Antituberculosis Druginduced Hepatotoxicity

The Role of Hepatitis C Virus and the Human Immunodeficiency Virus

JAIME R. UNGO, DENIS JONES, DAVID ASHKIN, ELENA S. HOLLENDER, DAVID BERNSTEIN, ANTHONY P. ALBANESE, and ARTHUR E. PITCHENIK
The University of Miami School of Medicine, Division of Pulmonary Diseases and Critical Care Medicine, Division of Gastroenterology, Department of Internal Medicine, Miami; A. G. Holley State Hospital, Lantana; Addiction Treatment Program Mount Sinai Medical Center, Miami Beach; V. A. Medical Center, Miami; The Florida Bureau of Tuberculosis Control and Prevention, Tallahasse, Florida; The Division of Gastroenterology, Winthrop University Hospital, Mineola, New York

Until recently it was thought that age greater than 35 yr was the main risk factor for the development of drug-induced hepatitis (DIH) in patients receiving antituberculosis therapy. We conducted a study to determine whether infection with either the hepatitis C virus or the human immunodeficiency virus (HIV) were significant risk factors for the development of DIH in patients receiving antituberculosis therapy. Our study consisted of two parts. In the first part, 134 consecutive patients admitted for the treatment of tuberculosis (TB) were followed for the development of DIH. All of these patients were also screened for the presence of hepatitis C and HIV. In the second part of the study, those patients who were hepatitis C positive and who developed DIH on repeated reintroduction of the anti-TB drugs were offered a liver biopsy. If active inflammation, which may be suggestive of hepatitis C infection, was present on the biopsy specimen, treatment with alpha-interferon was begun and the anti-TB drugs were subsequently reintroduced. During the 18 mo of the study, 22 patients developed DIH. The relative risk of developing DIH if the patient was hepatitis C or HIV positive was fivefold and fourfold, respectively (p 0.05). If a patient was coinfected with both hepatitis C and HIV the relative risk of developing DIH was increased 14.4-fold (p 0.002). In the treatment part, four patients were treated with alpha-interferon, and all were able to undergo the reintroduction of anti-TB therapy without reoccurrence of DIH. Infection with hepatitis C and HIV are independent and additive risk factors for the development of DIH during TB therapy. The treatment of hepatitis C with alpha-interferon may allow the reintroduction of anti-TB agents in those who previously developed DIH when exposed to these drugs. Ungo JR, Jones D, Ashkin D, Hollender ES, Bernstein D, Albanese AP, Pitchenik AE. Antituberculosis druginduced hepatotoxicity: the role of hepatiAM J RESPIR CRIT CARE MED 1998;157:18711876. tis C virus and the human immunodeficiency virus.

The United States has experienced a resurgence of tuberculosis (1). Despite the availability of effective chemotherapeutic agents that exist to treat this illness, hepatotoxicity from firstline drugs such as isoniazid (INH), rifampin (RIF), and pyrazinamide (PZA) is common and may limit their use (28). Alcohol use, increasing age, and the presence of chronic liver disease have been reported to increase the risk of developing anti-TB drug-induced hepatotoxicity (DIH) (3, 6). The exact mechanism of this DIH remains unclear. Populations such as those at risk for human immunodeficiency virus (HIV) infection, the elderly, substance abusers, and immigrants from countries with a high incidence of TB

(Received in original form November 11, 1997 and in revised form February 19, 1998) Presented in part at the ATS/ALA International Conference, New Orleans, 1996. Correspondence and requests for reprints should be addressed to David Ashkin, M.D., A. G. Holley State Hospital, 1199 W. Lantana Road, Lantana, FL 33462. Am J Respir Crit Care Med Vol 157. pp 18711876, 1998 Internet address: www.atsjournals.org

are more likely to develop active TB (4). Similarly, these same groups have also been observed to have an increased incidence of infection with the hepatitis C virus (HCV) (9, 10). Recently, alpha-interferon has been recommended as a first-line treatment for patients with hepatitis C infection who show pathologic evidence of active inflammation (11). In studies, patients infected with hepatitis C have experienced a 47% biochemical and 73% histologic response when treated with this agent (12). We postulated that alpha-interferon may have a therapeutic role in those patients coinfected with HCV and TB who develop elevated liver transaminase levels with pathologic evidence of active inflammation that may be indicative of damage caused by HCV. We conducted a prospective study to (1) investigate the possible associations between the presence of HCV infection and/or HIV infection with the occurrence of anti-TB medication DIH, and (2) investigate the possible role that alpha-interferon 2a may have in allowing the reintroduction of anti-TB medications in patients with HCV infection who experience recurrent elevated transaminase levels while receiving anti-TB medication.

