Infected Bilomas in Liver Transplantation • CID 2004:39 (15 August) • 517
M A J O R A R T I C L E
Infected Bilomas in Liver Transplant Recipients:
Clinical Features, Optimal Management, and Risk
Factors for Mortality
Nasia Safdar,1Adnan Said,2Michael R. Lucey,2Stuart J. Knechtle,3Anthony D’Alessandro,3Alexandru Musat,2
John Pirsch,3John McDermott,4Munci Kalayoglu,3and Dennis G. Maki1
of Surgery, and
1Infectious Diseases and
4Section of Interventional Radiology, Department of Radiology, University of Wisconsin, Madison
2Gastroenterology and Hepatology, Department of Medicine,
3Section of Transplant Surgery, Department
plantation. Limited data exist on management and outcome of biloma.
We report a cohort study of 57 liver transplant recipients with posttransplantation bilomas un-
dertaken to identify the clinical features of biloma, management strategies, and outcome.
Fever (44%) and abdominal pain (40%) were the most common presenting symptoms, but one-
third of patients were asymptomatic; 79% had elevated hepatic enzyme levels. Patients without hepatic artery
thrombosis (HAT) had the highest rates of resolution with percutaneous drainage and anti-infectivetherapy(64%).
Retransplantation was necessary in 64% of patients with HAT and biloma. Independent predictors of resolution
with nonsurgical therapy were absence of HAT (odds ratio [OR] 7.69;
9.09;) or enterococcal infection (OR, 7.69;P p .02P p .03
mortality (Cox proportional hazard ratio [HR], 2.38;P p .008
). Predictors of mortality by multivariable analysis included renal insufficiency (OR, 12.51;.0001
infection with Candida species (OR, 4.93;) or gram-negative bacilli (OR, 9.12;P p .03
Posttransplantation biloma should be suspected in patients with fever or abdominal pain or
abnormalities of hepatic enzymes, and it can be confirmed by computerized tomography and radiographically
guided aspiration. Bilomas are most likely to be successfully treated nonsurgically in patients without HAT and
without Candida or enterococcus infection.
Infected hepatic fluid collections (bilomas) are a major infectious complication of liver trans-
) and absence of Candida (OR,P p .01
). Patients with bilomas had significantly greater
, by log rank test) and graft loss (HR, 4.31;P !
) orP p .02
).P p .01
Despite major advances in liver transplantation over
the past 2 decades, biliary complications—inparticular,
infected intrahepatic or parahepatic bilious collections
(bilomas)—present a major challenge in the treatment
of liver transplant recipients . Posttransplantation
bilomas are associated with greatly increased costs, a
high rate of retransplantation, and significantmortality
The traditional approach to the management of he-
patic biloma has been early retransplantation [4–6].
Limited experience from recent small studies suggests
Received 8 January 2004; accepted 31 March 2004; electronically published 3
Reprints or correspondence: Dr.DennisG.Maki,H4/572ClinicalSciencesCenter,
University of Wisconsin Hospital and Clinics, 600 Highland Ave., Madison, WI
Clinical Infectious Diseases 2004;39:517–25
? 2004 by the Infectious Diseases Society of America. All rights reserved.
that some patients with bilomas can be successfully
treated nonsurgically with percutaneous drainage and
prolonged anti-infective therapy, without retransplan-
tation [1, 2, 7–9]. Clinical features that indicate that
this conservative approach will succeed have yet to be
We describe the clinical features of bilomaandreport
patient and graft outcomes, includingindependentpre-
dictors of mortality in patients with bilomas. Finally,
we propose a management strategy, identifyingpatients
in whom nonsurgical management is most likely to be
Using a prospectively maintained database, we identi-
fied all patients undergoing liver transplantation at the
University of Wisconsin (Madison) from 1 January
1994 through 31 October 2001 whodeveloped?1post-
transplantation biloma. Two uninfected livertransplant
by guest on November 7, 2015
518 • CID 2004:39 (15 August) • Safdar et al.
subjects) were randomly selected for each patient with biloma.
