was zero (zero of 77 patients) compared with 13% (two of 15 patients) in ASA III patients (p < 0.05). Postoperative complications occurred in 14 patients (15%), including one patient with hemorrhage from a duodenal ulcer and 13 patients (14%) with local complications. This included eight patients (9%) with minor local complications: four with biochemical pancreatitis (without radiologic features of pancreatitis on computed tomography), one patient with transient cholestasis, one with transient biliary

726

Fig. 2. Limitations of the transcystic route during laparoscopic common bile duct (CBD) exploration.

CBDS. There is a need for a carefully selected surgical strategy during laparoscopic CBDE, with specific indications for either the transcystic route or choledochotomy. Limiting factors of success with TCD-CBDE (Fig. 2) include anatomic features related to the cystic duct, such as small size, tortuous duct, obstructive cystic valve, rupture of the cystic duct during instrumentation, and low level of insertion of the cystic duct on the CBD, especially in the case of a very low suprapapillary insertion of the duct, which makes grasping of a stone impossible. Other limiting factors are related to excessively large size of stones and location of stones above the cysticocholedochal junction, which makes it impossible to grasp the stone. The optimal circumstances for transcystic stone extraction are small-size stones (diameter smaller than or equal to the size of the cystic duct), a limited number of stones, location of the stones in the lower portion of the CBD below the cysticocholedochal junction, and a patent cystic duct for instrumentation (Fig. 3). Performance of intraoperative cholangiography at the beginning of the procedure is crucial to explore the feasibility of and any anatomic limitations for TCD-CBDE, except for the problem of obstructive cystic valve, which can only be detected during instrumentation. Early selection of the optimal approach will decrease the technical morbidity and the operative time observed in our series, especially when an inadequate initial approach was chosen and repeatedly used before changing to an intraoperative surgical strategy. The unexpectedly high incidence of CBD injury (5.4%) in our serieswhich is significantly higher than the incidence with laparoscopic cholecystectomy for the two senior authors, 0.1% (p < 0.001)was also the consequence of an inadequate choice of the transcystic route, and related again to anatomic particularities of the cystic duct, mainly the presence of obstructive valves at the level of the cysticocholedochal junction. Repeated and blind Dormia basket introduction in such difficult cases was responsible for cystic duct or posterior CBD wall perforation, as was the use of transcystic dilatation using a semirigid angioplasty balloon catheter. Several recommendations to prevent biliary tract injury during laparoscopic CBDE can be made: (1) Avoid repeated instrumentation when obstructive valves are detected. (2) Introduce all instruments through the cystic duct into the CBD under fluoroscopic guidance. (3) Insert all instruments as parallel as possible with the CBD, avoiding a perpendicular

Fig. 3. Operative surgical strategy for laparoscopic common bile duct (CBD) exploration. ES, endoscopic sphincterotomy.

approach. (4) Use a soft, atraumatic Dormia basket or balloon catheter. (5) Avoid dilatation of nondilated cystic duct and use laparoscopic choledochotomy in such cases when feasible. (6) Perform routinely control cholangiographic examination at the end of the procedure to detect any contrast material extravasation which might signal the presence of a biliary tract injury. It is also important to point out that with intraoperative detection of these injuries in our series, primary repair was optimal, and no patient had serious postoperative complications. The final 81.5% stone clearance rate achieved in this series during laparoscopic CBDE is certainly partially related to the learning curve required with this approach. However, limitations of TCD-CBDE related to anatomy of the cystic duct and of the cysticocholedochal junction and the characteristics of CBDS have already been highlighted, explaining a limited final success rate of 62%. As previously reported [11, 16, 22], laparoscopic choledochotomy carried a higher final stone clearance rate (90%). But two prerequisites for laparoscopic choledochotomy are mandatory: an accessible hepatoduodenal ligament with no severe inflammatory process (acute cholecystitis) and the presence of a CBD diameter of at least 7 mm, which has to be determined on the basis of the intraoperative cholangiography. The surgeon must be aware that with the magnification obtained with video-assisted surgery, one could inadvertently open a duct of 5 mm in sizeas we did, without consequence, on one occasion. Another advantage of laparoscopic choledochotomy is the ease with which one may assess stone clearance with choledochoscopy, exploring the whole extra- and intrahepatic biliary tract. In comparison, complete choledochoscopic stone clearance assessment during TCD-CBDE was successful in only 15% of cases, requiring routine control cholangiographic examination to achieve stone clearance assessment. Again, anatomic features of the cysticocholedochal junction were responsible for the difficulties in accessing the common hepatic duct. To improve stone clearance, we selectively used papillary balloon dilatation for multiple small stones and electrohydraulic lithotripsy for impacted papillary stones. These procedures were

727 Table 2. Problems and complications related to biliary drain insertiona Approach Transcystic Problems Complications Without biliary drain With biliary drain

2 drain out 2 drain too strangled 1 transient cholestasis (balloon dilatation) 1 abdominal pain after biliary drain extraction 1 transient biliary leak 1 drain pulled out reintervention 5.5% 43% p < 0.004 0 0 0 1 ES for T-tube disruption 1 postradiologic cholangitis 8%

Choledochotomy Problems Complications

ES, endoscopic sphincterotomy

perfectly well tolerated in our experience. But a 15% incidence of pancreatitis has been reported with papillary balloon dilatation [6, 8, 30] and electrohydraulic lithotripsy required a careful use in close contact with the CBDS to avoid hemobilia or bile duct perforation [3, 7, 11]. Finally, in this series, insertion of a biliary drain accounted for a longer hospital stay, partly because of a longer initial postoperative stay and partly because of readmission for cholangiographic assessment and biliary drain removal. During TCD-CBDE, placement of a biliary drain is also associated to a higher incidence of related local problems or complications. Nevertheless, laparoscopic choledochotomy even with a T-tube involved a shorter postoperative stay compared to that for a converted patient. Alternatives to external biliary drainage should thus be searched, such as internal biliary drainage by endoprosthesis or laparoscopic anterograde sphincterotomy, described recently by De Paula and associates [10, 11] and Curet and colleagues [9]. Since completing the series reported here, we have begun to use an internal biliary drainage with a fine transpapillary prosthesis inserted through the choledochotomy, which has resulted in a reduced postoperative hospital stay and in no complications. However, this alternative carries a potential risk of pancreatitis and the need for postoperative duodenoscopy 3 weeks postoperatively for prosthesis removal [11]. Another disadvantage of this technique is that one cannot perform postoperative cholangiography to detect residual CBDS. In patients previously treated by endoscopic sphincterotomy, but treated for residual CBDS, primary closure of the choledochotomy could be considered. Finally, we emphasize the importance of the routine use of intraoperative cholangiography in patients with suspicion of CBDS. In conclusion, laparoscopic CBDE is feasible following the same rules as in open biliary surgery. A selective intraoperative surgical strategy is mandatory with specific indications for TCD-CBDE or choledochotomy, depending on biliary anatomic features and stones characteristics. Early performance of intraoperative cholangiography will allow a prompt choice of the most appropriate technique for CBDS removal, reducing operative time and in TCD-CBDE the rate of local complications. When technically feasible, laparoscopic choledochotomy has the best stone clearance rate and allows an easy endoscopic stone clearance assessment,

with a low complications rate. For this reason, this approach appears in our experience to be the technique of choice for CBDS clearance, when technically feasible. The procedure is safe in patients who are less than 75 years old and who have no preoperative risk factors. The ultimate limitations for the use of laparoscopic CBDE in the surgical community in Europe pertain to (1) the level of surgical expertise required (including meticulous performance of endoperitoneal suturing techniques), which is greater than for laparoscopic cholecystectomy, (2) the technological requirements, and (3) the cost of the procedure (especially with regard to the need for fragile and expensive endoscopes).
Acknowledgment. We are deeply grateful to Dr. Werner Heidel and Dr. Jon-Arne Soreide for reviewing the manuscript and to Nadine Thiebaut for secretarial assistance.

References
1. American Society of Anesthesiologists (1963) New classification of physical status. Anesthesiology 24: 111 2. Arregui ME (1992) The author replies. Surg Endosc 6: 267268 3. Arregui ME, Davis CJ, Arkush AM, Nagan RF (1992) Laparoscopic cholecystectomy combined with endoscopic sphincterotomy and stone extraction or laparoscopic choledochoscopy and electrohydraulic lithotripsy for management of cholelithiasis with choledocholithiasis. Surg Endosc 6: 1015 4. Bagnato VJ (1993) Laparoscopic choledochoscopy and choledocholithotomy. Surg Laparosc Endosc 3: 164166 5. Barkun AN, Barkun JS, Fried GM, Ghitulescu G, Steinmetz O, Pham C, Meakins JL, Goresky CA, the McGill gallstone treatment group (1994) Useful predictors of bile duct stones in patients undergoing laparoscopic cholecystectomy. Ann Surg 220: 3239 6. Carroll BJ, Phillips EH, Chandra M, Fallas M (1993) Laparoscopic transcystic duct balloon dilatation of the sphincter of Oddi. Surg Endosc 7: 514517 7. Carroll B, Chandra M, Papaioannou T, Daykhovsky L, Grundfest W, Phillips E (1993) Biliary lithotripsy as an adjunct to laparoscopic common bile duct stone extraction. Surg Endosc 7: 356359 8. Carroll BJ, Fallas MJ, Phillips EH (1994) Laparoscopic transcystic choledochoscopy. Surg Endosc 8: 310314 9. Curet MJ, Pitcher DF, Martin DT, Zucker KA (1995) Laparoscopic antegrade sphincterotomy: a new technique for the management for complex choledocholithiasis. Ann Surg 221: 149155 10. DePaula AL, Hashiba K, Bafutto M, Zago R, Machado MM (1993) Laparoscopic antegrade sphincterotomy. Surg Laparosc Endosc 3: 157160

728 11. DePaula AL, Hashiba K, Bafutto M (1994) Laparoscopic management of choledocholithiasis. Surg Endosc 8: 13991403 12. Fan ST, Lai ECS, Mok FPT, Lo CM, Zheng SS, Wong J (1993) Early treatment of acute biliary pancreatitis by endoscopic papillotomy. New Engl J Med 328: 228232 13. Ferzli GS, Massaad A, Kiel T, Worth MH Jr (1994) The utility of laparoscopic common bile duct exploration in the treatment of choledochlithiasis. Surg Endosc 8: 296298 14. Fink AS (1993) To ERCP or not to ERCP? That is the question. Surg Endosc 7: 375376 15. Franceschi D, Brandt C, Margolin D et al. (1993) The management of common bile duct stones in patients undergoing laparoscopic cholecystectomy. Am Surg 59: 525532 16. Franklin ME Jr, Pharand D, Rosenthal D (1994) Laparoscopic common bile duct exploration. Surg Laparosc Endosc 4: 119124 17. Frazee RC, Roberts J, Symmonds R et al. (1993) Combined laparoscopic and endoscopic management of cholelithiasis and choledocholithiasis. Am J Surg 166: 702706 18. Graham SM, Flowers JL, Bailey RW, Zucker KA, Imbembo AL (1992) Utility of planned perioperative endoscopic retrograde cholangiopancreatography and sphincterotomy in the era of laparoscopic cholecystectomy. Endoscopy 24: 788789 19. Graham SM, Flowers JL, Scott TR et al. (1993) Laparoscopic cholecystectomy and common bile duct stones: the utility of planned perioperative endoscopic retrograde cholangiography and sphincterotomy: experience with 63 patients. Ann Surg 218: 6167 20. Hunter JG (1992) Laparoscopic transcystic common bile duct exploration. Am J Surg 163: 5358 21. Lai ECS, Mok FPT, Tan ESY et al. (1992) Endoscopic biliary drainage for severe acute cholangitis. New Engl J Med 326: 15821586 22. Lezoche E, Paganini A, Feliciotti FS, Chan R (1993) Laparoscopic suture technique after common bile duct exploration. Surg Laparosc Endosc 3: 209212 23. Mitchell SA, Jacyna MR, Chadwick SF (1993) Common bile duct stones: a controversy revisited. Br J Surg 80: 759760 24. Neoptolemos JP, Carr-Locke DL, Fossard DP (1987) Prospective randomised study of preoperative endoscopic sphincterotomy versus surgery alone for common bile duct stones. Br Med J 294: 470474 25. Neoptolemos JP, London NJ, James D, Carr-Locke DL, Bailey IA, Fossard DP (1988) Controlled trial of urgent endoscopic retrograde cholangiopancreatography and endoscopic sphincterotomy versus conservative treatment for acute pancreatitis due to gallstones. Lancet 2: 979983 26. Neuhaus H, Feussner H, Ungeheuer A, Hoffmann W, Siewert JR, Classen M (1992) Prospective evaluation of the use of endoscopic retrograde cholangiography prior to laparoscopic cholecystectomy. Endoscopy 24: 745749 27. Petelin JB (1993) Laparoscopic approach to common duct pathology. Am J Surg 165: 487491 28. Phillips EH (1993) ERCP in conjunction with LC. Surg Endosc 7: 393394 29. Phillips EH, Carroll BJ, Pearlstein AR, Daykhovsky L, Fallas MJ (1993) Laparoscopic choledochoscopy and extraction of common bile duct stones. World J Surg 17: 2228 30. Phillips EH, Rosenthal RJ, Carroll BJ, Fallas MJ (1994) Laparoscopic trans-cystic-duct common-bile-duct exploration. Surg Endosc 8: 13891394 31. Stain SC, Cohen H, Tsuishoysha M, Donovan AJ (1991) Choledocholithiasis: endoscopic sphincterotomy of common bile duct exploration. Ann Surg 213: 627634 32. Stiegmann GV, Goff JS, Mansour A, Pearlman N, Reveille RM, Norton L (1992) Precholecystectomy endoscopic cholangiography and stone removal is not superior to cholecystectomy, cholangiography and common duct exploration. Am J Surg 163: 227230 33. Surick B, Washington M, Ghazi A (1993) Endoscopic retrograde cholangiopancreatography in conjunction with laparoscopic cholecystectomy. Surg Endosc 7: 388392 34. Widdison AL, Longstaff AJ, Armstrong CP (1994) Combined laparoscopic and endoscopic treatment of gallstones and bile duct stones: a prospective study. Br J Surg 81: 595597

Surg Endosc (1997) 11: 754757

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic-assisted colon surgery by abdominal wall lifting with newly developed lifting bars
H. Nisii, T. Hirai, H. Ohara, Y. Masuda
Department of Surgery, Fujieda Heisei Memorial Hospital, Fujieda, Japan Received: 18 March 1996/Accepted: 12 December 1996

Abstract Background: The aim of the study is to evaluate the efficacy of laparoscopic-assisted colon surgery by lifting the abdominal wall with newly developed lifting bars. Methods: We have made and used two kinds of lifting bars: type I and type T. Two I-type lifting bars are used in transverse colectomy and right hemicolectomy. One I-type lifting bar and one T-type bar are used in sigmoid colectomy and low anterior resection. After the intestine is dissected and the mesenterium is treated under laparoscopy, a small laparotomy wound about 4 to 6 cm long is made, and the intestine is pulled out of the body for extracorporeal anastomosis. Results: The mean operating time was 153.8 51.9 min, and no particular complications were noted. Conclusions: Since postoperative pain is mild and postoperative recovery is rapid, this method is considered to be an effective surgical procedure. Key words: Laparoscopic-assisted colon surgery Abdominal wall lifting method Pneumoperitoneum

dominal wall tenting by combining our unique lifting bars. In the present study we performed laparoscopic-assisted colon surgery by the lifting method in 22 patients and obtained satisfactory results. Materials and methods Subjects
Laparoscopic surgery was performed at our hospital in 22 patients with intestinal disease in the 18 months between July 1994 and December 1995. The patients consisted of nine men and thirteen women with a mean age of 60.7 years (range: 41 to 80). Malignant and benign diseases were noted in sixteen and six of them, respectively. Malignant diseasesmucosal cancer of the transverse colon, submucosal cancer of the transverse colon, submucosal cancer of the ascending colon, advanced cancer of the ascending colon, and submucosal cancer of the sigmoid colonwere noted in five, one, four, three, and three patients, respectively. As benign diseases, polyps in the transverse and sigmoid colons were noted in one and three patients, respectively. These were all wide-base polyps over 2 cm in size for which EMR was considered difficult. In addition, two patients with symptomatic rectosigmoid diverticulosis were included. Table 1 shows the details of surgical procedures used.

Pneumoperitoneum has generally been used for laparoscopic colon surgery. However, it may be accompanied by hypercapnia, gas embolism, subcutaneous emphysema, and other relevant conditions. An increase of intraperitoneal pressure due to pneumoperitoneum leads to insufficient venous return from the inferior vena cava, resulting in a decrease of cardiac output, hypotension, and sometimes pulmonary embolism [1]. We have performed laparoscopic surgery without pneumoperitoneum by lifting the full layer of the abdominal wall using two unique kinds of lifting bars that we developed. Almost the same surgical field as by pneumoperitoneum can be obtained without causing ab-

Surgical procedures

Correspondence to: H. Nisii, 123-1 Mizukami, Fujieda, Shizuoka 426, Japan

Right hemicolectomy. Laparoscopic-assisted right hemicolectomy plus dissection of D2 and D3 [4] was performed in patients with cancer of the ascending colon. After a small incision is made in the left umbilical margin or in the outer margin of the left straight muscle of the abdomen at the umbilical level, type-I lifting bars (Fig. 1) are inserted into the right hypochondrial and right lateral regions and pulled. The type-I lifting bar is a stainless-steel rod 5 mm in diameter, which is bent as shown in Fig. 1. The straight portion of this bar is inserted intraperitoneally, while the hookshaped portion is connected to a tractor. The type-I lifting bar is inserted from point A intraperitoneally through the small wound made by open laparoscopy. The bar is then advanced slowly along the abdominal wall up to point B while checking the resistance. A laparoscope is also inserted through the small wound, and the I-type lifting bar is inserted carefully under laparoscopy. Various sizes of lifting bars are available, and a slightly longer lifting bar is inserted into the right hypochondrial region so that it can be advanced beyond the costal arch, while a slightly shorter lifting bar

755 Table 1. Clinical profile of the patients and 22 laparoscopic procedures by abdominal wall lifting using original lifting bars Age (mean SD, years) Sex (M/F) Weight (mean SD, kg) Partial transverse colectomy Transverse colectomy D2 Right hemicolectomy D2 Right hemicolectomy D3 Partial sigmoid colectomy Sigmoid colectomy D2 Low anterior resection 60.7 13.3 9/13 51.8 13.7 6 1 4 3 3 3 2

Fig. 1. The type-I lifting bar.

is inserted into the right lateral region. When the lifting bars are connected to the tractor which is attached to the right side of the patient and pulled, the upper-right and lower-right parts of the abdomen are lifted uniformly, and a good surgical field can be obtained (Figs. 2 and 3). All trocars used are 10 mm in size (Fig. 4). Since there is no fear of leakage from a pneumoperitoneum, the trocars are removed after insertion, and surgery is performed in the presence of the gripping devices alone. The area near the lesion is marked with metal clips during preoperative endoscopy, and the lesion is confirmed under fluoroscopy during surgery. A clip or a thread is applied under fluoroscopy to the serous membrane of the intestine immediately above the lesion so that it can be used as a guide to dissection. After the patient is placed with the right side high, the ascending colon is dissected from the retroperitoneum and mobilized sufficiently. The hepatocolic and gastrocolic ligaments are also dissected and divided. They may also be dissected and divided under direct vision through a small laparotomy wound about 6 cm long, which is made in the upper-right part of the abdomen. The ileum is divided after the intestine is pulled out of the body through the small wound made in the upper-right part of the abdomen (Fig. 5). After the blood vessels around the margin of the ileum are divided, an incision is made in the mesentery and is extended toward the inflow point of the ileocolic vein (Fig. 6). After dissection along the surgical trunk, the ileocolic artery and vein, right colic artery and vein, and the right branches of the transverse colicartery and vein are divided at their origin and D3 dissection is performed (Fig. 7). The transverse colon is divided, and side-to-side ileotransversestomy is performed by instrumental anastomosis. After the defect of the mesenteric membrane is closed by suturing, the lifting method is reinstituted, and sufficient irrigation and aspiration are performed through the small laparotomy wound and the port for laparoscopy.

Fig. 2. Lifting method in right hemicolectomy. I-type lifting bars are inserted into the right hypochondrial and lateral regions. Fig. 3. Surgical field in right hemicolectomy. The surgical field obtained by the lifting method is satisfactory and almost comparable to that obtained by the pneumoperitoneum method.

is reconstructed by hand-suturing and end-to-end anastomosis and brought back to the peritoneal cavity. With the lifting method, intraperitoneal hemostasis and irrigation can be performed using aspirating instruments which are used in general laparotomy, and there is no fear of poor surgical fields due to aspiration. Finally, the wound is closed.

Transverse colectomy. A small midline incision is made by open laparoscopy in the lower part of the abdomen about 2 cm caudal to the umbilicus, and one I-type lifting bar is inserted into each of the right and left hypochondrial regions (Fig. 8). After the greater omentum and transverse mesocolon are dissected, the transverse colon is mobilized sufficiently. If the lesion is close to the flexure of the liver or spleen, the hepatocolic and splenocolic ligaments are dissected and divided. After the transverse colonespecially the site of lesionis mobilized sufficiently, a small incision about 4 cm long is made in the upper part of the abdomen, and the intestine is taken out of the body. After the transverse colon is divided, it

Sigmoid colectomy, and low anterior resection. In colon surgery in the lower part of the abdomen, abdominal wall tenting tends to occur, resulting in a poor surgical field. This can be ascribed to the absence of intrinsic supportive tissue, or the ribs, unlike in the upper part of the abdomen. With our original T-type lifting bar (Fig. 9), abdominal wall tenting is unlikely to occur because the abdominal wall can be lifted as a plane. This method is therefore very useful for surgery in the lower part of the abdomen. The T-type lifting bar is inserted intraperitoneally from point (a) through a small wound made by open laparoscopy. It is advanced along the abdominal wall toward the right as viewed by the surgeon while checking the resistance. It is then advanced toward the left as viewed by the surgeon and

756

Fig. 4. Location of lifting bars and trocars in right hemicolectomy. Fig. 8. Location of lifting bars and trocars in transverse colectomy.

Fig. 9. The T-type lifting bar.

Fig. 5. A small incision about 6 cm long is made in the upper-right part of the abdomen, and the intestine is taken out of the body. Fig. 6. After the blood vessels around the margin of the ileum are divided, an incision is made in the mesentery and is extended toward the inflow point of the ileocolic vein. Fig. 7. After dissection along the surgical trunk, the ileocolic artery and vein, right colic artery and vein, and the right branches of the transverse colic artery and vein are divided at their origin and D3 dissection is performed.

inserted up to point (c). It is further advanced straight up to point (d), and insertion is completed. A laparoscope is also inserted through the same wound, and the lifting bar is inserted carefully under laparoscopy. A small wound is made by open laparoscopy in the upper-right part of the abdomen, and T- and I-type lifting bars are respectively inserted into the lower-left part of the abdomen and left lateral region and pulled (Figs. 10 and 11). The sigmoid colon is dissected from the retroperitoneum, and the descending colon is mobilized sufficiently. The left urethra and testic (ovarian) artery and vein are identified, and care must be taken not to damage them. After the lesion is identified, the mesenteric vessels are divided, and dissection of D2 is performed. The origin of the sigmoid and superior rectal arteries may also be divided under direct vision after making a small incision about 6 cm long in the lower-left part of the abdomen. With the lifting method, it is also possible to utilize both the laparoscopic port and a small laparotomy wound. Briefly, surgical instruments which are used in general laparotomy, including forceps such as Listers forceps and a Babcock clamp for holding and pulling tissue and intestine, can be inserted through the small laparotomy wound, while dissection and treatment of blood vessels can be performed under laparoscopy through the port for laparoscopy. Furthermore, gauze, swabs, and an aspiration tube as used in general laparotomy can easily be inserted through the small laparotomy wound. It is also possible to hold the tissue through the port for laparoscopy and to insert instruments such as a Kellys clamp through the small wound to perform tissue dissection and clipping for vascular treatment and hemostasis. After the intestine and mesenteric membrane including the lesion are taken out of the body through the small laparotomy wound, additional D2 dissection and enterectomy are performed. Reconstruction is achieved by instrumental side-to-side anastomosis. The mesenteric defect is closed by suturing, and the intestine is repositioned into the peritoneal cavity, followed by irrigation and aspiration of the surgical field. The drain is retained, and the wound is closed. In patients with rectal disease, a small midline incision is made in the lower part of the abdomen, and the intestine is pulled out of the peritoneal cavity. It is brought back into the peritoneal cavity after making pursestring sutures to the adoral stump of the sigmoid colon followed by anvil

757

hospital stay was 13.0 4.9 days and was shorter than with our traditional open laparotomy. We recognize that this period is longer than that in North America, but it reflects the hospital stay for patients undergoing colonic surgery according to Japanese standards, which allow for later discharge because of the public health system. Discussion Laparoscopic surgery using abdominal wall lifting has obvious advantages over surgery using pneumoperitoneum. Briefly, this method has no effect on the respirocirculatory system and causes no phlebothrombosis. Owing to these advantages, it can be used, without particular worry, on the aged and patients with poor systemic conditions. In laparoscopic-assisted colon surgery using the abdominal wall lifting method for colorectal disease, tissue dissection, taking in and out of instruments, and irrigation and aspiration can be done through a small laparotomy wound, which is made immediately after the start of surgery. Furthermore, the use of conventional surgical instruments provides an advantage with this technique. On the other hand, there is still some controversy as to whether or not a surgical field which is comparable to that obtained by the pneumoperitoneum method can be obtained by the lifting method. However, a surgical field which is almost comparable to that obtained by the pneumoperitoneum method can be obtained by our lifting method. In the present study we experienced 11 cases of colectomy accompanied with D2 and D3 dissection. Based on the results presented, our lifting method is useful, because lymphatic dissection which is the same as in general laparotomy may also be done fully under direct vision through a small laparotomy wound. Conclusion We performed laparoscopic-assisted colon surgery by the abdominal wall lifting method in 22 patients and obtained favorable results. This paper presented the results together with some discussion. References
1. Cottin V, Delafosse B, Viale J-P (1996) Gas embolism during laparoscopy. Surg Endosc 10: 166169 2. Fowler DL (1991) Laparoscopy assisted sigmoid resection. Surg Laparosc Endosc 1: 183188 3. Jacobs M (1991) Minimally invasive colon resection. Surg Laparosc Endosc 1: 144150 4. Japanese Society for Cancer of the Colon and Rectum (1994) General rules for clinical and pathological studies on cancer of the colon, rectum and anus 5. Kinbara Press, Tokyo 5. Nisii H, Hirai T, Fujita S (1992) Procedure of cholecystectomy using lifting bars under an electron endoscope. Rinsho Geka (Jpn J Clin Surg) 47: 12271230

Fig. 10. Location of lifting bars and trocars in low anterior resection and sigmoid colectomy.

Fig. 11. Surgical field obtained by the lifting method in low anterior resection and sigmoid colectomy. The surgical field from the lower-left part of the abdomen to the lateral region is favorable.

insertion. After the rectum is cut at the level of endo G1A, ECS29 or ECS31 is inserted per anum, and intracorporeal anastomosis is performed under laparoscopy. The rectum may be cut using a roticulator through the small laparotomy wound.