1872

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 157

1998

METHODS
The following protocols were reviewed and approved by the A. G. Holley State Hospital Institutional Review Board. A.G. Holley Hospital is a 50-bed, state-funded and operated, inpatient tuberculosis sanatorium located in Lantana, Florida. The patient population it serves is composed of patients with TB who have been court-committed to inpatient care because of noncompliance with outpatient therapy (approximately 40% of the patients) as well as voluntarily admitted patients whose complex underlying medical, psychiatric, or social ailments preclude outpatient TB treatment (approximately 60% of cases).

Association between HCV and/or HIV Infection and the Development of Drug-induced Hepatitis
All patients with tuberculosis admitted to A. G. Holley Hospital between December 1994 and May 1996 were considered eligible for the study. All patients had liver chemistries (alanine aminotransferase [ALT], aspartate aminotransferase [AST], and total bilirubin), IgM, and IgG hepatitis A antibodies, hepatitis B surface antigen, hepatitis B surface antibody, IgG and IgM hepatitis B core antibodies, hepatitis C antibody, HIV antibody, and a urine drug screen performed as part of the routine admission procedure. Liver chemistries were performed at the State of Florida Department of Health and Rehabilitation Services at West Palm Beach. A commercial EIA-2 assay was used to determine the presence of hepatitis C antibodies. All patients who tested positive for the hepatitis C antibody underwent qualitative and quantitative polymerase chain reaction testing (Amplicor; Roche Laboratories, Nutley, NJ) at the University of Miami School of Medicine, Division of Hepatology. HIV ELISA testing with confirmatory Western blot was performed by the State of Florida laboratory in Miami. Liver chemistries were performed on a monthly basis after the initiation of anti-TB treatment. If any abnormalities in these liver chemistries were detected, monitoring was performed at least biweekly. In an attempt to ensure that patients did not use alcohol and/or other substances of abuse, random drug testing was performed throughout the study period. The following criteria were used to determine the development of DIH. 1. Normal liver chemistries prior to starting an anti-TB drug regimen. 2. No use of alcohol or other substances of abuse for at least 10 d prior to starting anti-TB medications. 3. Patients had to be receiving INH, RIF/rifabutin (RBT) or PZA, in standard doses, alone or in combination for at least 5 d prior to the development of abnormal liver chemistries. 4. While receiving anti-TB treatment, there had to be an increase in ALT and/or AST to 120 IU/L (normal 40 IU/L) and/or an increase in total bilirubin to 1.5 mg/dl (normal 1.5 mg/dl). These levels were chosen based upon common clinical practices and the recommendation of many studies and authorities to closely observe and consider the discontinuation of medications in patients receiving antituberculosis therapy once transaminase levels reach three times normal in order to decrease the incidence of severe liver injury (1317). 5. No other apparent cause for the elevation of liver chemistries. 6. Removal of the medications resulted in a normalization or at least a 50% improvement of the abnormal liver chemistries.

cutaneously injected, three times a week was started. Complete blood counts, thyroid function, and autoimmune markers were monitored during treatment. Once liver chemistries normalized on interferon treatment, INH, RBT, and PZA (if appropriate) were reintroduced, one at a time. Once INH, RBT, and PZA were tolerated, EMB, ofloxacin, and streptomycin were discontinued. Liver chemistries were followed weekly after reintroduction of anti-TB therapy. Short-course anti-TB chemotherapy (INH/RBT/PZA for 2 mo followed by INH/ RBT for an additional 4 mo) was completed if no further significant elevations in liver chemistries were encountered. Patients who were HIV-positive were not enrolled in the alphainterferon treatment protocol because of an association between interferon therapy and increased HIV replication (18).

Statistics
Statistical analysis was performed by the Department of Biostatistics at the University of Miami. Chi-square and Fishers exact tests were utilized in the analysis of data.