Most liver transplantations were performed using piggyback
surgical technique; venovenous bypass was used in only 33
patients. In most patients (72%), a duct-to-duct biliary anas-
tomosis was constructed; percutaneous biliary drainage (T
tube) was established at surgery in 82%. In general, T tubes
were removed 3–4 months after transplantation. Ceftriaxone
was given preoperatively and for 48 h after transplantation;
selective digestive decontamination was not used. The standard
immunosuppressive protocol consisted of tacrolimus or cy-
closporine and prednisone and has been detailed elsewhere
. Ultrasonography was performed on the firstpostoperative
day to assess hepatic artery patency; subsequent examinations
were obtained as indicated clinically.
“Hepatic artery thrombosis” was defined as
a lack of arterial flow determined by doppler ultrasonography
and, in most cases, confirmed by angiography. “Hepatic artery
stenosis” was defined as reduced flow, as shown by increased
velocities on doppler ultrasonography and, in most cases, con-
“Biloma” was defined as an intrahepatic or parahepatic fluid
collection with typical appearance on ultrasonography, CT, or
cholangiography. Parahepatic fluid collections were typically
associated with extrahepatic bile duct leaks (most often anas-
tomotic leaks). All intrahepatic and parahepatic bilious collec-
tions associated with concurrent biliary abnormalities, such as
strictures, were included. Parahepatic collections detected in
the first 2 weeks after liver transplantation, most of which were
infected hematomas and postoperative parahepatic abscesses,
Abnormal liver function test results were defined as alkaline
phosphatase, aspartate aminotransferase, or alanine amino-
transferase levels 11.5-fold greater than the upper limit of nor-
mal or as a total bilirubin level of 12 mg/dL.
The outcomes recorded were resolution after retransplan-
tation, resolution with nonsurgical management (completeres-
olution of the fluid collection on imaging, with no recurrence
during 11 year of observation, afterdiscontinuationofdrainage
and antimicrobial therapy), persistence of biloma without graft
loss (continued catheter drainage and antimicrobial therapy at
the end of the study), or death.
All patients presenting with bilomas un-
derwent early diagnostic percutaneous aspiration, with Gram
stain and culture of the fluid sample. Initial anti-infectivether-
apy, usually guided by the results of the Gram stain of the
aspirate, was modified if necessitated by the results of cultures
and susceptibility tests.
Drainage of noncommunicating collections was achieved
percutaneously under fluoroscopic guidance by an interven-
tional radiologist. If the abscess cavity communicated with the
biliary tree, transbiliary drainage was attempted by advancing
the percutaneous catheter through the abscess cavity into the
bile ducts and duodenum . The initial goal was to manage
the biloma with drainage and prolonged (?4-week) anti-
infective therapy. The most frequentlyusedantibioticsfortreat-
ment of biloma were third-generation cephalosporins, fluor-
oquinolones, and extended-spectrum penicillins. The choice of
antimicrobial agent used was based on the findings of theGram
stain and the results of susceptibility tests.
The patency of drainage catheters was confirmed fluoro-
scopically routinely every 2–4 weeks (or earlier, if clinically
indicated). Drainage catheters were routinely replaced every 4–
8 weeks. Drainage was continued until abscess cavities resolved
or retransplantation was considered necessary. If bilomas re-
curred after removal of the drainage catheter and prolonged
anti-infective therapy, catheter drainage was reestablishedusing
the same protocol. Endoscopic retrograde cholangiopancrea-
ticographic stenting was used for extrahepatic biliary strictures
or leaks . In patients with fulminant sepsis at the outset of
treatment who had evidence of progressive graft failure or con-
tinued signs of active infection (especially bacteremia or can-
didemia), retransplantation was undertaken in most cases.
SAS software, version 8.1 (SAS Institute). For patient and graft
survival after transplantation, Kaplan-Meier statistics (log rank
test) were used to test for significant differences between pa-
tients with bilomas and control subjects. Life table analysis
using Kaplan-Meier statistics was also used to assess factors
potentially associated with decreased survival in patients with
bilomas; risk factors with a P value of ?.1 onunivariateanalysis
were entered into a Cox proportional hazards multivariable
model to identify risk factors independently associated with
Factors that predicted successful resolution of biloma with
nonsurgical management were first assessed by univariate
analysis using x2and Fisher’s exact testsforcategoricalvariables
and Student’s t test for continuous variables. A multivariable
logistic regression model was constructed to identify factors
independently predictive of success with nonsurgical manage-
ment, incorporating variables found to be significantpredictors
in the univariate analysis (
P ? .1
els was assessed by the Hosmer-Lemeshow test . Two-sided
P values of ?.05 were considered to be statistically significant.