Results The mean operating time was 153.8 51.9 min. The nasogastric tube was usually removed on postoperative day 2, as bowel movements and liquid diet began on the same day. A regular diet was usually tolerated by postoperative 4. All patients tolerated the procedure, and following our technique, they experienced less pain compared with traditional colectomy. They were given on average only two intravenous injections of pentazosin (15 mg), which was much less than that required after open colectomy in our hospital. The postoperative course was uneventful in all patients. No particular complications were noted. The mean postoperative

Original articles
Surg Endosc (1997) 11: 703706

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Lobectomy with extended lymph node dissection by video-assisted thoracic surgery for lung cancer
S. Kaseda,1 N. Hangai,1 S. Yamamoto,2 M. Kitano2
1 2

Department of Thoracic Surgery, Saiseikai Kanagawa-ken Hospital, 6-6 Tomiya-cho, Kanagawa-ku, Yokohama 221, Japan Department of Surgery, Saiseikai Kanagawa-ken Hospital, 6-6 Tomiya-cho, Kanagawa-ku, Yokohama 221, Japan

Received: 10 May 1996/Accepted: 19 November 1996

Abstract. Background: Between September 1992 and September 1996, we performed 88 VATS (video-assisted thoracic surgery) lobectomies and two VATS pneumonectomies. Methods: The indications for surgery were 68 cases of lung cancer, nine cases of bronchiectasis, six cases of tuberculosis, and seven cases of benign lesions. Of the 68 cases of lung cancer, 36 were treated by VATS lobectomy with extended lymph node dissection for clinical stage I lung cancer, making full use of recently developed devices for thoracoscopic surgery, such as roticulating endoscissors, miniretractors, endoclips, and harmonic scalpels. Results: Twenty-four lymph nodes were resected on average (range, 10 to 51) by VATS. This number was comparable to lymph nodes resected in open thoracotomy during the same period. Among the 36 patients who underwent extended lymph node dissection, 20 showed no lymph node metastasis postoperatively (stage I), while 16 had N1 or N2 cancer. All patients with stage I cancer have survived 4 to 36 months (median: 17 months) with no signs of recurrence. Conclusions: This survival of stage I lung cancer after VATS is comparable to that of open thoracotomy. We thus believe that VATS lobectomy with extended lymph node dissection can be an alternative to standard posterolateral thoracotomy for stage I lung cancer. Key words: Thoracoscopy Lung neoplasms Lobectomy Pneumonectomy Video-assisted Lymph node

results comparable to open thoracotomy [3]. We accomplished this by using newly developed thoracoscopic instrumentation such as roticulating endoscissors, miniretractors, endoclips, and harmonic scalpels, in addition to instruments for conventional surgery and endostapler devices. In this paper, we report on the method and results of our VATS lobectomy and extended lymph node dissection for lung cancer.

Materials and methods

Between September 1992 and September 1996, 88 VATS lobectomies and two VATS pneumonectomies were performed. The patients consisted of 68 cases of lung cancer, nine cases of bronchiectasis, six cases of tuberculosis, and seven cases of benign lesions. Among 68 patients with lung cancer, 66 were evaluated as stage I cancer because they did not have lymph nodes larger than 1 cm in size on CT scan. Of these 66 patients with stage I lung cancer, 36 underwent extended lymph node dissection as well. Thirty-five of these were peripherally located adenocarcinoma (n 30) and squamous cell carcinoma (n 5), and one was a carcinoid tumor originating from the orifice of the right middle lobe. Remaining 30 patients received only lymph node sampling because they were older than 75 years old or had major complications like cerebral infarction, asthma, emphysema, or other malignant diseases other than lung cancer. Under full flexion in a decubitus position which made spreading with a retractor unnecessary, a thoracoscope was introduced through the seventh intercostal space at the midaxillary line. A small anterolateral access thoracotomy (58 cm) was made in the fourth intercostal space for an upper lobectomy or pneumonectomy, or in the fifth intercostal space for a middle or lower lobectomy. The size of the incision was decided by the size of the primary tumor (1245 mm, median 31 mm) or the volume of the specimen to be removed. Two access holes 510 mm in size were made at the postaxillary line, which enabled the operator and the first assistant to make full use of their hands. Separation of blood vessels from the surrounding tissue was performed by using forceps for conventional thoracic surgery. Blood vessels with a diameter of 10 mm or more were transected with an Endo GIA stapler (United States Surgical Corporation, Norwalk, CT: USCC) after ligation with silk suture at the proximal portion. In cases where it was difficult to position the Endo GIA stapler through the operating trocar, we either created another access port for the stapler or applied a Roticulator 30-V3 stapler (USSC) through the thoracotomy opening. The stump of the bronchus was closed with a Roticulator 30-3.5 stapler (USSC) during lobectomy. During pneumonectomies, an appropriate portion of the main bronchus was cut with a Roticulator 30-4.8 stapler (USSC)

After performing a large volume of less-demanding thoracoscopic procedures, we began to apply the video-assisted thoracic surgery (VATS) technique for lobectomy and pneumonectomy. Furthermore, we extended the application of VATS to include extended lymph node dissection with
Correspondence to: S. Kaseda

704

Fig. 1. The right main bronchus is retracted with an ENDO RETRACT MAXI (USSC) and lymph nodes are grasped with long grasping forceps. Fig. 2. Separation of lymph nodes with harmonic scalpel.

prior to lymph node dissection. After lymph node dissection, the bronchus was again transected with a Roticulator 30-4.8 stapler near the carina. On the right side, extended lymph node dissection was started at the lower mediastinum. During dissection at the carina, the esophagus and the bronchus were compressed with a miniretractor (ENDO MINI-RETRACT: USSC), and the bronchial artery was either clipped with endoscopic clips (ENDO CLIP II: USSC) and cut or transected with electrocautery or harmonic scalpel (Ethicon, Inc., Cincinnati). Under the effective guidance of miniretractors, we could extend the dissection of lymph nodes to the contralateral hilus. During dissection at the upper mediastinum, the azygos vein was ligated with suture and transected, and the right main bronchus was then retracted laterally using endo loops (Ethicon). The vagus nerve was taped with silk thread which was brought outside the thorax with an ENDO CLOSE (USSC), and the right main bronchus was retracted with ENDO RETRACT MAXI (USSC) to obtain a good operative view (Fig. 1). We used longer forceps than those for conventional thoracic surgery in the extended resection of mediastinal lymph nodes. Holding lymph nodes with a long grasping forceps, we separated them from the trachea with harmonic scalpel (Fig. 2), and then ligated small vessels with endoscopic clips and cut (Fig. 3). When dissecting at the most upper mediastinum, we avoided using electrocautery, so as not to damage the recurrent laryngeal nerve (Fig. 4). On the left side, we could also perform extended dissection at the carina

by compressing the esophagus and pericardium with retractors. In addition, we were able to dissect pretracheal and tracheobronchial lymph nodes by compressing the aorta with an ENDO RETRACT MAXI (USSC) or taping it to obtain a good operative view.

Results Operating time for VATS lobectomy and extended lymph node dissection was 161 to 395 min (295 on average). In the earlier cases, it took longer time to perform extended lymph node dissection because of the immaturity of the technique. However, in the latest ten consecutive cases, we were able to operate within 234 min on average by doing extended lymph node dissection within 60 to 90 min with good cooperation between operator and assistants. For VATS lobectomy or pneumonectomy, 4.6 cartilage for autosuture were used on average. This number was comparable to that in open thoracotomy. In the early period of our experience,

705

Fig. 3. Removal of lymph nodes using scissors after clipping them at the proximal portion. Fig. 4. The appearance of upper mediastinum after completion of extended lymph node dissection.

a massive hemorrhage from the pulmonary artery occurred due to failure of staples to close in one patient. In another patient, pyothorax occurred because of infection from the drainage port. There were no other serious complications noted postoperatively. The number of resected lymph nodes ranged from ten to 51 (24 on average) in 36 patients who underwent extended lymph node dissection under the diagnosis of clinical stage I lung cancer. Although all 36 patients, who underwent extended lymph dissection were preoperatively diagnosed with N0, pathological examination revealed N1 (stage II) in six patients and N2 (stage IIIA) in ten patients. Patients with N0 and N1 were discharged from the hospital after confirming the absence of bronchial fistula by bronchoscopy 2 weeks after operation. However, patients with N2 underwent chemotherapy. Consequently, median hospitalization time after operation was 24 days on average. Of the 20 patients with N0 cancer (stage I), no recurrence has oc-

curred after 4 to 36 months (median: 17 months) follow-up. However, recurrence has occurred in one patient with N1 and six patients with N2.

Discussion After performing approximately 300 less-demanding thoracoscopic surgical procedures, we began to apply VATS techniques to lobectomy and pneumonectomy [2, 4, 7]. Since the instrumentation for thoracoscopic surgery is not adequate to perform dissection around the hilar vessels, we often use forceps for conventional thoracic surgery. While small blood vessels are ligated with silk threads using the forcepsto cut costs, among other reasonsvessels with a larger diameter are safely cut with an endostapler, such as the Endo GIA. The Endo GIA endostapler clasps the tissue with six

706

rows of staples and simultaneously transects between staple rows with a knife blade. In the early period of our experience, a massive hemorrhage was encountered from the pulmonary artery of one patient due to failure of staples to close properly. In that case, the pulmonary artery was clamped promptly with a vascular clamp and successfully sutured. Although operation on the bronchus is not so critical, insufficient stapling of the artery is often lifethreatening [9]. After this incident, we started to apply a vascular Endo GIA after ligating large vessels at the proximal portion in case the vessel is cut without being properly stapled. We have not experienced any trouble in operating on the pulmonary artery or vein since this alteration in our surgical technique. However, we think that VATS lobectomy should not be done by an inexperienced surgeon who cannot cope with sudden bleeding from major vessels. The diameter of the Endo GIA staple is 0.21 mm. This staple is only strong enough to clasp the rather thin bronchial wall. The Roticulator 30-3.5 and 30-4.8 endostaplers have staples with diameters of 0.23 mm and 0.28 mm, respectively. These staples are suitable for clasping thicker bronchial walls. At present, we usually use a Roticulator endostapler to operate on the bronchus. The thoracotomy opening should be over 5 cm long for sufficient maneuverability of a Roticulator endostapler. It will be feasible to perform VATS through a smaller opening, when a roticulator-type Endo GIA stapler is developed in the future. In the early period of our experience with VATS lobectomy and pneumonectomy, we thought that it was not possible to perform extended lymph node dissection, as in open thoracotomy. Therefore, we converted VATS to open thoracotomy to perform lymph node dissection once interoperative pathologic examination confirmed the diagnosis of lung cancer. Since the development of roticulating endoscissors, miniretractors, and endoclips for thoracoscopy, it has become possible to perform not only VATS lobectomy and pneumonectomy but also VATS lymph node dissection. As a result, ten to 51 (24 on average) lymph nodes were resected by VATS in 36 patients with clinical stage I lung cancer. This number is not significantly different from

the lymph node range of 16 to 36 (22 on average) resected from patients with clinical stage I lung cancer who underwent conventional open thoracotomy during the same period. Thus, clinical stage I lung cancer really has become a candidate for thoracoscopic surgery. Although there are controversies over the significance of extended lymph node dissection in the case of lung cancer [1, 5, 6, 8], thorough pathologic examination of resected lymph nodes at least provides accurate staging and a reliable prediction of the prognosis. As for postoperative pain, VATS lobectomy is far less invasive in comparison with open thoracotomy. In conclusion, VATS lobectomy with extended lymph node dissection can be carried out safely and may be a strong alternative to standard posterolateral thoracotomy for stage I lung cancer. References
1. Daly BDT, Mueller JD, Faling LJ, et al. (1993) N2 lung cancer: outcome in patients with false-negative computed tomographic scans of the chest. J Thorac Cardiovasc Surg 105: 904911 2. DeCamp MM, Jaklitsch MT, Mentzer SJ, et al. (1995) The safety and versatility of video-thoracoscopy: a prospective analysis of 895 consecutive cases. J Am Coll Surg 113120 3. Kaseda S, Kitano M (1994) Lobectomy and lymph node dissection under thoracoscopic guidance. Surg Endosc 8: 632 4. McKenna RJ (1994) Lobectomy by video-assisted thoracic surgery with mediastinal node sampling for lung cancer. J Thorac Cardiovasc Surg 107: 879882 5. Pearson FG, Delarue NC, Ilves R, et al. (1982) Significance of positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J Thorac Cardiovasc Surg 83: 111 6. Pearson FG (1993) Staging of mediastinum: role of mediastinoscopy and computed tomography. Chest 103: 346s348s 7. Roviaro G, Varoli F, Rebuffat C, et al. (1993) Major pulmonary resections: pneumonectomies and lobectomies. Ann Thorac Surg 56: 779 783 8. Shield TW (1990) The significance of ipsilateral mediastinal lymph node metastasis (N2 disease) in non-small cell carcinoma of the lung. J Thorac Cardiovasc Surg 90: 4853 9. Yim APC, Ho JKS (1995) Malfunctioning of vascular staple cutter during thoracoscopy lobectomy. J Thorac Cardiovasc Surg 109: 1252

Surg Endosc (1997) 11: 774777

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Consequences of lost gallstone

M. P. McDonald, J. L. Munson, L. Sanders, J. Tsao, J. Buyske
Department of General Surgery, Lahey Clinic Foundation, 41 Mall Road, Burlington, MA 01805, USA Received: 22 March 1996/Accepted: 3 June 1996

Abstract. Laparoscopic cholecystectomy has become the treatment of choice in the management of calculus gallbladder disease. Intraperitoneal gallstone loss is not uncommon; it occurs in up to 40% of cases. Often, the stones are left unretrieved and are thought to be inconsequential. We present a series of patients who have had serious sequela from gallstones in the peritoneal cavity. We performed a retrospective study of the management of six patients with complications from intraperitoneal gallstones. The patients presented with a variety of complaints, from fevers to pneumonia to a colo-cutaneous fistula. Presentation ranged from immediately postoperatively to 18 months after surgery. Diagnosis included perihepatic abscesses and colo-biliary fistula. General anesthesia was usually necessary for removal of the stones. All patients have resolved following the removal of the gallstones. Our recommendation is to attempt to avoid spillage through careful dissection and retrieve any lost stones. The defect in the gallbladder can be closed with a clip. Whether the procedure should be converted to an open one to retrieve all the stones remains open to debate. The surgeon should be aware of the possible consequences of the lost gallstone. Key words: Laparoscopic cholecystectomy Complications Lost gallstones

lost stones and irrigate the abdomen. This is generally much more difficult to do in laparoscopic procedures. Often, the stones are left unretrieved and are thought to be inconsequential. Welch and colleagues have reported studies in which there were no adverse effects from gallstones that were left behind in the peritoneal cavity [10]. Others, however, have reported that an occasional patient has had a complication related to a retained stone left free in the abdomen. This excludes cases where stones or debris contaminated the wound as the gallbladder was removed from the abdomen. We present a series of six patients who have had serious sequelae from gallstones left behind in the peritoneal cavity. Materials and methods
This is a retrospective chart review of six patients who presented to our institution for management of various complications found to have been related to gallstones free in the peritoneal cavity. There were two females and four males. The ages ranged from 48 to 67 years old. Two patients had heart disease, one had diabetes, and four were described as healthy patients.

Results There were six patients with complications from gallstones left behind at laparoscopic cholecystectomy. Descriptions of the initial cholecystectomy were varied. Review of the operative reports revealed that five out of six reports described the gallbladder as inflamed. The sixth patient was described as having an uneventful cholecystectomy, and it was specifically noted that no stones were spilled. One patient had a necrotic gallbladder, and another had pus drained from the gallbladder prior to removal of the specimen. A third patient suffered an injury to the common bile duct and underwent immediate reconstruction with a hepaticojejunostomy. Time from operation to patient presentation fell from within a few days of surgery to 18 months later. Patient complaints included weakness, chills and malaise, pleuritic chest pain, right shoulder pain, and right back pain. Two patients, both of whom presented within 2 weeks of surgery, had bile drainage from their trocar or drain sites. Comput-

Laparoscopic cholecystectomy has become the treatment of choice in the management of calculus gallbladder disease. It is a safe procedure, achieving a complication rate of 35%, which is similar to that of open cholecystectomy. Intraperitoneal gallstone loss is not uncommon, occurring in 10 40% of laparoscopic cases that report such as a complication [3, 11]. In open cases, it is not difficult to retrieve these

Correspondence to: J. Buyske, Hospital of University of Pennsylvania, 4 Silverstein Pavilion, 3400 Spruce St, Philadelphia, PA 19104, USA

775

Fig. 1. This CT scan demonstrates an intrahepatic abscess with a small gallstone at its base. Fig. 2. A subhepatic abscess collection with a free-floating gallstone as its nidus.

erized tomography scan-guided drainages were performed in most of these patients. Of those patients requiring CTguided drainage of an abscess, multiple drainage procedures were often required to achieve success. One patient had an intrahepatic abscess from a gallstone embedded in the liver parenchyma (Fig. 1). He underwent CT-guided drainage of the abscess, followed by percutaneous catheter-guided intrahepatic lithotripsy to crush and remove the offending stone. He temporarily improved; however, a year later he complained of vague right upper quadrant pain and underwent a CT scan. The scan revealed a residual subphrenic abscess which required surgery for debridement. The cavity, though it drained purulent material, was sterile by culture. His symptoms have now resolved. Another patient presented with right upper quadrant pain, scapular pain, fevers, and chills 2 weeks after initial surgery. A CT revealed a subphrenic abscess, which was treated by CT-guided drainage. Gallstones were found in the drain effluent. A third patient developed a subhepatic abscess that required CT-guided drainage. That patient presented 12 days after surgery with malaise, chills, and a fever to 101F. A CT scan revealed the subhepatic abscess (Fig. 2). Again, a gallstone was found as the nidus for the abscess. Later, a flank abscess developed at the site of the CT drain. This required an incision and drainage in the office. Once the gallstones and debris were removed, the patient recovered. Two patients developed fistula originating from a lost gallstone. One patient had a common bile duct injury during a laparoscopic cholecystectomy at an outside hospital. She subsequently underwent a hepaticojejunostomy at that institution and developed a biliary fistula postoperatively. Percutaneous transhepatic cholangiograms revealed no evidence of an anastomotic leak. Twenty months later, after her referral to this hospital, a sinogram revealed a sinus tract

originating near the anastomosis. A gallstone was seen at the base of the tract (Fig. 3). She underwent an exploratory laparotomy to lay open the sinus tract; a 1.5-cm pigmented gallstone lay at the base of the tract and was removed. The biliary-enteric anastomosis was left intact. She recovered without further intervention. A second patient presented with pus from a trocar site 10 days postoperatively. This was incised and drained. He then drained bile and gallstone debris from that site. An endoscopic retrograde cholangiogram revealed a colo-biliary-cutaneous fistula (Fig. 4). Percutaneous drains were placed under CT guidance. He resolved but later reopened the fistula. He underwent exploratory laparotomy and resection of his colonic-biliarycutaneous fistula. Several gallstones and debris were recovered along the tract. Though he did well after that surgery, he re-presented with a flank abscess several weeks later. He failed CT-guided drainage and required another exploratory laparotomy to drain the abscess, revealing pus and multiple 24-mm gallstones. He ultimately recovered from his difficulties. The final patient in this series was a 48-year-old male who underwent a laparoscopic cholecystectomy. Eighteen months after his surgery he presented to his medical doctor with dyspnea on exertion and right pleuritic pain. A chest x-ray revealed an elevated right hemidiaphragm and right pleural consolidation. He was treated with antibiotics for a presumed pneumonia. After failing to improve, he underwent a CT scan of the chest which demonstrated continued right lower lobe consolidation. A bronchoscopy was performed and though the cultures from this were negative, he was again treated for pneumonia with 2 months of antibiotics. Finally a CT of the abdomen was performed which revealed a subphrenic abscess with a gallstone at its center. He underwent surgery to debride the abscess and multiple stones were found and removed. The cultures from this

776

Fig. 3. A colo-biliary-cutaneous fistula with a gallstone as the nidus, as demonstrated by this ERCP. The clip can be seen across the cystic duct. Fig. 4. A biliary-cutaneous sinus. This is a sinogram which shows the leak from the biliary tract, leading to the free gallstone, then to a sinus cavity, and finally out to the skin. A previously performed percutaneous transhepatic cholangiogram demonstrated that the biliary anastomosis was clearly patent into the small bowel.

Table 1. Patient presentations Day of Patient presentation 1 2 3 4 5 6 12 days Symptoms signs Diagnosis Studies 5 CT 1 U/S PTC Fistulagram 3 CT 1 ERCP Fistulagram 4 CT Number of procedures 2 CT drain 1 I and D None Number of operations 0 1

Immediate; referred at 20 months 10 days Bile, pus, stool from trocar site 10 months 2 weeks 18 months Total

Weakness, malaise, Subhepatic abscess, temp 101F, chills flank abscess Fistula Fistula Colo-biliary-cutaneous fistula; subcut. aneous abscess Liver abscess

R shoulder scapular pain, dyspnea on exertion RUQ pain, scapular Subphrenic abscess pain, fevers, chills Dyspnea on exertion, Subphrenic abscess R pleuritic pain

4 CT CXR 2 CT Bronchoscopy CXR 17 CT 5 CT drains 1 ERCP 2 I and D 1 U/S 1 bronchos2 fistulagrams copy

Nasobil. stent 2 CT drain 1 I and D CT drain 2 percutaneous lithotripsy CT drain 0 1 6

RUQ, right upper quadrant; CT, CAT scan; U/S, ultrasound; PTC, percutaneous transhepatic cholangiogram; I and D, incision and drainage; CXR, chest x-ray; CT drain, CT-guided drainage.

abscess were negative. His pulmonary manifestations have resolved since surgery. The patient presentations are summarized in Table 1. Among these six patients significant suffering, morbidity, lost days, and expense resulted from lost gallstones. Ten hospital admissions were required, with multiple office visits as well. At least 17 CT scans, three fistulagrams or sinograms, several ultrasounds, chest radiographs, and multiple cultures were performed. Short- or long-term antibiot-

ics were required in all cases. Six exploratory laparotomies were needed and an additional general anesthesia for the percutaneous lithotripsy. Eight additional procedures were needed, such as CT-guided drainages. A bronchoscopy was performed on a patient for whom the true diagnosis was not suspected. All patients have recovered from their complications save one gentleman who continues to drain small amounts of granular material 10 months after his subhepatic abscess was drained.

777

Discussion It has been traditionally taught that lost gallstones are harmless when left in the abdomen [10]. More recently, it has been established that these stones do not resolve and in fact, are often found upon later exploration [1]. In experimental studies, dense adhesions are found around gallstones when free stones were implanted in experimental animals [6], or when human patients were reexplored for obstructive symptoms [4]. At least two studies have found that gallstones can become adherent to adjacent solid organs and could imbed themselves in these organs, causing significant structural damage [2, 11]. This could involve the liver parenchyma, the colon, and/or the abdominal wall as demonstrated in several of our cases. Adherence to a solid structure caused fistula or sinus and abscess formation. Most patients illnesses resolved when the stones were removed. The stone apparently does not need to be infected to cause these complications, as one of our patients had persistently negative culture results. Several case reports have discussed the rare incidence of complications from lost gallstones [5, 8, 9, 12]. It is not known how often complications occur. If stones really are lost in 1040% of all laparoscopic cases, then the percent causing complications is extremely low. As most of the patients in this series were referrals to our institution, we cannot estimate the overall incidence of complications. However, when complications do occur, they can cause terrible morbidity for these patients, as well as great expenditure in time and medical resources. This series represents the largest to date reporting major complications from the lost gallstone. Our recommendation is to attempt to avoid spillage through careful dissection and to safely retrieve any lost stones. If a perforation does occur, the defect should be closed with a clip or an endoloop, and the gallbladder should be removed in a bag to prevent further extrusion of stones. Spilled stones should be retrieved by using graspers, irrigation, and suction, or by placing

stones in a bag and removing them from the abdomen. Whether the procedure should be converted to an open one to retrieve all the stones is subject to debate. The Southern Surgeons Club reports a conversion rate of 0.1% to retrieve lost stones [7]. The surgeon must be aware of the possible complications of the lost gallstone and this entity should be considered in the differential diagnosis of any patient who fails to thrive after a laparoscopic cholecystectomy.

References
1. Cline RW, Poulos E, Clifford E (1994) An assessment of potential complications caused by intraperitoneal gallstones. Am Surg 60: 303 305 2. Cohen TV, Pereira PRB, de Barros MV, Ferreira EAB, Tolosa EM (1994) Is the retrieval of lost peritoneal gallstones worthwhile? Surg Endosc 8: 1360 3. Peters JH, Gibbons GD, Innes JT et al. (1991) Complications of laparoscopic cholecystectomy. Surgery 110: 769778 4. Ponce J, Cutshall E, Hodge MJ, Browder W (1995) The lost laparoscopic stone. Arch Surg 130: 666669 5. Regoly-Merei J, Ihasz M (1995) The sequelae of retained or lost stones. Surg Endosc 9: 931932 6. Sax HC, Adams JT (1993) The fate of the spilled gallstone. Arch Surg 128: 469 7. The Southern Surgeons Club (1991) A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 324: 10731078 8. Tschmelitsch J, Glaser K, Klinger P, Bodner E (1993) Late complication caused by stone spillage during laparoscopic cholecystectomy. Lancet 342: 369 9. Van Brunt PH, Lanzafame RJ (1994) Subhepatic inflammatory mass after laparoscopic cholecystectomy. Arch Surg 129: 882883 10. Welch N, Hinder RA, Fitzgibbons RJ, Rouse JW (1991) Gallstones in the peritoneal cavity: a clinical and experimental study. Surg Laparosc Endosc 1: 246247 11. Wetscher F, Schwab G, Fend F, Glaser K, Ladurner D, Bodner E (1994) Subcutaneous abscess due to gallstones lost during laparoscopic cholecystectomy. Endoscopy 26: 325326 12. Wilton PB, Andy OJ, Oeters JJ, Thomas CF, Patel VS, Scott-Conner CEH (1993) Laparoscopic cholecystectomy: leave no (spilled) stone unturned. Surg Endosc 7: 537538

News and notices

Surg Endosc (1997) 11: 793795

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

New Address for the European Association for Endoscopic Surgery (E.A.E.S.)
Effective January 1, 1997, the new correspondence, telephone, and fax numbers of the E.A.E.S. office are: E.A.E.S. Office, c/o Mrs. Ria Palmen Luchthavenweg 81 Unit 1.42 5657 EA Eindhoven The Netherlands or: P.O. Box 335 5500 AH Veldhoven The Netherlands Tel: +31 40 2525288 Fax: +31 40 2523102

Fellowships in Minimally Invasive Surgery The University of Pittsburgh Medical Center Pittsburgh, PA, USA
One year fellowships in advanced minimally invasive surgery in both general and thoracic surgery are being offered at the University of Pittsburgh Medical Center beginning on July 1, 1997. Requirements include completion of residency training programs in the desired area. The fellowships include a competitive salary and travel allowance. Interested candidates should send a letter of inquiry with curriculum vitae to: Philip R. Schauer, MD (General Surgery) or James Luketich, MD (Thoracic Surgery) The University of Pittsburgh Medical Center 3471 Fifth Avenue Suite 300 Pittsburgh, PA 15213-3221

Volunteer Surgeons Needed Northwestern Nicaragua Laparoscopic Surgery Teaching Program, Leon, Nicaragua
Volunteer surgeons are needed to tutor laparoscopic cholecystectomy for this non-profit collaboration between the Nicaraguan Ministry of Health, the National Autonomous University of Nicaragua, and Medical Training Worldwide. The program consists of tutoring general surgeons who have already undergone a basic laparoscopic cholecystectomy course. Medical Training Worldwide will provide donated equipment and supplies when needed. For further information, please contact: Medical Training Worldwide Ramon Berguer, MD, Chairman Tel: 707-423-5192 Fax: 707-423-7578 e-mail: berguer.r@martinez.va.gov

Fellowships in Laparoscopic Surgery Staten Island University Hospital Staten Island, NY USA
A one year fellowship, to start July 1, 1997, in advanced laparoscopic surgery is being offered at Staten Island University Hospital. The selected fellow will be exposed to many advanced general laparoscopic surgeries including: hiatal hernia repair, splenectomy, adrenalectomy, bowel resection, and others. Participation in research projects will be encouraged. For further information, please contact: Barbara Coleman Coordinator, Surgical residency program Tel: 718-226-9508

Fellowship in Minimally Invasive Surgery George Washington Medical Center Washington, DC USA
A one-year fellowship is being offered at the George Washington University Medical Center. Interested candidates will be exposed to a broad range of endosurgical Education and Research Center. Active participation in clinical and basic science research projects is also encouraged. For further information, please contact: Carole Smith 202-994-8425 or, send curriculum vitae to: Dr. Jonathan M. Sackier Director of Endosurgical Education and Research George Washington University Medical Center Department of Surgery 2150 Pennsylvania Avenue, N.W. 6B-417 Washington, DC 20037, USA

Essentials of Laparoscopic Surgery Surgical Skills Unit University of Dundee Scotland, UK

Under the direction of Professor A. Cuschieri the Surgical Skills Unit is offering a three-day practical course designed for surgeons who wish to undertake the procedures such as laparoscopic cholecystectomy. This intensely practical program develops the necessary operating skills, emphasizes safe practice, and highlights the common pitfalls and difficulties encountered when starting out. Each workshop has a maximum of 18 participants who will learn both camera and instrument-manipulation skills in a purpose-built skills laboratory. During the course there is a live demonstration of a laparoscopic cholecystectomy. The unit has a large library of operative videos edited by Professor Cuschieri, and the latest books on endoscopic surgery are on display in our Resource area. Course fee including lunch and course materials is $860.

