2417
Rate of Bilirubin Regression After Stenting in
Malignant Biliary Obstruction for the Initiation
of Chemotherapy
How Soon Should We Repeat Endoscopic Retrograde Cholangiopancreatography?
Brian R. Weston, MD1
William A. Ross, MD1
Robert A. Wolff, MD2
Douglas Evans, MD3
Jeffrey E. Lee, MD3
Xuemei Wang4
Lian-chun Xiao4
Jeffrey H. Lee, MD1
1
Department of Gastroenterology, Hepatology, and
Nutrition, The University of Texas M. D. Anderson
Cancer Center, Houston, Texas.
2
Department of Gastrointestinal Medical Oncology,
The University of Texas M. D. Anderson Cancer
Center, Houston, Texas.
3
Department of Surgical Oncology, The University
of Texas M. D. Anderson Cancer Center, Houston,
Texas.
4
Department of Biostatistics and Applied Mathe-
matics, The University of Texas M. D. Anderson
Cancer Center, Houston, Texas.
Address for reprints: Jeffrey H. Lee, MD, Depart-
ment of Gastroenterology, Hepatology, and Nutri-
tion, The University of Texas M. D. Anderson
Cancer Center, 1515 Holcombe Boulevard,
FC10.2028, Houston, TX 77030; Fax: (713) 563-
4398; E-mail: jefflee@mdanderson.org
Received August 14, 2007; revision received
November 12, 2007; accepted December 17,
2007.
ª 2008 American Cancer Society
DOI 10.1002/cncr.23454
Published online 10 April 2008 in Wiley InterScience (www.interscience.wiley.com).
BACKGROUND. This study was conducted to evaluate the rate of regression of bili-
rubin after stent placement for malignant biliary obstruction.
METHODS. Records were reviewed from October 2002 to September 2005 for
patients who underwent endoscopic retrograde cholangiopancreatography with
stent placement. The time to achieve a bilirubin level
2 mg/dL was the primary
endpoint because this is the level required by most chemotherapy protocols.
Patient variables included type of cancer, liver metastasis, recent chemotherapy,
baseline creatinine, and international normalized ratio (INR). Stent variables
included type, dimension, stricture location, and sphincterotomy.
RESULTS. In total, 156 patients were included in the analysis: Ninety-three patients
achieved a poststent bilirubin level
2 mg/dL, 29 patients failed because of stent
failure, and 34 patients failed because of inadequate follow-up. The time required
for 80% of patients to achieve normalization was more than doubled in those who
had prestent bilirubin levels 10 mg/dL (6 weeks) compared with those who had
prestent bilirubin levels <10 mg/dL (3 weeks). The following variables were identi-
fied as statistically significant: prestent bilirubin level, stricture location, liver me-
tastasis, and INR. The cancer type, recent chemotherapy, stent type and diameter,
and sphincterotomy were not statistically significant variables.
CONCLUSIONS. The rate of bilirubin normalization after biliary stenting was
highly dependent on the prestent bilirubin level. Endoscopic intervention should
be considered in patients who fail to achieve adequate normalization of serum
bilirubin in 6 weeks if prestent bilirubin level was 10 mg/dL and in 3 weeks if
their prestent bilirubin level was <10 mg/dL. Independent variables, such as
diffuse liver metastases, stricture outside the common bile duct, and elevated
INR had predictive value for bilirubin normalization. Cancer 2008;112:2417–23.