For transplant recipients, the following potential risk factors
for decreased survival were tested in univariate life table
analyses: age at transplantation, sex, race, Model for End-Stage
Liver Disease (MELD) score before transplantation,Childscore
before transplantation, etiology of liver disease, previous liver
transplantation, cytomegalovirus (CMV) status of recipient,
high-risk CMV status (CMV-positive donor with a CMV-
negative recipient), pretransplantation comorbid conditions
). Goodness-of-fit of the mod-
by guest on November 7, 2015
Infected Bilomas in Liver Transplantation • CID 2004:39 (15 August) • 519
oma in 57 liver transplant recipients.
Signs and symptoms of bil-
Signs or symptom
Fever and/or chills
Nausea or vomiting
aDefined as ?2 of the following signs: tem-
perature of 138?C or !36?C, heart rate of 190
beats/min, respiratory rate of 120 breaths/min,
and WBC count of 112,000 or !4000 cells/mm3
or more than 10% bands .
(i.e., diabetes mellitus, coronary artery disease, chronic pul-
monary disease, deep venous thrombosis, pulmonary embo-
lism, coagulation abnormalities, peripheral vascular disease,
preexisting renal insufficiency [serum creatinine level, 11.8mg/
dl], and hepatocellular carcinoma [both pretransplantation
or found at transplantation]), and alcohol abuse after
For donors, the following potential risk factors for decreased
survival were tested in univariate life table analyses: age at do-
nation, sex, race, ABO blood group, and CMV infection status.
The following surgical and perioperative risk factors for de-
creased survival were tested in univariate life table analyses:
cold ischemia time, surgical time, type of biliary anastomosis,
hepatic artery reconstruction, T tube use and duration, and
blood products used intraoperatively.
The following postoperative risk factors for decreased sur-
vival were tested in univariate life table analyses: hepatic artery
thrombosis and stenosis, biliary tree abnormalities, hepatic ar-
tery thrombosis management (i.e., surgical, angioplasty, or
other), ursodeoxycholic acid use, anticoagulation therapy (i.e.,
aspirin, clopidogrel, or warfarin), other sites or types of infec-
tion (including bacteremia), immunosuppressive therapy
(maintenance, induction, and rejection), number of rejections
(total and in first month and first year), time from transplan-
tation to biloma formation, and number of biloma collections
Of the 492 patients who underwent livertransplantationduring
1994–2001, a total of 57 patients(11.5%)developed?1biloma.
Ninety-eight percent of the entire study population(57patients
with bilomas and 114 control subjects) had undergone ortho-
topic transplantation, and only 2 patients received liver trans-
plants from living-related donors.
Clinical and laboratory features of biloma at presenta-
The most common presenting features of bilomas were
fever (in 44% of patients) and abdominal pain (in 40%) (table
1). A septic presentation  was uncommon (14%). Persis-
tently elevated liver enzyme levels were seen in 79% of patients
at diagnosis, although 68% of subjects had intermittent ele-
vations weeks to months before the diagnosis of biloma (table
2). Twenty (35%) of patients with bilomas were asymptomatic
and were identified only because of unexplained elevations of
hepatic enzymes. Hyperbilirubinemia was infrequent (28%).
Elevated WBC counts (50%) and leukopenia (18%) had little
correlation with fever or signs of sepsis.
The most common primary infecting path-
ogens, on the basis of the initial percutaneous aspirate of the
biloma, were enterococci (37% of the 57 patients)—48% of
which exhibited resistance to vancomycin (VRE)—coagulase-
negative staphylococci (26%), and Candida species (26%).