794 For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042 recreational pursuits which can be arranged by the facility to suit your personal agenda. For further details please contact: Carole Smith: Department of Surgery 2150 Pennsylvania Avenue NW 6B Washington, DC 20037, USA Tel: (202) 994-8425

Advanced Endoscopic Skills Surgical Skills Unit University of Dundee Scotland, UK

Each month Professor Cuschieri Surgical Skills Unit offers a 412 day course in Advanced Endoscopic Skills. The course is intensely practical with hands on experience on a range of simulated models. The program is designed for experienced endoscopic surgeons and covers advanced dissection techniques, extracorporeal knotting techniques, needle control, suturing, internal tying technique, stapling, and anastomotic technique. Individual workstations and a maximum course number of 10 participants allows for personal tuition. The unit offers an extensive collection of surgical videos and the latest books and publications on endoscopic surgery. In addition, participating surgeons will have the opportunity to see live advanced laparoscopic and/or thoracoscopic procedures conducted by Professor Cuschieri and his team. The course is endorsed by SAGES. Course fee including lunch and course materials is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Courses at the Royal Adelaide Centre for Endoscopic Surgery

Basic and Advanced Laparoscopic Skills Courses are conducted by the Royal Adelaide Centre for Endoscopic Surgery on a regular basis. The courses are limited to six places to maximize skill development and tuition. Basic courses are conducted over two days for trainees and surgeons seeking an introduction to laparoscopic cholecystectomy. Animal viscera in simulators is used to develop practical skills. Advanced courses are conducted over four days for surgeons already experienced in laparoscopic cholecystectomy who wish to undertake more advanced procedures. A wide range of procedures are included, although practical sessions can be tailored to one or two procedures at the participants request. Practical skills are developed using training simulators and anaesthetised pigs. Course fees: $A300 ($US225) for the basic course and $A1,600 ($US1,200) for the advanced course. For further details and brochure, please contact: Dr. D. I. Watson or Professor G. G. Jamieson The Royal Adelaide Centre for Endoscopic Surgery Department of Surgery Royal Adelaide Hospital Adelaide SA 5000 Australia Tel: +61 8 224 5516 Fax: +61 8 232 3471

The Practical Aspects of Laparoscopic Fundoplication Surgical Skills Unit University of Dundee Scotland, UK
A three-day course, led by Professor Cuschieri, designed for experienced laparoscopists wishing to include fundoplication in their practice. The course covers the technical details of total and partial fundoplication using small group format and personal tuition on detailed simulated models. There will be an opportunity to observe one of these procedures live during the course. Maximum course number is six. Course fee including lunch is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Advanced Laparoscopic Suturing and Surgical Skills Courses MOET Institute San Francisco, CA, USA
Courses are offered year-round by individual arrangement. The MOET Institute is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians and designates these CME activities for 2040 credit hours in Category 1 of the Physicians Recognition Award of the American Medical Association. These programs are also endorsed by the Society of Gastrointestinal Endoscopic Surgeons (SAGES). For further information, please contact: Wanda Toy, Program Administrator Microsurgery & Operative Endoscopy Training (MOET) Institute 153 States Street San Francisco, CA 94114, USA Tel: (415) 626-3400 Fax: (415) 626-3444

Courses at George Washington University Endosurgical Educational and Research Center

George Washington University Endosurgical Educational and Research Center is proud to offer a wide range of surgical endoscopy courses. These courses include advanced laparoscopic skills such as Nissen fundoplication, colon resection, common bile duct exploration, suturing, as well as subspecialty courses. Individual surgeons needs can be met with private tuition. The Washington D.C. area is a marvelous destination to visit for

Courses at WISE Washington Institute for Surgical Endoscopy Washington, DC, USA
The Washington Institute of Surgical Endoscopy is pleased to offer the following courses:

795 Laparoscopic antireflux and hiatal hernia surgery (July 1415, 1997); Laparoscopic management of the common bile duct and difficult cholecystectomy (May 1516, August 1112, November 1011, 1997); Laparoscopic colon and rectal surgery (June 2021, September 1516, December 45, 1997). Also, courses for operating room nurses and technicians will be run on a monthly basis and personal instruction and preceptorship is available. For further information, please call: Carole Smith Washington Institute of Surgical Endoscopy 2150 Pennsylvania Avenue, N.W. Washington, DC 20037 Tel: 202-994-9425

9th International Meeting Society for Minimally Invasive Therapy July 1416, 1997 Kyoto, Japan
Scientific program to include: Plenary, Parallel, Poster, and Video sessions. Host Chairman: Professor Osamu Yoshida, Department of Urology, Kyoto University, 54 Shogoin Kawahara-sho, Sakyo, Kyoto 606, Japan. Phone: +81 75 751-3328, Fax: +81 75 751-3740. This meeting coincides with the Gion Festival in Kyoto, one of the greatest festivals in Japan. For further information, please contact: Secretariat of SMIT 9th Annual International Meeting c/o Academic Conference Planning 383 Murakami-cho Fushimika, Kyoto 612 Japan Tel: +81 75 611-2008 Fax: +81 75 603-3816

Call for Abstracts Society of American Gastrointestinal Endoscopic Surgeons (SAGES) 1998 Annual Meeting April 14, 1998 Seattle, WA, USA
Abstract deadlines: Oral and Poster abstracts: September 12, 1997 Video Submissions: September 18, 1997 For further information, or to obtain an abstract form, please contact: SAGES Program Committee Society of American Gastrointestinal Endoscopic Surgeons Suite #3000 2716 Ocean Park Boulevard Los Angeles, CA 90405 Tel: (310) 314-2404 Fax: (310) 314-2585 e-mail: SAGESMail@AOL.com

Colorectal Disease in 1998 February 1921, 1998 Fort Lauderdale, FL, USA Symposium Director: Steven D. Wexner, MD
Cleveland Clinic Florida presents its ninth annual postgraduate course. Provides an intensive, in-depth, analytical review of all aspects of colorectal disease, including laparoscopy; colorectal carcinoma screening and genetics, inflammatory bowel disease; and pouch surgery. There will be a review of both basic and advanced principles of diagnosis and management of disease. Video techniques will be shown as well. The faculty is internationally represented and includes leading experts in the field. Simultaneous Spanish and Italian translation is available. For more information, please contact: Cleveland Clinic Florida Department of Education 2950 West Cypress Creek Road Fort Lauderdale, FL 33309-1743 Tel: 800-359-6101, ext. 6066 Fax: 954-978-5539

European Course on Laparoscopic Surgery (English language) November 1821, 1997 Brussels, Belgium
Course director: G.B. Cadiere For further information, please contact: Administrative Secretariat Conference Services s.a. Avenue de lObservatoire, 3 bte 17 B-1180 Bruxelles Tel: (32 2) 375 16 48 Fax: (32 2) 375 32 99

Courses Offered at the University of Minnesota Minneapolis, Minnesota, USA September 17, 1997: Fourth Annual Conference, Molecular Biology of Colorectal Cancer September 17, 1997: Sixth Annual Conference, Endorectal Ultrasonography September 1820, 1997: Sixtieth Annual Conference, Principles of Colon and Rectal Surgery
For further information, please contact: Continuing Medical Education University of Minnesota 615 Washington Avenue SE, Suite 107 Minneapolis, MN 55414 Tel: 800-776-8636 Fax: (612) 626-7766

6th World Congress of Endoscopy Surgery Roma 98 6th International Congress of European Association for Endoscopic Surgery June 36, 1998 Rome, Italy
The program will include: the latest, original high quality research; symposia; plenary lectures; abstract presentations (video, oral, and posters); EAES and SAGES postgraduate courses, OMED postgraduate course on therapeutic endoscopy; working team reports; educational center and learning corner; meeting of the International Society of Nurses and Associates; original and non original scientific reports; and a world expo of new technology in surgery. For further information, please contact: Congress Secretariat: Studio EGA Viale Tiziano, 19 00196 Rome, Italy Tel: +39 6 322-1806 Fax: +39 6 324-0143

Surg Endosc (1997) 11: 769771

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Prosthetic reinforcement of posterior cruroplasty during laparoscopic hiatal herniorrhaphy

C. T. Frantzides, M. A. Carlson
Department of Surgery, Medical College of Wisconsin, 9200 West Wisconsin Avenue, Milwaukee, WI 53226, USA Received: 5 March 1996/Accepted: 3 June 1996

Abstract. Symptomatic gastroesophageal reflux after Nissen fundoplication may occur if the wrap herniates into the thorax. In an attempt to prevent recurrent hiatal hernia we employed polytetrafluoroethylene (PTFE) mesh reinforcement of posterior cruroplasty during laparoscopic Nissen fundoplication and hiatal herniorrhaphy. Three patients with symptomatic gastroesophageal reflux and a large (8 cm) hiatal defect underwent laparoscopic posterior cruroplasty and Nissen fundoplication. The cruroplasty was reinforced with a PTFE onlay. No perioperative complications occurred, and in follow-up (11 months) the patients are doing well. When repairing a large defect of the esophageal hiatus during fundoplication, the surgeon may consider reinforcement of the repair with PTFE mesh. Key words: Laparoscopic Nissen fundoplication Gastroesophageal reflux Hiatal hernia Polytetrafluoroethylene Prosthetic Antireflux procedure

posterior cruroplasty in three patients with symptomatic GER and large hiatal hernia. Case reports
Our technique of laparoscopic Nissen fundoplication and posterior cruroplasty follows a previous description [5]. Cefazolin (2 g IV) is given with induction of general anesthesia. Carbon dioxide pneumoperitoneum and five 10-mm trocars are used. The lesser omentum and phrenoesophageal ligament are incised, a 50 Fr bougie is placed within the esophagus, and a window is created posterior to the esophagus with blunt dissection. The hiatal hernia is reduced with sharp and blunt dissection of the hernia sac and with gentle traction on the stomach. The short gastric vessels are ligated with a right-angle clip applier or an ultrasonic scalpel (Ehicon). The esophagus is circumferentially mobilized until the lower 5 cm is intrabdominal. A typical hiatal hernia for which mesh only was utilized is shown in Fig. 1. Prior to the onlay posterior cruroplasty is performed with interrupted sutures of 2-0 polyester into the right and left bundles of the right crus, ensuring that full-thickness bites are taken. The mesh onlay is an oval sheet (15 10 0.1 cm) of fenestrated PTFE (MycroMesh GoreTex, W. L. Gore and Associates). A radial slot with a 3-cm defect in the center (keyhole) is cut into the mesh. The mesh is pushed through a 10-mm trocar into the peritoneal cavity and placed around the gastroesophageal junction with the esophagus coming through the 3-cm defect and the radial slot oriented anteriorly (Fig. 2). The PTFE is stapled to the diaphragm with a straight hernia stapler (Ethicon); the two leaves of the keyhole are stapled to each other. Finally, a 3-cm-long 360 fundoplication is created loosely around the 50 Fr bougie with three sutures of 2-0 polyester, taking bites of stomach wall only (Fig. 3). The most cephalad stitch incorporates the mesh and the anterior arch of the right crus. This procedure was performed on three patients who all had severe heartburn, reflux, a partially intrathoracic stomach, and an esophageal hiatus 8 cm in diameter. No perioperative complications occurred. At 5, 16, and 18 months of follow-up the patients are doing well and either barium meal or endoscopy has confirmed a successful operation in each.

After several years of follow-up, the success of laparoscopic Nissen fundoplication in controlling the symptoms of gastroesophageal reflux (GER) appears to be 8090% [5, 6, 7, 10]. A possible mechanism of Nissen fundoplication failure is herniation of the stomach and wrap into the chest [11], which can produce recurrent reflux. We have noted failure of laparoscopic Nissen fundoplication in some patients who had a large defect (>8 cm) of the esophageal hiatus repaired with posterior cruroplasty. Subsequent barium meal in these patients confirmed a recurrent hiatal hernia with demonstrable reflux. In an attempt to prevent reherniation of the stomach into the chest, we reinforced the posterior cruroplasty with polytetrafluoroethylene (PTFE) onlay in the next patients with large hiatal hernia. Here we report laparoscopic Nissen fundoplication with mesh reinforcement of

Discussion The cause(s) of recurrent reflux after an antireflux procedure may be revealed by examining the reoperation for reflux. In one series of 87 patients undergoing operation for a failed antireflux procedure, 72% had recurrent hiatal hernia [11]. Despite the problem of recurrence, the technique of

Correspondence to: C. T. Frantzides

770

Fig. 1. Intraoperative view of an enlarged esophageal hiatus which was repaired with posterior cruroplasty and PTFE onlay. (A) liver, (B) diaphragm, (C) stomach, (D) esophageal hiatus, (E) gastrohepatic omentum.

Fig. 2. Intraoperative view of the hiatus after posterior cruroplasty and PTFE onlay, but before the fundoplication: (A) inflatable balloon retractor, (B) PTFE stapled to diaphragm, (C) esophagus, (D) liver.

Fig. 3. Intraoperative view after completion of Nissen fundoplication: (A) liver, (B) inflatable balloon retractor, (C) PTFE stapled to diaphragm, (D) fundoplication.

771

hiatal herniorraphy has not changed appreciably since its inception. Cruroplasty has been done by suturing the crura with 0 or 2-0 nonabsorbable material anterior and/or posterior to esophagus [12]. Cushieri et al. [3] has reported eight patients with large hiatal hernia who were treated with laparoscopic posterior cruroplasty, with a good short-term result. Precedent for the use of prosthetic at the esophageal hiatus is limited; one example is the Angelchik and Cohen prosthesis [1], a C-shaped ring of silicone gel contained in a silicone elastomer shell which is placed around the gastroesophageal junction. The enthusiasm for this device was ended by reports of complications of prosthesis migration, lumenal penetration, and dysphagia [2]. The indication for the purpose of PTFE placement in the present study are different from those for placement of an Angelchik prosthesis, so comparison of these two techniques is difficult. Migration of the PTFE is unlikely since it is stapled to the diaphragm. We do not as yet have long-term follow-up. Edelman [4] has reported a series of 5 patients with paraesophageal hernia who were treated with laparoscopic hiatal herniorrhaphy with polypropylene mesh, gastropexy, and gastrostomy. Pitcher et al. [9] has reported a series of 12 patients with paraesophageal hernia who underwent laparoscopic repair; two of these required polypropylene mesh to close a large hiatus. Since polypropylene mesh has been associated with enterocutaneous fistula secondary to the polypropylene eroding into the bowel lumen [8], we employed PTFE for the onlay because there have been no reports of bowel erosion secondary to PTFE. Determination of the efficacy and safety of PTFE reinforcement of posterior cruroplasty for the large hiatal defect will require 5075 patients observed over 1015 years. The short-term result in our first three patients has been satisfactory. We do not, however, recommend routine use of PTFE in hiatal herniorrhaphy. At the present we are placing

PTFE in patients with a large hernia sac and whose hiatal diameter is 8 cm or greater. The diameter indication was arrived at empirically; we do not have data specifying the hiatal diameter at which the risk for cruroplasty disruption is increased. We feel that the PTFE onlay provides a buttress where tissue is healing under tension and is subjected to stress from coughing, straining, retching, or obesity. We propose the use of PTFE when confronted with a large defect of the hiatus. References
1. Angelchik JP, Cohen R (1979) A new surgical procedure for the treatment of gastroesophageal reflux and hiatal hernia. Surg Gynecol Obstet 148: 246248 2. Crookes PF, DeMeester TR (1994) The Angelchik prosthesis: what have we learned in fifteen years? Ann Thorac Surg 57: 13851386 3. Cuschieri A, Shimi S, Nathanson LK (1992) Laparoscopic reduction, crural repair, and fundoplication of large hiatal hernia. Am J Surg 163: 425430 4. Edelman DS (1995) Laparoscopic paraesophageal hernia repair with mesh. Surg Laparosc Endosc 5: 3237 5. Frantzides CT, Carlson MA (1995) Laparoscopic versus conventional fundoplication. J Laparoendosc Surg 5: 137143 6. Hinder RA, Filipi CJ, Wetscher G, Neary P, DeMeester TR, Perdikis G (1994) Laparoscopic Nissen fundoplication is an effective treatment for gastroesophageal reflux disease. Ann Surg 220: 472483 7. Jamieson GG, Watson DI, Britten-Jones R, Mitchell PC, Anvari M (1994) Laparoscopic Nissen fundoplication. Ann Surg 220: 137145 8. Kaufman Z, Engelberg M, Zager M (1981) Fecal fistula: a late complication of Marlex mesh repair. Dis Colon Rectum 24: 543544 9. Pitcher DE, Curet MJ, Martin DT, Vogt DM, Mason J, Zucker KA (1995) Successful laparoscopic repair of paraesophageal hernia. Arch Surg 130: 590596 10. Rattner DW, Brooks DC (1995) Patient satisfaction following laparoscopic and open antireflux surgery. Arch Surg 130: 289294 11. Stirling MC, Orringer MB (1986) Surgical treatment after the failed antireflux operation. J Thorac Cardiovasc Surg 92: 667672 12. Wichterman K, Geha AS, Cahow CE, Baue AE (1979) Giant paraesophageal hiatus hernia with intrathoracic stomach and colon: the case for early repair. Surgery 86: 497506

Surg Endosc (1997) 11: 762765

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic colposuspension
Is it cost-effective?
K. Loveridge, A. Malouf, C. Kennedy, A. Edgington, A. Lam
Strathfield Private Hospital, University of Sydney, 3 Everton Road, Strathfield NSW 2135, Australia Received: 19 August 1996/Accepted: 20 December 1996

Abstract Background: The laparoscopic approach must be shown to be cost-effective as well as safe and technically effective before being widely adopted. A review of 54 consecutive patients who underwent open and laparoscopic colposuspension is presented and a cost-analysis is performed comparing the two approaches. Methods: This study was a retrospective controlled review of patient records and accounts of in-hospital costs incurred at a private hospital. Results: Theater costs were significantly greater in the laparoscopic group but this was balanced by a shorter length of stay and subsequent reduced accommodation cost. There was no difference in the overall in-hospital costs between the two groups. Conclusion: The laparoscopic surgical approach is safe and effective and by no means more expensive than the open approach. In the future, the laparoscopic approach can only become more cost efficient; techniques will improve and there will be earlier returns to work and, subsequently, greater productivity. Key words: Laparoscopic Colposuspension Cost

advantages of the laparoscopic approach: easier access, improved magnified view, minimal intraoperative blood loss, shorter length of stay, and earlier return to normal lifestyle [27]. Other reports in the literature have also proven laparoscopic colposuspension to have similar advantages [2, 31, 33, 40]. Aims While the laparoscopic technique may provide a significant improvement in patient management it must also be shown to be cost-effective before being widely adopted. The aim of this paper is to compare the in-hospital cost effectiveness between the laparoscopic and open Burch colposuspension. Methods
This is a retrospective consecutive case-control study of all patients who underwent colposuspension at one private hospital between March 1992 and March 1995. Patients were divided into two groups. In the study group were all those patients who had a laparoscopic colposuspension and in the control group were all patients who had an open procedure. Patients who had a major procedurenamely, a total abdominal hysterectomy performed concomitantly with the colposuspension were excluded. One surgeon performed the laparoscopic procedures but seven different surgeons were responsible for the open cases. Data was collected from hospital records. The two groups were matched for age, comorbidity, weight, previous abdominal surgery, previous bladder neck surgery, concomitant minor surgery, and the incidence of complications. Theater time, length of stay, and all resource usage were recorded. Resource usage included all investigations performed, medications and fluids given, instruments utilized, operations performed, and specialist services consulted. In-hospital costs were categorized into five areas: investigatory, medicinary, accommodation, theater, and consultant fees. Investigatory costs included pathology and radiology services and cross-consultation fees. Medications included analgesics, antibiotics, and miscellaneous drugs such as antiemetics, anticoagulants, and ted stockings. Regular medications were not included. These two areas were allocated standardized 1995 prices per unit used. Accommodation costs were based upon length of stay and the standard cost of a private room in 1995. Theater costs involved a fee for disposable

Urinary incontinence is a common problem affecting 812% of all women. Of these, 5% will have daily incontinence [1, 22, 48]. Many different operations have been performed to treat urinary incontinence [3, 24, 25, 35, 38, 42, 45]. The suprapubic approach first described by Burch has been regarded as the operation of choice due to its long-term success rates [8, 9, 11, 12, 46]. Since 1992 this particular operation has been routinely performed laparoscopically by one of the surgeons at Strathfield Private Hospital. In an earlier paper this surgeon reported upon the

Correspondence to: K. Loveridge

763 Table 1. Patient demographics Laparoscopic Age (years) Weight (kg) Comorbidity ASA 12 (no. of patients) 57 14 70 16 26 (100%) Open 48 8 64 26 23 (100 ) p 0.007 0.318

items identified on patient accounts and a standardized fee allocated from private health funds for an open or laparoscopic procedure. Consultant fees for both the anesthetist and surgeon were calculated according to the 1995 Medicare rates. Costs that were not taken into consideration were overheads or indirect costs which were incurred in support of clinical services. This included building and equipment depreciation, administration, labor and wages (hence, theater and nursing time requirements were not accountable), meals, laundry/cleaning, and energy costs. Medications excluded were anesthetic gases and drugs, regular routine medications, bandages/dressings, syringes, giving sets, needles, etc. Theater costs, including operation fees and costs of disposable instruments directly incurred as a result of other minor procedures performed at the time of surgery, were not included. The two groups were matched statistically using chi-squared tests for age, weight, comorbidity, and number of patients who had previous abdominopelvic surgery or who had other minor procedures at the time of surgery. The cost analysis was performed using Student t-tests.

Fig. 1. Previous surgery.

Results There were a total of 54 patients, of whom five patients were excluded as they had hysterectomies at the time of surgery. Of the remaining 49 patients, 26 underwent laparoscopic and 23 open operations. Patient demographics and previous surgery performed are outlined in Table 1. The laparoscopic group was significantly older (p 0.007); however, as this difference occurred within the sixth decade and all patients were women, the clinical significance was thought to be negligible. Sixteen of the laparoscopic and 18 of the open patients had previous abdominopelvic surgery ( p 0.06). The patients that had open colposuspension tended to have a history of a greater number operations per person (Fig. 1). Figure 2 records the type of surgery previously performed. The number of patients having previous bladder neck surgery between the two groups was comparable. As described in Table 2, the laparoscopic group contained a greater number of patients who had other minor procedures which were performed at the time of colposuspension ( p 0.02). There was no significant difference in complication rate as reported in Table 3 ( p 0.214). In the laparoscopic group the length of stay was significantly shorter with a mean of 3.7 2.3 days compared with 6.3 3.0 days ( p 0.001). Table 4 outlines the cost analysis results. In-hospital costs were comparable in terms of medications, investigations, and consultant fees. Theater costs, however, were greater in the laparoscopic group ( p <0.005) but were counterbalanced by the reduced accommodation costs, which were less due to the significantly shorter length of stay in the laparoscopic group. There was no significant difference in the overall cost ( p 0.53). Discussion Laparoscopic colposuspension has been already proven to reduce length of stay and postoperative pain requirements,

Fig. 2. Nature of previous surgery.

