Orthotopic liver transplantation in human immunodeficiency virus (HIV)-positive patients: Outcome of 7 patients from the Bonn cohort



The outcome and clinical features of 7 HIV-positive patients who were liver transplanted at Bonn University in the era of highly active antiretroviral therapy (HAART) between 1997 and 2004, analyzed by retrospective chart review, are reported. Reasons for orthotopic liver transplantation (OLT) were end-stage liver disease due to chronic hepatitis C (n = 4) or hepatitis B (n = 1) or acute liver failure due to fulminant hepatitis B (n = 2). Immunosuppression was based on cyclosporine A and prednisone. HAART was reinitiated 1 month after transplantation, and immunosuppression was carefully adapted to account for drug-drug interactions between cyclosporine A and protease inihibitors. Prednisone was withdrawn 5 months (median) after OLT when immunosuppression had been reliably established in the presence of HAART. One patient died 95 days after OLT due intrathoracic hemorrhage, whereas 6 patients were alive at a median of 24 months. A single episode of acute rejection was observed. The spectrum of postoperative complications was no different from HIV-negative patients apart from Kaposi's sarcoma and multicentric Castleman's disease in a single patient. Recurrent hepatitis B infection was efficiently prevented, whereas hepatitis C reinfection occurred in all 4 patients who had preexisting hepatitis C. Earlier reports on fatal courses of recurrent hepatitis C infection, high rates of organ rejection, and HAART-related liver toxicity were not observed in our patients. In conclusion, even though preliminary, our data suggest that outcomes after liver transplantation of HIV-infected patients can be improved. (Liver Transpl 2005;11:1515–1521.)

Highly active antiretroviral therapy (HAART) of human immunodeficiency virus (HIV) infection effectively prevents opportunistic diseases and has greatly enhanced overall life expectancy in HIV-positive patients.1, 2 On the other hand, several large cohort studies have shown that morbidity due to chronic infection with hepatitis B (HBV) and hepatitis C (HCV) viruses has become a leading cause of death in HIV-infected patients with concomitant chronic viral hepatitis.3, 4 Thus, orthotopic liver transplantation (OLT) may be a rational therapeutic option for selected HIV patients with concomitant liver disease. In the HAART era, orthotopic liver transplantations in HIV-infected patients have been performed at various centers in the United States and Europe, with roughly 100 of these cases reported in the literature worldwide.5–13 Under the premise of careful patient selection, overall results have been promising, with survival rates and graft function comparable to HIV-negative patients.

However, high rates of HAART-associated toxicity,7 fatal recurrent hepatitis C,14 and acute organ rejections13 have been troubling transplantation centers and highlight the difficulties encountered in the management of this special patient population. Here, we report on 7 HIV-positive patients who had been listed for OLT at Bonn University between 1997 and 2004. Based on our experiences and comparison with patient management and outcome of other groups, we have derived some suggestions as to how management of HIV-infected liver transplanted patients can be improved.


HIV, human immunodeficiency virus; OLT, orthotopic liver transplantation; HAART, highly active antiretroviral therapy; HBV, hepatitis B virus; HCV, hepatitis C virus; CyA, cyclosporine A; HHV-8, human herpes virus 8; NRTI, nucleoside reverse transcriptase inhibitor.

Materials and Methods

Seven HIV-positive patients have been liver transplanted within the Bonn University transplant program. Criteria for accepting a candidate for OLT were a CD4 count > 100 cells/μl and an HIV-RNA lower than the limit of detection or at least a reasonable option for efficient HAART, in addition to end-stage liver disease or acute liver failure. An AIDS-defining event prior to listing of OLT was not an exclusion criterion per se if the above criteria were met.