2008 American Cancer Society.
KEYWORDS: biliary stent, malignant biliary obstruction, serum bilirubin level,
chemotherapy.
O bstructive jaundice is a frequent complication of many malig-
nant processes. Many chemotherapy drugs require intact
mechanisms of bilirubin excretion and bile drainage to prevent tox-
icity.1–3 Typically, a bilirubin level <2 mg/dL is required before the
initiation of therapy at our institution. Endoscopic biliary stent
insertion is a well established method for providing biliary drainage
in patients with malignant obstructive jaundice.4 In addition to
symptom relief, biliary stenting may provide the opportunity for
further treatment in the form of systemic chemotherapy.
2418 CANCER June 1, 2008 / Volume 112 / Number 11
The time required for normalization of serum
bilirubin levels in patients with malignant biliary
strictures after stent placement has important impli-
cations with respect to the timing of chemotherapy
implementation and/or the need for repeat endo-
scopic intervention. To our knowledge, few studies to
date have adequately examined the rate of this pro-
cess. The objective of the current study was to evalu-
ate the rate of regression of serum bilirubin levels in
patients with stent placement for malignant biliary
obstruction in addition to identification of patient
and stent variables that may be predictors for biliru-
bin normalization. This information will be used to
guide future endoscopic practices when approaching
malignant biliary strictures in patients who are
awaiting chemotherapy.
MATERIALS AND METHODS
The study was approved by the Institutional Review
Board at the University of Texas M. D. Anderson Can-
cer Center. A retrospective chart review was per-
formed from an electronic database of consecutive
patients from October 1, 2002 through September 30,
2005 who underwent endoscopic retrograde cholan-
giopancreatography (ERCP) with biliary stent place-
ment for malignant obstructive jaundice.
The presence of an underlying biliary obstructive
process was suggested by abnormal biochemical liver
tests and/or preprocedure imaging (ie. biliary dila-
tion on abdominal ultrasound, computed tomogra-
phy, or magnetic resonance imaging studies). The
presence of a dominant biliary stricture amenable to
endostent therapy was identified by cholangiography
at the time of ERCP. A dominant biliary stricture on
cholangiography was defined by the presence of a
focal, discrete narrowing of either extrahepatic or in-
trahepatic bile ducts with dilation proximal to the
stricture.1 Endoscopic biliary stent placement was
performed in standard fashion with or without
sphincterotomy in all patients.4,5
The time to achieve a serum bilirubin level
2
mg/dL was the primary study endpoint, because this
is the level required by most chemotherapy protocols
before initiation of chemotherapy.6 Prestent total se-
rum bilirubin levels were recorded at presentation; in
all patients, this was drawn within 7 days of endo-
scopic intervention. In most instances, the duration
of abnormal bilirubin elevation before intervention
was not known. Then, bilirubin levels were reviewed
retrospectively at variable intervals after stent place-
ment until bilirubin normalization (total serum biliru-
bin
2 mg/dL) or nadir. Duration until normalization
was recorded in days. We included only those
patients who met predefined criteria for serial labo-
ratory data. The maximal interval during which se-
rum bilirubin measurements must have been
obtained to be included for analysis was as follows:
serum bilirubin level
5 mg/dL, 2 weeks; serum bili-
rubin levels from 6 mg/dL to 10 mg/dL, 3 weeks; se-
rum bilirubin levels from 10 mg/dL to 15 mg/dL,
4 weeks; serum bilirubin levels from 16 mg/dL to
20 mg/dL, 5 weeks; and serum bilirubin levels >20
mg/dL, 6 weeks. Patients who had inadequate serial
data were excluded by using these parameters.
Patients who had prestent bilirubin levels
2 mg/dL
also were excluded, because they already were eligi-
ble for chemotherapy.
The primary patient variables examined included
type of cancer, presence and extent of liver metasta-
sis (none, focal, or diffuse as evident radiographically
at the time of ERCP), recent chemotherapy (within
6 months), baseline creatinine, and international
normalized ratio (INR). Demographic information,
including patient age and sex, was recorded. Patient
disposition, including subsequent receipt of chemo-
therapy or other outcome (surgery, hospice, and/or
death before bilirubin normalization) also was noted.
The primary stent variables examined included
stent type (plastic vs metal), size (diameter only),
and number placed. The location and extent of the
dominant stricture was documented as follows: left
and/or right hepatic duct, common hepatic duct,
common bile duct (CBD), or multifocal. Sphincterot-
omy also was noted when it was performed. Need
for endoscopic revision before normalization of bili-
rubin was recorded. The time to bilirubin normaliza-
tion was considered from placement of the second
stent in all patients when applicable. Stent failure
was defined as nonnormalization of bilirubin despite
successful stent deployment across a stricture with
or without revision.
Exclusion criterion included any history of prior
biliary manipulation (ie, prior stent placement,
sphincterotomy, percutaneous biliary intervention, or
other hepatobiliary surgery). Patients with a known
history of chronic liver disease or cirrhosis were
excluded. Verification of prior normal liver tests and
creatinine before onset of obstructive jaundice was
confirmed when possible.
Statistical Methods
Descriptive statistics were used to summarize patient
characteristics and stent-related factors. The Kaplan-
Meier method was used to determine the time at
which at least 80% of patients achieved poststent bili-
rubin normalization (serum level
2 mg/dL) for
groups with different prestent bilirubin levels (<5
 Bilirubin Regression after Stenting/Weston et al. 2419