Table 2. Concomitant procedures Laparoscopic (n 26) Suprapubic catheter Cytoscopy Endometrial ablation Posterior/anterior vaginal repair Salpingoophorectomy Diathermy tubal ligation Drainage of cysts Dilatation & curretage Removal or insertion of IUD; scar revision No other procedure 23 3 1 6 2 1 3 2 1 Open (n 23) 12 6 1

enable early return to work, and achieve equally good shortterm results. In addition, this approach has shown no rise in the complication rate [2, 27, 28, 30, 31, 33, 40]. Before embracing this procedure, however, it is important to assess the cost implications. To date no cost-utility study has been performed with respect to laparoscopic colposuspension. Laparoscopic cholecystectomy has been assessed clinically and there are now many studies analyzing the cost benefit of the procedure, particularly in relation to different fractions within the health system [5, 18, 19, 39, 43, 47]. Likewise, there are a few papers that analyze the costs involved in laparoscopic appendectomy, fundoplication, and herniorrhaphy [4, 10, 15, 29]. Despite this, there is still a paucity of papers which analyze costs in detail, so it is difficult to draw comparisons between studies. Most studies report a significantly lower cost associated with laparoscop-

764 Table 3. Complications Type UTI Urinary retention Home with SPC Cannula site infect Detrusor instability Wound infection PUO Paralytic ileus Bladder perforation Conversion to open 2nd to bleeding Hypoglycemia Antibiotic reaction Total Laparoscopic 4 1 1 1 1 1 1 1 1 11 Open 1 4 3

N/A 1 10

Table 4. Cost analysis per patient Mean difference Accommodation Theater Disposables Operation Medications Investigations Consultations Total $1,084.00 $280.77 $347.90 ($2.00) ($34.00) $246.00 $246.00 p 0.001 <0.005 0.001 0.920 0.365 0.533

ic surgery, especially cholecystectomy, when compared with the open technique regardless of their methodology [57, 1719, 33, 39, 43]. A few studies, however, report no such significant difference, and some even report an increase in laparoscopic costs [23, 47]. These differences in reporting may be attributable to methodology. Some papers fail to define how costs were derived [47]. Most papers rely on hospital charges because of the accessibility of details without taking into account all other indirect costs or overheads. In addition, certain components, in particular overheads or indirect costs for any one operation, differ between hospitals depending on a multitude of given variables [20, 32]. Ideally, all costs need to be considered in any accurate cost analysis, but to date there is no widely accepted formula for estimating these costs. Despite this, the experience gained and trends noticed from laparoscopic cholecystectomy can be extrapolated and compared in principle to other forms of laparoscopic surgery. In our study as with others, one of the major factors which determined the overall cost of the procedure was length of stay. The cost of new technology in our study as in others balanced any advantage gained by a shorter length of stay [16, 21, 23, 26, 41, 47]. However, in the future, with further improvement in length of stay, economic advantages can be anticipated with the laparoscopic approach to colposuspension. Previous research has shown that over the last decade, length of stay has not changed dramatically in those patients who had an open colposuspension [40]. This suggests that the managed care of patients having an open procedure has already been optimized and is unlikely to significantly change in the future. In contrast, as more expertise is gained in the laparoscopic approach, length of stay

should decrease. This has been the experience with laparoscopic cholecystectomy in this country as hospital bed days were reduced by 1425% within a 2-year period of its introduction [34]. As with laparoscopic cholecystectomy, day-stay surgery may still become a possibility. Cholecystectomy rates have increased since the inception of the laparoscopic approach, and as Fletcher suggests, this rise has negated any economic benefit gained by a shorter length of stay for health funders [13]. In the public sector, a shorter length of stay can produce major savings provided that more aspects of treatment are transferred to the outpatient setting and beds closed [14]. In the private sector this may also be achieved provided bed occupancy rates are maintained. One might anticipate a similar rise in the rate of laparoscopic colposuspension as a less-invasive procedure with fewer complications offers an opportunity to patients who might not otherwise have been considered suitable for surgery. The other major area of in-hospital costs in our study and in others was theater-related expenses, which included operation costs and disposable instruments [16]. As mentioned previously, these costs counterbalanced any advantage gained by a shorter length of stay in our study. Fortunately, this is another area in which there should be improvement with time and where costs can be consciously minimized. The use of nondisposable instruments is a necessity in order to achieve this. Although theater costs at our hospital were not based upon time spent in theater, reduction in theater time secondary to experience will also improve overheads associated with theater costs. The difference in costs between the two approaches is therefore only likely to increase, showing a significant cost benefit associated with the laparoscopic approach. Most studies agree that the major cost differences occur after discharge. In the case of laparoscopic cholecystectomy, as with colposuspension, most patients are within the working age group, and those that have a laparoscopic procedure return to work or resume normal activity on average 46 weeks earlier than those undergoing an open procedure. This increases national productivity and decreases costs incurred by employers and insurers dramatically [18, 36, 37, 44, 49, 50]. Our study did not consider this; hence, when costs incurred secondary to a delay in return to work are also considered, the benefits of the laparoscopic approach should become even more apparent. Patient satisfaction in terms of decreased waiting lists, shorter length of stay, and earlier return to normal activity are also factors which are invaluable advantages. A prospective randomized study detailing each item/ resource used and overhead costs involved would be the best means of accounting for all costs accurately. But we believe this study to be adequate in terms of identifying major differences in in-hospital costs between procedures and proving the null hypothesis. Laparoscopic colposuspension is certainly not significantly more costly than an open procedure. Other studies have proven laparoscopic surgery to be a safe and effective mode for the management of urinary stress incontinence. This study now proves that it is also a feasible option.
Acknowledgment. Surgeons: Boulas J, MBBS FRACS; Lalak A, MBBS

765 FRCS(Ed.) FRACS; Maher P, MBBS FRCS(Eng.) FRACS; OToole V, MBBS FRACOG FRACDG; Sheldon J, MBBS FRACOG; Wong J, BSc MBBS FRACS. Statistician: Hurst T. Research Assistant: Gane M, RN. 25. Kelly HA (1913) Incontinence of urine in women. Urol Cutan Rev 17: 291 26. Kurzawinski T, Hayter B, Tate J et al (1992) The cost implications of open laparoscopic versus open cholecystectomy. Gut (Suppl) 33: S64 27. Lam AM, Jenkins GJ, Hyslop RS (1995) Laparoscopic Burch colposuspension for stress incontinence: preliminary results. MJA 162: 18 21 28. Langebrekke A, Dahlstrom B, Eraker R, Urnes A (1995) The laparoscopic Burch procedure. A preliminary report. Am J Obstet Gynecol Scand 74: 153155 29. Laycock WS, Oddsdottir M, Franco A, Mansour K, Hunter J (1995) Laparoscopic/Nissen fundoplication is less expensive than open Belsey Mark IV. Surg Endosc 9: 426429 30. Liu CY (1993) Laparoscopic retropubic colposuspension (Burch procedure). A review of 58 cases. J Reprod Med 38(7): 526530 31. Liu CY, Peak W (1993) Laparoscopic retropubic colposuspension (Burch procedure). Gynacol Laparosc 1: 3134 32. Macario A, Vitez T, Dunn B, McDonald T (1995) Where are the costs in perioperative care? Analysis of hospital costs and charges for inpatient surgical care. Anaesthesiology 83: 6, 11381144 33. McDougall EM, Klutke CG, Clayman RV, Cornell T (1994) Comparative analysis of vaginal (Raz) and laparoscopic bladder neck suspension for type 1 or type 2 stress urinary incontinence (abstract 1085). J Urol 151: 499A 34. Marshall D, Hailey D, Hirsch N et al (1994) The introduction of laparoscopic cholecystectomy in Canada and Australia. Australian Institute of Health and Welfare, Canberra 35. Marshall VF, Marchetti AA, Krantz KE (1961) The correction of stress incontinence by simple vesicourethral suspension. Surg Gynaecol Obstet 88: 509518 36. Nathanson LK, Shimi S, Cuschieri A (1991) Laparoscopic cholecystectomy the Dundee technique. Br J Surg 78: 155159 37. Neugebauer E, Troidl H, Spangenberger W et al (1991) The Cholecystectomy Study Group. Conventional versus laparoscopic cholecystectomy and the randomised control trial. Br J Surg 78: 150154 38. Pereyra AJ (1959) A simplified surgical approach for the correction of stress incontinence in women. West J Surg Obstet Gynaecol 67: 223 226 39. Peters JH, Ellison EC, Innes JT, et al (1991) Safety and efficacy of laparoscopic cholecystectomy: a prospective analysis of 100 patients. J Laparoendosc Surg 1: 193196 40. Polascik TJ, Moore RG, Rosenber MT, Kavoussi LR (1995) Comparison of laparoscopic and open retropubic urethropexy for treatment of stress urinary incontinence. Urology 45(4): 647652 41. Prasad, Foley (1994) Br J Surg 81: 777 42. Richardson AC, Edmonds PB, William NL (1981) Treatment of stress urinary incontinence due to paravaginal fascia defect. Obstet Gynaecol 57: 357362 43. Schirmer B, Dix J (1991) Cost effectiveness of laparoscopic cholecystectomy. Gastroenterology 100 (5 part 2): A17 44. The Southern Surgeons Club (1991) A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 324: 10731078 45. Stamey TA (1973) Endoscopic suspension of the bladder vesical neck for urinary incontinence. Surg Gynecol Obstet 136: 547 46. Stanton SL, Cardozo LD (1979) A comparison of vaginal and suprapubic surgery in the correction of incontinence due to uretral sphincter incompetence. Br J Urol 51: 497499 47. Stoker ME, Vose J, OMara PO, Maini BS (1992) Laparoscopic cholecystectomy. A clinical and financial analysis of 280 operations. Arch of Surg 127: 589595 48. Thomas TM, Plymat KR, Blannin J, Meade TW (1980) Prevalence of urinary incontinence. Br Med J 281: 12431245 49. Vandenbergh HC, Wilson T, Adams SE, Inglis MJ (1995) Laparoscopic cholecystectomy its impact on national health economics. Med J Aust 162: 587590 50. Wenner J, Graffner H, Lindell G (1995) A financial analysis of laparoscopic and open cholecystectomy. Surg Endosc 9: 702705

References
1. Abrams P, Blaivas JG, Stanton SL, Anderson JT (1990) The standardisation of terminology of lower urinary tract function. Br J Obstet Gynaecol (Suppl) 97: 116 2. Albala DM, Schuessler WW, Vancaillie TG (1992) Laparoscopic bladder neck suspension. J Endourol 6: 137141 3. Aldridge AH (1942) Transplantation of fascia for relief of urinary stress incontinence. Am J Obstet Gynae 44: 398 4. Apelgren KN, Molnar RG, Kisala JM (1992) Is laparoscopic better than open appendectomy? Surg Endosc 6: 298301 5. Baird DR, Wilson JP, Mason EM et al (1992) An early review of 800 laparoscopic cholecystectomies at a university affiliated community teaching hospital. Am Surg 58: 206210 6. Barkun J, Caro J, Barkun A, Trindale E (1995) Cost-effectiveness of laparoscopic and mini-cholecystectomy in a prospective randomised trial. Surg Endosc 9 (11): 12211224 7. Bass E, Pitt H, Lillemoe K (1993) Cost-effectiveness of laparoscopic cholecystectomy versus open cholecystectomy. Am J Surg 165: 466 471 8. Bergnan A, Ballard CA, Konnings PP (1989) Comparison of three different procedures for genuine stress incontinence: prospective randomised study. Am J Obstet Gynecol 160: 11021106 9. Bhatia NM, Bergman A (1985) Modified Burch versus Pereyra retropubic urethropexy for stress incontinence. Obstet Gynaecol 255261 10. Brooks DC (1994) A prospective comparison of laparoscopic and tension-free open herniorrhaphy. Arch Surg 129 (4): 361366 11. Burch JC (1961) Urethrovaginal fixation to Coopers ligament for correction of stress incontinence, cystocele and prolapse. Am J Obstet Gynecol 81: 281290 12. Feyereisl J, Dreher E, Haenggi W, Zikmund J, Schneider H (1994) Long term results after Burch colposuspension. Am J Obstet Gynecol 171: 647652 13. Fletcher D (1995) Laparoscopic cholecystectomy. What national benefits have been achieved and at what cost? MJA 163: 535538 14. Fletcher D (1995) Laparoscopic cholecystectomy in Australia outcomes and costs. Surg Endosc 9 (11): 12301235 15. Fritts LL, Orlando R (1993) Laparoscopic appendectomya safety and cost analysis. Arch Surg 128: 521528 16. Fullarton GM, Darling K, McMillan R, Bell G (1994) Evaluation of the cost of laparoscopic and open cholecystectomy. Br J Surg 81: 124126 17. Gilchrist BF, Vlessis AA, Kay GA, Swartz K, Dennis D (1991) Open versus laparoscopic cholecystectomy: an initial analysis. J Laparoendosc Surg 1: 193196 18. Grace PA, Quereshi A, Coleman J, et al (1991) Reduced postoperative hospitalisation after laparoscopic cholecystectomy. Br J Surg 78: 160 162 19. Graves HA, Ballinger JF, Anderson WJ (1991) Appraisal of laparoscopic cholecystectomy. Ann Surg 213: 655664 20. Hardy K, Miller H, NcNeil J, Shulkes A (1994) Measurement of surgical costs: a clinical analysis. Aust N Z J Surg 64: 607611 21. Hirsch NA (1992) Laparoscopic cholecystectomy. Aust Clin Rev 12: 2327 22. Hoist K, Wilson DP (1988) The prevalence of female urinary incontinence and reasons for not seeking treatment. N Z Med J 101: 756 758 23. Jordan AM (1991) Hospital charges for laparoscopic and open cholecystectomy. JAMA 226: 3425 24. Karram MM, Bhatia NM (1989) Transvaginal needle bladder neck suspension procedures for urinary stress incontinence: a comprehensive review. Obstet Gynaecol 73: 906914

Surg Endosc (1997) 11: 745749

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic colectomy
G. A. Fielding, J. Lumley, L. Nathanson, P. Hewitt, M. Rhodes, R. Stitz
Royal Brisbane Hospital, Herston, Brisbane, Australia Received: 26 March 1996/Accepted: 15 October 1996

Abstract Background: Laparoscopic colectomy has developed with the explosion of technology that has followed laparoscopic cholecystectomy. Accumulation of skills in general laparoscopic surgery has made complex surgery, such as colectomy, feasible. Methods: Three hundred fifty-nine laparoscopic cases were prospectively studied. Data has been kept on benign and malignant cases, operative results, hospital stay, and morbidity. Special care has been taken to follow malignant cases, looking for recurrence of disease. Results: There were 359 cases (206 females, 153 male) average age 58.8 years (1894), and 149 patients had malignancy. All types of resections were performed, including 151 anterior resections, 66 right hemicolectomies (RHC), 36 total colectomies, and 22 rectopexies. Operating times fell with experiencethe last 20 cases of anterior resection took 150 min (110240) and of RHC took 130 min (65 210). Twenty-six (7%) cases were converted to open surgery. Hospital stays for anterior resection lasted 57 days (233); in the last 20 cases the average stay was 4 days. Morbidity included seven leaks (2.7%), four strictures (1.2%), 12 wound infections (3.3%), and nine ileus (2.5%). There were six deaths within 30 dayssepsis, myocardial infarction, aspiration pneumonia, and disseminated liver metastases. One hundred forty-nine cancer cases have had ten recurrences: one pelvic recurrence, six liver metastases, two para-aortic nodal, and one case of disseminated disease. Average time of recurrence was 33 months (1546 months). Conclusions: Laparoscopy in the hands of experienced laparoscopic surgeons is a safe, efficient procedure. All types of procedures are possible. Early results in 149 malignancies are encouraging and recurrence rates are low. Prospective studies, now that skills are developed to a level comparable to that of open surgery, are now being performed to further assess laparoscopys possible role in treating cancer. Key words: Laparoscopy Colectomy Laparoscopic colectomy surgery

Following the rapid acceptance of laparoscopic cholecystectomy to treat disease of the gallbladder and bile duct, surgeons with an interest in laparoscopic surgery turned their attention to numerous other procedures such as Nissen fundoplication, splenectomy, hernia repair, and colectomy. There was hope that the benefits experienced with cholecystectomy would be extended to these more complex procedures and allow more rapid return to health after major intra-abdominal surgery. This paper reports our experience with 359 laparoscopic-assisted colorectal cases between June 1991 and January 1996. We have used the technique for both benign and malignant conditions (Table 1). In this paper we have sought to address its application, and the effect on operating time, hospital stay, complications, and return to bowel function. We have analyzed results for benign and malignant conditions and attempted to compare laparoscopic-assisted colorectal surgery, in context, with open surgery. Materials and methods
Between June 1991 and January 1996 we performed 359 consecutive cases of laparoscopic-assisted colorectal procedures. We came to laparoscopic surgery with significant experience in laparoscopic cholecystectomy, having done well over 500 cholecystectomies before commencing colorectal surgery. We instituted a policy of assisting each other early on to rapidly gain exposure to as many cases as possible and to allow transmission of skills learned from cholecystectomy by some members of the unit to the colorectal specialists involved who had had no prior laparoscopic exposure. This allowed for rapid accumulation of skills by the colorectal specialists, and they were soon performing the laparoscopic colorectal procedures without assistance by other members of the unit. Data has been kept prospectively on all patients. The aim was to perform exactly the same operation laparoscopically as done at open surgery. This included exposure of the ureters and dissection of the pedicle well above the pelvic brim for anterior resection and exposure of the ureter, pancreas, and duodenum for right hemicolectomy. We initially chose benign diseases and, as experience grew, turned our attention to malignancy. Initial malignant cases were either palliative procedures or large adenomas that were not resectable at colonoscopy. Certainly, early on in our experience, no difficult cases were attempted, nor were curative resections for carcinoma. As our experience grew though, this changed, and we now perform resections with curative intent for carcinoma and difficult benign cases of Crohns disease and diverticulitis. Data on patient age, diagnosis, type of colorectal procedure, operative time, return to bowel function, discharge from hospital, complications, cancer staging, and mortality were kept. Follow-up has been maintained for

Correspondence to: G.A. Fielding, Level 2, Wesley Medical Centre, Auchenflower, Brisbane, Australia

746 Table 1. Laparoscopic colorectal surgeryindications for surgery Carcinoma Polyps Crohns disease Slow transit constipation Bypass Ulcerative colitis Prolapse Diverticular disease Diverticular fistula Diverticular perforation Volvulus 149 14 46 23 4 8 26 69 10 8 2 Table 3. Laparoscopic colorectal surgeryanterior resection (151) Carcinoma Benign tumour Diverticular disease Diverticular fistula Diverticular perforation Sigmoid volvulus Total 48 14 69 10 8 2 151

Table 4. Operating time (minutes) First 20 cases Table 2. Laparoscopic colorectal surgeryprocedures performed Procedures Stoma RHC Anterior resection APR Total colectomy Total and reservoir Sling - rectopexy Resect - rectopexy Hartmanns Reverse Hartmanns Bypass Ileal resection LHC Total 31 66 151 36 31 7 4 22 3 2 2 1 3 359 Malignant 23 39 48 35 1 146 RHC Anterior resection 180 240 Last 20 cases 130 150

all colorectal malignancies, looking for both local recurrence and death from cancer. With respect to colorectal malignancy we initiated a policy from the beginning of using a no touch technique with minimal handling of the tumor and bagging of all specimens prior to removal through the abdominal wall. In the very early part of our experience we converted the first four right hemicolectomies performed for tumor to check that we had done an adequate resection by our normal standards, and this was the case. Furthermore, the lymph node count in specimens was the same at laparoscopy and open surgery. We were then happy to continue with laparoscopic resection for possible and known cancers.

Results There were 359 patients, 206 females, 153 males, average age 58.8 years, range 1894 years. Procedures were as listed in Table 2. Thirty-six patients had abdominoperineal resection and this reflects the referral base to this unit for very low colorectal tumors (Table 2). Three patients had a Hartmanns procedure for perforated diverticular disease, two having had their Hartmanns reversed laparoscopically. Two patients have had a left hemicolectomy. The major group in the series was anterior resection (Table 3). Twenty-six patients were converted to an open procedure. These include ten with failure to progress, particularly early on in our experience; three where the anatomy was not identified due to severe adhesions or inflammatory disease; three where there was uncertainty about the stage of disease,

three for equipment failure; two due to hemorrhage; one suffering perforation; and four with second pathology requiring other treatment. Operating times improved with experience (Table 4). This improvement in operating time was reflected also in comparison of the first 100 and last 100 patients. In the first 100 there were ten conversions and 26 complications, compared to the last 100, where there were five conversions and 16 complications. Hospital stay shows the same improvement with experience (Table 5). Overall for anterior resection (151 cases), the median hospital stay is 5.7 days (233 days). Forty-four percent of patients were discharged within 5 days and 29% were discharged within 4 days. Hospital stay for total colectomy reflects the underlying total bowel transit problems these patients have, with an often-prolonged ileus in the small bowel postoperatively. Return to bowel function occurred at 2 days with a range of 2 days (15) for anterior resection, 2 days (14) for right hemicolectomy, 2.93 days (16) for total colectomy, and 1.89 days (13) for rectopexy. Morbidity is listed in Table 6. There were seven leaks (2.7% of anastomoses). Five were treated by stoma diversion and two by conservative management with drain and total parenteral nutrition. Four patients developed stricture requiring balloon dilatation. There were six deaths within 30 days (Table 7). One was after necrotizing fasciitis developed following a right hemicolectomy in a 75-year-old. The second was after a myocardial infarction following a right hemicolectomy in a 65year-old female. Two patients, 76 and 78 years old, one a rectal prolapse repair and one with dementia and incontinence, treated with a stoma, died of aspiration pneumonia. Two patients died within 30 days of their disseminated tumor, both after palliative stomas. There were no other delayed procedural deaths. Ten further patients have died of disseminated cancer in the intervening period and these will be discussed later. One hundred forty-nine patients were treated for malignancy (Table 8). (Staging of these tumors is listed in Tables 9 and 10). Table 6 lists tumor staging according to procedures performed.

747 Table 5. Hospital stay (days) First 20 cases RHC Anterior resection 5 (37) 6 (433) Last 20 cases 5 (311) 4 (318) Table 8. Laparoscopic colorectal surgerymalignancy Stoma RHC Anterior resection APR Bypass Hartmanns Total 23 39 48 35 2 1 149

Table 6. Laparoscopic colorectal surgerymorbidity Anastomotic stricture Anastomotic leak Ileus Wound infection Haemorrhage: wound/intra-abdominal Intra-abdominal sepsis Respiratory: atelectasis/pneumonia/PE Urinary: infection/retention Myocardial infarction Bowel obstruction Total 4 7 (1.9%) 9 (2.5%) 12 (3.3%) 8 2 8 9 3 5 67

Table 9. Laparoscopic colorectal surgerystages and malignancy Stage A B C D Anal Total Malignancy no. 30 40 32 35 10 149

Table 7. Laparoscopic colorectal surgeryperi-operative deaths (6) Necrotising fasciitis Myocardial infarct Aspiration pneumonia Aspiration pneumonia Disseminated tumour Disseminated tumour RHC RHC Prolapse Stoma End stoma Loop stoma Table 10. Laparoscopic colorectal surgerymalignancy A Stoma APR RHC Anterior resection 6 11 13 B 6 13 18 C 12 8 11 D 16 4 6 6 Anal 3 7

These patients are being followed carefully for recurrent cancer, particularly in view of concern about port site recurrence, wound recurrence, and pelvic recurrence (Table 11). To date no patients with stage A (30) have had recurrence. Four patients with stage B have had tumor recurrence in each case, liver metastases. No stage B tumor has had pelvic wound or port site recurrence. Four patients with stage C disease (32) have had further cancer. One was a pelvic recurrence in a 65-year-old man; one was a paraaortic nodal recurrence and two were liver metastases. One stage C patient developed a port site recurrence after a right hemicolectomy. This recurrence was part of more general para-aortic and liver recurrence. One stage D patient (35) has developed a port site recurrence as part of disseminated intra-abdominal malignancy developing from a carcinoma of the rectum already presenting with liver and peritoneal metastases at the time of presentation. Recurrences have occurred, ranging from 15 to 46 months with a median of 33 months after the initial procedure. Two of the ten patients have died of their recurrent disease, both with liver metastases, including the patient who presented with the port site recurrence concurrent with liver and peritoneal metastases. No patients with stage A or B disease have died from their malignancy to date. Discussion Following a cautious start using often rudimentary instruments, there has been an explosion of technology that has

Table 11. Laparoscopic colorectal surgerymalignancy (149) and recurrence (10) Pelvic Liver Nodal Ports 1 5 2 2 (1 D, 1 with liver)

a Recurrence at 1546 months, average 33 months; two have died of recurrence (liver)

allowed more complex laparoscopic procedures to be carried out. One of these is colorectal surgical resection. The biggest hurdle facing those commencing laparoscopic colorectal surgery is the acquisition of general laparoscopic skills that will allow them to deal with this very complex procedure. It is essential to be able to retract, dissect, control bleeding, suture if necessary, and manipulate multiple instruments in a small field of view. Our unit commenced laparoscopic colorectal surgery early but our enthusiasm was based on an already-wide exposure to laparoscopic skills. We believe this has been an essential part of the development of our program of laparoscopic colorectal resection. The approach has been to extend our exposure to all benign colorectal disease and to the majority of malignancy. Hesitation has been only at the prospect of transverse colon carcinomas. There is difficulty in mobilizing omentum and in minimizing risk of entering into disease covered by omentum. There is risk also in low tumors, where stapling

748

of the distal rectum without undue manipulation of the tumour is very difficult with existing instruments [4, 6]. We believe that careful selection, particularly early on, and maintaining a careful policy and good surgical technique allows good results with minimal complications. We have performed a full range of colorectal procedures with laparoscopic assistance. Of particular interest has been the ability to perform complex procedures for diverticular disease. Our policy has been one of aggression after a second presentation with acute diverticulitis due to the overwhelming evidence of further complications and mortality from repeat attacks of acute diverticulitis, particularly if associated with peritonitis or stricture [3, 4]. For diverticular disease, our resection goes into the upper third of the rectum to ensure removal of all sigmoid diverticular disease. Fistulas can be tracked and resected from the bladder, and the bladder defect can be repaired with suturing and covered with omentum. Acute diverticulitis can also be dealt with by doing a standard Hartmanns resection with a limited resection of the diseased portion of bowel and bringing out an end stoma. At the time of Hartmanns hookup a wider resection of involved diverticular disease to the upper rectum is performed, as would be done at standard operation for diverticular disease. There is nothing new or radical in this approach to colorectal surgery. All that is different is the delivery system using laparoscopic techniques rather than open surgery. As others have noticed, as skills are accumulated operating time falls significantly, as is reflected in the operating times for the last 20 cases each of right hemicolectomy and anterior resection that we have studied [4, 9, 12]. An anterior resection performed in under 3 h is on equal terms with any open series in the literature. This has been seen by Wishner et al. [12]. They show a significant decrease in mean operative time from 250 to 156 min over the first 50 cases and have leveled out at a median of 140 min. This is also similar to the results of Tucker et al., who reported 114 cases with a median operating time of 172 min [8]. The learning curve is a difficult phase. We believe that the diminished operating time is just a reflection of accumulation of skills and increased confidence with exposure to more cases. Our conversion rate of 26 cases (7.3%) is low, and again we believe this reflects our general laparoscopic experience at the time of commencing colorectal surgery. It is certainly consistent with reports in numerous publications [4, 8, 9, 12]. The most contentious issue in laparoscopic colorectal surgery is whether laparoscopic-assisted surgery is acceptable in malignant cases [1, 5, 6, 11]. Franklin et al. addressed this with a nonrandomized prospective study of laparoscopic vs open colon resection with a 3 1/2-year followup of 194 patients from several centers [1]. They found that laparoscopic surgery allowed a similar resection with an equal number of mesenteric nodes, the same margin with short follow-up, comparable survival and disease-free interval with a mean follow-up of 22 months. Seven percent had recurrent disease in that time after open surgery, 8% after laparoscopic surgery. When looking at stage-based comparisons there is no difference between open and laparoscopic surgery. This

group felt that laparoscopic colorectal surgery was a valid alternative to open surgery without claiming that it was better. Another randomized study was performed by Lacey et al. in three large departments of surgery in Spain, and they again showed no difference between lymph nodes removed, pathological staging, and margins [10]. They had fewer complications in their laparoscopic patients than in open in a study of 52 patients. Tait et al. in another prospective randomized trial in small numbers with 11 laparoscopic and 14 open cases found no difference in the histological/pathological specimen [2]. The findings in our study reflect these three randomized trials, and we are confident that we are performing the same procedure laparoscopically as we do at open surgery. This is reflected in our recurrence rates, both local recurrence and late disease. We have had no deaths from cancer in Dukes A and B and only one pelvic recurrence in 48 anterior resections for carcinoma. Wexner has raised the concern of port site recurrence [7]. Our own incidence has been very low. Both were in late-stage disease, one of whom had peritoneal and liver metastases at the time of the original operation. We believe that by not manipulating the tumor, bagging the specimen, and protecting of all wounds before extraction, this complication can be minimized. Concerns still exist as to the possibility of spreading shed cells through the pneumoperitoneum due to the increased pressure. This has not been reflected in our own experience of 148 malignancies treated laparoscopically. Having accumulated laparoscopic skills to a level that we feel is comparable with our open skills, we feel it is now the time to commence a randomized trial with malignant cases and have done so as part of a national trial in Australia. At present we will not perform a curative resection for a carcinoma of the transverse colon due to the difficulty in being certain we are not breaching the tumor when clearing omentum. This is also the case for low rectal tumors, where it is still, with existing equipment, impossible to get below very low tumors. However, in this situation laparoscopy can be used to mobilize the splenic flexure and the left colon, divide the vessels, and dissect down to the tumor to allow the final resection and anastomosis to be performed through a Pfannenstiel incision, thus reducing morbidity. These patients are covered with a covering ileostomy after resection. Laparoscopic-assisted colorectal surgery is a valid alternative to open surgery. We believe it has benefit over open surgery for benign disease and palliative malignancy. Our earlier results for the management of carcinoma are encouraging and will be further evaluated using a randomized prospective trial as mentioned above. Surgeons contemplating undertaking this surgery should be competent in laparoscopic techniques. It is of enormous benefit to be assisted by a skilled laparoscopic surgeon and to work as a team. Surgeons should choose easy cases such as resection of large polyps in the right colon and high anterior resection for diverticular disease, early on, before progressing to more complex cases. We feel that, with the exception of ultralow anterior resection for carcinoma, most cases can be done laparoscopically. Further instrument technology will make this surgery even more straightforward.