RESULTS
A total of 134 patients were admitted to A. G. Holley Hospital between December 1994 and May 1996, of which six did not undergo hepatitis testing because of either refusal or early transfer to another facility. These six were excluded from the study. One hundred twenty-eight patients were included in the study (96 male and 32 female). The average age was 44 yr (range, 13 to 72 yr). Ninety-nine (77%) patients had a history of alcohol abuse (defined by DSM IV criteria) (19). Thirty-two (25%) of the 128 patients developed abnormal liver chemistries at some point in their inpatient course. From this group, 10 did not meet the criteria for DIH for the following reasons: five had elevated liver chemistries on admission, two patients had abdominal TB, and of the remaining three, one each had exposure to another hepatotoxic medication, recent crack cocaine use, or concomitant cytomegalovirus and cryptococcus infection.
Association between Viral Hepatitis and/or HIV Infection and the Development of Drug-induced Hepatitis

Twenty-two (19%) of the 128 patients had DIH according to our criteria, with the average age of these patients being 42.1 yr (range, 27 to 66 yr). Fifteen patients with DIH had a history of alcohol use but not within 10 d of starting anti-TB medications. The average age, sex, and incidence of alcohol use among the DIH population was not significantly different from those of the total population. There was no statistically significant difference in age or alcohol use in those patients who developed DIH compared with those who did not develop DIH (Table 1). Patients older than 35 yr of age showed no significant increase in DIH over those patients younger than 35 yr of age (Table 2). However, those older than 35 were more likely to be infected with HCV, whereas those younger than 35 were more likely to be infected with the HIV (p 0.05) (Table 2).

Alpha-Interferon Treatment
INH, rifampin, and PZA were discontinued in all HIV-negative patients with HCV who developed DIH and these patients were begun on an anti-TB regimen of ethambutol (EMB), ofloxacin, and streptomycin. After discontinuation, liver chemistries were followed biweekly. When the liver chemistries normalized or improved by at least 50%, INH, RBT, and PZA were reintroduced, one at a time. If DIH recurred, the hepatotoxic medication was discontinued. All patients infected with HCV who developed this recurrent elevation of liver chemistries were asked to participate in the study, and consent obtained. A percutaneous liver biopsy, performed with computerized tomography guidance, was used to determine the extent and nature of the liver damage. If active inflammation was seen on the biopsy, alphainterferon-2a (Roche Pharmaceuticals, Nutley, NJ), 3 million units, sub-

TABLE 1 DEMOGRAPHICS
Average Age (yr) Total population, n 128 DIH positive, n 22 DIH negative, n 106 p Value*
* Chi-square test.

Male:Female 3.3:1 2.1:1 3.6:1 0.05

Alcohol Abuse (%) 79 77 79

44.0 42.1 44.4 0.05

Ungo, Jones, Ashkin, et al.: Antituberculosis Druginduced Hepatotoxicity

TABLE 2 AGE AS A RISK FACTOR
DIH Age Age 35 35 7 15 p No DIH 23 83 NS HCV ( ) 4 36 p 0.02 HCV ( ) 26 62 HIV ( ) 16 28 p 0.02 HIV ( ) 14 70 HCV ( HCV ( HIV ( HIV ( ) ) ) )

1873
TABLE 3 RISK FACTORS FOR DEVELOPING DRUG-INDUCED HEPATITIS (n 128)
DIH Patient No. 40 88 44 84 (n) 12 10 12 10 (%) 30 11 27 12 p Value* 0.021 0.046

Definition of abbreviations: DIH

drug-induced hepatitis; HCV

hepatitis C virus.

Twelve of the 22 patients (55%) who developed DIH were HCV positive. No patients with DIH were IgM anti-HAV positive. One patient who developed DIH was positive for both HBSAg and anti-HCV. All anti-HCV positive patients who developed DIH were HCV-RNA positive by qualitative PCR testing. Two patients who developed DIH were EIA anti-HCV negative and HCV-RNA positive by qualitative PCR testing. Twelve of the 22 patients (55%) who developed DIH tested positive for HIV by ELISA and the results were confirmed by Western blot assay. Five (4%) of the 128 patients in the study were positive for HBSAg. Of these, only one (20%) developed DIH. This patient, as stated previously, was also coinfected with HCV. There was no statistically significant association between being a carrier of hepatitis B SAg and the development of DIH, though these results may have been limited by the small number of hepatitis B carriers in this study. Thirty-one (24%) of the 128 patients were positive for hepatitis B antibody (HBSAb), of which one (3%) developed DIH. This patient was also infected with HIV. There was no statistically significant association between prior infection with hepatitis B [HBSAb ( )] and the development of DIH. As seen in Table 3, the presence of HCV or HIV were significant independent risk factors for the development of DIH. In our population, the relative risks of developing DIH if the patient was either HCV or HIV positive were fivefold and fourfold, respectively. The relative risk for DIH with HCV and HIV coinfection was 14.44 (Table 4).
Alpha-Interferon Treatment