TABLE 1 TABLE 2
Summary of Categorical Patient Characteristics Summary of Categorical Stent Characteristics

No. of patients (%) No. of patients (%)

Success Failure Success Failure

Variable Total subset* subsety Variable Total subset* subsety

Total no. of patients 156 (100%) 93 (100%) 29 (100%) Total no. of patients 156 (100%) 93 (100%) 29 (100%)
Mean age [range], y 60.2 [27–84] 61.8 [27–84] 58.8 [40–77] Stent type
Sex Plastic 75 (48) 45 (48) 15 (52)
Men 87 (56) 55 (59) 17 (59) Metal 81 (52) 48 (52) 14 (48)
Women 69 (44) 38 (41) 12 (41) No. of stents
Cancer diagnosis 1 143 (92) 88 (95) 25 (86)
Pancreatic adenocarcinoma 63 (40) 41 (44) 8 (27) 2 11 (7) 5 (5) 3 (10)
Cholangiocarcinoma 16 (10) 7 (8) 4 (14) >2 2 (<1) 0 1 (4)
Gallbladder 7 (5) 3 (3) 2 (7) Stent width
Metastatic 45 (29) 23 (25) 13 (45) 6 mm 1 (<1) 0 0
Other{ 25 (16) 19 (20) 2 (7) 7 Fr 7 (5) 2 (2) 4 (14)
Liver metastasis 8 mm 11 (7) 8 (9) 2 (7)
None 56 (36) 44 (47) 4 (14) 8.5 Fr 7 (5) 6 (6) 1 (3)
Focal lesion 13 (8) 12 (13) 0 10 mm 69 40 12
Multiple/diffuse 87 (56) 37 (40) 25 (86) 10 Fr 61 37 10
Prior recent chemotherapy 75 (48) 42 (45) 16 (55) Stricture location
Prestent bilirubin, mg/dL Left and/or right hepatic duct 6 (4) 5 (5) 1 (3)
<5 41 (26) 28 (30) 6 (21) Common hepatic duct 25 (16) 12 (13) 7 (24)
6–10 53 (34) 32 (34) 9 (31) Common bile duct 96 (61) 68 (73) 10 (35)
>10 62 (40) 33 (36) 14 (48) Multifocal 29 (19) 8 (9) 11 (38)
Baseline creatinine, mg/dL Sphincterotomy 87 (56) 53 (57) 12 (41)
<1.4 146 (94) 89 (96) 26 (90) Endoscopic revision
1.5–2 7 (5) 4 (4) 1 (3) Yes 15 (10) 3 (3) 7 (24)
>2 3 (1) 0 2 (7) No 137 (88) 90 (97) 19 (66)
Baseline INR ERCP repeat w/o stent change 4 (3) 0 3 (10)
<1.4 112 (72) 75 (81) 17 (59)
1.5–2 30 (19) 13 (14) 9 (31) ERCP indicates endoscopic retrograde cholangiopancreatography; w/o, without.
>2 6 (4) 2 (2) 3 (10) * The subset that achieved bilirubin normalization postbiliary stent placement.
y
NA 8 (5) 3 (3) 0 The subset that failed to achieve bilirubin normalization postbiliary stent placement.
Disposition
Chemotherapy 82 (53) 75 (81) 2 (7)
Surgery 6 (4) 6 (7) 0 ing prestent bilirubin levels, were log-transformed
Hospice care 45 (29) 11 (12) 18 (62)
Death (before bilirubin normal) 13 (8) 0 3 (10)
before fitting Cox proportional-hazards models.
Percutaneous biliary drain 5 (3) 0 5 (17)
NA 5 (3) 1 (<1) 1 (4)

NA indicates not available; INR, international normalized ratio.