Gram-negative organisms were identified in only 16% of cases;
one-half of the organisms were identified as Pseudomonas spe-
cies. Anaerobic bacteria were recovered from only 5% of
One or more new “superinfecting” pathogens (156 isolates)
were later cultured from the percutaneous drainage catheter in
95% of cases. A much larger proportion of the organisms re-
covered from later cultures were multidrug resistant, including
VRE (44% of patients), Candida species (79%), and drug-
resistant gram-negative bacteria (35%; Citrobacter freundii,
Stenotrophomonas maltophilia, and extended-spectrum b-lac-
tamase–producing Escherichia coli).
Diagnosis of biloma.
The median interval from trans-
plantation to the diagnosis of biloma was 9.1 weeks, with a
wide range (2–377 weeks); 95% of the cases were identified in
the first year after transplantation. The median duration of
follow-up after biloma resolution was 143 weeks (2.7 years);
no patients were lost to follow-up.
A total of 102 bilomas were detected in the 57 affected pa-
tients; 68 (67%) of the bilomas were intrahepatic,and34(33%)
were parahepatic (i.e., subhepatic or parahepatic; they were
most often found in the region of the porta hepatis). The
number of collections ranged from 1–4 per patient (mean, 1.8
collections). Thediameterof bilomaswas1–11cm,withamean
diameter of 5 cm.
Biphasic contrast-enhanced CT was the initial imaging test
of choice for the diagnosis of bilomas in 48 (84%) of 57 cases
(figure 1). Ultrasonography was the initial diagnostic imaging
modality in only 9 cases (16%).
Bilomas on CTs appeared as low attenuation lesions (figure
1) . Associated biliary tree abnormalities were detected in
37 patients (65%) and ranged from anastomotic strictures to
ischemic-type biliary strictures (multiple intrahepaticstrictures
with extravasation and poor contrast filling). The cholangio-
graphic modality used to delineate these abnormalities were T-
by guest on November 7, 2015
520 • CID 2004:39 (15 August) • Safdar et al.
57 liver transplant recipients.
Laboratory abnormalities associated with bilomas in
Percentage of patients with abnormal
results of ?1 liver function testa
Specific laboratory abnormality
Mean ? SD
Alkaline phosphatase, IU/mL
Mean ? SD
gGlutamyl transferase, IU/mL
Mean ? SD
Aspartate aminotransferase, IU/mL
Mean ? SD
Alanine aminotransferase, IU/mL
Mean ? SD
International normalized ratio
Mean ? SD
Mean ? SD
Mean ? SD
WBC count, cells ?106/mm3
Mean ? SD
Mean ? SD
Platelet count, platelets ?106/mm3
Mean ? SD
1.8 ? 0.3
235.0 ? 18
353.0 ? 48
188.0 ? 25
124.0 ? 17
1.1 ? 0.03
3.0 ? 0.1
1.4 ? 0.1
10.6 ? 0.8
13.4 ? 1.9
198.0 ? 17
aValue of 11.5-fold greater than the upper limit of normal.
plant recipient. Biloma is seen as a low attenuating lesion without an
CT image showing an intrahepatic biloma in a liver trans-
tube cholangiography (8 cases), endoscopic retrograde cholan-
giopancreaticography (10 cases), percutaneous transhepatic
cholangiography (17 cases), or hepatoiminodiacetic acid scan
Outcome with different management strategies.
57 patients with bilomas, 47 (82%) were treated initially with
a trial of percutaneous drainage and prolonged anti-infective
therapy (figure 2). Five patients (9%) were successfully treated
with prolonged anti-infective therapy after a single aspiration
of the fluid collection, without long-term catheter drainage; all
had extrahepatic bilomas and no evidence of ischemic cholan-
gitis or central biliary necrosis.
Five patients (9%) underwent early retransplantation with-
out a trial of prolonged percutaneous drainage and antimicro-
bial therapy; 4 patients presented at the outset with frank sepsis
and massive hepatic necrosis. These 5 patients also had a very
early onset of biloma after transplantation (mean time after
transplantation, 4 weeks), compared with the rest of the cohort
(mean time after transplantation, 44 weeks). Ischemic cholan-
gitis and hepatic artery thrombosis (HAT) were seen in 3 of
these patients. Despite early retransplantation, 3 of the 5 pa-
Of the remaining 47 patients who were treated with a trial
of long-term percutaneous catheter drainage and anti-infective
therapy, 9 (19%) died, and 18 (38%) subsequently underwent
retransplantation. Of the 18 who underwent retransplantation,
3 developed recurrent bilomas, and 1 patientsubsequentlydied.