Data were collected from the patients' records by means of a structured questionnaire, which was composed of general demographic data (gender, age, body weight), characteristics of HIV infection (date of diagnosis, CDC stadium,15 transmission risk factor, CD4 cell count, HIV-RNA, use of HAART) as well as details of concomitant liver disease (type and date of diagnosis, reason for OLT, biochemical and clinical signs of terminal liver disease). Data corresponding to the time of listing for OLT, the time of OLT itself, (re)initiation of HAART after OLT, and last follow-up were evaluated.

The model for end-stage liver disease (MELD) score16 was used for assessing the severity of liver disease in patients with end-stage liver disease, and a score ≥ 15 was required for listing (patients with hepatocellular carcinoma were listed regardless of their MELD scores). Patients with acute liver failure were evaluated according to the Kings's College criteria.17

Immunosuppression was based on cyclosporine A (CyA) plus prednisone for induction. CyA was dosed according to target trough levels (150 - 200 ng/ml for weeks 0 - 2, 100 - 150 ng/ml for weeks 3 - 11, 75 - 125 ng/ml for week 12 and thereafter). Prednisone was given as boluses of 500 mg on days 0 and 1 and then tapered starting at 1 mg/kg bodyweight. Complete steroid withdrawal was aimed for 12 weeks after OLT. However, prednisone was continued at 10 mg/d until stable CyA levels after initiation of HAART were achieved. CyA trough levels were considered stable if they were repeatedly (at least 4 measurements over a 2-week period) within the desired target range without the need to modify the daily oral CyA dose.

Initiation of HAART was delayed until transaminases had normalized to lower than 3 times the upper limit of normal, and bilirubin was lower than 2 mg/dl. Also, stable CyA blood levels within the desired range (100 - 150 ng/ml) had to be established as defined above. After introduction of HAART, dose adjustments of CyA were made while maintaining a twice daily dosing scheme as described in detail elsewhere.18 Patients were seen at least on a biweekly schedule for the first 6 months and monthly thereafter.

Pegylated interferon ribavirin combination therapy was initiated in patients who had a preexisting hepatitis C virus infection as soon as stable immunosuppression after initiation of HAART was established. Pegylated interferon-alfa-2a (Pegasys®, Hoffmann-La Roche AG, Grenzach-Wyhlen, Germany) was dosed 180 μg once weekly, and ribavirin (Copegus®, Hoffmann-La Roche AG) was dosed according to HCV genotype and body weight between 800 and 1,200 mg per day divided into two doses.

Patients with a preexisting hepatitis B infection were treated with HBs-hyperimmunoglobulin (Hepatect®, Biotest Pharma GmbH, Dreieich, Germany) 10,000 IU i.v. once daily for 1 week starting in the anhepatic phase of transplantation, followed by maintenance therapy by monthly infusions to keep serum levels of anti-HBs > 100 IU/ml. Lamivudine (Epivir®, GlaxoSmithKline GmbH & Co. KG, Munich, Germany) 300 mg once daily was added for patients requiring HAART.

Liver biopsies of the liver graft were not performed per protocol but were done if clinically indicated.

All laboratory tests were performed at the Bonn University Department of Clinical Chemistry and the Institute of Microbiology and Immunology within the clinical routine. In detail, serum/plasma levels of HIV-RNA, HCV-RNA and HBV-DNA were measured by bDNA assays (Bayer Vital GmbH Diagnostika, Fernwald, Germany). Hepatitis C virus genotypes were determined via a second-generation hybridization assay (INNO-LiPA HCV II Kit, Innogenetics, Heiden, Germany). CyA whole-blood levels and alanine aminotransferase (ALT) serum levels were determined via immuno- and enzymatic assays, respectively (Dimension, Dade Behring Inc., Newark, DE). CD4 cell counts were analyzed on a FACS calibur flow cytometer (Becton Dickinson, Heidelberg, Germany).

The study was conducted according to good clinical practice guidelines and in agreement with the declaration of Helsinki and its subsequent revisions.