* The subset that achieved bilirubin normalization postbiliary stent placement.
y
The subset that failed to achieve bilirubin normalization postbiliary stent placement.
{
Other cancer diagnosis included the following malignancies: neuroendocrine (N 5 10), lymphoma
(N 5 5), sarcoma (N 5 4), and ampullary mass (N 5 6).
mg/dL, 5–10 mg/dL, and >10 mg/dL). A univariate
Cox proportional-hazards model was fit to assess the
association between bilirubin normalization with
patient and stent variables. The variables with P
values <0.2 in univariate analysis (UVA) were
included in the multivariate analyses (MVA). Then,
variables were removed 1 by 1 until all variables that
remained in the model were statistically significant at
a .05 significance level. Continuous variables, includ-
RESULTS
In total, 217 patients who underwent first-time ERCP
with biliary stent placement for malignant biliary
obstruction were identified during the 36-month
study period. Of the 156 patients who were included
in the final analysis, 93 patients achieved a poststent
bilirubin level (PSB)
2 mg/dL, 29 patients failed to
achieve PSB
2 mg/dL because of stent failure, and
34 patients failed to achieve a documented PSB
2
mg/dL because of inadequate follow-up. Sixty-one of
217 patients were excluded from the final analysis,
including 34 patients who achieved a PSB
2 mg/dL
but had inadequate serial laboratory data and 27
patients who had prestent bilirubin levels
2 mg/dL.
A summary of categorical patient and stent char-
acteristics is presented in Tables 1 and 2. Patients
who had prestent bilirubin levels 10 mg/dL
2420 CANCER June 1, 2008 / Volume 112 / Number 11

TABLE 3 TABLE 4
Prestent Bilirubin and Time to Normalization Result of Univariate Cox Proportional Hazards Model

Prestent bilirubin, mg/dL Normalization, % Time to normalization, wk RR (95% CI),

Variable N 5 156 P
<5 80 2.6
5–10 80 2.7 Cancer diagnosis
10 80 5.6 Pancreatic adenocarcinoma 1
Cholangiocarcinoma (vs pancreatic adenocarcinoma) 0.59 (0.26–1.3) .19
Gallbladder (vs pancreatic adenocarcinoma) 0.77 (0.24–2.48) .66
Metastatic lesion (vs pancreatic adenocarcinoma) 1.04 (0.62–1.74) .88
Other (vs pancreatic adenocarcinoma) 1.13 (0.65–1.95) .66
Prestent bilirubin (n 5 146; log scale) 0.34 (0.24–0.49) <.0001
Stent diameter (continuous) 1.1 (0.85–1.42) .46
Stricture location
Common bile duct (baseline)
Left and or right hepatic duct (vs common bile duct) 0.92 (0.29–2.96) .89
Common hepatic duct (vs common bile duct) 0.59 (0.33–1.05) .07
Multifocal (vs common bile duct) 0.45 (0.22–0.94) .03
Common bile duct (baseline)
Other 0.56 (0.35–0.88) .01
Stent type
Plastic (baseline)
Metal 1.03 (0.69–1.56) .87
Liver metastasis
None (baseline)
Focal 1.30 (0.69–2.47) .42
Diffuse 0.82 .0009
FIGURE 1. Bilirubin normalization over time postbiliary stent placement. Age (continuous) 0.99 (0.98–1.02) .88
E indicates the number of patients who achieved bilirubin normalization Sex
(<2 mg/dL); N, the number of patients in that category. Men (baseline)
Women 1.06 (0.7–1.61) .78
INR
<1.4 (baseline)
required approximately twice the time (5.6 weeks) to 1.5 0.51 (0.29–0.88) .02
achieve normalization compared with patients who Creatinine, mg/dL
had prestent bilirubin levels <10 mg/dL (2.7 weeks) <1.4 (baseline)
when 80% of patients achieved bilirubin normaliza- >1.5 0.82 .67
tion; the results are presented in Table 3 and in
RR indicates relative risk; 95% CI, 95% confidence interval; INR, international normalized ratio.
Figure 1. Results of UVA and MVA (when applicable)
to assess the association between bilirubin normali-
zation with patient and stent variables are presented
in Tables 4 and 5. We were able to draw the following TABLE 5
Result of Multivariate Cox Proportional Hazards Model
conclusions based on the analyses: Patients with
high prestent bilirubin levels were less likely to Full model Reduced model
achieve bilirubin normalization than patients with
lower prestent bilirubin levels (UVA: relative risk Variable RR (95% CI) P RR (95% CI) P
[RR], 0.34; P < .0001; MVA: RR, 0.30; P < .0001).
Prestent bilirubin, 0.31 (0.21–0.45) <.0001 0.30 (0.20–0.43) <.0001
Patients with diffuse liver metastasis (UVA: RR, 0.82;
log transformed
P 5 .0009; MVA: RR, 0.39; P < .0001) were statistically Stricture location, 0.79 (0.46–1.34) .38
less likely to achieve adequate bilirubin normaliza- other vs CBD
tion. Patients who had strictures outside the CBD or INR
multifocal lesions had less chance to achieve biliru- <1.4
1.5 0.81 (0.45–1.44) .47
bin normalization than patients who had lesions iso-
Liver metastasis
lated anywhere (proximal, middle, or distal) in the None
CBD (UVA: RR, 0.56; P 5 .01). Patients who had a Focal 1.58 (0.81–3.07) .18 1.42 (0.74–2.69) .29
higher protime (INR1.5; suggestive of either pro- Diffuse 0.43 (0.25–0.74) .003 0.39 (0.24–0.62) <.0001
longed cholestasis and/or underlying hepatic dys-
RR indicates relative risk; 95% CI, 95% confidence interval; CBD, common bile duct; INR, interna-
function) were less likely to achieve bilirubin
tional normalized ratio.
normalization than patients who had a normal pro-