Twenty patients (43%) were successfully treated nonsurgically,
all of whom had complete resolution of bilomas.
Overall, with this management strategy, bilomas in 35 (61%)
of the 57 cases resolved completely; two-thirds (23 of 35) of
these cases resolved without need for retransplantation. In 12
cases (21% of the entire population), bilomas were cured only
Twenty (35%) of the 57 patients died, 11 of 23 (48%) fol-
lowing retransplantation and 9 of 34 (26%) during nonsurgical
management. The same survival differences were observed
when the analyses were restricted to the 38 patients with in-
trahepatic bilomas: 29 (76%) underwent prolonged percuta-
and 12 (32%) died. Of the 19 patients who underwent retran-
splantation, 6 (32%) died, and of the 19 treated with drainage
and antimicrobial therapy alone, 6 (32%) died.
Patient and graft outcomes.
Cumulative mortality after
by guest on November 7, 2015
Infected Bilomas in Liver Transplantation • CID 2004:39 (15 August) • 521
Outcomes for 57 patients with post–orthotopic liver transplantation (OLT) biloma, by treatment strategy
were associated with greater mortality in the first 5 years after transplantation (
Rate of survival after liver transplantation for 57 patients with bilomas and 114 control subjects, by Kaplan-Meier analysis. Bilomas
P p .008 , by log-rank test).
liver transplantation was much greater inpatientswithbilomas,
compared with transplant recipients without bilomas (hazard
ratio [HR], 2.38; 95% CI, 1.22–4.63;
test) (figure 3). The 1-year mortality rate was 12% among
patients with bilomas, compared with 6% among control
subjects (HR, 2.32; 95% CI, 0.78– 6.90;
mortality rate was 21% among patients with bilomas and 11%
, by log-rankP p .008
). The 3-yearP p .13
among control subjects (HR, 2.03; 95% CI, 0.91–4.52; P p
Cumulative graft loss was also higher in recipients of liver
transplants whose cases were complicated by bilomas (HR,
4.31; 95% CI, 2.45–7.58;P ! .0001
year graft loss rate was 39% among transplant recipients with
bilomas, compared with 7% among control subjects (HR, 6.62;
, by log rank test). The 1-
by guest on November 7, 2015
522 • CID 2004:39 (15 August) • Safdar et al.
were associated with greater graft loss in the first 5-years after liver transplantation (
Rate of graft survival after liver transplantation for 57 patients with bilomas and 114 control patients, by Kaplan-Meier analysis. Bilomas
, by log-rank test).P ! .0001
control subjects after liver transplantation.
Causesof death amongatotalof57patientswithbilomaand114matched
Cause of death
Intra-abdominal infection, sepsisa
Complications after liver retransplantationb
Relapse of primary liver disease
No. (%) of patients who died
Patients with biloma
(n p 20)
(n p 15)
bPostretransplantation hemorrhage (in 2 patients) and graft failure (in 3 patients).
95% CI, 2.96–14.83;
was 44% among transplant recipients with bilomas and 13%
among control subjects (HR, 4.28; 95% CI, 2.22–8.24; P !
) (figure 4). .0001
Causes of death are summarized in table 3. In transplant
recipients with bilomas, sepsis was the most frequent cause of
death (50%) (relative risk [RR] for death due to sepsis in pa-
tients with bilomas, 4.81; 95% CI, 1.03–22.57;
Predictors of successful medical therapy without retran-
As noted, 34 patientsweretreatednonsurgically,
without the need for retransplantation; of these, 23 (68%) had
complete resolution of the bilomas. In a multivariable logistic
regression model (table 4), patients who did not have HAT
(OR, 7.69; 95% CI, 1.49– 33.33;
not have bilomas infected with enterococci (OR, 7.69; 95% CI,
); at 3 years, the rate of graft lossP ! .0001
).P p .04
) and those who didP p .01
therapy. The Hosmer-Lemeshow test confirmed the validity of
this model (, by Pearson x2test). Of note, parahepaticP p .90
biloma was not associated with agreaterlikelihoodofsuccessful
nonsurgical management than was intrahepatic biloma (OR,
0.55; 95% CI, 0.16–1.89;P p .34
Predictors of survival.