For definition of abbreviations, see Table 2. * Fishers exact test.

ofloxacin, and streptomycin was utilized. All patients showed an improvement in their liver chemistries with therapy (range, 10 to 43 d). When the liver chemistries improved, INH and RBT were sequentially reintroduced to all four patients, who were then able to complete short-course chemotherapy without further significant elevations in the liver chemistries. With interferon therapy, two patients showed a decline in their viral levels, one (Patient 2) from a pretreatment level of 11 106 to 7.9 106 after 3 mo of therapy and the other (Patient 3) from 106 to 200,000 after 4 mo of a pretreatment level of 3.2 therapy. Patient 4 had an increase in his HCV viral level, from a pretreatment level of 2.4 106 to 5.4 106 after 90 d of therapy. Patient 1 had an initial decline in his HCV viral load, 200,000 after 3 mo of therapy, only to infrom 579,000 to 106 after 9 mo of therapy. Interestingly, decrease to 1.05 spite the increase in viral load, a repeat liver biopsy done after 9 mo of therapy on this patient showed an improvement in the degree of inflammation present (Table 5).

DISCUSSION
The utilization of multidrug regimens for the treatment of TB such as the combination of INH, RIF, and PZA have been associated with an increased incidence of DIH when compared with INH monotherapy used for TB prophylaxis (13, 21, 22). The reported risk factors for the development of hepatotoxicity secondary to anti-TB therapy include the presence of chronic liver disease, increasing age, and active alcohol use (3, 6). Although chronic liver disease is known to increase the risk of DIH, the relative risk of each specific etiology of chronic liver disease for the development of DIH remains unclear. Infection with hepatitis B and/or C viruses are common causes of the chronic liver disease that is frequently seen in populations at risk for TB infection. However, the relationship between the presence of chronic viral hepatitis and the development of DIH after the use of anti-TB medications is not well defined in the literature (14, 23). In contrast to current beliefs, our study did not show a heightened risk for anti-TB medication DIH in patients older

Of the 22 patients who developed DIH, six developed recurrent DIH on reintroduction of the anti-TB medications. Four were treated with alpha-interferon and are described below. The other two patients were coinfected with HCV and HIV and were thus excluded from receiving alpha-interferon. Both developed elevated hepatic transaminase levels upon reintroduction of the anti-TB medications. One of these patients needed his INH to be discontinued and RBT substituted for RIF and was successfully treated with RBT/EMB/PZA biweekly for 6 mo. The other patient could not be given INH, RIF, or PZA without a resultant elevation in transaminases and required prolonged therapy with RBT and EMB for 1 yr. Seven of the 29 patients who were HCV positive and HIV negative developed DIH (24%). Of these seven, three did not develop DIH upon reintroduction of INH and RBT with or without PZA and were able to complete short-course chemotherapy without further complication. Four to (29 to 59 yrs of age) developed DIH on reintroduction of INH and RBT. Three of these four had pansensitive TB and one was resistant to INH. All four consented to participate in the study and underwent a liver biopsy. All were found to have active inflammation, attributed at least in part to hepatitis C (Table 5). As per the protocol, all were begun on subcutaneous alpha-interferon 2a, 3 million units three times a week (20). While initiating interferon therapy, an anti-TB regimen consisting of EMB,

TABLE 4 RELATIVE RISKS FOR DEVELOPING DRUG-INDUCED HEPATITIS

Patients Viral Serologies HCV ( HCV ( HIV ( HIV ( ) HIV ( ) HIV ( ) HCV ( ) HCV ( ) ) ) ) (n) 55 29 33 11 (%) 43 23 26 9 DIH (n) 3 7 7 5 (%) 5 24 21 45 Relative Risk 1 5 4 14.44 95% Confidence Limits* 1.30523.311 1.11419.541 2.74076.135

p Value 0.028 0.036 0.002

* Significant when confidence interval does not include 1. Fishers exact test. Significant when compared with patients without risk factors.