time (INR<1.4; UVA: RR, 0.51; P 5 .02). Other vari-
ables, such as type of cancer, recent chemotherapy,
stent type (plastic vs metal), stent size (diameter),
and sphincterotomy, were not identified as statisti-
cally significant (P > .05). There was inadequate data
distribution available to comment on renal function
that was not significantly different among the
groups. Approximately 80% of patients (75 of 93
patients) who achieved bilirubin normalization after
stent placement subsequently could receive chemo-
therapy.
DISCUSSION
Hepatobiliary obstruction complicates many malig-
nant processes typically as a result of locally
advanced primary or metastatic disease. Resultant
hyperbilirubinemia may be associated with signifi-
cant morbidity, including malabsorption, coagulopa-
thy, progressive hepatocellular dysfunction, cholangitis,
and renal impairment, in addition to clinical symptoms
of jaundice, pruritus, malaise, anorexia, and weight
loss.7–9 Prospective studies have demonstrated that a
serum bilirubin level >1.5 mg/dL predicts decreased
survival and/or significant impairment in quality of life
in these patients.1,8–12
Hyperbilirubinemia also may prevent the initia-
tion of chemotherapy. Intact mechanisms of bilirubin
excretion and bile drainage are needed to prevent
drug toxicity.1–3 It has been established that several
chemotherapeutic agents alter metabolism in the set-
ting of hyperbilirubinemia (ie, gemcitabine, doxoru-
bicin, irinotecan, and the taxanes). Data to guide the
administration of chemotherapy in the setting of
resolving hyperbilirubinemia after biliary decompres-
sion are limited. The majority of current chemother-
apeutic regimens require a serum bilirubin level <2
mg/dL.1,6 For instance, because gemcitabine is typi-
cally the foundation of systemic therapy for pancre-
atic cancer, we wait until the bilirubin level is from
2 mg/dL to 2.5 mg/dL and is falling before initiation
of treatment with this agent. This approach is based
on published results demonstrating that gemcitabine
led to increased hepatotoxicity in patients with liver
dysfunction manifested by hyperbilirubinemia.6
Endoscopic biliary stent placement often is per-
formed to relieve the symptoms of biliary obstruction
and may provide patients with an opportunity to
receive systemic chemotherapy. Effective and rapid
biliary decompression is the therapeutic goal for
lesions that affect the extrahepatic and large intrahe-
patic bile ducts. The rate of serum bilirubin regres-
sion in patients with stent placement for malignant
biliary obstruction has important implications with
Bilirubin Regression after Stenting/Weston et al. 2421
respect to the timing of chemotherapy implementa-
tion and/or the need for repeat endoscopic interven-
tion in these patients. The objective of the current
study was to evaluate the rate of regression of serum
bilirubin levels in patients with stent placement for
malignant biliary obstruction and to identify patient
characteristics or other factors that mat be predictors
for bilirubin normalization. This information will
assist physicians in anticipating when to start chem-
otherapy and may prevent unnecessary, repeat, inva-
sive procedures.
Our results demonstrate that patients with pres-
tent bilirubin levels 10 mg/dL require approxi-
mately twice the time (6 weeks vs 3 weeks) to
achieve normalization compared with patients who
have prestent bilirubin levels <10 mg/dL when 80%
of patients achieved normalization (bilirubin
2 mg/
dL). Patients with high prestent bilirubin levels were
less likely to achieve bilirubin normalization than
patients with lower prestent bilirubin levels. Inde-
pendent variables, such as the presence and extent
of liver metastases, stricture outside the CBD, and
elevated INR, also had predictive value in bilirubin
normalization.
Serial serum bilirubin levels provide a practical
marker with which to follow patients for adequate
biliary decompression and may be used by oncolo-
gists to determine the timing of chemotherapy. How-
ever, many complex interacting factors may
significantly affect the rate of serum bilirubin regres-
sion. When cholestasis is relieved, serum bilirubin
values fall slowly to normal; this is caused in part by
the formation of bilialbumin in which bilirubin and
albumin (half-life, 3 weeks) are bound covalently.13,14
Serial data for fractionated (direct and indirect) se-
rum bilirubin and d-bilirubin levels, although of
greater potential benefit than total serum bilirubin
alone, were not available in the vast majority of
patients and, thus, could not be analyzed in this
study. On the basis of history and baseline laboratory
data, we excluded patients who had known underly-
ing chronic liver disease and/or cirrhosis from the
current study; however, occult disease could not be
excluded entirely. Patients who had evidence of
diffuse hepatic metastases (approximately 1 in 3
patients in our study population) were included in
our analysis as long as they had evidence of a domi-
nant biliary stricture that was amenable to endo-
scopic stent placement. It may be concluded that
patients with extensive hepatic parenchymal tumor
involvement have concomitant small bile duct dis-
ease and/or possible hepatic insufficiency as a com-
ponent of their cholestasis. Incomplete relief of
obstructed liver segments by endostent placement in
2422 CANCER June 1, 2008 / Volume 112 / Number 11
these situations also may impair the rate of biliary