hazards model (table 5), independent predictors of mortality
in liver transplant recipients with biloma included pretran-
splantation renal disease (OR, 12.51; 95% CI, 1.60–98.20;
) and infection with either gram-negative bacilli (OR,P p .02
9.12; 95% CI, 1.61–51.53;P p .01
4.93; 95% CI, 1.17–20.75;P p .03
was not associated with a lower survival rate, the retransplan-
) or Candida species (OR, 9.09; 95% CI,
) were more likely to respond to nonsurgical
P p .03
P p .02
) or Candida species (OR,
) (table 6). Although HAT
by guest on November 7, 2015
Infected Bilomas in Liver Transplantation • CID 2004:39 (15 August) • 523
Predictors of successful resolution of bilomas in 34 patients treated nonsurgically, by mul-
Crude OR (95% CI)Adjusted OR (95% CI)P
Absence of hepatic artery thrombosis
Absence of biloma infection due to enterococci
Absence of biloma infection due to Candida species
aPatients with hepatic artery thrombosis or with Candida or enterococcal infection of the biloma were far less likely to
resolve their biloma with nonsurgical therapy alone.
Data are from goodness-of-fit models; P p .90, by Pearson x2test; receiver operating characteristic c-statistic,
bilomas, by Cox proportional hazards modeling.
Independent predictors of mortality in patients with
Pretransplantation renal diseasea
Adjusted OR (95% CI)P
aCreatinine level, 11.8 mg/dL.
tation rate was much higher in patients with HAT (64%), and
overall cure was achieved in only 32%. In patients with hepatic
artery stenosis or with no discernible hepatic artery abnor-
malities, retransplantation rates were 28% and 14%, respec-
tively, and resolution was achieved nonsurgically in 72% and
When 23 patients with biloma who underwent retransplan-
tation were compared with 34 patients who were treated non-
surgically, patients undergoing retransplantation had a worse
outcome than did patients treated nonsurgically (5-year mor-
tality rate, 21% vs. 35% for patients treated nonsurgically and
those who underwent transplantation, respectively), although
this difference did not reach statistical significance (HR, 0.50;
95% CI, 0.18–1.38; ) (figure 5).P p .18
Bilomas after liver transplantation are associated with greatly
increased costs resulting from repeated hospitalizations and,
frequently, the need for retransplantation [16, 17]. Our data
show that bilomas after liver transplantationareassociatedwith
worse patient and graft survival (figures 3–5).
Bilomas most frequently present with fever or abdominal
pain; however, in one-third of cases, the patient is asymptom-
atic but has unexplainedabnormalitiesinhepaticenzymelevels,
investigation. The majority (68%) of our patients had interim
elevations in liver enzyme levels long before biloma was rec-
ognized. Patients with fever or other signs of cryptogenic in-
fection or with unexplained hepatic enzyme abnormalitiesafter
liver transplantation should undergo an imaging study, such as
ultrasonography or CT.
If there is cholangiographic evidence of free biliary com-
munication with the abscess cavity, endoscopic stenting is
often helpful in establishing drainage (11 patients [19%] were
stented successfully). Endoscopic stenting is also helpful for
the management of associated biliary tractabnormalities,such
as extrahepatic strictures or leaks. For abscesses that do not
communicate with the biliary tree on cholangiography, per-
cutaneous catheter drainage of each biloma can be achieved
in most cases by an interventional radiologist.
Hepatic artery patency is very importantindecidingbetween
early retransplantation and nonsurgical management. When
HAT is present, the likelihood of successful resolution of the
biloma with nonsurgical management is much poorer; nearly
two-thirds of our patients with HAT-associated biloma ulti-
mately needed retransplantation. However, more than one-
third (36%) of patients with HAT had resolution of thebilomas
nonsurgically, and patients with bilomas infected with micro-
organisms other than enterococci or Candida species were far
more likely to be cured nonsurgically. Enterococcal infections
are characteristically persistent, even if the intrinsic virulence
of enterococci  is generally considered to be less than that
of Candida  or gram-negative bacilli . With hepatic
artery stenosis, perhaps because of preservation of some flow
to the ischemic parenchyma and biliary system, the resultswith
nonsurgical therapy appear to be much better, and our retran-
splantation rate was only 28%.