749

References
1. Franklin ME, Rosenthal D, Norem RF (1995) Prospective evaluation of laparoscopic colon resection versus open colon resection for adenocarcinoma. Surg Endosc 9: 811816 2. Lacy AM, Garcia-Valdecasas JC, Pique JM, Delgado S, Campo E, Bordas JM, Taura P, Grande L, Fuster J, Pacheco JL, Visa J (1995) Short-term outcome analysis of a randomized study comparing laparoscopic vs open colectomy for colon cancer. Surg Endosc 9: 1101 1105 3. Liberman MA, Phillips EH, Carroll BJ, Fallas M, Rosenthal R (1996) Laparoscopic colectomy vs traditional colectomy for diverticulitis. Surg Endosc 10: 1518 4. Lumley J, Fielding G, Rhodes M, Nathanson L, Siu S, Stitz R (1996) Laparoscopic assisted colorectal surgery. Lessons learned from 240 consecutive patients. Dis Colon Rectum 5. Monson JFT, Hill ADK, Darze A (1995) Laparoscopic colonic surgery. Br J Surg 82: 150157

6. Stitz R, Lumley J, (1995) Laparoscopic resection for colorectal cancerthe Australian perspective. Semin Laparosc Surg 2: 235241 7. Tate JJT, Kwok S, Dawson JW, Lau Wy, Li AKC (1993) Prospective comparison of laparoscopic and conventional anterior resection. Br J Surg 80: 13961398 8. Tucker JG, Ambroze WL, Orangio GR, Duncan TD, Mason EM, Lucas GW (1995) Laparoscopically assisted bowel surgery. Surg Endosc 9: 297300 9. Vernava AM, Liebscher G, Longo WE (1995) Laparoscopic restoration of intestinal continuity after Hartmann procedure. Surg Laparosc Endosc 5(2): 129132 10. Wexner SD, Cohen SM (1995) Port site metastases after laparoscopic colorectal surgery for cure of malignancy. Br J Surg 82: 295298 11. Williams N, Kapila L (1993) Laparoscopic surgery for colorectal cancer. Br J Surg 80: 12291230 12. Wishner JD, Baker JW, Hoffman GC, Hubbard GW, Gould RJ, Wohlgemuth SD, Ruffin WK, Mellick CF (1994) Laparoscopic-assisted colectomythe learning curve. Surg Endosc 9: 11791183

EndoScope: world literature reviews

Surg Endosc (1997) 11: 778781

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Original articles from a wide range of international surgical journals are selected by our editors and presented here as a structured summary and critical review. EndoScope serves as a quick and comprehensive survey of the expansive endoscopic literature from all the corners of the globe.

Section Editor: J. M. Sackier

Gastroesophageal reflux disease The impact of laparoscopic cholecystectomy on the management and outcome of biliary tract disease in North Carolina: a statewide, population-based, time-series analysis
Rutledge R, Fakhry SM, Baker CC, Meyer AA J Am Coll Surg (1996) 183(1): 3145 Patients with the admitting diagnosis of biliary tract disease from 1988 through 1993 were evaluated from the statewide hospital discharge database of all 157 hospitals in North Carolina. Laparoscopic cholecystectomy has become the major procedure, accounting for over 7,800 cases in 1993. This figure has not been associated with an increase in the overall rate of cholecystectomy; in fact, it declined after 1990. Unfortunately, bile duct repairs rose from 13 to a peak of 36 by 1992, showing a significant correlation between the rate of laparoscopic cholecystectomy and the rate of bile duct repairs (p 0.0001). Even after stratification by age and type of gallbladder disease, hospital charges and the length of stay were significantly less for the laparoscopic procedure. Finally, the authors report of initial reluctance, then gradual acceptance, by the older and non-boardcertified surgeons in adopting laparoscopic cholecystectomy. In terms of insurance coverage, health maintenance organizations (HMOs) and newer-type managed-care plans had the highest rates of laparoscopic cholecystectomy in the state. On the other hand, Medicare, Blue Cross, and Blue Shield had some of the lowest rates. Length of stay is shorter for laparoscopy than for open cholecystectomy (mean 1.7 days vs 4.3, respectively, for elective admissions), and even after stratification by age and severity of illness, patients who had laparoscopic cholecystectomy had lower total hospital charges. Interestingly enough, when individual components of the hospital bill were analyzed, the charges for laparoscopy remained lower than the open method in all categories except for the cost of supplies. Even in this category, the difference was only a matter of $169 in 1993. From the data presented in this paper, the HMOs tendency to back laparoscopy appears to be a prudent one: Laparoscopic cholecystectomy in North Carolina is associated with the shortest hospital stay and service charge without increased utilization. On the other hand, lower cost was obtained at the expense of increased bile duct repair rate, which needs to be factored in for the long-

Kahrilas PJ JAMA (1996) 276(12): 983988 A literature review examining the management of gastroesophageal reflux disease (GERD) in adults is reported, placing particular emphasis on controlled therapeutic trials. Articles were selected by Medline search from 1991 to 1996 or by cross-citation from other review papers. To avoid repeating the commendable paper, only several features will be discussed. Mild GERD can be safely treated with medication and lifestyle modifications. The treatment of severe GERD and its complications is not so well defined. The timing of evaluation for GERD and its complications, such as esophagitis, esophageal stricture, and Barretts metaplasia, varies widely but some guidelines can be followed. Endoscopy should be performed when there is dysphagia, GI bleeding, chronic symptoms lasting more than 5 years, or refractory symptoms despite therapy. After establishing the diagnosis of esophagitis, long-term use of H2 blockers for mild disease and proton-pump inhibitors for more severe disease should be instituted. For Barretts metaplasia, close surveillance with endoscopy is required for its dysplastic potential since neither medical nor surgical therapies reduce the risk of carcinoma. Reflux-induced asthma and otolaryngological complications such as subglottic stenosis, laryngitis, pharyngitis, or cancer can occur without evidence of esophagitis. Interestingly, asthma resolves particularly well after antireflux surgery whereas most otolaryngological manifestations can be controlled with medications. There are not data available with which to confidently assess the superiority of laparoscopic antireflux surgery over medical therapy, and hence, by default, the author recommends medical treatment. Certainly in capable hands, as the author points out, laparoscopic antireflux operations have shown excellent results. There are no data comparing laparoscopic repair with maintenance therapy using protonpump inhibitors since both methods are relatively new. Long-term outcome of medical therapy is still unknown. Reports of possible risk of carcinoid with omeprazole use and atrophic gastiritis, with its implication in gastric carcinoma, need further evaluation.

779

term follow-up cost analysis. Hopefully, as the learningcurve phase of this procedure is passed, the bile duct injury rate will return to the prelaparoscopic age.

safe and effective for benign adrenal tumors under 6 cm in size.

Laparoscopic adrenalectomy compared to open adrenalectomy for benign adrenal neoplasms

Brunt LM, Dohery GM, Norton JA, Soper NJ, Quasebarth MA, Moley JF J Am Coll Surg (1996) 183(1): 110 A retrospective review from 1988 to 1995 comparing laparoscopic adrenalectomy with the open approach is examined to emphasize the advantages of the minimally invasive technique. A total of 66 consecutive patients with benign adrenal neoplasms are divided into three groups based on the operative approach: group I (n 25), anterior abdominal approach; group II (n 17), retroperitoneal approach; and group III (n 24), laparoscopic flank approach. Among the study groups presented, there are no significant differences noted in terms of age, gender, American Society of Anesthesiologists class, tumor location, and size. Mean operative time was longer in duration (p < 0.001) for the laparoscopic resection when compared to either open approach in unilateral adrenalectomy, whereas this statistical significance is lost when bilateral adrenalectomy operative times are compared. Significant advantages for the laparoscopic resection are as follows: less operative blood loss (p < 0.001) and lower incidence of perioperative blood transfusion; reduced perioperative parenteral pain medication requirements (p < 0.001); more rapid resumption of regular diet (p < 0.01); and shorter postoperative length of stay (p < 0.01) without significantly higher total hospital charges. The average recovery time after laparoscopy is 10.6 4.9 days and the mean return to work time is 16.0 6.1 days. Comparable data in the open groups are not available. Thus, the paper advocates laparoscopic adrenalectomy to be the preferred approach for small, benign adrenal neoplasms. The mean operative time for unilateral adrenalectomy is 142 38 min in group I, 136 34 min in group II, and 183 35 min in group III. The vast majority of open adrenalectomies are performed by highly experienced endocrine surgeons, while the laparoscopic procedures are performed by three different attending surgeons and sometimes by chief residents. Hence, the longer operating time in group III reflects the learning curve required and will undoubtedly diminish with experience. Another advantage touched upon by the authors may occur in cases of Cushings syndrome, where extensive and prolonged steroid replacement therapy is required. The minimal incisional wound in laparoscopy may minimize wound-healing complications in these patients. The results of this retrospective evaluation between open and laparoscopic adrenalectomies favor the less invasive procedure as the preferred method of resection. Of the 24 laparoscopic adrenalectomies performed in St. Louis, the authors have shown the minimally invasive procedure to be

Pulmonary function after laparoscopic cholecystectomy in the elderly

Milheiro A, Castro Sousa F, Oliveira L, Joao Matos M Br J Surg (1996) 83: 10591061 Pulmonary function tests as reflected by pre- and postoperative FVC and FEV1 values are evaluated for 52 patients older than 69 years of age. A total of 390 patients undergoing laparoscopic cholecystectomy over a 2-year period are examined, out of which 52 (13%) are identified as being 70 or older. At the time of operation, the majority were found to have gallstone(s), 23% had acute cholecystitis, and 13% had acute pancreatitis of biliary origin. Pulmonary function tests were assessed prospectively in 20 consecutive patients before operation, 24 h postop, and on the 7th day after surgery. There was one death (2%) and 14% morbidity; conversion to laparotomy was 15%. Of the 338 patients younger than age 70, pulmonary function tests were performed in 30 consecutive patients not significantly different in perioperative conditions in order to compare to the older group. Preoperative values of FVC and FEV1 were significantly lower in the elderly. The values of FVC, FEV1, and forced expiratory flow at 50%, 24 h after surgery, were less depressed in the elderly. They also recovered more quickly than the younger group at 7 days postop. The results show that patients 70 years or older will actually tolerate laparoscopic cholecystectomy better than the younger group, as judged by comparing pulmonary function values. Although the numbers studied are small, this paper has pointed out an interesting trend that is commonly observed clinically in laparoscopynamely, that the elderly appear to tolerate laparoscopy better than the younger age group despite having lower preoperative pulmonary reserve. When feasible, laparoscopic cholecystectomy should be the treatment of choice for gallbladder disease especially in the elderly.

Wound metastases following laparoscopic and open surgery for abdominal cancer in a rat model
Mathew G, Watson DI, Rofe AM, Baigrie CF, Jamieson, GG Br J Surg (1996) 83: 10871090 A rat model employing implanted tumors is used to assess the risk of spread to abdominal wounds in an attempt to study the effect of tumor implantation in laparoscopy. Mammary adenocarcinoma cell suspensions (2 108 200 l) are injected into the left flank of 42 rats where tumor usually grows in size to about 2025 mm after 7 days. Three

780

control groups of six rats each are categorized as follows: (1) no surgery, (2) blunt manipulation of tumor laparoscopically, and (3) blunt manipulation of the tumor at laparotomy; 24 additional rats underwent surgical laceration of the tumor capsule at either laparoscopy (n 12) or laparotomy (n 12). One week later, the rats are terminated and examined histologically for metastasis. Primary tumor size is greater with operation than without and is greatest after laparotomy. Wound metastasis is five times more likely after laparoscopic tumor laceration than after laparotomy laceration. No tumor metastases developed if the tumor capsule remained intact. Thus, this study shows a high rate of wound metastases in rats after laparoscopy only 1 week after surgery. There is a growing body of evidence that a laparoscopic procedure somehow promotes tumor spread to instrument port sites. The animal model presented in this paper reinforces this implication. The ultimate question that may never be answered is how the port-site metastases will affect the overall survival of advanced carcinomas. Given the minimal invasive nature of laparoscopy in general, the minimized operative morbidity may outweigh the risk of tumor seeding, for which no survival impact is known or can be known.

Heller myotomy via minimal access surgery: an evaluation of antireflux procedure

Raiser F, Perdikis G, Hinder RA, Swanstrom LL, Filipi CJ, McBride PJ, Katada N, Neary PJ Arch Surg (1996) 131: 593598 Objective: The author examined the question of surgical approach for Heller myotomy and choice of fundoplication in the setting of minimal access surgery. Methods: Between 1992 and 1995, 39 patients were diagnosed as having achalasia. The patients underwent Heller myotomy via either thoracoscopy or laparoscopy with either a Dor or Toupet fundoplication according to surgeons preference. Short-term (39 months, mean 4 months) and long term (1146 months, mean 26 months) follow-up evaluations were performed using a symptom questionnaire and chart review. Four patients underwent a thoracoscopic approach and 35 patients underwent laparoscopic abdominal approach. Each patient underwent surgical esophagocardiomyotomy of the distal esophagus to a point 23 cm onto the anterior wall of the stomach. The mucosa was allowed to bulge between the distracted muscle over 90100 of the circumference of the esophagus. Results: Surgical complications were three esophageal perforations among the patients who underwent laparoscopic procedure (9%) and no perforation among the four thoracoscopic cases. In the early laparoscopic experience four patients were encountered with gastric perforation due to aggressive retraction of the stomach. All perforations were recognized at the time of surgery and repaired laparoscopically. The patients who underwent laparoscopic surgery were discharged on postoperative day 1 whereas two patients who underwent thoracoscopy were discharged on day

3 (n 1) and day 4 (n 1) due to slower postoperative recovery. The laparoscopic group was also faster in resuming normal activity, averaging 12 days compared to 42 days for the thoracoscopy group. In the early postoperative period none of the patients had a positive DeMeester reflux score; however, 22% of the patients in the Toupet group and 67% of the patients in the Dor group complained of heartburn. At long-term follow-up 27% of the patients who underwent Toupet procedure had at least occasional heartburnlike symptoms compared to 57% in the Dor group. On short-term follow-up, dysphagia was encountered in 44% of patients in the Toupet group and in 78% of patients in the Dor group. Residual dysphagia was brought out by the questionnaire in 33% of the Toupet group and 71% of the Dor group on long-term follow-up; however, the severity of symptoms was quite mild. Three patients required further revision surgery. Conclusion: A minimally invasive approach for achalasia of the esophagus is safe and effective. Also, there were few esophageal and gastric perforations, all of which were found and managed intraoperatively without late sequelae. A laparoscopic approach is better than the thoracoscopic approach because it facilitates extension of the myotomy onto the anterior wall of the stomach, involves less postoperative pain, and shortens postoperative recovery. The Toupet procedure proved to be superior to the Dor based on postoperative symptoms. Editorial comment: This is an important study which once again shows that minimally invasive surgery can apply to a variety of traditional open procedures. Although this series was not randomized the authors conclusions are rational and based on the data. We agree with the authors about their choice of procedure. In fact, we also have a series of 35 patients who underwent laparoscopic Heller myotomy and Toupet fundoplication and our results are similar to the series. We have found that megaesophagus or redo Heller myotomy is associated with a poorer outcome.

Laparoscopic adrenalectomy: a comparison of the lateral and posterior approach

Duh QY, Siperstein AE, Clark OH, Schecter WP, Horn JK, Harrison MR, Hunt TK, Way LW Arch Surg (1996) 131: 870876 Background: There are three basic laparoscopic approaches to adrenalectomy: the anterior abdominal approach, the lateral transabdominal approach, and the posterior approach. This study was conducted to compare the two most common preferred approaches: the lateral transabdominal approach and the posterior approach. Methods and patients: The authors, in an unrandomized fashion, divided 36 patients who required adrenalectomy into two groups. The first group underwent lateral transabdominal approach (n 23) and the second group underwent posterior approach (n 14). Adrenal diseases being treated laparoscopically included pheochromocytoma, Cushings, and other tumor and disorders. In the lateral

781

transabdominal approach the author positions four 10-mm trocars at the midclavicular, anterior axillary, midaxillary, and posterior axillary lines one to two finger breadths below the costal margin. The liver is detached from the triangular ligament and rotated medially, and usually it is unnecessary to take down the hepatic flexure. On the left, the spleen, pancreas, and splenic flexure of the colon are dissected from the retroperitoneal attachments and rotated medially. The authors avoided handling the gland directly by grasping the periadrenal tissue with an atraumatic grasper or by lifting it between the open jaws of an atraumatic grasper. In the posterior approach the patient is placed in the prone jackknife position. Because of difficulty in anatomy, often laparoscopic ultrasound is used for the retroperitoneal approach. Three 10-mm trocars are placed posteriorly between the costal margin and the ileac crest. A balloon dissector is used to facilitate exposure of the retroperitoneal space. Results: The tumor excised by the lateral approach averaged 5.3 cm and the tumor excised by the posterior approach averaged 2.6 cm. The average operative time for the lateral approach was 3.8 h compared to 3.4 h with the posterior approach. One patient, a 66-year-old man with ectopic ACTH syndrome, died from multisystem organ failure and diffuse hemorrhage. The patients who underwent lateral transabdominal approach stayed an average of 2.2 days in

the hospital compared to 1.5 days in the posterior approach group. Discussion: Posterior and lateral laparoscopic approaches for adrenalectomy have comparable results. Each has specific advantages. A lateral transabdominal laparoscopic approach should be employed in tumors of medium size (615 cm). It is easier to learn and is probably the procedure of choice for unilateral tumors. The patient must be repositioned for bilateral tumors if they are performed through the lateral approach. The posterior retroperitoneal approach is preferred for small tumors (<6 cm) and is excellent for bilateral tumors. Open procedures should be considered for large (>15 cm) tumors. This approach provides the best exposure for carcinoma. The authors suggest that these cases be considered advanced laparoscopic procedures which require laparoscopic expertise. Adrenocortical cancer should not be resected laparoscopically because of the risk of fracturing and seeding the cancer. Also, radical dissection is very difficult and tumor recurrence at the trocar site is of concern.

Reviewers for this issue: F. Chae, J. M. Sackier, H. Spivak, J. G. Hunter

Surg Endosc (1997) 11: 734736

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic treatment of nonparasitic cysts of the liver with omental transposition flap
A. Emmermann,1 C. Zornig,1 D. M. Lloyd,2 M. Peiper,1 C. Bloechle,1 C. E. Broelsch1
1 2

Department of General Surgery, Universittskrankenhaus Eppendorf, Martinistr. 52, 20246 Hamburg, Germany Leicester Royal Infirmary, Infirmary Square, Leicester LE 15 WW, United Kingdom

Received: 19 January 1996/Accepted: 26 August 1996

Abstract Background: Between 1991 and November 1994, 18 patients with large, solitary, nonparasitic liver cysts underwent laparoscopic deroofing; the last 13 of them also received an omental transposition flap in addition. Methods: Using three to four trocars, the cystic contents were first aspirated, and the cyst derooted widely using diathermia. An omental transposition flap was fashioned and stapled into the cyst cavity itself. Results: Postoperative complications included one case of pulmonary atelectasis. Another patient developed a subhepatic bile collection which was aspirated percutaneously. On average, patients were discharged on the 4th (214) postoperative day. Follow-up was performed with abdominal ultrasound for 243 months (mean 19 months). There were two early cyst recurrences, both in cases without an omental transposition flap (overall recurrence rate, 11%; in patients with omental flap, 0). Conclusions: Deroofing in combination with an omental transposition flap is a safe and effective therapy for symptomatic solitary liver cysts and can be performed using minimal-access surgical techniques. Key words: Nonparasitic liver cysts laparoscopic deroofing omental transposition flap

ture, cholestasis, compression of the vena cava, and even portal hypertension are extremely rare [12, 15, 17]. Even in large cysts surgical intervention is only necessary in patients who are symptomatic or in those patients with a rapidly increasing cyst volume. The development of minimal-access surgical techniques over the past few years has proved to be extremely beneficial for the majority of patients in that postoperative pain, early postoperative mobilization, convalescence, and cosmetic result are all very much improved. This is particularely emphasized where the intraoperative procedure itself is relatively limited but access usually necessitates a very large incision. In this series standard surgical techniques for deroofing simple liver cysts combined with an omental transposition flap were employed together with the advantages of the laparoscopic approach. To evaluate the results we analyzed our patients in a retrospective study.

Patients and methods

Between May 1991 and November 1994, 18 patients with large, solitary liver cysts were operated on laparoscopically at University Hospital of Hamburg-Eppendorf. All patients were female with a mean age of 57 (3973) years. All patients were symptomatic. Sixteen patients presented with pain in the right upper quadrant and a feeling of pressure or early satiety. In two patients the cyst volume increased rapidly. Preoperative ultrasound and CT scans of the abdomen were performed to evaluate the size and location of the cysts as well as the relation to large intrahepatic vessels and the hepatoduodenal structures. Calcification of the cyst wall was considered to be an indication of a parasitic infestation. A negative serological finding for Echinococcus was necessary to help exclude an infective cause. Patients who were thought to have parasitic cysts, biliary collections, adenoma, or even adenocarcinoma of the liver were excluded from having a laparoscopic approach and were not entered into this study. Because patients with peptic ulcer disease present with similar symptoms of pain in the right upper quadrant, an upper-GI endoscopy was routinely performed in all cases. Follow-up data were required from all patients by direct questioning, physical examination, and repeat ultrasound. The patients were operated upon under general anesthesia, usually in an anti-Trendelenburg position. Normally three, sometimes four, trocars were placed in the abdominal cavity following formation of a pneumoperito-

Nonparasitic, dysontogenetic liver cysts are considered a rare clinical entity. The prevalence is about 1% in the adult population [3]. The detection of cysts has increased with the routine use of ultrasound and CT scan, but the majority of cases are asymptomatic. Clinical symptoms occur in about 5% of patients. They include abdominal pain and a feeling of fullness, or of pressure within the abdomen, or of early satiety. Complications such as intracystic hemorrhage, rup-

Correspondence to: A. Emmermann

735 Table 1. Characteristics of 18 patients with dysontogenetic liver cysts Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

a

Age 57 63 65 73 61 39 62 47 56 54 57 44 55 59 51 53 62 69

Size (cm) 14 18 10 15 11 11 18 12 9 10 13 10 9 15 10 9 10 15

Volume (ml) 1,200 2,400 800 1,400 800 900 2,500 1,100 800 900 1,200 900 800 1,400 900 800 1,000

Localization liver lobe Right Right Right Right Right Right Right Right Right Right Right Left Left Right Right Left Left Right

Symptoms Pain Pain, satiety Pain, satiety Pain Pressure Satiety Pain Satiety Pain Pressure Increasing cyst volume Pain Pain Pressure Pain, dyspnea Pain Increasing cyst volume Pain

Operation time 65 150a 70 130 50 100 80 90 110 130 85 90 90 145a 150a 110 80 75

Omental flap No Yes Yes Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes No Yes No

Complications No Bleeding during ChE No Bilioma No No No No No No No Dystelectasis No No No No No No

P.o. hospitalization time (days) 5 7 2 13 3 4 3 5 4 3 2 14 4 3 4 3 2 11

Recurrence No No No No No No No No No No No Yes No No No Yes No No

G.A. R.A. R.A. A.B. N.D. H.H. L.K. R.K. H.L. M.M. M.M. I.P. E.R. M.S. K.S. A.W. P.W. B.Y.

Including cholecystectomy (ChE) simultaneously

neum, in positions similar to those used for patients undergoing laparoscopic cholecystectomy. We found the use of a 45 angled telescope extremely useful in the visualization of the superior aspect of the liver surface. Following aspiration of the cystic contents, wide deroofing of the cyst was performed close to the border of the normal liver parenchyma. Bleeding from the edge of the cyst wall was controlled with a combination of both monopolar and bipolar electrocautery, and when necessary, titanium clips were used on larger vessels. In one particular case, a linear stapler was used to transect a thick part of the cyst wall. Specimens up to 12 cm in diameter could be obtained using this technique. In two early patients in whom no omental flap was performed the endothelial lining was obliterated with argon beam electrocautery used cautiously. Thirteen patients had an omental transposition flap fashioned and placed inside the cyst cavity itself. The omentum was dissected free from adhesions and from the transverse colon. As much omentum was mobilized as possible and secured with ligaclips or an endostapler at the liver margin. In all patients a 14-French silicone drain was placed inside the cyst cavity for 1 day.

Results All patients were female, with a mean age of 57 years. The solitary liver cysts were localized in the right lobe (n 14) or left lobe (n 4) of the liver with diameters between 9 and 18 cm and volumes of 8002,500 ml (Table 1). In three patients preoperative ultrasound confirmed the presence of cholelithiasis, and these patients underwent simultaneous laparoscopic cholecystectomy. Mean operation time was 100 (50150) min. Conversion to the open technique was necessary in one case of simultaneous cholecystectomy. A branch of the portal vein located directly behind the gallbladder was injured during mobilization of the gallbladder. There were no other intraoperative complications and no patient received blood transfusion. The histopathological findings proved all specimens to be part of simple, benign liver cysts with a single layer of cuboidal cells. There were no signs of parasitic cysts, cystadenoma, or malignancy. Postoperatively, one patient developed pulmonary atelectasis which improved with conservative management. A

second patient developed a subhepatic bile collection after loosening the drain by accident. It was drained percutaneously, following which complete resolution resulted. In this particular patient it was thought that the bile collection was caused by transection of a small biliary radical at the edge of the cyst wall as there was no evidence of bile within the cyst cavity at laparoscopy. This seems to be preventable only by careful dissection of the cystic wall either by electrocautery or mechanically with clips or linear stapler. The overall morbidity rate was 11%, and no patient died following this procedure. Postoperative stay ranged from 2 to 14 days (mean, 5 days). Among the first patients there were two early relapses within 2 months. In both these patients, a transposition flap was not performed. One patient is asymptomatic. The other became symptomatic and underwent laparotomy where the edge of the cysts adhered to the anterior abdominal wall and right hemidiaphragm, causing the cyst to recur. This patient developed a further recurrence after conventional surgery, but is currently asymptomatic. In the follow-up period of 243 (mean, 19) months there were no further cyst recurrences. The other 16 patients were asymptomatic; clinical examination and ultrasound were unremarkable. The overall recurrence rate therefore was 11%; in patients with omental transposition flap it was zero.

Discussion Surgical therapy of nonparasitic liver cysts is indicated in cases of clinical symptoms. It is, of course, mandatory to exclude infestation by Echinococcus, and all patients should have negative serology. Furthermore, it is necessary to exclude the very rare causes of liver cysts such as cystadenoma and cystadenocarcinoma [4, 7]. In these circumstances, cyst septations, papillary structures, or multiloculated cystic formation confirmed on ultrasound or CT scan should alert

736

the physician. These lesions require radical excision and are not amenable to laparoscopic surgery at the present time. Even benign conditions (cystadenoma) should be radically resected to prevent recurrence or malignant transformation [4, 7]. The theories of Moschowitz and von Mayenburg postulate that solitary liver cysts result from accumulation of fluid in aberrant bile ducts which do not communicate to the main biliary tree [11, 14]. One can therefore conclude that the cysts should be lined by a single layer of cuboidal epithelium and should not contain any bile. In all our 18 patients this was confirmed and has been corrobarated by several other authors [4, 6]. Furthermore, this theory of pathogenesis gives a plausible explanation for the high incidence of recurrences following simple aspiration as the lining of the cyst wall continues to secrete cyst fluid [4, 9, 16]. In fact, Saini et al. reported 100% recurrence within 2 years in a series of 13 large liver cysts when treated by simple aspiration alone [16]. The combination of aspiration and installation of absolute alcohol used as a sclerosing agent is described in a small number of cases [1, 2]. However, no detailed knowledge exists regarding the toxic effects of this type of treatment on the liver, and in one particular report, the long-term efficacy was not detailed as the term regression was used without a mention of reduction in size or recurrences of the cysts [1]. Therefore we do not recommend this type of treatment. Drainage of the cyst by formation of a jejunal loop sutured to the cyst wall has been described but leads to a high incidence of hepatic complications due to the infection of the cyst contents [9, 10]. This procedure therefore should be limited to those cysts which have communication with the biliary tree. Liver resection including lobectomy or hemihepatectomy is extremely rarely indicated for the management of these cysts and should only be done in exceptional circumstances. Indeed, this radical approach, although it eliminates any cyst recurrence, is naturally associated with a higher postoperative morbidity and mortality. Destruction of the endothelium of the cyst cavity by ablation with diathermy electrocurrent, although theoretically sound, could be associated with major thermal injury to deep hepatic structures and structures at the hilum. The use of argon beam electrocautery may be safer but the depth of thermal injury may still be so variable as to preclude its routine use. The treatment of choice is excision of the roof of the cyst. There is no problem in dissection with a thick layer of liver tissue, because cysts which need therapy are always so large that they become part of the surface of the liver. Deroofing results in a maximum sparing of normal liver tissue and is associated with a low rate of complications and with a rate of zero to 38% of recurrence [3, 9, 17]. To improve this recurrence rate and as a result of personal experience of early relapses among our first laparoscopically treated patients, we decided to combine the deroofing with an omental transposition flap. The omentum has the function of keeping the cyst cavity open to the abdomen and is competent itself to resorb fluid which is produced by the cystic endothelium.