normalization.15,16 Our results reflect this, because all
patients with diffuse hepatic metastasis and multifo-
cal large bile duct disease outside the CBD were sta-
tistically less likely to achieve adequate bilirubin
normalization. Further extrapolation from our data
suggests those with prolonged protime also had less
chance for bilirubin normalization. Prolonged pro-
thrombin time/INR, although nonspecific, may
reflect hepatic synthetic dysfunction and/or pro-
longed severe cholestasis in the absence of other
nonhepatic etiologies.13,14 Several patients in this
subgroup failed to achieve bilirubin normalization
and predictably may have greater difficulty tolerating
subsequent chemotherapy. Prolonged duration of
hyperbilirubinemia (in particular with high concen-
trations; ie, >15 mg/dL) often suggests a component
of secondary hepatocellular dysfunction in addition
to obstruction; this may be manifested further by a
lagging fall in bilirubin level.13,14 The duration of
hyperbilirubinemia before stent placement was not
available in most instances. Intrahepatic cholestasis
or hemolysis related to such factors as infection,
hepatotoxic drugs (ie, chemotherapy, antibiotics,
total parenteral nutrition, etc), or blood product
transfusion, etc, all are common in our patient popu-
lation and are possible uncontrolled confounding
variables in the absolute rate of bilirubin regression.
Clinically apparent cholangitis or sepsis syndrome
was observed in only a small minority of patients
(estimate, <5%) at the time of ERCP. A majority of
patients were on multiple medications. It is note-
worthy that no significant difference was observed in
the rate of bilirubin regression in the patients who
received recent chemotherapy before ERCP (approxi-
mately half of the study group) compared with the
patients who did not. Renal function also has a well
described correlation with serum bilirubin levels.
Inadequate distribution of data was available to com-
ment on renal function, which was not significantly
different between groups. However, we already know
that patients with significant renal impairment will
have inhibited bilirubin clearance related to inability
to excrete the conjugated form.
Stent type, size, and sphincterotomy were not
identified as statistically significant variables (P > .05).
The number of stents placed also was not identified
as a significant variable; however, multiple stents
were used in only a minority of patients (<8%). A
relatively greater percentage of patients in the failure
group had multifocal biliary lesions (38% vs 9%). The
most common cancer diagnosis in the subset that
achieved bilirubin normalization was pancreatic
adenocarcinoma (44% vs 27% other than pancreatic
adenocarcinoma). The most common cancer diagno-
sis in the subset that failed to achieve bilirubin nor-
malization was metastatic disease (45% vs 25% in
patients without metastasis).
The impact of any of the variables described
above may significantly affect the rate of bilirubin
regression. Susceptibility to any of these factors may
be difficult to predict, and such patients need to be
managed on an individual basis. The need for repeat
ERCP and/or stent revision will depend largely on
the finding of persistent biliary dilation by radio-
graphic imaging or cholangiography and clinical
judgment. In disease characterized by large bile duct
obstruction, the proximal bile ducts generally are
dilated, especially if the bilirubin concentration is
>10 mg/dL and if the patient has been jaundiced for
>2 weeks, which is typically the case in patients with
malignant obstruction. Acute large bile duct obstruc-
tion does not cause dilation of the bile ducts imme-
diately.13,14
Early initiation of chemotherapy to optimize out-
come and minimization of unnecessary invasive pro-
cedures clearly are desirable in this patient
population. Despite potential limitations, the tempo-
ral rate of total serum bilirubin decline provides a
practical surrogate marker to predict the need for
further endoscopic intervention or evaluation. The
current results demonstrate that the rate of bilirubin
normalization after biliary stenting depends highly
on the prestent bilirubin level. This study supports
the conclusion that evaluation for endoscopic inter-
vention should be considered in patients with malig-
nant biliary obstruction who fail to achieve adequate
normalization of serum bilirubin within 6 weeks if
their prestent bilirubin level was 10 mg/dL and
within 3 weeks if their prestent bilirubin level was
<10 mg/dL.
REFERENCES
1. Martin JA, Slivka A, Rabinovitz M, et al. ERCP and stent
therapy for progressive jaundice in hepatocellular carci-
noma: which patients benefit, which patient don’t. Dig Dis
Sci. 1999;44:1298–1302.
2. Eklund JW, Trifilio S, Mulcahy MF. Chemotherapy dosing in
the setting of liver dysfunction. Oncology (Williston Park).
2005;19:1057–1063.
3. Donelli MG, Zucchetti M, Munzone E, et al. Pharmacoki-
netics of anticancer agents in patients with impaired liver
function. Eur J Cancer. 1998;34:33–46.
4. Adler DG, Baron TH, Davila RE, et al. Standards of Practice
Committee of the American Society for Gastrointestinal
Endoscopy. ASGE guideline: the role or ERCP in diseases of
the biliary tract and the pancreas. Gastrointest Endosc.
2005;62:1–8.
5. Leung J. Fundamentals of ERCP. In: Cotton PB, Leung J,
eds. Advanced Digestive Endoscopy: ERCP. Malden, Mass:
Blackwell Publishing, Ltd.; 2006:17–80.