In the absence of graft failure, provided that sepsis can be
treated adequately with drainage and anti-infective therapy,
nonsurgical therapy can obviate the need for retransplantation
in many patients with bilomas, even in the presence of HAT.
If there is no improvement clinically or if there is worsening
of graft function during nonsurgical therapy, then retransplan-
tation is necessary. Our data showed no difference in survival
rates between patients who underwent retransplantation early
versus those who underwent retransplantation after an initial
trial of drainage therapy was unsuccessful.
Although patients with bilomas fare worse as a group, to our
not been previously described. In our cohort, patients with
by guest on November 7, 2015
524 • CID 2004:39 (15 August) • Safdar et al.
Table 6.Outcome for 57 patients with bilomas, by infecting microbial species.
No. of patients cured/
no. of patients treated (%)
Coagulase-negative Staphylococcus species
patients (91%) were initially treated nonsurgically, 5 (9%) underwent immediate retransplantation, and 18 (32%) were initially treated
nonsurgically but later underwent retransplantation.
The 2 groups of patients (those treated nonsurgically and those treated by retransplantation) are not mutually exclusive: 52
for patients treated surgically versus those treated conservatively (
Rate of patient survival for 57 patients with bilomas, by type of management. No statistically significant difference in survival was found
).P p .18
bilomas who had preexisting renal disease or infection with
gram-negative bacilli or Candida species had greatly increased
mortality (table 5). Preexistent renal disease has long been as-
sociated with worse outcomes in patients with native liver dis-
ease and hepatic abscess , as well as in liver transplant
recipients [22, 23]. Infections with gram-negative bacilli or
Candida species have also been shown to portend a poor out-
come in liver transplant recipients . Our resultssuggestthat
prevention of Candida infection and gram-negative bacterial
infection would yield a substantial survival benefit. Although
controversial, some liver transplant centers have reported re-
duced rates of gram-negative bacterial and fungal infections in
hepatic transplant recipients with the use of selective digestive
decontamination [25, 26]; however, the long-term graft and
patient survival in our center, which does not use selective
digestive decontamination, is as high or higher than in those
centers that use it.
Our study has several limitations. First, our study focused
on bilomas and was not designed to describe all manifestations
of liver damage caused by HAT or to analyze risk factors for
HAT. It is well known that, early in the posttransplantation
course, complete occlusion of the hepatic artery can result in
massive necrosis of the liver and acute graft loss . In these
cases, early retransplantation is the best management strategy
by guest on November 7, 2015
Infected Bilomas in Liver Transplantation • CID 2004:39 (15 August) • 525
, and we do not advocate nonsurgical management in this
circumstance. Secondly, our patients with bilomas were not
randomized to receive nonsurgical management or early re-
transplantation, and differences in outcomes may have been
influenced by uncharacterized differences in the study popu-
lations rather than by the intervention itself. However, in our
multivariable analysis, we were unable to find a difference in
survival rates between patients treated nonsurgically and those
who required surgical management.
We would like to thank Barbara Voss and Karen Armstrong for database
Oscar Rennebohm Foundation of Madison, Wis-
consin (unrestricted gift for research; to D.G.M.).
Conflict of interest.
All authors: No conflict.
1. Tachopoulou OA, Vogt DP, Henderson JM, Baker M, Keys TF. Hepatic
abscess after liver transplantation: 1990–2000. Transplantation 2003;
2. Kaplan SB, Zajko AB, Koneru B. Hepatic bilomas due to hepatic artery
thrombosis in liver transplant recipients: percutaneous drainage and
clinical outcome. Radiology 1990;174:1031–5.
3. Margarit C, Hidalgo E, Lazaro JL, Murio E, Charco R, Balsells J. Biliary
complications secondary to late hepatic artery thrombosisinadultliver
transplant patients. Transpl Int 1998;11(Suppl 1):S251–4.