Furthermore, we take great care to excise as much of the roof of the cyst as possible, which may represent one-third of the cyst wall itself. With this operative technique, so far, we have not seen cyst recurrence in 13 patients. We believe that it is the combination of radical deroofing together with omental flap interposition which eliminates cyst recurrence, as simple incision and placement of omentum inside the cyst has been associated with high recurrence [18]. The conventional surgical technique of deroofing simple cysts requires a large incision to gain sufficient exposure to the superior aspect of the liver. It therefore seems that the laparoscopic approach is well suited for this particular operation as the patients benefit from the minimal-access incisions without compromising the surgical outcome [5, 8, 13]. The results of our series of 13 patients show that the laparoscopic technique of deroofing solitary liver cyst combined with an omental transposition flap is a safe and effective method which takes all available advantages of the minimal invasive approach. References
1. Andersson R, Jeppsson B, Lunderquist A, Bengmark S (1989) Alcohol sclerotherapy of non-parasitic cysts of the liver. Br J Surg 76: 254255 2. Bean WJ, Rodan BA (1985) Hepatic cysts. Treatment with alcohol. AJR Am J Roentgenol 144: 237241 3. Benhamou JP, Menu Y (1988) Non parasitic cystic diseases of the liver and intrahepatic biliary tree. In: Blumgart LH (ed) Surgery of the liver and biliary tract, vol 2. Churchill Livingston, New York, pp 10131025 4. Edwards JD, Eckhauser FE, Knol JA, Strodel WE, Appelmann HD (1987) Optimizing surgical management of symptomatic solitary hepatic cysts. Am Surg 53: 510514 5. Fabiani P, Katkouda N, Jovine L, Mouiel J (1991) Laparoscopic fenestration of biliary cysts. Surg Laparosc Endosc 5: 175179 6. Flagg RS, Robinson DW (1967) Solitary nonparasitic hepatic cysts. Arch Surg 95: 964973 7. Ishak KG, Willis GW, Cummins SD, Bullock AA (1977) Biliary cystadenoma and cystadenocarcinoma. Report of 14 cases and review of the literature. Cancer 38: 322338 8. Lange V, Meyer G, Rau H, Schildberg FW (1992) Minimal-invasive Eingriffe bei solitren Lebercysten. Chirurg 63: 349352 9. Litwin DEM, Taylor BR, Greig P, Langer B (1987) Nonparasitic cysts of the liver. The case for conservative surgical management. Ann Surg 205: 45 10. Longmire WP, Trout HH, Greenfield J, Tompkins RK (1974) Elective hepatic surgery. Ann Surg 179: 712721 11. von Mayenburg H (1918) ber die Cystenleber. Beitr Pathol Anat 64: 477480 12. Morgenstern L (1959) Rupture of solitary non-parasitic cysts of the liver. Ann Surg 150: 167171 13. Morino M, De Giuli M, Festa V, Garrone C (1994) Laparoscopic management of symptomatic nonparasitic cysts of the liver. Ann Surg 219: 157164 14. Moschowitz E (1906) Non-parasitic cysts (congenital) of the liver with a study of aberrant bile ducts. Am J Med Sci 131: 674680 15. Rashed A, May RE, Williamson RCN (1982) The management of large congenital liver cysts. Postgrad Med J 58: 536541 16. Saini S, Mller PR, Ferrucci JT, Simeone JF, Wittenberg J, Butch RJ (1983) Percutaneous arpiration of hepatic cysts does not provide definitive therapy. AJR Am J Roentgenol 141: 559560 17. Sanchez H, Gagner M, Rossi RL, Jenkins RL, Lewis WD, Munson JL, Braassch JW (1991) Surgical management of nonparasitic cystic liver disease. Am J Surg 161: 113119 18. Wellwood JM, Madara JL, Cady B, Haggitt RC (1978) Large intrahepatic cysts and pseudocysts. Am J Surg 135: 5764

Technique
Surg Endosc (1997) 11: 782784

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Endoscopically guided percutaneous repair of inguinal hernia through a 2-cm incision

Minihernia repair
A. Darzi, C. C. Nduka
Academic Surgical Unit, St. Marys Hospital, Praed Street, Paddington, London W2 1NY, United Kingdom Received: 21 May 1996/Accepted: 8 August 1996

Abstract Background: The laparoscopic repair of inguinal hernia is still controversial. Transabdominal preperitoneal repair violates the peritoneal cavity and may result in visceral injuries or intestinal obstruction. The laparoscopic extraperitoneal approach has the disadvantage of being technically demanding and requires extensive extraperitoneal mobilization. The Lichtenstein repair gives good long-term results, is easy to learn, can be performed under local anesthesia, but requires a larger incision. Methods: We describe a novel percutaneous tension-free prosthetic mesh repair performed through a 2-cm groin incision. The inguinal canal is traversed with the aid of a 5-mm video-endoscope and the canal is widened using specially designed balloons. Spermatic cord mobilization, identification and excision of the indirect sac, and posterior wall repair are carried out under endoscopic guidance. Results: Between October 1993 and July 1995, 85 primary inguinal hernia repairs (48 indirect and 33 direct) were performed on 81 patients (80 men, one woman) by the author (A.D.). The mean age was 41 years (range 1783 years). Six repairs were performed under local anesthetic. Mean operative time was 42 min (range 2574). Mean hospital stay was 1.2 days (03 days). The mean return to normal activity was 8 days (210 days). Eight complications have occurred: a serous wound discharge, two scrotal hematomas, a scrotal swelling that resolved spontaneously, wound pain lasting 2 weeks, an episode of urinary retention, and two recurrences early in the series (follow-up 122 months). Conclusion: The endoscopically guided percutaneous hernia repair avoids the disadvantages of laparoscopy (i.e., lack of stereoscopic vision, reduced tactile feedback, unfamiliar anatomical approach, risk of visceral injury), yet the use of endoscopic instrumentation allows operation through a

2-cm incision. The minihernia repair thus combines the virtues of an open tension-free repair with minimal access trauma. Key words: Inguinal hernia repair Laparoscopic Mesh repair

Correspondence to: A. Darzi, Minimal Access Surgery Unit, St. Marys Hospital, Praed Street, Paddington, London W2 1NY, United Kingdom

The surgical approach to inguinal hernias continues to undergo technical modifications and the optimal management of inguinal hernia is still being debated. Traditional repairs of McVay, Bassini, and Shouldice involve approximation of nonanatomically opposed tissues under tension [3, 15, 17, 21]. This may account for the reported recurrence rates of up to 21% for primary repairs [2, 4, 10, 19] and explain the lengthy and painful recovery periods. Laparoscopic techniques applied to hernia repair have shown short-term advantages over traditional approaches [7, 8, 20], though they are technically demanding, require violation of the peritoneal cavity, and have unknown long-term results. With a reported recurrence rate as low at 0.7% [11, 12], the tension-free hernioplasty with prosthetic mesh described by Lichtenstein is a simple and effective technique that has probably become the new standard against which new hernia repairs should be compared. The authors describe a novel endoscopically assisted tension-free repair [5] which employs the principles of the Lichtensteins repair but is carried out through a single 2-cm skin incision. This technique allows the benefits of an open surgical approach such as hands-on manipulation, 3-D vision, familiar anatomical approach, and use of conventional instruments to be combined with the advantages derived from the use of laparoscopic instrumentationnamely, minimized tissue trauma and improved cosmesis.

783

Operative technique
The procedure is performed in a day-case surgery unit under local or general anesthesia. The patient is placed in a supine position and draped as for an open hernia repair. A 2-cm incision is made in the skin at the level of the internal inguinal ring. The subcutaneous tissues are bluntly dissected down to the external oblique aponeurosis and a 1-cm incision is made in it, parallel to its fibers. The index finger is inserted into the inguinal canal to separate the spermatic cord from the external oblique aponeurosis down to the level of the external inguinal ring. Following this a 10-mm laparoscopic fan retractor (Laparofan Origin Medsystems, Inc., Menlo Park, California) is inserted into the inguinal canal to elevate the anterior wall of the canal both upward and laterally. The two blades of the fan are spread and locked in an open position. The Laparofan itself is connected to a Laparolift (Origin Medsystems, Inc., Menlo Park, California) system which consists of a hydraulic arm placed on the side of the surgical bed and of a telescoping arm that extends over the patients abdomen. The Laparolift is then elevated up to 15 mmHg of upward pressure. This allows the exposure of the inguinal canal and creates a working space where a 5-mm laparoscope, which is connected to routine video-imaging equipment, is inserted. This arrangement allows the inguinal canal to be visualized directly using light from the laparoscope and also allows a magnified image of the canal to be viewed on the monitor. Using conventional grasping and dissecting instruments the spermatic cord is mobilized from the floor of the inguinal canal. The spermatic cord is grasped and a window is created just beneath the vas at the level of the pubic tubercle. When elevating the spermatic cord, great care should be taken to include the external spermatic vessels and the ilio-inguinal nerve with the cord. This ensures that the genital nerve is preserved. With the aid of a conventional right-angle instrument, a nylon tape is inserted into the canal and passed through the window around the cord. Using endoshears the canal is further dissected to mobilize completely the cord from the posterior wall and to expose the conjoint tendon in all its length. The laparoscopic fan retractor is removed and the spermatic cord is delivered through the skin incision. An incision is made in the spermatic fascia and the cord is inspected for the presence of a hernial sac. If a sac is identified it is mobilized, inspected, ligated, and divided in the usual manner. The laparoscopic fan retractor is reinserted and placed below the spermatic cord and the external oblique aponeurosis. The retractor is once again raised, thus providing good exposure of the inguinal canal. A sheet of monofilament polypropylene mesh (AutoSuture Company, Ascot, UK) measuring 9 by 15 cm is fashioned to match the varying sizes of the inguinal floor. The mesh is inserted into the inguinal canal under laparoscopic guidance with aid of two anchoring sutures. Using the VersaTack stapler (AutoSuture Company, Ascot, UK), the mesh is anchored under direct vision to the pubic tubercle. Two or three staples are then placed along the conjoint tendon and inguinal ligament, thus securing the mesh medially and laterally. A slit in the mesh at the internal ring allows emergence of the spermatic cord and creates two tails. The tails of the mesh are crossed over without tension, wrapped around the cord, and stapled together just above the cord to fashion a new internal ring. After the mesh has been secured in place, the retractor is then released the mesh should buckle slightly. This laxity is desirable to ensure a true tension-free repair and is taken up when the patient strains postoperatively. The external oblique aponeurosis is closed over the cord using absorbable suture. The wound is sprayed with antiseptic solution and the skin is closed with a single suture or steristrips. Patients are discharged within 24 h of the operation with minimal postoperative pain for which mild analgesics are prescribed. Unrestricted activity is encouraged.

that resolved spontaneously, wound pain lasting 2 weeks, an episode of urinary retention, and two recurrences early in the series (follow-up 122 months). Discussion The surgical treatment of inguinal hernias has greatly progressed in the last few years and continues to undergo technical modifications. However, the optimal management of inguinal hernia must still be defined. Traditional procedures such as McVay, Bassini, and Shouldice repairs performed outside specialist centers may give rise to high recurrence rates [2]. The application of laparoscopic surgery [7, 8] to the treatment of inguinal hernias has been much debated. The principal advantages of this technique are no muscle or fascial incision, less pain, and quick return to work [20]. This approach also has the advantage of facilitating the repair of recurrent hernia and the ability to treat bilateral hernia using the same access incisions. On the other hand, the laparoscopic hernia repair is susceptible to a number of potential complications as a result of the intraabdominal approach [13]. These include injuries to the urinary bladder, major nerves, intestine, or major vessels. Laparoscopy also necessitates a general anesthetic, may be unsuitable in patients with cardiorespiratory disease [9], and may be associated with postoperative small-bowel obstruction associated with trocar-site hernia [16] or adhesions [1]. Laparoscopic extraperitoneal hernia repairs have been advocated to avoid some of the pitfalls of the transabdominal approach, but this is an even more technically demanding procedure [6, 14]. The tension-free hernioplasty using a prosthetic mesh described by Lichtenstein [11, 12, 18] results in a simple and effective technique with recurrence and complication rates of less than 1% and has probably become the new gold standard for hernia repair. Most importantly, it can be performed under local anesthetic using a familiar anatomical approach without violating the peritoneal cavity. However, the Lichtenstein repair still requires an incision of up to 8 cm and division of the external oblique aponeurosis. In contrast, the minihernia repair described above only requires an incision of 12 cm through the skin and external oblique to allow access to the inguinal canal. As the minihernia repair shares the same theoretical principal as the Lichtenstein repair one may expect similar excellent longterm results. But because it can be performed through a much smaller incision it is hoped that there will be reduced wound pain and better cosmesis without the cost or hazards of an intraperitoneal approach. Furthermore, the operation is carried out under 3-D vision while maintaining the important sensory cues of touch and feel, and the anatomical approach employed is familiar to the surgeon. A prospective trial is underway to confirm the efficacy of this procedure. Nevertheless, our early experience indicates that the minihernia repair may represent an important advance in minimal access hernia repair. References
1. Arregui ME, Navarrete J, Davis CJ, Castro D, Nagan RF (1993) Laparoscopic inguinal herniorrhaphy: techniques and controversies. Surg Clin North Am 73: 513527

Results Between October 1993 and July 1995, 85 primary inguinal hernia repairs (48 indirect and 33 direct) were performed on 81 patients (80 men, one woman) by the author (A.D.). The mean age was 41 years (range 1783 years). Six repairs were performed under local anesthetic. Mean operative time was 42 min (range 2574). Mean hospital stay was 1.2 days (03 days). The mean return to normal activity was 8 days (210 days). Eight complications have occurred: a serous wound discharge, two scrotal hematomas, a scrotal swelling

784 2. Asmussen T, Jensen FV (1983) A follow-up study on recurrence after inguinal hernia repair. Surg Gynecol Obstet 156: 198 3. Bassini E (1888) Spora 100 casi di cura radicali dellernia inguinale operata col. methodo dellautore. Arch Atti Soc Ital Chir 5: 315 4. Cahlin E, Weiss L (1980) Results of postoperative clinical examination of inguinal hernia after three years. Acta Chir Scand 146: 421 5. Darzi A, Bouchier-Hayes D, Menzies-Gow N, Nduka CC (1995) Endoscopically guided surface repair of inguinal hernia. Br J Surg 82: 515517 6. Ferzli GS, Massad A, Albert P (1992) Extraperitoneal endoscopic inguinal hernia repair. J Laparoendosc Surg 2: 281286 7. Ger R (1991) The laparoscopic management of groin hernias. Contemp Surg 39: 1519 8. Ger R, Monroe K, Duvivier R, et al (1990) Management of indirect inguinal hernias by laparoscopic closure of the neck of the sac. Am J Surg 159: 320323 9. Holzman M, Sharp K, Richard W (1992) Hypercarbia during carbon dioxide insufflation for therapeutic laparoscopy: a note of caution. Surg Laparosc Endosc 2: 1114 10. Ingimersson O, Spak I (1983) Inguinal and femoral hernias: long term results in a community hospital. Acta Chir Scand 149: 291 11. Lichtenstein IL (1987) Herniorrhaphy: a personal experience with 6321 cases. Am J Surg 153: 553559 12. Lichtenstein IL, Shulman AG, Amid PK, Montlor MM (1989) The tension-free herninoplasty. Am J Surg 157: 188193 13. MacFadyen BV Jr, Arregui ME, Corbitt JD Jr, et al (1993) Complications of laparoscopic inguinal hernia repair. Surg Endosc 7: 155158 14. McKerman JB, Lewis HL (1993) Laparoscopic repair of inguinal hernias using a totally extraperitoneal psthetic approach. Surg Endosc 7: 2628 15. McVay CB (1948) Inguinal and femoral hernioplasty: anatomical repair. Arch Surg 57: 524 16. Phillips E, Carroll B (1993) Laparoscopic inguinal hernia repair. Gastrointest Endosc Clin North Am 3: 297307 17. Shouldice EE (1953) The treatment of hernia. Ontario Med Rev 114 18. Shulman AG, Amid PK, Lichtenstein IL (1992) The safety of mesh repair for primary inguinal hernias: results of 3,019 operations from five diverse surgical sources. Am Surg 58: 255257 19. Sisley JF, Scarbourough CS, Morris RC, et al (1987) Shouldice hernia repair: results of a teaching institution. Am Surg 53: 495496 20. Stoker DL, Spiegelhalter DJ, Singh R, Wellwood JM (1994) Laparoscopic versus open inguinal hernia repair: randomised prospective trial. Lancet 343: 12431245 21. Wantz GE (1988) Shouldice repair. Contemp Surg 33: 1521

Surg Endosc (1997) 11: 750753

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Elective laparoscopic-assisted sigmoid resection for diverticular disease

Q. A. J. Eijsbouts, M. A. Cuesta, L. M. de Brauw, C. Sietses
Department of Surgery, Academic Hospital, De Boelelaan 1117, Vrije Universiteit, Amsterdam, The Netherlands Received: 26 August 1996/Accepted: 26 November 1996

Abstract Background: Although the laparoscopic-assisted approach to colorectal cancer remains controversial, its use for benign diseases can have important advantages. The purpose of this study is to determine the feasibility of this approach for the treatment of elective diverticular disease and to identify preoperative and perioperative factors which can help to select the best procedure for each patient: either assisted laparoscopic resection (ALR) or dissection-facilitated laparoscopic resection (DLR). Methods: From November 1991 to the present, we conducted a prospective study of 41 patients approached electively for diverticular disease. Results: Twenty-nine patients underwent an ALR, seven were approached by DLR, and another five patients were converted to laparotomy (15%). Morbidity was 17.5% and there was no mortality in this series. The mean hospital stay after operation was 6.5 days. Conclusions: Because of the complexity of this inflammatory process, choice of either an assisted or a more invasive laparoscopic facilitated approach is necessary. The decision is based on the technical difficulty as determined by data collected both preoperatively and during laparoscopy. Key words: Diverticular disease Laparoscopic colonic resection Colorectal cancer

technically difficult to work on. A prospective study was therefore begun in November 1991 to analyze the feasibility of the laparoscopic-assisted approach in elective surgery of diverticular diseases. Of 83 patients approached laparoscopically for benign colorectal diseases (Table 1), elective sigmoid resection due to diverticular disease was performed in 41 of them. Surgical indications for elective diverticular disease are [12]: resection following recurrent attacks of diverticulitis (after two attacks or after a single attack in either young patients or patients requiring chronic immunosuppressive therapy); diverticulitis associated with fistulas to the bladder or vagina; and elective resection following previously percutaneously drained pericolic abscesses (Hinchey stage I and II) [3]. Materials and methods
Forty-one patients (mean age 65 years) were included in this prospective study (Table 2). There were 19 males and 22 females. Twenty-seven patients had a recurrent attack of diverticulitis; another four patients (younger than 50 years) had had a single attack of diverticulitis; one patient had a fistula to the vagina and another four had fistulas to the bladder. Diverticulitis in combination with adenomatous polyps in the sigmoid was present in four patients and an elective resection of the sigmoid was indicated after a transrectally drained pelvic abscess in the last patient (Hinchey stage II).

Preoperative workup Laparoscopic colonic resection performed by laparoscopy remains controversial in cases of colorectal cancer [14]. The appearance of cancer implants in trocar and wound sites warrants prospective studies [5, 13]. Nevertheless, its application in benign colorectal diseases is becoming an option due to the advantages of endoscopic procedures. In most cases, however, benign surgery of the colon and rectum involves inflammatory processes and these can be
Presented at the annual meeting of the European Association of Endoscopic Surgery (EAES), Trondheim, Norway, 2326 June 1996 Correspondence to: M. A. Cuesta Diagnosis of diverticulitis was confirmed by sigmoidoscopy and barium enema performed approximately 6 weeks after the attack. Sigmoidoscopy will help to differentiate diverticulitis from a sigmoid cancer and the barium enema is a tool for assessment of location of the process and the length of the sigmoid loop. Starting in 1993 we added an enhanced computer-assisted tomography (CT scan) with oral and rectal contrast in all patients who are candidates for surgery in order to establish the diagnosis and anticipate technical difficulties. All these studies together are an important aid in planning the operative approach, as the length of the sigmoid loop and the presence of any fistulas or phlegmon need to be determined.

Operating technique

Pilot study. Based on the initial experience obtained from the first six patients operated, two laparoscopic techniques were developed for the

751 Table 1. Laparoscopic approach for benign colorectal diseases in 83 patients Ileocoecal resection (Crohns disease) Left hemicolectomy (Crohns disease) Total colectomy (Crohns disease) Total colectomy + ileo-anal anastomosis (1 polyposis coli, 4 ulcerative colitis) Rectopexy Sigmoid resection in diverticulitis Sigmoid resection (polyps) Stoma creation Take down Hartmann 4 1 1 5 17 41 6 7 1

Table 2. Laparoscopic approach diverticular disease in 41 patients After drainage of abscess After one attack in young patients Combination of polyp and diverticulitis Fistulas Colovesical Colovaginal Recurrent diverticulitis 1 4 4 4 1 27

Fig. 1. Technique for assisted laparoscopic resection.

sigmoid resection. Which of the two was used for subsequent patients depended on the data obtained during the preoperative study and operative laparoscopic findings. These data were obtained during the preoperative assessment: Presence of residual inflammatory phlegmon Presence of fistulas Short sigmoid loop We have classified the laparoscopic findings based on the degree of difficulty of dissecting the sigmoid completely free and identifying the left ureter: Grade I: the presence of a relatively short segment of inflammation with loose adhesions to the abdominal wall or pelvis Grade II: an extensive segment of inflammation sticking to the abdominal wall or pelvis; difficult dissection of the ureter Grade III: the presence of woody diverticulitis with a very strong fibrosis around the inflammatory process; nearly impossible dissection of the ureter Grade IV: the presence of a fistula The data obtained from these two pre- and peroperative assessments will from the start determine the grade of technical difficulty of the procedure and facilitate the choice of laparoscopic approach.

Technique No. 2 (dissection-facilitated laparoscopic resection, DLR). In cases where mobilization of the most affected segment of the sigmoid was not possible by assisted laparoscopy (ALR) because of the existence of a large phlegmon or fibrotic fistula track or woody diverticulitis or lack of identification of the left ureter, we proceeded with a dissectionfacilitated laparoscopic resection (Fig. 2). After mobilization by laparoscopy of the major part of the sigmoid, descending colon, and splenic flexure, the sigmoid was devascularized using clips. For mobilization of the splenic flexure the patient is placed in the reverse Trendelenburg position and slightly tilted to the right side. The optic is changed to port No. 4 in the left lower abdomen and the surgeon dissects the flexure free using the umbilical and right abdominal ports (No. 1 and 2) while the assistent helps to hold the optic and the instruments through the upper left abdominal trocar (No. 5). After this we proceeded to perform a Pfannenstiel incision of about 7 cm in length. Through this incision we could continue with the rest of the mobilization of the sigmoid and proximal rectum and eventually identification of the ureter. After resection of the sigmoid, the anastomosis was performed extracorporeally either manually or using a double-stapled anastomosis.

Choice of operative technique

After the first six patients the choice of the laparoscopic approach was as follows. We performed an assisted laparoscopic sigmoid resection in the following cases: Without mobilization of the splenic flexure: Grade I laparoscopic findings With mobilization of splenic flexure: A short sigmoid loop Grade II laparoscopic findings Facilitated laparoscopic dissection was performed in the following cases (with mobilization of the splenic flexure if necessary): Grade III and IV laparoscopic findings Grade II if impossible to dissect free from the pelvic wall

Technique No. 1 (assisted laparoscopic resection, ALR). We used the fouror five-cannula approach depending on whether mobilization of the splenic flexure was necessary (Fig. 1). After mobilization of the sigmoid, the left ureter was identified; the most distal part of the descending colon and proximal rectum were then mobilized. In those cases where extensive inflammatory processes or fistulas were found, mobilization of the splenic flexure was performed and the procedure was continued as a dissectionfacilitated laparoscopic resection (see technique No. 2). Devascularization took place after this dissection, clipping and cutting the sigmoid vessels, and once the distal level of resection was determined, the arcade vessels were cut. In the majority of cases the proximal rectum was cut with the endostapler (Endo-gia blue, USSC, Norwalk, CT, USA) at the promontorium level. Then a small incision of 5 cm was made at the level of the superior iliac spine in the left lower abdomen and through it the specimen was retrieved. Resection was performed followed by introduction of the anvil in the proximal end. After its reintroduction into the abdomen and closure of the wound and reinsufflation, a double-stapled anastomosis was performed [1]. The anastomosis was checked by injecting methylene blue into the rectum. Finally the abdomen was drained and the ports were closed.

Postoperative treatment
Patients were treated prophylactically with 3,075 IU subcutaneous heparin (Fraxiparine, Sanofi Winthrop Co.) the night before the operation; treatment continued once a day until discharge from hospital. Antibiotics (Flagyl, 3 500 mg iv, Rhone Poulenc Co.; and Zinacef, 2 750 mg, Glaxo) were given prophylactically for 24 hours, starting with the premedication. The nasogastric tube was removed following the operation. If peristalsis was present, liquids were given to the patients on the

752

tients had a wound abscess, one patient developed a venous thrombosis in the popliteal vein, one had a small bile leakage due to an associated laparoscopic cholecystectomy (limited to 2 days), and another patient had bleeding at the trocar site. Another two patients were reoperated during the postoperative course, one having developed peritonitis due to a small-bowel perforation after laparoscopic adhesiolysis and one having a leakage of the anastomosis 3 days after discharge, which was treated by a Hartmann procedure. No mortality was registered in this series.

Discussion Uncomplicated diverticulitis means the presence of peridiverticulitis or phlegmon which implies a limited inflammatory process which must be treated conservatively. Under complicated diverticulitis we understand one of the following conditions: the presence of obstruction, a free perforation in the abdominal cavity, the presence of a pericolic abscess or the development of a fistula (either to the bladder60% of the casesor to the vaginal stump following resection of the uterus or to the abdominal wall) [12]. It is obvious that the existence of a free perforation or an obstruction implies an acute surgical intervention. In the case of a pericolic abscess this can be drained percutaneously and the patient prepared conservatively for elective intervention [11]. Fistulas are treated surgically during an elective procedure. Surgical indications for elective procedures in diverticular disease are (1) recurrent attacks of uncomplicated diverticulitis (two attacks of diverticulitis or a single attack in patients younger than 50 years); (2) diverticulitis associated with fistulas in which the existence of a carcinoma has to be excluded; (3) elective resection after a previously drained pericolic abscess; and (4) one attack in a patient requiring chronic immunosuppressive therapy [12, 10]. Basic guidelines for this surgical resection include the excision of the sigmoid loop and anastomosis in the proximal rectum as a high anterior resection in order to avoid an anastomosis with the distal sigmoid where intraluminal high pressure can induce the recurrence of diverticulitis. More debate centers on the proximal level of the resection. No compelling data are available to suggest that all diverticula must be removed; rather, the involved segment of colon, which is identified by serosal changes of prior inflammation, has to be resected along with all bowel distal to the rectum [10]. Only in exceptional cases will more colon be resected if there is obvious gross inflammatory involvement in the descending colon or the presence of giant diverticulum. The anastomosis has to be performed without any resulting tension, occasionally necessitating mobilization of the splenic flexure. In the case of an extensive inflammatory process or the presence of a short sigmoid loop, the mobilization will be done before the anastomosis and in the rest of the cases once the anastomosis is performed, if any tension is observed. Laparoscopic colon resection was introduced in 1990 for cancer and inflammatory processes [4]. In almost all published series, a combination of cases involving malignancies and diverticulitis has been reported. Also, a small

Fig. 2. Technique for dissection-facilitated laparoscopic resection.