Seven HIV-positive patients were liver transplanted at Bonn University between 1997 and 2004. Reasons for OLT were acute fulminant hepatitis B infection (n = 2) and end-stage liver disease due to chronic hepatitis B (n = 1) and chronic HCV infection (n = 4) (Table 1). Three patients were incidentally diagnosed to be HIV-positive during workup for OLT, whereas 4 patients were known to be HIV-positive prior to listing. At the time of listing, all patients had CD4 counts > 100/μl and HIV-RNA levels lower than the limit of detection or, in case of detectable HIV-RNA, had viable treatment options to achieve durable suppression of HIV-RNA below the level of detection. One of the 7 patients had a history of an AIDS-defining event in terms of esophageal candidiasis prior listing for OLT (patient no. 7). With HAART, this AIDS-defining event had resolved 3 months prior to listing.

Table 1. Patient characteristics at the time of listing for OLT, complications, and outcome after OLT
Patient No.Age (yr)GenderTransmission Risk FactorCDCCD4Cells [/μl]HIV-RNA [/ml]Liver DiseaseHCV GenotypeKnown Duration of Chronic Hepatitis [yr]Clinical Symptoms of ESLDMELD-ScoreComplications Post-OLTXOutcome
  • Abbreviations: OLT, orthotopic liver transplantation; CDC, Centers for Disease Control and Prevention AIDS-Classification 199315; ESLD, end-stage liver disease; MELD, model for end-stage liver disease16; MSM, men who have sex with men; HCV, hepatitis C; HBV, hepatitis B; HCC, hepatocellular carcinoma; HU, high urgent; n.a., not applicable; CMV, cytomegalovirus.

  • *

    Relative CD4 cell count because the absolute CD4 count was not available.

  • Incidental discovery of HIV infection during the workup for OLT.

  • After emergency thoracic surgery, hypoxic brain edema with apallic state was diagnosed, and exitus letalis followed 95 days after OLT.

143MHemophiliaA221020,828Chronic HCV2a/2c20Fatigue, continuous deterioration of liver function tests15CMV viremia, recurrent cholestatic hepatitis C infectionAlive
234MMSMA150%*81Acute HBVAcute liver failureHUPrimary graft nonfunction, CMV viremiaAlive
352MMSMA2223>500,000Acute HBVAcute liver failureHUKaposi's sarcoma, multicentric Castleman's diseaseAlive
457MMSMA3162180,000Chronic HBV20Ascites, hepatic encephalopathy, peripheral edema25Intrathoracic hemorrhage after placement of a thoracic drainDead
558MHemophiliaA3168639Chronic HCV HCC3a29HCC detected by screeningn.a.Leakage of the biliary tract, occlusion of the hepatic artery with partial infarction of right liver lobeAlive
631MHemophiliaA3239<50Chronic HCV1a13Ascites16CMV viremia, acute graft rejectionAlive
731MHemophiliaC33201,127Chronic HCV1a15Ascites, recurrent pleural effusion, variceal hemorrhage, hepatic encephalopathy, arterioportal shunting with ischaemic bowel damage15Ascites, recurrent cholestatic hepatitis C infectionAlive

Transplantation was done immediately in the two patients with a high-urgency status due to fulminant hepatitis B infection. Patients with an elective listing status due to end-stage liver disease were transplanted after a median of 11 months (range, 1 - 13) on the waiting list. After transplantation, in-hospital care was required for a median of 65 days (range, 14 - 130). Six patients remain alive with a median follow-up of 24 months (range, 6 - 92); one patient died 3 months after OLT from severe intrathoracic hemorrhage after the placement of a thoracic drain (Table 2).

Table 2. Follow-up of the 7 transplanted patients after OLT
Patient No.Follow-up [months]HAARTImmunosuppressionCyA Dose [mg/d]CyA trough level [ng/ml]CD4 Cells [/μl]HIV-RNA [/ml]
  • OLT, orthotopic liver transplantation; HAART, highly active antiretroviral therapy; CyA, cyclosporine A; LPV/r, lopinavir/ritonavir; SQV, saquinavir; 3TC, lamivudine; MMF, mycophenolate mofetil; ABC, abacavir; NFV, nelfinavir; TDF, tenofovir DF; AZT, zidovudine.