1874

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

TABLE 5 IN THE FOUR PATIENTS WHO UNDERWENT LIVER BIOPSIES, ACTIVE INFLAMMATION WAS FOUND, ATTRIBUTED, AT LEAST IN PART, TO HEPATITIS C
Patient No. 1, pretherapy 1, 9 mo posttherapy 2, pretherapy 3, pretherapy 4, pretherapy Liver Biopsy Results Micronodular cirrhosis, portal triaditis without piecemeal necrosis. Moderate to severe lymphoplasmacytic infiltrate. No Mallory bodies, fatty changes, or acidophilic bodies Cirrhosis with moderate inflammatory activity, improved from prior study. No fatty change, absence of metastatic tumor, granulomas, bile stasis, and liver cell dysplasia Cirrhosis severely active with extensive piecemeal necrosis and active lobular inflammation. Intralobulary moderate acidophilic bodies and scattered foci of hepatocellular necrosis Nonspecific acute and chronic mild-to-moderate inflammation with occasional epithelioid, nonspecific granulomata. AFB negative. No fatty changes or cirrhosis Moderate chronic hepatitis with piecemeal necrosis and portal fibrosis expansion. Mild fatty changes and rare acidophilic bodies. Marked dense packing of inflammatory cells, mainly lymphoplasmacytic

VOL 157

1998

than 35 yr of age (3, 6). This may have been due to the presence of HIV coinfection in patients younger than 35 yr of age, which may predispose this population to DIH and thus increase the number of patients in this age group who developed DIH. Previous large studies that showed age as a risk factor for DIH were performed before HIV was prevalent (3). The prevalence of HCV in the United States is estimated to be 1 to 2% of the total population (9, 10, 24). At least 85% of those infected with HCV will develop chronic liver disease (11). The role of HCV in the development of DIH secondary to anti-TB medications is unclear. Our study suggests that HCV infection is an independent risk factor for the development of anti-TB DIH. The incidence of TB in the United States has increased dramatically in the last decade for many reasons, none of which are more important than HIV infection and AIDS (1). In addition to predisposing to TB, AIDS may predispose to the development of DIH secondary to the use of anti-TB agents (25, 26). Ozick and colleagues showed a link between inner city AIDS patients and the development of anti-TB DIH (25). Our study also shows a link between anti-TB DIH and HIV. Ozick and colleagues (25), however, did not take into account HCV as a possible confounding variable. Our study suggests that HIV and HCV are independent and additive risk factors for the development of anti-TB DIH. The natural course of liver chemistries in hepatitis C infection may be a confounding factor when evaluating for the presence of anti-TB medication DIH. Transaminases in HCV may be normal or wax and wane over time in what has been described as a sine wave pattern (27). Despite very strict inclusion criteria for DIH, it is possible that the increase in liver enzymes seen in some of our patients may have merely been the natural history of the disease as opposed to DIH. Two other possible confounding variables in DIH are alcohol and antiretroviral agents. In both our study and the literature, the potential hepatotoxic effects of antiretroviral agents used in the treatment of HIV were not taken into account. Although this link has not been well established, two case studies have been reported in the recent literature (28). Chronic, active alcohol abuse is a known risk factor for the development of anti-TB medication toxicity (3, 6). None of our patients who developed DIH were found to be positive during random urine drug and alcohol screens. Some of our patients with HCV were unable, despite repeated attempts, to complete a course of INH and RIF or RBT because of elevations in liver chemistries. Rifabutin, a rifamycin derivative, has been shown to be effective in the treatment of TB and to be less hepatotoxic than rifampin (29). However, in four patients who developed DIH secondary to