2nd and 3rd postoperative days, followed by another 2 days of bland diet and finally a normal diet.

Results According to the laparoscopic findings we classified 18 patients as having grade I diverticulitis, 15 patients as having grade II, three patients with diverticulitis grade III, and finally five patients with a fistula and therefore as grade IV. In grade I, all patients were operated using ALR (conversion grade 0%). In grade II, 11 of the 15 patients underwent an ALR and one a DLR, another three patients being converted to laparotomy (conversion rate 20%). A temporary loop ileostomy was created in the patient operated on by means of a DLR to protect the anastomosisnecessary because a pelvic abscess (Hinchey stage II) was preoperatively drained transrectally. In grade III, two patients underwent a DLR and the third was converted. In grade IV, four patients were operated on using a DLR and another one was converted at the outset because of cancer which had fistulized into the bladder. The procedure was converted to a median laparotomy in a total of six patients of 41 (15%) due to: a stapler perforation of the rectum (2); failure to identify the left ureter (2) (bladder perforation also occurred in one of these patients); perforation of the transverse colon during mobilization of the splenic flexure (1); and a fistulized sigmoid cancer into the bladder (1). The mean operating time was 3 h 15 min. In the first 20 patients this time was 3 h 55 min and in the last 20 patients 3 h 5 min. Mean blood loss was 225 ml. The converted patients were transfused with 3 units of blood on average. Only one of the grade I patients was transfused (two units of blood) due to bleeding from the sigmoid vessels during the laparoscopic procedure. In three patients the procedure was associated with a laparoscopic cholecystectomy because of symptomatic cholelithiasis and in a fourth with an extensive adhesiolysis. The mean hospital stay was 6.5 days after the operation. Morbidity was found in seven patients (17.5%). Two pa-

753

number of papers have been published concerning diverticulitis and fistulas exclusively [2, 8]. Liberman et al. [6] recently published one retrospective study comparing laparoscopic-treated diverticulitis with the conventional procedure, concluding that the laparoscopic approach resulted in a more rapid return of bowel function and shortened hospital stay. In spite of higher operating room charges, total hospital charges and general costs were significantly reduced. The general advantages of laparoscopic surgery have been described for this procedure. In our series, significant safeguarding of patient comfort along with preservation of peristalsis, quick reassumption of food intake, and respect of the integrity of the abdominal wall are considered the most important advantages to the patient. Not all so-called laparoscopic colon resections are technically the same. We need to distinguish between the dissection-facilitated laparoscopic resection where the dissection and sometimes the devascularization have taken place laparoscopically and the assisted laparoscopic resection where even the resection takes place laparoscopically with retrieval of the specimen through a small incision, the anastomosis being performed intracorporeally [7]. Although laparoscopic colorectal resection for cancer remains controversial because of the possibility of cancer implants at the port sites and incision places [13], laparoscopic resection for inflammatory colon processes has been performed with more caution because it is technically more demanding due to the inflammatory reaction and absence of planes. Complication rates can be high and so, too, the rate of consequent conversions [9]. In our series we have seen that this technique is feasible, having a low conversion and complication rate, if the surgeon plans and approaches this laparoscopic resection selectively. A thorough preoperative study together with the laparoscopic findings is important in planning the laparoscopic approach. Whereas a short inflammatory segment of the sigmoid may be approached by a assisted laparoscopic sigmoid resection, the presence of extensive inflammatory residual phlegmon or a short sigmoid loop will require a mobilization of the splenic flexure. In the same way a difficult local dissection caused by a fibrotic woody diverticulitis or the presence of a fistula to the bladder or vagina will be facilitated by mobilization of the sigmoid and proximal rectum by a Pfannenstiel incision in order to solve the local problem. With respect to the mobilization of the splenic flexure, a clear policy is difficult to plan. Only in

extensive inflammatory processes or when a barium enema shows a short sigmoid loop to be present will the splenic flexure be mobilized prior to the completion of the anastomosis. Conclusion The laparoscopic-assisted sigmoid resection is a technically feasible procedure. With a good pre- and peroperative study, adoption of a complete or dissection-facilitated laparoscopic approach will solve the unpredictible difficulties of this important inflammatory disease. References
1. Cuesta MA, Borgstein PJ, Paul MA, de Jong D (1992) Surgery of the distal colon assisted by laparoscopy. Video Rev Surg 9: 1021 2. Hewett PJ, Stitz R (1995) The treatment of internal fistulae that complicate diverticular disease of the sigmoid colon by laparoscopically assisted colectomy. Surg Endosc 9: 411413 3. Hinchey EJ, Schaal PG, Richards GK (1978) Treatment of perforated diverticular disease of the colon. Adv Surg 12: 85109 4. Jacobs M, Verdeja JC, Goldstein HS (1991) Minimally invasive colon resection (laparoscopic colectomy). Surg Laparosc Endosc 1: 144150 5. Lacy AM, Garcia-Valdecasas JC, Pique JM, Delgado S, Campo E, Bordas JM, Taura P, Grande L, Fuster J, Pacheco JL (1995) Short-term outcome analysis of a randomized study comparing laparoscopic vs open colectomy for colon. Surg Endosc 9(10): 11011105 6. Liberman MA, Phillips EH, Carroll BJ, Fallas M, Rosenthal R (1996) Laparoscopic colectomy vs traditional colectomy for diverticulitis. Outcome and costs. Surg Endosc 10: 1518 7. Phillips EH, Rosenthal RJ (1995) Nomenclature in laparoscopic colon surgery. In: Phillips EH, RJ Rosenthal (eds) Operative strategies in laparoscopic surgery. Springer-Verlag, New York, pp 215218 8. Puente I, Sosa JL, Utpal Desai BS, Sleeman D, Hartmann R (1994) Laparoscopic treatment of colovesical fistulas: technique and reports of two cases. Surg Laparosc Endosc 4: 157160 9. Reissman P, Salky BA, Pfeifer J, Edye M, Jagelman DG, Wexner SD (1996) Am J Surg 171(1): 4750 10. Schoetz DJ (1993) Uncomplicated diverticulitis. Indications for surgery and surgical management. Surg Clin North Am 73: 965974 11. Stabile BE, Puccio E, vanSonnenberg E, Neff CC (1990) Preoperative percutaneous drainage of diverticular abscesses. Am J Surg 159: 99 105 12. Standard Task Force American Society of Colon and Rectal Surgeons (1995) Practice parameters for sigmoid diverticulitis-supporting documentation. Dis Colon Rectum 38: 126132 13. Wexner SD, Cohen SM (1995) Port site metastases after laparoscopic colorectal surgery for a cure of malignancy. Br J Surg 82: 295298 14. Wexner SD, Yoansen OB (1992) Laparoscopic bowel resection advantages and limitations. Ann Med 24(2): 105109

Surg Endosc (1997) 11: 711713

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The effects of sterile or infected bile and dropped gallstones in abdominal adhesions and abscess formation
zs G. R. Soybir,1 F. Ko ksoy,1 C. Polat,1 A. O eker,1 O. Yalc ln,1 Y. Aker,2 C. Topuzlu3
1 2

Department of Surgery, Istanbul Taksim State Hospital, Siraselviler Cad., 80060 Istanbul, Turkey Department of Surgery, Celal Bayar University, 45000 Manisa, Turkey 3 Oncology Institute, Istanbul University, 34390 Istanbul, Turkey Received: 29 March 1996/Accepted: 26 August 1996

Abstract Background: The effects of gallstones and sterile or infected bile on postoperative adhesions and abscess formation were investigated in Sprague Dawley rats. Methods: The first three groups were injected intraperitoneally with serum saline, sterile bile, or infected bile, respectively. Laparotomy was adjusted to the next seven groups. Serum saline, sterile bile, and infected bile were injected in the fourth through sixth groups intraperitoneally, respectively. Gallstones were placed intraabdominally to the seventh through ninth groups. The injections of sterile bile in group 7 and of infected bile in group 8 were added to laparotomies. Only laparotomy was carried out in group 10. Results: Abscess formations were seen only in infectedbile-injected groups. Significant adhesion formations were seen in infected-bile groups. Sterile bile and/or gallstones had no significant effect in abscess or adhesion formation. Conclusions: Results suggest that noninfected gallstones and sterile bile, even in combination, do not increase postoperative intraabdominal complications in rats. Key words: Gallstone Bile Adhesion Abscess Rat

of laparoscopic cholecystectomies from a center with considerable experience noted a bile leakage rate of 13% [8]. The grave consequences of peritonitis secondary to leakage of large volumes of bile have been well documented [6, 11]. However, the long-term effects of lesser volumes of bile and infected bile, or of lost gallstones on the peritoneal cavity, are unknown. The aim of this study was to evaluate the effects of sterile or infected bile alone and in combination with gallstone on the peritoneal cavity of the rat.

Materials and methods

One hundred six male Sprague Dawley rats weighing 250350 g were obtained from Charles River Breeding Laboratories (Germany). They were brought to our laboratory at least 1 week before use. All rats were housed in polypropylene cages on a 12-h light, 12-h dark cycle and in a temperature-controlled room. All animals were fed a standard laboratory chow diet and drinking water ad libitium. Animals were divided into ten groups consisting of ten or thirteen rats. The groups of animals studied are listed in Table 1. Groups 1 to 3 received an intraperitoneal injection (2 ml) of saline, sterile bile, and infected bile, respectively. Bile was obtained by the aspiration of choledoc of the patients with normal bile systems undergoing an abdominal operation. Bile was sterilized with a 0.2-m bacteria filter (Becton-Dickinson-Seitz). Sterility was confirmed by standard enrichment procedures performed on the day of injection. Infected bile was obtained by reinoculation of sterile bile with Escherichia coli to provide a concentration of 1 1067 colony-forming units (cfu)/ml. Cholesterol gallstones had been recovered from the patients who had undergone cholecystectomy for cholelithiasis and were sterilized. Under general anesthesia, groups 4 to 10 received a 46-mm lower midline incision. Sterile saline, sterile bile, and infected bile were injected intraperitoneally near the gallbladder of the rats in groups 4 to 6, respectively. Gallstones 2 to 3 mm in diameter were placed using a forceps into the right upper abdominal quadrant of the rats in groups 6 to 9. Groups 7 and 8 received an intraperitoneal injection (2 ml) of sterile bile and infected bile, respectively. An empty forceps was passed to the right upper quadrant of the rats in group 10. The wounds in groups 4 to 10 were closed with two layers of 4/O polyglactin 910. At the end of the experiment all animals were sacrificed by cervical dislocation at 4th weeks. The abdomen was opened through a full-length incision slightly to the right of the previous midline incision. The stones were located in all cases, intraabdominal abscess and lesions were noted,

Cholecystectomy is one of the most common operations in general surgery and much of the literature about this operation suggests that it is associated with a relatively high incidence of septic complications [13]. Bile and gallstone spillage at the time of laparoscopic cholecystectomy is a common problem that occurs in approximately 3240% of procedures [9, 10]. A recent report

Correspondence to: G. R. Soybir, Kaptanpas a Mah. Ziya Tu rkkan Cad., zu O mit B Blok D:56 Okmeydani, 80250 Istanbul, Turkey

712 Table 1. Groups and intraabdominal findingsa Groups 1 St/Saline 2 St/Bile 3 Inf.Bile 4 Lap+St.Saline 5 Lap+St.Bile 6 Lap+Inf.Bile 7 Lap+GS+St.Bile 8 Lap+GS+Inf.Bile 9 Lap+GS 10 Lap
a b

n 10 10 10 10 10 13 10 10 10 13

Mean adhesion scoreb 1.00 0 1.20 0.42 2.50 1.08 1.20 0.42 1.20 0.42 2.61 1.04 1.80 1.13 3.90 1.19 1.60 0.96 1.23 0.43

Abscess formation 1 2 2

St: sterile, Inf: infected, Lap: laparotomy, GS: gallstone Adhesions were scored with Mazuji Classification [7]

Table 2. Comparison of the adhesion scores of groupsa Groups St.Saline St.Saline St.Bile Lap+St.Saline Lap+St.Saline Lap+St.Bile Lap+GS+St.Bile Lap+GS+Inf.Bile Lap+GS+St.Bile Lap+Inf.Bile
a b

Median St.Bile Inf.Bile Inf.Bile Lap+St.Bile Lap+Inf.Bile Lap+Inf.Bile Lap+GS Lap+GS Lap+GS+Inf.Bile Lap

pb 0.2368* 0.0003* 0.023 1.0 0.0069* 0.0052 1.0 0.0052 0.0198 0.0052*

%95 CI 0.48 to 0.08 0.78 to 2.21 0.50 to 2.09 0.39 to 0.39 0.68 to 2.14 0.68 to 2.14 0.79 to 1.19 1.27 to 3.32 1.00 to 3.19 0.73 to 2.03

1.5 1 2 1 2 3

St: sterile, Inf: infected, Lap: laparotomy, GS: gallstone, CI: confidence interval * Fisher exact test, Median 2 test

and culture studies were performed for abscesses. The adhesions were evaluated according to the Mazuji Classification [7]. Mean scores and the standard deviations were calculated for each group. Statistical evaluations were conducted with the Statistical Package for Social Sciences (SPSS; Chicago, Illinois, USA). Statistical tests used included the median 2 and Fisher Exact tests.

Discussion Leakage of bile and gallstones may occur during laparoscopic cholecystectomy, especially when the gallbladder is being removed through the abdominal wall [5]. Bile and gallstones in the peritoneal cavity are associated with an increased risk of intraabdominal adhesion and abscess formation [8]. Authors have documented considerable morbidity and significant mortality in association with free intraabdominal leakage of large volumes of bile [3, 6]. Bile is deleterious when given intraperitoneally with E. coli but not when it is given alone or intravenously [12]. Toxic effects of bile salts absorbed into the systemic circulation [1], loss of fluid into the peritoneal cavity [5], and impairment of host defense mechanism(s) against bacteria [2] are the theories which have been proposed to explain the adverse effects of bile in the peritoneal cavity [1, 12]. Bile may diminish the local host defense by a dilution effect from peritoneal exudate accumulation [4]; or by impaired chemotactic, engulfing, or killing activities of peritoneal phagocytes; or by increased resistance of bile-treated bacteria to phagocytosis and/or killing; or by changes in the rate of absorption of bacteria from the peritoneal cavity [1]. Possibly, bile-induced accumulation of peritoneal fluid and the increased absorption of bacteria from the peritoneum into the blood may have contributed to the rapid development of abscess and adhesions. It is not clear whether (sterile or infected) bile or gallstones are responsible in postoperative complications. In

Results The results are summarized in Table 1. Intraabdominal abscess formation was seen in five rats and all of them were in infected-bile-injected groups (groups 3, 6, and 8). Abscesses were directly adjacent to the gallstones in the right upper abdomen and neatly enclosed in omentum. The bacteria grown in culture studies of abscesses were identical to that E. coli that had been inoculated, showing identical antibiotic susceptibility patterns. The mean adhesion scores according to the Mazuji Classification and standard deviations of all groups are listed in Table 1. All adhesions in the groups were in the right upper quadrant and between omentum, intestine, liver, and abdominal wall, and in those groups with gallstones they were adjacent to the gallstones. The results of Median 2 and Fisher Exact tests of the mean adhesion scores of the groups are summarized in Table 2. There were significant differences between groups of concerning infected bile injection and the other groups. The groups with sterile bile or laparotomy, or gallstones or combinations thereof did not show significant differences when compared to the groups excluding infected bile.

713

studying the effects of a gallstone and bile combination on postoperative adhesions and abscess formation, infected gallstone and culture studies of abscess were not taken into account [5]. In our study, in contrast to Johnstons study [5], it has been shown that sterile bile and gallstone without infection, alone or together, did not have any effect on postoperative adhesions and abscess formation. However, postoperative abscess formation and significant adhesions were seen in groups including infected bile. The present study may help to expand knowledge concerning dropped gallstones and bile spillage in laparoscopic cholecystectomy. However, the authors believe that new clinical studies are necessary in order to determine the applicability of this finding in patients. Although the relative importance of the factors discussed here remain speculative, the effect of infected bile alone or in combination with gallstone is certain in intraabdominal adhesions and possible in the formation of abscess. However, noninfected gallstones and sterile bile, even in combination, do not increase postoperative intraabdominal complications in rats. References
1. Andersson R, Tranberg KG, Bengmark S (1990) Roles of bile and bacteria in biliary peritonitis. Br J Surg 77: 3639

2. Conn JH, Chavez CM, Fain WR (1970) Bile peritonitis: an experimental and clinical study. Am Surg 36: 219224 3. Dale G, Solheim G (1975) Bile peritonitis in acute cholecystitis. Acta Chir Scand 141: 746748 4. Dunn LD, Barke LA, Ahrenholltz DH, Humphrey EW, Simmons RL (1984) The adjuvant effect of peritoneal fluid in experimental peritonitis: mechanisms and clinical implications. Ann Surg 199: 3743 5. Johnston S, Kieran OM, McEntee G, Grace P, Smyte E, BouchierHayes D (1994) The need to retrieve the dropped stone during laparoscopic cholecystectomy. Am J Surg 167: 608610 6. Kyle J, Carey L, Fallon RJ, (1989) Scientific foundations in surgery. 4th ed. Heimeman, Oxford, pp 608619 7. Mazuji K, Kalambahati K, Powar B (1964) Prevention of adhesions with polyvinylpyrolidone. Arch Surg 89: 10111015 8. Nathanson LK, Shimi S, Cuschieri A (1991) Laparoscopic cholecystectomy: the Dundee experience. Br J Surg 78: 155159 9. Peters JH, Gibbsons GD, Innes JT, Nichols KE, Front ME, Roby SR, Ellison EC (1991) Complication of laparoscopic cholecystectomy. Surgery 110: 769778 10. Soper NJ, Dunnegan DL (1991) Does intraoperative perforation influence the early outcome of laparoscopic cholecystectomy? Surg Laparosc Endosc 1: 156161 11. Thomas CG, Tindall B (1988) Medical microbiology. 16th ed. Churchill Livingstone, New York, pp 154155 12. Thoren L (1957) Experimental biliary peritonitis. Acta Chir Scand 113: 494496 13. Willis RG, Lawson WC, Hoare EM, Kingston RD, Sykes PA (1984) Are bile bacteria relevant to septic complications following biliary surgery. Br J Surg 71: 845849

Surg Endosc (1997) 11: 785788

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The knot before loop technique for the endoscopic ligation and suture-instrument set
Knot formation and application
G. Schaller,1 V. Paolucci,2 C. Gutt,2 B. Weber,1 B. C. Manegold3
1 2 3

Black Forest Endosurgery Institute, D-79110 Freiburg, Am Lusbu hl 32a, Freiburg, Germany Department of Surgery Prof. Enke, Johann Wolfgang Goethe, University of Frankfurt, Frankfurt, Germany Department of Surgical Endoscopy, Mannheim, University of Heidelberg, Heidelberg, Germany

Received 24 July 1996/Accepted: 12 December 1996

Abstract Background: In minimally invasive surgery intracorporal knot-tying is complicated by a limited field of vision and depth perception. Methods: The knot before loop technique aims to reduce intraabdominal movements in number and space needed. A grasping instrument 3 mm in diameter guides a slipfit hollow knot pusher with a notch to hold the thread, when extracorporally forming the knot on the instrument tip and an axial slot. The loop is finished under endoscopic vision, yet a second loop is created along the thread. The knot is tightened and secured by closing the second loop without troublesome instrument change. Results: The strength of the knot was tested and the feasibility of the instrument set was proven in pigs and 25 cholecystectomies and hernia repairs in humans. Conclusions: Endoscopic application of a secured slip knot is simplified by the knot before loop technique. The independent formation of the knot by the assisting personnel allows quick application, equivalent to the use of clips and staples. The benefit in cost saving is high. Key words: Endoscopic suture Modified knot pusher Knot before loop technique

Minimally invasive surgery enforced major changes in the surgical techniques. The degradation of task efficiency is caused by impaired visual perception and kinematic restriction [1]. This negatively effects intracorporeal knot tying, performed by rather complicated twisting and opposite traction. New technical developments suggest standardization
Correspondence to: G. Schaller

of intraabdominal manipulations and reduction of number and space needed [5, 7]. The extracorporal knot, however, is formed by hand, using well-trained skills. A pushing rod is then applied to internally place and tighten it. Gas leakage, distraction of attention from the operational field, and harmful traction on vulnerable internal structures are negative [4]. Recent publication demonstrated differences in knot strength and security with laparoscopic procedures. The best known extracorporeally tied Roeder slipknot was failed at less than 30 N because of slippage [3]. Yet different kind of knots, e.g., surgical square knot, and additional half hitches improved the reliability of these knots significantly [2]. Pretied Roeder knots and metal or resorbable clips and staples provided by industry at high cost limit future surgical capabilities [6]. In our analysis we recognized two logically different steps. A knot creates by definition a high friction between the ends of threads. Thus the loop around a vessel or through tissue is held in place. Defining the ideal technique, we stated that the knot must be formed extracorporeally by hand prior to the intraabdominal formation of the loop as a second step. Combining grasping instrument and slipfitting knot pusher allows an easier extracorporeal formation of a knot. To improve the strength of the slipknot three loops are created. With the combined instrument the knot of a ligation or suture can be immediately closed and tightened intracorporeally without troublesome instrument change. Slightly modifying the technique, the slipknot can be fixed by a fourth loop. Furthermore, the knot before loop technique allows one to form a knot on a second instrument independently from the operational steps.

Materials and methods

The instrument set comprises a common needle driver 3 mm in diameter and 360 mm in length. A hollow, movable knot pusher is guided on the

786

Fig. 1. Hingeless needle driver, OD 3 mm, with knot attached; hollow, slipfitting knot pusher, OD 5 mm; instrument set in trocar; knot training device. Fig. 2. AD Extracorporeal formation of the knot (hand actions, gray lines). Fig. 3. Suture with needle and knot ready for intraabdominal insertion.

grasping instrument, when it is introduced internally. The instrument protrudes by 10 mm to attach and hold the extracorporeally formed knot. The insertible length of the instrument set is 320 mm, the outer diameter 5 mm (Fig. 1). At the knot pusher, a notch facilitates the knot formation and an axial slot allows to retract even larger needles. To extracorporeally form the knot the active end of the thread, eventually carrying a needle, and the instrument itself are held with the left hand. The other, passive end is crossed over the instrument shank, coiled around it one time, and simultaneously fixed at the notch (Fig. 2A, hand actions, gray lines). Then the instrument and the active end are held with the right hand to form a loop and two fingers of the left hand are introduced therein. With a right turn of the hand the loop is placed over the instrument tip (Fig. 2B, hand actions, gray lines). Immediately the passive end is grasped and another loop is formed, guided through the first, and placed over the instrument tip (Fig. 2C, hand actions, gray lines). The resulting knot is fixed on the tip of the instrument and the passive end is cut to 10 mm (Fig. 2D). Then the needle is grasped and the instrument is ready for intraabdominal insertion (Fig. 3).

Under endoscopic vision tissue is sutured to achieve approximation or closure. Then an additional loop is created along the active end of the thread, grasping it as far as possible away from the needle. The knot is pushed forward from the instrument on this loop; it is kept fairly loose so it can easily slip (Fig. 4A, instrument actions, gray lines). By pulling on the passive end the knot is approximated (Fig. 4B, instrument actions, gray lines) and tightened with the knot pusher (Fig. 4C). A ligation is created similarly (Fig. 4Ba, instrument action, gray lines). After guiding the instrument through the persisting loop, the passive end is grasped and the loop is closed by pulling at the active end (Fig. 4D, instrument actions, gray lines). Now the slipping knot is additionally secured. The above is visualized with the instrument in a model (Fig. 5AE). Ligations can be placed and later be closed at any time (Fig. 6).

Results The feasibility of the instrument set and the strength and the security of the knot were tested with a tensiometer. Con-

787

Fig. 4. AE Application of ligation and suture (instrument actions, gray lines).

Fig. 5. AE Application of the suture in a model.

788

Fig. 6. Balloon-like ligation can be placed and then be closed at any time.

ventional thread materials (Vicryl, Dexon, Maxon, Seralon, Mersilk) were tested with 3/0 monofilament and braided materials. With the additional fourth loop no slippage occurred and knots broke at forces between 30 N to 50 N, equivalent to the strength of knots tied in open surgery. During applications in five pigs the operational steps were practiced and the sutures were judged after sacrifying the animals. In humans 25 cholecystectomies and hernia repairs were performed. The extracorporal formation of the knot was learned in 30 min. Initially the intracorporal application required 65 s; this was reduced to 25 s after 4 months. The intraabdominal manipulations were judged to be easy compared to known techniques: This is supported by reduction of operational steps and space needed. The technique proved superior in overcoming tension between tissues to be approximated. The possibility of forming additional knots independently from the surgeons work on a second instrument by the assisting personnel compensates for the longer duration of the proceduresup to 20 min in hernia repaircompared to those with the use of single loaded clips or staples. The use of conventional suture materials reduced the cost of the hernia repair by a factor of nine ($225 to $24)significantly. Discussion Even though no statistical data are available, many surgeons seem to avoid endoscopic suturing with conventional thread

materials. Metal and resorbable clips and staples along with multifire equipment are more convenient. The resulting lack of surgical skills and the high cost with the endoscopic procedures are most negative effects. After analyzing the known knotting techniques, we developed a new instrument set for the knot before loop technique. The knot is formed easily extracorporally and is attached at the tip of the inner instrument. A special hollow slipfitting knot pusher is guided thereon. After intracorporal insertion the loop is formed by suturing tissue. Then the knot is forwarded on an additionally created loop, approximated, tightened by the knot pusher, and immediately secured by an additional loop. Tests proved the high security of the knot. The quick formation and application of the knot by surgeons and assisting personnel suggest advantages of the new instrument set. As an economical benefit all conventional suture material can be used. More time is needed for the procedures compared to the use of clips and staples, especially in hernia repair. Intensive training, however, will compensate this negative effect. Further reduction in size of the instrument set to 3 mm outer diameter allowed suturing in endoscopic training in rats, suggesting, e.g., future application in heart surgery. References
1. Crosthwaite G, Chung T, Dunkley P, Shimi S, Cuschieri A (1995) Comparison of direct vision and electronic two- and three-dimensional display on surgical task efficiency in endoscopic surgery. Br J Surg 82: 849851 2. Dorsey JH, Sharp HT, Chovan JFD, Holz PM (1995) Laparoscopic knot strength: a comparison with conventional knots. Obstet Gynecol 86: 536540 3. Kadirkamanathan SS, Shelton JC, Hepworth CC, Laufer JG, Swain CP (1996) A comparison of the knot strength of knots tied by hand and at laparoscopy. J Am Coll Surg 182(1): 4554 4. Licici MM, Melzer A, Taddeo FS, Buess G, Angelini L (1994) Tissue approximation in minimal invasive surgery. End Surg 2: 4754 5. Nathanson LK, Easter DW, Cuschieri A (1991) Ligation of the structures of the cystic pedicle during laparoscopic cholecystectomy. Am J Surg 161: 350354 6. Nelson MT, Nakashima M, Mulvihill SJ (1992) How secure are laparoscopically placed clips? An in vitro and in vivo study. Arch Surg 127: 718720 7. Szabo, Z, Hunter J, Berci G, Sackier J, Cuschieri A (1994) Analysis of surgical movements during suturing in laparoscopy. End Surg 2: 5561

Surg Endosc (1997) 11: 758761

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Application of rectal stents for palliation of obstructing rectosigmoid cancer

M. Dohmoto, M. Hu nerbein, P. M. Schlag
Department of Surgery and Surgical Oncology, Robert-Ro ssle Hospital, Humboldt University, Berlin, Lindenbergerweg 80, 13122 Berlin, Germany Received: 10 May 1996/Accepted: 11 November 1996

Abstract Background: The rationale of palliative endoscopic treatment is to avoid a colostomy in patients with advanced disease and limited life expectancy. This study was conducted to evaluate the role of endoscopic stent implantation for palliation of obstructing rectal cancer. Methods: Overall, 19 patients (aged 4787 years) with nonresectable or metastatic rectal cancer were treated by stent insertion after laser recanalization or dilation. Three types of stents, i.e., plastic tubes (n 8), self-expanding mesh stents (n 6), and endocoil stents (n 5), were used to maintain luminal patency. Results: Endoscopic stent implantation was successfully performed in all 19 patients. Long-term luminal patency and satisfactory bowel function were achieved in 16 of 19 patients (84%). After a median follow-up of 6 months, eight of the patients have died and eight are still alive without evidence of recurrent obstruction. Dislocation of the endoprosthesis occurred in two of eight plastic tubes and one of five mesh stents. Recurrent obstruction due to tumor ingrowth was only observed in patients treated with self-expanding mesh stents (n 2). In spite of reinsertion and laser therapy a colostomy was required in three of 19 patients. There was no evidence of treatment failure in five patients who received endocoil stents. None of the patients experienced serious complications related to the endoscopic procedure. Conclusions: Endoscopic stent implantation seems to be a safe and efficient palliative approach to selected patients with obstructing rectal cancer. Currently, self-expanding coil stents are superior to other devices because of lower risk of dislocation and tumor ingrowth. Key words: Endoscopic palliation Rectal cancer Selfexpanding metal stent Endoprosthesis

Radical resection of rectal cancer is not feasible in 3040% of the patients because of distant metastases, extensive local tumor infiltration, fatal medical conditions, or refusal of the patient [1, 6]. These patients require palliative treatment to improve symptoms of their disease and to prevent livethreatening complications such as obstructive ileus. Although palliative resection provides excellent control of the symptoms of rectal cancer, even limited surgery may be associated with a considerable morbidity and mortality [8, 10]. High-dose radiotherapy can produce local control of the disease in 6070% of the patients. However, the response is slow and the complications increase with dose [11, 17]. Endoscopic treatment, including laser photocoagulation, endoscopic transanal resection, electrocoagulation, and cryotherapy, has been used increasingly for palliative treatment of patients with advanced rectal cancer. These techniques allow rapid and efficient tumor destruction and are associated with low morbidity and mortality. Furthermore, endoscopic treatment can be performed on an outpatient basis, which is another important determinant to the quality of life in cancer patients with limited life expectancy. The aim of this study was to investigate the value of endoscopic stent implantation for palliative treatment of advanced rectal cancer. For this purpose three types of stents, i.e., plastic tubes, self-expanding mesh stents, and endocoil stents, were evaluated.