  • *

    Under concurrent liposomal doxorubicin therapy.

  • All therapeutic measures were reduced to supportive fluid maintenance after patient had been recognized as brain dead.

  • Under concurrent pegylated interferon/ribavirin therapy.

138LPV/r, SQV, 3TCMMF + CyA1047367<50
331LPV/r, ABC, 3TCMMF + CyA965271*<50
43PausedPaused  84n.d.
524NFV, 3TC, TDFCyA70132123<50
613TDF, ABC, AZTCyA375160131<50
76TDF, 3TC, LPV/rCyA357550<50

Immunosuppression and HAART

All patients received immunosuppression with CyA and prednisone for induction after OLT. Prednisone was stopped in all patients after a median of 23 weeks (range, 12 - 59). HAART was initiated at a median of 33 days after OLT (range, 18 - 53 days) in patients receiving protease inhibitors, whereas patients receiving triple NRTI therapy started HAART immediately after OLT due to lack of drug-drug interactions. One patient spontaneously had persistently low HIV-RNA and high CD4 counts without indication for antiretroviral therapy. Initial HAART consisted of two NRTIs (lamivudine, tenofovir disoproxilfumarate [DF], abacavir, or zidovudine) and boosted lopinavir in 4 patients; two patients received triple NRTI therapy (patient no. 4: stavudine, lamivudine, abacavir; patient no. 6: tenofovir DF, abacavir, zidovudine). No HAART-associated clinical hepatic toxicity was observed. In 4 patients, follow-up biopsies were available, which did not show HAART-related microvesicular steatosis at 2 (patient no. 7), 4 (patient no. 5), 8 (patient no. 1), and 16 weeks (patient no. 6) after start of HAART, respectively. Two patients required modifications of their HAART regimens because ritonavir intolerance in the first patient necessitated a switch to nelfinavir, and a rise in serum creatinine required an exchange of tenofovir DF for saquinavir in the second patient. Immune reconstitution under HAART was observed in all patients, leading to a mean increase of 170 CD4 cells/μl at last time of follow-up (patient nos. 6 and 7, both on interferon therapy, not counted). CyA doses had to be reduced to 5 - 20% of the original doses after initiation of HAART to avoid toxic drug levels. Calcineurin-inhibitor-associated nephrotoxicity was observed in two patients; both patients were subsequently treated with mycophenolate mofetil (doses 1.5 - 2 g/d) and reduced CyA doses (target trough levels 45 - 65 ng/ml) to prevent further progression of renal insufficiency.

Complications after OLT

Primary graft nonfunction occurred in one patient (no. 2) and required immediate re-OLT. Acute graft rejection was observed in a single patient (no. 6) 4 months after transplantation (2 months after initiation of interferon ribavirin combination therapy) but was successfully treated with high-dose prednisone. This patient received triple NRTI therapy for his HIV infection, and thus no drug-drug interactions between HAART and CyA were noted. Leakage of the biliary tract and subcapsular hematoma with partial infarction of the left liver lobe required surgical revision in another patient (no. 5). CMV-viremia without clinical overt organ damage was noted in three patients (nos. 1, 2, and 6) and was preemptively treated with intravenous ganciclovir therapy. HIV-related complications were noted only in a single patient (no. 3). Cutaneous Kaposi's sarcoma was noted early after OLT in this patient (no. 3), which regressed transiently after reduction of immunosuppression. In the same patient, multicentric Castleman's disease occurred 18 months after liver transplantation and was subsequently associated with generalized recurrence of the Kaposi's sarcoma, although his CD4 counts had remarkably recovered at this time (Fig. 1). Analysis of frozen serum samples showed that the patient had been positive for HHV8-DNA prior to OLT, even though at that time the disease was not clinically overt. Currently — 31 months after OLT — the patient is treated with 4-weekly infusions of liposomal doxorubicin at 20 mg/m2 and has achieved complete remission of both Kaposi's sarcoma and multicentric Castleman's disease.