6. Venook AP, Egorin MJ, Rosner GL, et al. Phase 1 and phar-
macokinetic trial of gemcitabine in patients with hepatic
and renal dysfunction: Cancer and Leukemia Group B
9565. J Clin Oncol. 2000;18:2780–2787.
7. Abraham NS, Barkun JS, Barkun AN, et al. Palliation of malig-
nant biliary obstruction: a prospective trial examining impact
on quality of life. Gastrointest Endosc. 2002;56:835–841.
8. Pereira-Lima JC, Jacobs R, Meier M, et al. Endscopic stent-
ing in obstructive jaundice due to liver metastases: does it
have a benefit for the patient? Hepatogastroenterology.
1996;43:944–948.
9. Arguedas MR, Huedebert GH, Stinnett AA, et al. Biliary
stents in malignant obstructive jaundice due to pancreatic
carcinoma: a cost-effective analysis. Am J Gastroenterol.
2002;97:898–904.
10. Indar A, Lobo D, Gilliam A, et al. Percutaneous biliary
metal wall stenting in malignant obstructive jaundice. Eur
J Gastroenterol Hepatol. 2003;15:915–919.
Bilirubin Regression after Stenting/Weston et al. 2423
11. Rai R, Dick R, Doctor N, et al. Predicting early mortality
following percutaneous stent insertion for malignant bili-
ary obstruction: a multivariate risk factor analysis. Eur J
Gastroenterol Hepatol. 2000;12:1095–1100.
12. Ballinger AB, McHugh M, Catnach SM, et al. Symptom
relief and quality of life after stenting for malignant bile
duct obstruction. Gut. 1994;35:467–470.
13. Green RM, Flamm S. AGA technical review on the evalua-
tion of liver chemistry tests. Gastreoenterology. 2002;123:
1367–1384.
14. Sherlock S, Dooley J (eds.). Cholestasis. In: Diseases of the
Liver and Biliary System, 11th ed. Malden, Mass: Blackwell
Publishing, Ltd.; 2002:219–240.
15. Kamath PS. Clinical approach to the patient with abnormal
liver test results. Mayo Clin Proc. 1996;71:1089–1094.
16. Watanapa P. Recovery patterns of liver function after com-
plete and partial surgical biliary decompression. Am J Surg.
1996;171:230–234.