Materials and methods

Endoscopic implantation of endoprosthesis was performed in 19 patients with obstructing rectal cancer. Endoscopic treatment was only offered to patients with contraindications to surgery. Furthermore, patients were selected according to the localization and the tortuosity of the tumors. Stent implantation was abandoned in patients with very low rectal cancer or distorted tumors in the rectosigmoid flexion. In the first patients, modified plastic tubes (Wilson Cook, USA) were used to maintain luminal patency. In later patients two types of selfexpanding metal stents (SEMS) were favored because of the obvious advantages of these devices. Self-expanding metal stents were available as

Correspondence to: P. M. Schlag

759 Table 1. Patient data and treatment results of palliative treatment of rectal cancer with rectal endoprosthesisa Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
a

Sex M W M W M M W W M M W M M W M W W M W

Age 75 82 84 87 72 58 68 52 72 58 47 87 75 84 65 61 66 67 74

Stenosis (cm) 1017 512 59 410 710 510 1522 816 1218 1519 1019 817 914 913 1115 1320 1020 711 1519

Pretreatment Dilation Laser Laser Laser Laser Laser Laser Laser Laser Dilation Laser Laser Dilation Dilation Dilation Dilation Laser Dilation Dilation

Prosthesis Plastic Plastic Plastic Plastic Plastic Plastic Mesh Plastic Mesh Plastic Mesh Mesh Mesh Endocoil Endocoil Endocoil Mesh Endocoil Endocoil

Length (cm) 9.4 9.4 6.4 6.4 4.6 4.6 10 11 10 9.4 15 10 7 10 10 10 15 10 10

Complication Dislocation Pyrexia Dislocation Restenosis Occlusion Dislocation Occlusion

Follow-up (weeks) 12, + 8, + 40, + 80, + 40, + 16, + 28, + Surgery 20, + Surgery Surgery 8 16 24 28 28 4 * *

Follow-up examination not yet performed

nitinol mesh stent (Ultraflex, Boston Scientific, Watertown, USA) with a diameter of 1.8 cm or as nickel titanium coil stent (Endocoil, Instent, Eden Prairie, USA) with a diameter of 2.4 cm. The endocoil stent consists of a biocompatible flat wire which is wound in a coiled configuration. The stent contains a central section of adjacent coils and two flared ends 2 cm long. Both types of SEMS were mounted on thermoplastic delivery catheters with tapered, flexible tips and a maximum diameter of 10 mm. All stents were available in various lengths ranging from 7 to 15 cm. Prior to the examination, the rectum was cleaned by a phosphate enema. All endoscopic procedures were performed under mild intravenous sedation with midazolam (35 mg). Before stent insertion all patients were treated either by NdYAG laser photocoagulation (n 11) or mechanical dilatation (n 8) to obtain an appropriate lumen for the stent. The implantation technique of plastic tubes is comparable to that in the upper GI tract and has been described earlier [4]. Self-expanding metal stents were introduced as follows: First a guidewire was placed beyond the stenosis under fluoroscopic control. Then the endoscope was passed over the stenosis. Using fluoroscopy the proximal and distal borders of the stricture were marked with radiopaque markers on the abdomen of the patient. The length of the narrowed segment was measured when the endoscope was withdrawn. The appropriate length of the stent was calculated by adding 56 cm to the length of the stenosis. The lubricated delivery catheter was advanced over the guidewire until the stent was positioned across the stricture. If correct position of the stent was demonstrated by endoscopy and fluoroscopy, the self-expanding stent was released. After expansion of the stent, luminal patency was confirmed by passing the endoscope through the stent. Routine follow-up endoscopy was scheduled 46 weeks after the implantation, or earlier if complications occurred.

Results Endoscopic stent insertion was applied to 19 patients not suitable for palliative resection. The mean age of the patients was 70 years (range: 4787 years). All patients had symptoms of obstruction including abdominal distension with pain and constipation. The median diameter of the residual lumen was 6 mm and only eight of the tumors allowed passage of the coloscope beyond the stenosis. Indications for endoscopic stent implantation included nonresectable disease (n 4), metastatic disease and a high surgical risk (n 12), or refusal of the patient to undergo surgery (n 3). Three types of stents, including plastic tubes (n 8), self-expanding mesh stents (n 6), and

endocoil stents (n 5), were evaluated for the treatment of malignant rectal obstruction. The patient data and treatment results are listed in Table 1. Placement of the stents was successful in all patients at the first attempt after laser vaporization or mechanical dilation. The implantation of celestin tubes proved to be more difficult compared to selfexpanding metal tubes because of the rigidity and the larger diameter of the plastic tubes. After implantation some of the patients experienced transient dull lower abdominal pain presumably caused by the stiffness of the plastic tubes. Insertion of mesh and endocoil stents was comparably easy and safe due to the flexibility and the small diameter (10 mm) of the delivery catheter (Figs. 1 and 2). After stent insertion all patients experienced a significant relief of obstructing symptoms resulting in almost normal defecation if laxative was administered. In most patients luminal patency was maintained until death without further treatment. The median follow-up of the patients is 6 months. In the meantime eight of the patients have died (median survival 7.7 months) without signs of recurrent obstruction. Eight patients are still alive (median survival 4.5 months) with sufficient luminal patency. Obstruction of the stent by tumor ingrowth was observed in two patients with mesh stents. Dislocation occurred in two of eight patients with plastic tubes, probably due to the smooth surface of the endoprosthesis. Migration of mesh stents was noticed in one of six patients. Recanalization of the lumen by laser vaporization or stent reinsertion was possible in two patients, whereas three patients had to undergo surgery. No serious complications were observed related to the endoscopic procedure. Temporary incontinence was observed in two patients, which subsided after several days. Abdominal discomfort due to obstruction and distension was markedly improved in all patients. However, stent implantation had no influence on tumor-related pain. None of the patients experienced increased long-term pain after stent implantion. After the procedure most patients were treated with oral laxatives and the bowel movement frequency usually ranged from 2 to 5 per day.

760

Fig. 1. Radiograph showing full expansion of a self-expanding mesh stent in stenotic rectal cancer.

Discussion Presently Nd:YAG laser therapy is considered the treatment of choice for endoscopic palliation of advanced rectal carcinoma. Consistently, success rates of 8595% have been reported for palliative laser photocoagulation of obstructing rectal carcinoma [9]. However, a major disadvantage of laser therapy is that multiple treatment sessions are necessary to maintain luminal patency after initial recanalization. In order to improve the quality of life of cancer patients with limited life expectancy an endoscopic approach which obviates the need for repeated attendance of patients in hospital seems desirable. Recently, Steele et al. have reported encouraging results for transanal endoscopic microsurgery (TEM) of rectal tumors [16]. Complete resection was achieved in 70 of 77 adenomas and 22 of 23 carcinomas. The postoperative morbidity related to the procedure was low. Although this technique seems most appropriate for the treatment of adenoma it should also be considered in patients with rectal cancer who are high-risk candidates for major surgery. Although endoscopic implantation of stents is a generally accepted method for the treatment of malignant esophageal and biliary strictures, there is only limited experience with this technique in rectal cancer [5, 14]. To our knowledge we were the first to report that implantation of modified plastic tubes is also feasible for palliative treatment of obstructing rectal cancer [2, 3]. Recently, on the basis of case reports, encouraging results have been reported for the treatment of stenotic rectal carcinoma by endoscopic placement of self-expanding metal stents [7, 12, 15]. Rey et al. used self-expanding mesh stents for palliative treatment of 12 patients with advanced rectosigmoid cancer [13]. Stent insertion was successful in 11 patients and the median time before reobstruction by tumor ingrowth was 9.7 months. Migration of the stent was observed in three patients. In this study, three types of stents were compared for treatment of rectal cancer. Stent insertion was technically feasible in all 19 patients. Initially modified celestine tubes were used in

Fig. 2. Obstructing rectal cancer: after implantation of an endocoil stent luminal patency is demonstrated by unimpeded flow of contrast agent through the stent.

eight patients with some success. However, the rather difficult implantation technique and migration of tubes in two patients discouraged further application of plastic tubes. Therefore, in later patients SEMS were favored. Placement of self-expanding stents was comparably easy due to the flexibility and the small diameter of the delivery system. Mild sedation was sufficient for the procedure and the stents were well tolerated by the patients. Only few patients complained of transient pain caused by the rapid and powerful release of the endocoil stents. Restoration of luminal patency and symptomatic relief were provided in all of the 11 patients, although some mesh stents required 12 h for full expansion. The great elasticity of mesh stents proved to be especially advantageous for angular tumors in the rectosigmoid flection. One major drawback of mesh stents was recurrence of obstruction by tumor growth between the meshes, which occurred in two of six patients. In one of these patients laser vaporization allowed recanalization of the lumen, whereas a colostomy became necessary in the other patients. Neither obstruction by tumor ingrowth nor dislocation was observed in any of the endocoil stents. This must be attributed to the large diameter (24 mm) and the configuration of these devices. The endocoil stent consists of a flat titanium wire with a central section of adjacent coils and two flared ends. The central portion prevents tumor penetration while the strong radial expansion force of the ends warrants strong fixation of the stent. Currently the endocoil stent seems to be most appropriate for rectal use due to the large diameter and the low risk of occlusion or dislocation

761

(Table 2). Overall a colostomy was avoided in 16 of 19 patients (84%) with obstructing rectal cancer. None of the patients developed serious complications related to the procedure. One major disadvantage of self-expanding metal stents is the tenfold higher price compared to plastic tubes. Our encouraging preliminary results and the absence of serious complications justify further evaluation of endoscopic stent placement for palliation of rectal cancer. It seems likely that treatment results of this method can be further improved by technical advances, e.g., silicone-covered SEMS and the development of specially designed rectal stents [18]. However, controlled trials will be required to assess the value of stent implantation for palliation of rectal cancer compared to simple obliteration by laser therapy or transanal resection. References
1. Baigrie RJ, Berry AR (1994) Management of advanced rectal cancer. Br J Surg 81: 343352 2. Dohmoto M (1991) Endoscopic implantation of rectal stents in palliative treatment of malignant stenosis (in Japanese). Endosc Dig 3 (11): 15071512 3. Dohmoto M, Rupp KD (1991) Endoskopische Stent Implantation als Palliation beim Rektumkarzinom. Dtsch Med Wochenschr 115: 943 4. Dohmoto M, Rupp KD, Hohlbach G (1992) Neue endoskopische Verfahren zur Palliation beim Rektumcarcinom. In: Anonymous (ed) Jahrbuch der Chirurgie. Biermann, Verlag, pp 109115 5. Ell C, Hochberger J, May A, Fleig WE, Hahn EG (1994) Coated and uncoated self-expanding metal stents for malignant stenosis in the upper GI tract: preliminary clinical experiences with Wallstents. Am J Gastroenterol 89: 14961500

6. Gordon NL, Dawson AA, Bennett B, Innes G, Eremin O, Jones PF (1993) Outcome in colorectal adenocarcinoma: two seven-year studies of a population. BMJ 307: 707710 7. Itabashi M, Hamano K, Kameoka S, Asahina K (1993) Self-expanding stainless steel stent application in rectosigmoid stricture. Dis Colon Rectum 36: 508511 8. Lau PWK, Lorentz TG (1993) Results of surgery for malignant bowel obstruction in advanced, unresectable, recurrent colorectal cancer. Dis Colon Rectum 36: 6164 9. Mathus-Vliegen EM, Tytgat GN (1986) Laser photocoagulation in the palliation of colorectal malignancies. Cancer 57: 22122216 10. Ma kela J, Haukipuro K, Laitinen S, Kairaluoma MI (1989) Surgical treatment of recurrent colorectal cancer. Arch Surg 124: 10291032 11. Overgaard M, Overgaard J, Sell A (1984) Dose response relationship for radiation therapy of recurrent, residual and primarily inoperable colorectal cancer. Radiother Oncol 1: 217225 12. Raijman I, Siemens M, Marcon N (1995) Use of an expandable ultraflex stent in the treatment of malignant rectal stricture. Endoscopy 27: 273276 13. Rey JF, Romanczyk T, Greff M (1995) Metal stents for palliation of rectal carcinoma: a preliminary report. Endoscopy 27: 501504 14. Spinelli P, Cerrai FG, Mancini A, Meroni E, Pizzetti P (1991) Esophageal intubation for malignant fistulas. Surg Endosc 5: 127129 15. Spinelli P, Dal Fante M, Mancini A (1992) Self expanding mesh stent for endoscopic palliation of rectal obstructing tumors: a preliminary report. Surg Endosc 6: 7274 16. Steele RJC, Hershman MJ, Mortensen NJM, Armitage NCM, Scholefield JH (1996) Transanal endoscopic microsurgery. Initial experience from three centres in the United Kingdom. Br J Surg 83: 207210 17. Taylor RE, Kerr GR, Arnott SJ (1987) External beam radiotherapy for rectal adenocarcinoma. Br J Surg 74: 455459 18. Wu WC, Katon RM, Saxon RR, Barton RE, Uchid BT, Keller F, Ro sch J (1994) Silicone covered self expanding metallic stents for the palliation of malignant esophageal obstruction and esophagorespiratory fistulas. Gastrointest Endosc 40: 2233

Case reports
Surg Endosc (1997) 11: 766768

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic treatment of a benign splenic cyst

G. J. Sellers, P. M. Starker
Columbia Presbyterian Medical Center, Department of Surgery, 161 Fort Washington Avenue, New York, NY 10032, USA Received: 29 May 1996/Accepted: 26 July 1996

Abstract. With an understanding of the spleens important immunologic function, splenectomy for benign splenic disorders has given way to a variety of splenic conservation techniques. Treatment options for benign nonparasitic splenic cysts include partial splenectomy, total cystectomy, or partial cyst decapsulation. External cyst wall decapsulation is a simplified operative procedure that carries no increased risk of cyst recurrence. However, a conventional upper abdominal laparotomy may subject patients to significant morbidity. We successfully performed a laparoscopic partial cyst decapsulation, achieving meticulous hemostasis with use of a laparoscopic-GIA stapling device. The patient tolerated the procedure well and was discharged on postoperative day 2. Follow-up has demonstrated no evidence of recurrent cyst formation. Key words: Splenic cyst Splenic pseudocyst Management Laparoscopic treatment Decapsulization

technique of splenic cyst decapsulation with improved hemostatic control using the laparoscopic approach. Patient and operative technique Patient
A 34-year-old female, in otherwise-excellent health, presented with vague abdominal pain and early satiety. Physical exam revealed a fullness in the left upper quadrant. Computerized tomography revealed a large cystic mass within the spleen measuring approximately 12 cm (Fig. 1). Given the patients symptoms and the size of the splenic cyst, she agreed to surgical cyst decapsulation via a laparoscopic approach.

Surgical technique
Preoperatively the patient received Pneumovax vaccine. Standard preoperative maneuvers included preoperative antibiotics and placement of Venodyne pneumatic compression devices on both lower extremities. Under general anesthesia and endotracheal intubation the abdomen was prepared and draped in standard sterile fashion. An 11.5-mm laparoscopic port was placed in the infraumbilical position using an open cut-down technique with direct visualization of the peritoneum prior to insertion. A 1015 mmHg pneumoperitoneum was created, followed by insertion of a 0 laparoscope connected to a Stryker video system. A thorough inspection of the abdomen revealed a large cyst located along the superiolateral aspect of the spleen. The left lobe of the liver was overhanging the medial portion of the spleen. Three additional ports were inserted under direct vision. An 11.5-mm port was placed parallel to the umbilical port, along the anterior auxillary line. A 5-mm port was placed midline, approximately 5 cm cephaled to the umbilical port. A third port, 11.5 mm, was placed to the right of midline within the right upper quadrant. The 0 laparoscope was replaced with a 30 scope for better visualization within the left upper quadrant. The cyst was decompressed via a trocar drainage catheter inserted into the midportion of the cyst. Approximately 2,000 cc of greenish-brown cloudy fluid was evacuated and sent for cytology and culture. A grasping trocar and endo-shears were used to open the extrasplenic cyst from the site of needle drainage down to the base. The cyst was without septation; however, the inner lining of the capsule was trabeculated. Frozen-section examination of a portion of cyst wall revealed benign tissue. An Endo GIA stapling device (Autosuture Division, US Surgical Corp.) was fired repeatedly along the margin of the extrasplenic cyst wall. Care was taken to avoid resecting normal-appearing splenic tissue. The superior margin of the cyst was adherent to the left hemidiaphragm, and the medial margin to the left liver lobe. This portion of the cyst wall was

Benign splenic cysts are an infrequently encountered entity. With an understanding of the spleens important immunologic role, especially in the bodys defense against encapsulated bacteria, splenic conservation techniques for the treatment of benign splenic disorders and splenic trauma is now the goal of surgical treatment. However, because of the large size of most splenic cysts, with risk of intraoperative [14] or late postoperative hemorrhage [15], especially from an epidermoid cyst, splenectomy has until recently been the procedure of choice [21]. Recent reports have appeared which demonstrate the safety and success of alternative techniques such as partial splenectomy or cyst decapsulation [1315, 17, 21]. Laparoscopic approaches to the surgical treatment of splenic disorders have produced an added benefit in reducing postoperative patient discomfort and shortened hospitalization while achieving the same technical success as an open approach [4, 16, 20]. We report a

Correspondence to: G. J. Sellers

767

Fig. 1. Abdominal CT scan demonstrating splenic cyst.

mobilized by dividing fibrous bands between endo-clips and using cautery as appropriate. The entire extrasplenic cyst was thus resected, extracted via the umbilical port, and sent for pathology. At the conclusion of the resection the spleen was inspected for hemostasis. Greater omentum was packed into the intrasplenic cyst cavity. The three working ports were removed and the 11.5-mm fascial defects were closed with 0 Vicryl suture. The umbilical port was removed, the pneumoperitoneum was released, and the infraumbilical fascia was closed in a similar manner via sutures placed during the initial cut-down. The skin sites were closed in a subcuticular fashion using a 4-0 Monocryl suture. The patient tolerated the procedure well and there were no complications. Total duration of the procedure was approximately 90 min. Pathologic examination revealed a wall composed of both fibrous tissue and epithelial elements, consistent with a true cyst.

Discussion Although accurate information on the incidence is unknown, benign splenic cysts are thought to be an uncommon entity. The differential diagnosis for splenic cysts includes cystic neoplasms, benign cysts, and abscess. Benign cysts are subdivided into parasitic or nonparasitic. Worldwide, echinococcal and malarial parasites are more common; however, in Europe and North America parasitic cysts are rare. Nonparasitic cysts are further divided, based upon historical classification schemes, as epithelial-lined true cysts and fibrous-lined pseudocysts. Epithelial cysts represent 1025% [11, 12], while pseudocysts represent 5075% of all splenic cysts [8, 14, 18]. Splenic cysts present with signs and symptoms of local compression of adjacent structures (local or referred pain, early satiety) or are found incidentally during physical or radiologic exam done for other reasons. Cysts are also discovered after presentation for the complications of rupture, infection, or intracapsular hemorrhage [14]. Both epitheliallined and pseudocysts have been found in all age groups; however, epithelial cysts are thought to be more common in younger patients.

While parasitic cysts can often be identified due to their characteristic radiographic/sonographic appearance, preoperative differentiation between epithelial-lined and pseudocysts is not reliable [5]. Currently, controversy surrounds the etiology of splenic cysts. Current classification schemes are modifications from Fowlers original work in the 1940s [7]. These classification schemes differentiate benign splenic cysts based upon the presence or absence of an epithelial cellular lining within the cyst wall [8, 11, 12]. Bostick and Lucia proposed that epithelial cysts were developmental in origin, arising from incorporated mesothelial elements during organogenesis with subsequent squamous metaplasia [1]. Pseudocysts were thought to be the result of trauma to the spleen with intracapsular hemorrhage and organization of a fibrous cyst wall. Importantly, both epithelial and fibrous elements are often seen histologically within cyst wall specimens from patients who give a history of trauma [23]. With the advent of electron microscopy and immunohistochemical techniques the classification schemes based upon histologic divisions as well as the theorized mechanisms of pathogenesis have been obscured [2, 6, 10]. Indications for surgical treatment of splenic cysts include all symptomatic cysts and cysts larger than 5 cm [22], as spontaneous resolution is less likely. Treatment goals include elimination of the cyst and prevention of recurrence. Alternatives to splenectomy include partial splenectomy, cystectomy, or partial cyst decapsulation. More conservative treatments including percutaneous drainage with ultrasound or CT guidance have not been successful in preventing recurrence, infection, or bleeding [3]. Salky et al. in 1985 report the first laparoscopic treatment of a splenic cyst. This group decompressed a splenic cyst via a laparoscopic trocar puncture followed by creation of a cyst-peritoneal window [19]. Kum et al. in 1993 report a laparoscopically assisted splenectomy for a wandering splenic cyst [9]. Posta in 1994 reports on a splenic cyst treated laparoscopically with aspiration of the cyst contents and excision of a portion of external cyst wall [16]. Targarone et al. in 1995 report on a similar laparoscopic approach to treatment of a splenic cyst [20]. Our modification on previously reported cases is the use of a laparoscopic GIA device. Use of this device reduces the risk of intraoperative and postoperative bleeding from the resected cyst wall, especially in a vascularized epithelial lined cyst. The spleen is well visualized laparoscopically. The avoidance of an upper abdominal incision decreases the postoperative pain and discomfort for the patient and shortens length of hosptialization. In choosing an appropriate surgical procedure of low morbidity/low risk the laparoscopic approach provides safe, definitive treatment of benign splenic cysts.

References
1. Bostick WL, Lucia SP (1949) Nonparasitic, noncancerous cystic tumours of the spleen. Arch Pathol 47: 215222 2. Burrig KF (1988) Epithelial (true) splenic cysts. Am J Surg Pathol 12: 275281 3. Carpenter G, Cotter PW, Davidson JRM (1986) Epidermoid cyst of the spleen. Aust NZ J Surg 56: 365368

768 4. Carroll BJ, Phillips EH, Semel CJ, Fallas M, Morgenstern L (1992) Laparoscopic splenectomy. Surg Endosc 6: 183185 5. Dachman AH, Ros PR, Murari PJ, Olmsted WW, Lichtenstein JE (1986) Nonparasitic splenic cysts: a report of 52 cases with radiologicpathologic correlation. Am J Roentgenol 147: 537542 6. Enghardt MH, Allegra SR (1987) An immunohistologic study of a splenic cyst. Mil Med 152: 321323 7. Fowler RH (1940) Cystic tumours of the spleen. Int Abstr Surg 70: 213223 8. Fowler RH (1953) Collective review: nonparasitic benign cystic tumours of the spleen. Int Abstr Surg 96: 209227 9. Kum CK, Ngoi SS, Goh P, Lee YS, Gopalan R (1993) A rare wandering splenic cyst removed with laparoscopic assistance. Singapore Med J 34: 179180 10. Lifschitz-Mercer B, Open M, Kushnir I, Czernobilsky B (1994) Epidermoid cyst of the spleen: a cytokeratin profile with comparison to other squamous epithelia. Virchows Arch 424: 213216 11. Martin J (1958) Congenital splenic cysts. Am J Surg 96: 302308 12. McClure RD, Altemier VA (1942) Epithelial cysts of the spleen. Ann Surg 116: 98103 13. Millar JS (1982) Partial excision and drainage of post-traumatic splenic cysts. Br J Surg 69: 477478 14. Musy PA, Roche B, Belli D, Bugmann P, Nussle D, Le Coultre C (1992) Splenic cysts in pediatric patientsa report on 8 cases and review of the literature. Eur J Pediatr Surg 2: 137140 15. Pedersen VM, Geerdsen JP, Worning AM (1985) Conservative surgical treatment of splenic cysts. Acta Chir Scand 151: 397398 16. Posta CG (1994) Laparoscopic management of a splenic cyst. J Laparoendosc Surg 4(5): 347354 17. Rebeiro E, Soper R (1993) Treatment of benign splenic cyst with splenic conservation. J Tenn Med Assoc 86: 439440 18. Robbins RG, Yellin AE, Lingua RW, Craig JR, Turrill FL, Mikkelsen WP (1978) Splenic epidermoid cysts. Ann Surg 187: 231235 19. Salky B, Zimmerman M, Bauer J, Gelernt I, Kreel I (1985) Splenic cystdefinitive treatment by laparoscopy. Gastrointest Endosc 31(3): 213215 20. Targarona EM, Martinez J, Ramos C, Becerra JA, Trias M (1995) Conservative laparoscopic treatment of a posttraumatic splenic cyst. Surg Endosc 9: 7172 21. Touloukian RJ, Seashore JH (1987) Partial splenic decapsulation: a simplified operation for splenic pseudocyst. J Pediatr Surg 22(2): 135 137 22. Walz MK, Metz KA, Sastry M, Eigler FW (1994) Benign mesothelial splenic cyst may cause high serum concentration of CA 19-9. Eur J Surg 160: 389391 23. Williams RJL, Glazer G (1993) Splenic cysts: changes in diagnosis, treatment and aetiological concepts. Ann R Coll Surg Engl 75: 8789