Figure 1.

CD4 cell count, HIV-RNA, and occurrence of Kaposi's sarcoma/multicentric Castleman's disease after OLT in patient no. 3. KS, Kaposi's sarcoma; MCD, multicentric Castleman's disease.

Recurrence of Hepatitis

Patients with HBV infection received standard prophylaxis with i.v. anti-HBs hyperimmunoglobulins beginning in the anhepatic phase of OLT. Two patients additionally received lamivudine as part of their HAART. Under these measures, none of the patients suffered HBV recurrence throughout the follow-up.

In contrast, HCV recurrence occurred in all 4 transplanted patients with prior HCV infection and took an accelerated cholestatic course in two patients (nos. 1 and 7). Liver biopsies were performed in all patients to rule out organ rejection or other causes of hepatitis. METAVIR activity (A1 - A3) and fibrosis (F1 - F4) scores were A1/F0 in patient no. 1 at 8 weeks, A0/F1 in patient no. 5 at 4 weeks, A1/F0 in patient no. 6 at 3 weeks, and A1/F0 in patient no. 7 at 2 weeks post-OLT. As a consequence, pegylated interferon ribavirin combination treatment was started 6, 15, 10, and 5 weeks after OLT, respectively. Currently, all patients show virological treatment responses. Patient nos. 1 and 5 with HCV-genotype 2 and 3 infection, respectively, have achieved sustained treatment responses. Patients with HCV-genotype 1 infection showed an early treatment response with a reduction of HCV-RNA of > 2 log10 12 weeks after pegIFN/RBV therapy and remain on therapy (weeks 24 and 48, respectively, both with negative HCV-RNA).


Thus far, only small numbers of HIV-infected patients who have received a liver graft have been reported from single centers.5, 6, 8–10, 12, 13, 19 Taken together, these reports indicate that liver transplantation may be a reasonable treatment option for selected HIV-positive patients in the HAART era. The overall encouraging results in our small series also support this concept. However, difficult-to-treat problems have been encountered by some groups concerning the management of HIV-infected patients after liver transplantation. Antiretroviral drug toxicity was a considerable problem in a French series of liver-transplanted HIV patients,7 whereas frequent episodes of rejection and extremely rapid and fatal courses of recurrent hepatitis C have troubled liver-transplanted HIV patients in centers in the US13 and UK,14 respectively. Even though the low number of patients being transplanted does not allow final conclusions, in our patients these problems were not prominent features. A comparison of our results with those of other groups suggests that some of the problems seen after OLT in HIV-infected patients may be circumvented by improved patient management. This may be pivotal for the future management of patients and design of clinical trials.

HAART-associated liver toxicity was reported in 4 of 7 patients in a French cohort of transplanted HIV-positive patients7 but was not encountered at all in our group of patients. Histopathological assessment of the French cases revealed underlying mitochondrial toxicity in all 4 patients. Mitochondrial toxicity deserves particular attention when selecting antiretroviral agents because of the increased risk for mitochondrial toxicity and lactic acidosis, associated particularly with the use of stavudine.20 To prevent this problem, use of didanosine or stavudine after OLT was avoided in our patients with the exception of patient no. 4, who was transplanted in 1999 and had received stavudine for 10 weeks. Strict avoidance of these so-called d-drugs may explain the lack of liver toxicity observed with the HAART regimen used in our patients.