rifampin, therapy with RBT was attempted, but it was also subsequently associated with elevated transaminases. The exact mechanism by which HCV causes liver inflammation and damage is incompletely understood and is felt to involve both a direct cytotoxic effect as well as stimulation of the immune response (30). The liver biopsy findings in chronic hepatitis C or DIH may be difficult to differentiate. The presence of lymphoid aggregates or fat deposits are consistent with HCV, whereas necrosis in Zone 3 is more characteristic of DIH (31). All four patients had evidence of lymphocytic inflammation, which is more consistent with HCV infection. Three of the four patients who underwent biopsies had evidence of cirrhosis, which is indicative of chronic injury, as can occur in patients infected with HCV. DIH acutely caused by TB medications usually does not cause chronic liver disease. None of the biopsies had evidence of necrosis in Zone 3, which would have been more consistent with DIH. Many advocate the use of alpha-interferon in patients infected with hepatitis C when signs of active inflammation are present (11). Alpha-interferon is thought to act as both an antiviral and an immunomodulatory agent. As an antiviral agent, alpha-interferon is thought to work through a variety of mechanisms, from inhibiting viral hepatocyte penetration to blocking viral integration, synthesis, and release (32, 33). The immunomodulatory effects of interferon may be equally important and appear to be mediated through other cytokines, particularly interleukins (18, 34, 35), macrophages, natural killer cells, and cytotoxic CD8 cells (36). It is by these mechanisms that interferon is thought to regulate the identification and destruction of viral-laden hepatocytes. The ability of alpha-interferon to downregulate some aspects of the immune response has also been noted (37). This is thought to be, at least in part, related to an intracellular release of arachidonic acid, whose metabolite PGE2 has immunosuppressant properties (38, 39). This may indeed be the reason why some patients respond to interferon with improvement in the liver function tests and inflammation, without a concomitant decrease in viral load, as was seen in some patients in this study. The mechanisms of drug-induced hepatotoxicity from antituberculosis agents are thought to involve direct cytotoxicity (by the drug or its metabolites), but they are, to date, not fully understood. An immune-related component is believed to exist as well since both INH and rifampin have documented immunologic effects (4042). We suspect that the ability to treat these patients with TB successfully may stem from the dual action of alpha-interferon. It is the dramatic difference in cost, efficacy, and length of therapy between the first-line and the second-line antimycobacterial medications that underlies the efforts to maintain