4. Slooff MJ, Klompmaker IJ, Grond J, et al. Mortality after orthotopic
liver transplantation: an analysis of the causes of death in the first 50
liver transplantations in Groningen, The Netherlands. Transpl Int
5. Tzakis AG, Gordon RD, Shaw BW Jr, Iwatsuki S, Starzl TE. Clinical
presentation of hepatic artery thrombosis after liver transplantation in
the cyclosporine era. Transplantation 1985;40:667–71.
6. Pariente D, Riou JY, Schmit P, et al. Variability of clinical presentation
of hepatic artery thrombosis in pediatric liver transplantation: role of
imaging modalities. Pediatr Radiol 1990;20:253–7.
7. Rabkin JM, Orloff SL, Corless CL, et al. Hepatic allograft abscess with
hepatic arterial thrombosis. Am J Surg 1998;175:354–9
8. Gunsar F, Rolando N, Pastacaldi S, et al. Late hepatic arterythrombosis
after orthotopic liver transplantation. Liver Transpl 2003;9:605–11.
9. Stange BJ, Glanemann M, Nuessler NC, Settmacher U, Steinmuller T,
Neuhaus P. Hepatic artery thrombosis after adult liver transplantation.
Liver Transpl 2003;9:612–20.
10. D’Alessandro AM, Hoffmann RM, Knechtle SJ, et al. Liver transplan-
tation from controlled non–heart-beating donors. Surgery 2000;128:
11. Mueller PR, Ferrucci JT Jr, Simeone JF, et al. Detection and drainage
of bilomas: special considerations. AJR Am J Roentgenol 1983;140:
12. Macfarlane B, Davidson B, Dooley JS, et al. Endoscopic retrograde
cholangiography in the diagnosis and endoscopic management of bil-
iary complications after liver transplantation. Eur J Gastroenterol He-
13. Lemeshow S, Hosmer DW Jr. A review of goodness of fit statistics for
use in the development of logistic regression models. Am J Epidemiol
14. American College of Chest Physicians/SocietyofCriticalCareMedicine
Consensus Conference: definitions for sepsis and organ failure and
guidelines for the use of innovative therapies in sepsis. Crit Care Med
15. Hoffer FA, Teele RL, Lillehei CW, Vacanti JP. Infected bilomas and
hepatic artery thrombosis in infant recipients of liver transplants: in-
terventional radiology and medical therapy as an alternative to re-
transplantation. Radiology 1988;169:435–8.
16. Hollenbeak CS, Alfrey EJ, Souba WW. The effect of surgical site in-
fections on outcomes and resource utilization after liver transplanta-
tion. Surgery 2001;130:388–95.
17. Tzakis AG. Clinical presentation of hepatic artery thrombosisafterliver
transplantation in the cyclosporine era. Transplantation 1985;40:
18. Jett BD, Huycke MM, Gilmore MS. Virulence of enterococci. Clin
Microbiol Rev 1994;7:462–78.
19. Haynes K. Virulence in Candida species. Trends Microbiol 2001;9:
20. Young LS, Stevens P, KaijserB.Gram-negativepathogensinsepticaemic
infections. Scand J Infect Dis Suppl 1982;31:78–94.
21. Pitt HA, Zuidema GD. Factors influencing mortality in the treatment
of pyogenic hepatic abscess. Surg Gynecol Obstet 1975;140:228–34.
22. Chou FF, Sheen-Chen SM, Chen YS, Chen MC, Chen FC, Tai DI.
Prognostic factors for pyogenic abscess of the liver. J Am Coll Surg
23. Nair S, Verma S, Thuluvath PJ. Pretransplant renal function predicts
survival in patients undergoing orthotopic liver transplantation. He-
24. Singh N, Wagener MM, Marino IR, Gayowski T. Trends in invasive
fungal infections in liver transplant recipients: correlation with evo-
lution in transplantation practices. Transplantation 2002;73:63–7.
25. Wiesner RH.Selectiveboweldecontaminationforinfectionprophylaxis
in liver transplantation patients. Transplant Proc 1991;23:1927–8.
26. Wiesner RH, Hermans PE, Rakela J, et al. Selective bowel decontam-
ination to decrease gram-negative aerobic bacterial and Candida col-
onization and prevent infection after orthotopic liver transplantation.
by guest on November 7, 2015