Frequent rejection episodes in the US report may be related to problems in adjusting immunosuppression to account for drug-drug interactions caused by HAART and immunosuppressive drugs. In the combined series from Miami and Pittsburgh, HIV-infected patients had a 38% rate of acute rejections. This rate is much higher than in our series and the data reported by other groups. Patients in Miami and Pittsburgh were put on a weekly dosing scheme of tacrolimus when protease inhibitors were given, a strategy possibly more vulnerable to dosing errors than the daily dosing scheme of CyA provided to our patients. Also, steroids had been withdrawn by week 12 after liver transplantation in all Pittsburgh and Miami patients, which may have been too early in the presence of the complex pharmacokinetic interactions between HAART and calcineurin inhibitors. Finally, HAART was initiated immediately (usually by the second week) after transplantation in the Miami and Pittsburgh patients, which may have met still unstable immunosuppression in the early posttransplantation period.

In our patients, we allowed for establishment of stable CyA levels within desired ranges before initiating HAART to circumvent the problem of reaching steady-state pharmakokinetics of CyA while at the same time accounting for drug-drug interactions between CyA and HAART. In practice, patients had to have repeated CyA trough levels within the target range over at least two weeks without need for any CyA dose adjustment. After initiation of HAART, steroids were continued until reliable CyA levels under concomitant HAART had again been reached. Due to the extensive drug-drug interactions this took a minimum of approximately 4 weeks in our patients. Thus, corticosteroids were withdrawn at a median of 23 weeks, which was significantly later than recommended by the current standard of care. Of note, the only rejection episode observed in our patients occurred in a patient without drug-drug interactions between HAART and CyA. Based on these preliminary observations, it may be beneficial for patients to start a protease inhibitor containing HAART delayed at 4 weeks and also to continue prednisone therapy at doses of 10 mg/d to 6 months post-OLT.

Recurrent hepatitis C infection remains a serious problem in HIV-negative patients.21 Because the degree of immunosuppression is a key factor on the activity of recurrent HCV infection,22 there has been concern that HIV-infected organ recipients may be at particular risk of experiencing aggressive HCV reinfection. Indeed, 5 of the 7 HCV/HIV-positive patients transplanted at the King's College died within 2 years after OLT due to complications of recurrent hepatitis.14 However, 4 of these 5 patients had not received dose adjustments of tacrolimus despite concomitant antiretroviral therapy, which may have resulted in excessive immunosuppression. On the other hand, none of the 11 HCV-positive patients transplanted in Miami and Pittsburgh showed reduced rates of survival at 2 years when compared with HCV-negative patients.13 In these patients, the dosing and monitoring of immunosuppression was adjusted according to HAART to counterbalance pharmacokinetic interactions between tacrolimus and protease inhibitors. In our patients, HCV recurrence was noted in all patients with previous HCV infection, and 2 of these 4 experienced cholestatic hepatitis after reinfection. Similar to the findings in Miami and Pittsburgh, all of our patients responded to pegylated interferon/ribavirin combination therapy and presently exhibit sustained or early virological response. However, we had preferentially selected HIV/HCV-coinfected patients for liver transplantation who were infected with genotype 2 and 3 isolates, which can be well treated by interferon ribavirin combination therapy. Currently, sustained virological response data for both patients with genotype 1 infection are not available, which limits the interpretation of our observations. Nevertheless, the favorable experience from Miami/Pittsburgh and our own data are substantiated by a larger combined meta-analysis on 24 patients (including some patients from Miami, Pittsburgh, and King's College)11 and first prospective data on 6 patients.23 Both in the meta-analysis by Ragni et al.11 and the prospective multicenter study by Roland et al.,23 post-OLT overall and graft survival was similar between HIV-positive and HIV-negative patients up to 3 years of follow-up. Taken together, it appears that early fatal courses of HCV reinfection can be avoided by taking into account drug-drug interactions between immunosuppressive agents and HAART. Furthermore, HIV-positive patients with concurrent HCV infection should not be excluded from OLT, in particular those patients with HCV-genotype 2 or 3 infection. Although our data suggest that delayed initiation of HAART and avoidance of d-drugs as part of HAART may improve patient outcome, future establishment of an international database appears desirable to better evaluate management and outcomes in these challenging patients and validate our observations.