Ungo, Jones, Ashkin, et al.: Antituberculosis Druginduced Hepatotoxicity

1875
Kuo, and M. Houghlon. 1990. Epidemiology of hepatitis C virus: a preliminary study in volunteer blood donors. J.A.M.A. 263:4953. Alter, M. J. 1994. Transmission of hepatitis C virus: route, dose and titer. N. Engl. J. Med. 330:784. National Institutes of Health Consensus Conference on Management of Hepatitis C. Bethesda, MD, March 2224, 1997. NIH statement. U.S. Government Printing Office, Bethesda, MD. Carithers, Jr. R. L., and S. S. Emerson. 1997. Therapy of hepatitis C: meta-analysis of interferon alfa-2b trials. Hepatology 26:83S88S. Durand, F., J. Bernuau, D. Pessayre, D. Samuel, J. Belaiche, C. Deggott, H. Bismuth, J. Belghiti, S. Ehrlinger, B. Rueff, and J. P. Benhamou. 1995. Deleterious influence of pyrazinamide on the outcome of patients with fulminant or subfulminant liver failure during antituberculous treatment including isoniazid. Hepatology 21:929932. Turktas, H., M. Unsal, N. Tulek, and O. Oruc. 1994. Hepatotoxicity of antituberculous therapy (rifampicin, isoniazid and pyrazinamide) or viral hepatitis? Tuber. Lung Dis. 75:5860. Centers For Disease Control/American Thoracic Society. 1994. Treatment of tuberculosis and tuberculosis infection in adults and children. Am. J. Respir. Crit. Care Med. 149:13591374. Steele, M. A., R. F. Burk, and R. M. DesPrez. 1991. Toxic hepatitis with isoniazid and rifampin: a meta-analysis. Chest 99:465471. Mitchell, I., J. Wendon, S. Fitt, and R. Williams. 1995. Anti-tuberculous therapy and acute liver failure. Lancet 345:555556. Peters, M. 1996. Actions of cytokines on the immune response and viral interactions: an overview. Hepatology 23:909914. Diagnostic and Statistical Manual of Mental Disorders. 4 ed. 1994. American Psychiatric Association, Washington, DC. Lindsay, K. L., G. L. Davis, and E. R. Schiff. 1996. Response to higher doses of interferon alpha-2b in patients with chronic hepatitis C: a randomized multicenter trial. The Hepatitis Interventional Group. Hepatology 24:10341040. Harding, S. M., and W. C. Bailey. 1994. Chemotherapy of tuberculosis. In D. Scholssberg, editor. Tuberculosis, 3rd ed. Springer-Verlag, New York. 6988. Parasarthy, R., G. Raghupati Sarma, B. Janardhanam, P. Ramachandran, T. Santha, S. Siuasubramanian, P. R. Somasundaram, and S. P. Tripathy. 1986. Hepatic toxicity in South Indian patients during treatment of tuberculosis with short course regiments containing isoniazid, rifampin and pyrazinamide. Tubercle 67:99108. Turner, M., and R. Haskal. 1995. Drug-induced hepatotoxicity in an inpatient TB unit: Lemuel Shattuck Hospital (abstract). Am. J. Respir. Dis. Crit. Care Med. 151:A514. Dawson, G. J., R. R. Lesniewski, J. L. Stewart, K. M. Boardway, R. A. Gutierrez, L. Pendy, R. G. Johnson, X. Alcalde, K. U. Rote, S. G. Devare, W. G. Robey, and D. A. Peterson. 1991. Detection of antibodies to hepatitis C virus in US blood donors. J. Clin. Microbiol. 29:551556. Ozick, L. A., L. Jacob, G. M. Comer, T. P. Lee, J. Ben-Zvi, S. S. Donnelson, and C. P. Felton. 1995. Hepatotoxicity from isoniazid and rifampin in inner-city AIDS patients. Am. J. Gastroenterol. 90:19781982. Ungo, J. R., D. Ashkin, E. S. Hollender, S. D. Ryan, D. Bernstein, A. P. Albanese, and A. E. Pitchenik. 1996. Antituberculosis drug-induced hepatotoxicity (ATDIH): the possible role of hepatitis viruses (abstract). Am. J. Respir. Crit. Care Med. 153:A410. Alter, H. J., H. S. Margolis, K. Krawczynski, A. Maras, F. N. Judson, P. Y. Hu, W. J. Alexander, R. E. Sampliner, M. A. Gerber, G. Kuo, M. Houghton, and D. W. Bradley. 1989. Clinical outcome and risk factors associated with hepatitis C in the United States. Hepatology 10:581. Moyle, G. L., and S. E. Barton. 1995. HIV-protease inhibitors in the management of HIV-infection. J. Antimicrob. Chemother. 38:921925. McGregor, M. M., P. Olliaro, and L. Wolmarans. 1996. Efficacy and safety of rifabutin in the treatment of patients with newly diagnosed pulmonary tuberculosis. Am. J. Respir. Crit. Care Med. 154:14621467. Scheuer, P. J., P. Ashrafzadel, S. Sherlock, D. Brown, and G. M. Dushienko. 1992. The pathology of hepatitis C. Hepatology 15:567571. Gerber, M. A. 1997. Histopathology of hepatitis C virus infection. Clin. Liver Dis. 1:529541. Goodman and Gilmans: The Pharmacological Basis of Therapeutics. 1995. 8th ed. Macmillan, New York. 12111223. Baron, S., S. Tyring, W. R. Fleischmann, D. H. Coppenhauer, D. W. Niesel, G. R. Klimpel, G. J. Stanton, and T. K. Hughes. 1991. The interferons: mechanism of action and clinical application. J.A.M.A. 266: 13751383. Aizawa, Y., M. Zeniya, H. Takahashi, and G. Toda. 1996. Pattern of immune responses and the efficacy of IFN therapy in chronic hepatitis type C. Prog. Hepatol. 2:3953. Blatt, L. M., J. M. Davis, S. B. Klein, and M. W. Taylor. 1996. The bio-

first-line drugs at the forefront of a therapeutic regimen. The obvious benefit of completion of antituberculosis therapy has a positive impact at both the individual and the societal levels. In the paradigm of anti-TB DIH, the goal of interferon therapy in our study was to allow the safe continuation of these medications in patients infected with HCV until the TB was cured. The role of interferon in patients without HCV who develop DIH is unknown and was not studied in this protocol. It is important to mention that our population at A. G. Holley Hospital is unique as our hospital is a referral center for patients with TB who are either noncompliant or difficult to treat as an outpatient for reasons such as alcoholism, substance abuse, homelessness, or the presence of psychiatric disease or complex medical conditions. Our incidence of HCV infection is fifteenfold that of the general population (9, 24). This may, in part, account for the tenfold incidence increase of anti-TB DIH seen in our patients as compared with the general population. Larger studies, utilizing different patient populations, are needed to better define the influence of HIV and HCV on DIH. Thirty-three (83%) of the 40 patients who were infected with hepatitis C had normal transaminase levels on admission (prior to beginning TB medications), and had subclinical hepatitis C infection. In the seven patients who had elevated transaminases on admission, all of these abnormalities were attributed by the treating clinicians to active alcohol abuse. The majority of patients with hepatitis C would not have been suspected to be infected by conventional transaminase screening and required anti-HCV testing to uncover their increased susceptibility for the development of DIH. In summary, this study suggests that HCV and HIV infection may play an important role in the development of antiTB medication DIH. Considering the potential large number of patients infected with both TB and HCV who may need this treatment, this association may have important implications. Clinicians treating patients at high risk for HIV and/or HCV with anti-TB agents may need to consider serologic screening of these patients. Further studies are necessary to better elucidate this association.
Acknowledgement : The writers thank William Abraham, Ph.D., and Adam Wanner, M.D., for their technical support and assistance, and Sean Ryan, P.A., for his assistance with data acquisition.

10. 11.

12. 13.

14.

15.

16. 17. 18. 19. 20.

21.

22.

23.

24.

25.

References
1. Cantwell, M. F., D. E. Snider, G. M. Cauthen, and I. Onorato. 1994. Epidemiology of tuberculosis in the United States, 1985 through 1992. J.A.M.A. 272:535539. 2. Peters, B. S., E. Carlin, R. J. Weston, S. J. Loveless, J. Sweeney, J. Weber, and J. Main. 1994. Adverse effects of drugs used in the management of opportunistic infections associated with HIV Infection. Drug Saf. 10:439454. 3. Kopanoff, D. E., D. E. Snider, and G. J. Caras. 1978. Isoniazid-related hepatitis: A U.S. Public Health Service Cooperative Surveillance Study. Am. Rev. Respir. Dis. 117:9911001. 4. Bailey, W. C., S. L. Taylor, H. E. Dascomb, H. B. Greenberg, and M. M. Ziskind. 1973. Disturbed hepatic function during isoniazid chemoprophylaxis. Am. Rev. Respir. Dis. 107:523529. 5. Gellis, S. N., and R. V. Murphy. 1955. Hepatitis following isoniazid. Dis. Chest. 28:462. 6. Gronhagen-Riska, C., P. E. Hellstrom, and B. Froseth. 1978. Predisposing factors in hepatitis induced by isoniazid-rifampin treatment of tuberculosis. Am. Rev. Respir. Dis. 118:461466. 7. Durand, F., D. Pessayre, M. Fournier, J. Belghiti, S. Erlinger, and J. Bernuau. 1995. Antituberculous therapy and acute liver failure (letter). Lancet 345:10701072. 8. Van der Kooi, K., J. J. Mottet, and C. Regamey. 1994. Isoniazid is not always the cause of hepatitis during the treatment of tuberculosis (letter). Clin. Infect. Dis. 19:987988. 9. Stevens, C. E., P. E. Taylor, J. Pindyck, Q. L. Choo, D. W. Bradley, G.

26.

27.

28. 29.

30. 31. 32. 33.

34.

35.

1876

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 157

1998

36.

37. 38. 39.

logic activity and molecular characterization of a novel synthetic alpha interferon species: Consensus Interferon. J. Interferon Cytokine Res. 76:489499. Weinstock-Guttman, B., R. M. Ransohoff, R. P. Kinkel, and R. A. Rudick. 1995. The interferons: biological effects, mechanisms of action, and use in multiple sclerosis. Ann. Neurol. 37:715. Reiter, Z. 1993. Interferon: a major regulator on natural killer cell-mediated cytotoxicity. J. Interferon Res. 13:247257. Fried, M. W., and J. H. Hoofnagle. 1995. Therapy of hepatitis C. Semin. Liver Dis. 15:8291. Andreone, P., C. Cursaro, and G. Gasbarrini. 1993. Interferon-alpha in-

creases prostaglandin E2 production by cultured liver biopsy in patients with chronic viral hepatitis; can non-steroidal anti-inflammatory drugs improve the therapeutic response to interferon? J. Hepatol. 19:228231. 40. Rothfield, N. F., W. F. Bierer, and J. W. Garfield. 1978. Isoniazid induction of antinuclear antibodies. Ann. Intern. Med. 88:650652. 41. Graber, C. D., J. Jebaily, R. L. Galphin, and E. Doering. 1973. Light chain proteinuria and humoral immunocompetence in tuberculosis patients treated with rifampin. Am. Rev. Respir. Dis. 107:713717. 42. Gupta, S., M. H. Greco, and I. Siegel. 1975. Suppression of T-lymphocyte rosettes by rifampin. Ann. Intern. Med. 82:484488.