fax: (33) 1 45 59 38 57
Survival and recurrence of hepatitis C after liver transplantation in patients coinfected with human immunodeficiency virus and hepatitis C virus†
Article first published online: 20 DEC 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 47, Issue 2, pages 407–417, February 2008
How to Cite
Duclos-Vallée, J.-C., Féray, C., Sebagh, M., Teicher, E., Roque-Afonso, A.-M., Roche, B., Azoulay, D., Adam, R., Bismuth, H., Castaing, D., Vittecoq, D. and Samuel, D. (2008), Survival and recurrence of hepatitis C after liver transplantation in patients coinfected with human immunodeficiency virus and hepatitis C virus. Hepatology, 47: 407–417. doi: 10.1002/hep.21990
Potential conflict of interest: Dr. Didier is a consultant for Novartis and Astellas Pharma.
The members of the THEVIC study group are listed in the appendix.
- Issue published online: 26 JAN 2008
- Article first published online: 20 DEC 2007
- Manuscript Accepted: 16 AUG 2007
- Manuscript Received: 9 MAY 2007
- Agence Nationale de Recherches sur le Sida et les Hepatites Virales (ANRS-HC08 program)
Liver transplantation in patients coinfected with human immunodeficiency virus (HIV) and hepatitis C virus (HCV) is a recent indication. In a single center, we have compared the survival and severity of recurrent HCV infection after liver transplantation in HIV-HCV–coinfected and HCV-monoinfected patients. Seventy-nine patients receiving a first liver graft for HCV-related liver disease between 1999 and 2005 were included. Among them, 35 had highly active antiretroviral therapy–controlled HIV infection. All patients were monitored for HCV viral load and liver histology during the posttransplantation course. Coinfected patients were younger (43 ± 6 versus 55 ± 8 years, P < 0.0001) and had a higher Model for End-Stage Liver Disease (MELD) score (18.8 ± 7.4 versus 14.8 ± 4.7; P = 0.008). The 2-year and 5-year survival rates were 73% and 51% and 91% and 81% in coinfected patients and monoinfected patients, respectively (log-rank P = 0.004). Under multivariate Cox analysis, survival was related only to the MELD score (P = 0.03; risk ratio, 1.08; 95% confidence interval, 1.01, 1.15). Using the Kaplan-Meier method, the progression to fibrosis ≥ F2 was significantly higher in the coinfected group (P < 0.0001). Conclusion: The results of liver transplantation in HIV-HCV–coinfected patients were satisfactory in terms of survival benefit. Earlier referral of these patients to a liver transplant unit, the use of new drugs effective against HCV, and an avoidance of drug toxicity are mandatory if we are to improve the results of this challenging indication for liver transplantation. (HEPATOLOGY 2007.)
In this era of highly active antiretroviral therapy (HAART), hepatocellular carcinoma and end-stage cirrhosis due to hepatitis C virus (HCV) are increasing causes of mortality among human immunodeficiency virus (HIV)–infected individuals.1–4 The access of HIV-infected patients to organ transplantation may be their only option.
The preliminary, short-term results of liver transplantation (LT) in HIV-HCV–coinfected patients have been reported in recent series.5–14 Mortality was high (5/7 patients) in the series at King's College10 and in Essen (3/5 patients).11 In 3 other series involving both HCV and non-HCV indications,6, 12–15 short-term survival appeared to be as good, although the onset of severe recurrent hepatitis C was also reported. In our series concerning 7 HIV-HCV–coinfected patients who underwent transplantation in our center,8, 9 recurrent hepatitis and mitochondrial toxicity were observed in most patients. However, they suggested a more severe recurrence of HCV than that in non–HIV-infected patients.
Because of the shortage of liver grafts, a better determination of prognosis (survival and histology) and the factors contributing to the severity of recurrent liver disease in this new indication for LT are important. This article describes our experience with 35 consecutive HIV-HCV–coinfected patients who received a first liver graft in a single center between December 1999 and October 2005.
Patients and Methods
In our center, between January 1999 and October 2005, 79 consecutive hepatitis B surface antigen–negative patients received a first liver graft (without previous or concomitant transplantation of another organ) for HCV-related liver disease. Of these, 35 were coinfected with HIV. The 35 HIV-HCV–coinfected patients underwent transplantation for decompensated cirrhosis (n = 32) or hepatocellular carcinoma in a setting of liver cirrhosis (n = 3). One patient underwent combined liver and kidney transplantation. Three of them had received combined interferon-alfa and ribavirin therapy for more than 1 year before transplantation and were serum HCV RNA–negative. All had controlled HIV infection with a CD4 count > 100 cells/mm3, no previous AIDS events or opportunistic infections, and an undetectable HIV plasma viral load when they were placed on the waiting list. Treatment regimens were not standardized. The antiretroviral agents given could be any of the 15 drugs available, namely, zidovudine, didanosine (DDI), zalcitabine, stavudine (d4T), lamivudine, abacavir, tenofovir, nevirapine, efavirenz, saquinavir, ritonavir, indinavir, nelfinavir, amprenavir, and lopinavir. The dosage of antiretroviral agents was adapted for each patient as a function of the monitoring results of residual plasma concentrations of protease inhibitors or nonnucleoside reverse transcriptase inhibitors in order to ensure that plasma concentrations attained therapeutic levels and to maintain an undetectable plasma viral load. Among the 44 monoinfected patients, 15 underwent transplantation for hepatocellular carcinoma in a setting of cirrhosis. None of them had been responders to interferon therapy before transplantation. HCV RNA levels were determined with the Roche (Nutley, NJ) Cobas Amplicor Monitor 2.0 assay. The lower limit of detection was 600 IU/mL. Samples with higher titers were diluted as needed to obtain a value in the linear range of the assay.
LT and Immunosuppressive Therapy
Liver grafts were obtained from cadaveric or living donors. In the latter case, the grafts came from patients who underwent transplantation for familial amyloidotic polyneuropathy (domino program)16 or from related subjects. After transplantation, all patients received 100 mg of fluconazole and 600 mg of cotrimoxazole for prophylaxis against Pneumocystis carinii. Ganciclovir (10 mg/kg/day) was given for 3 months post-LT as prophylaxis against cytomegalovirus. Primary immunosuppression was based on tacrolimus, tacrolimus and mycophenolate mofetil (MMF), cyclosporine, and prednisone. In all patients, corticosteroids were initiated at a dose of 5 mg/kg/day, then reduced to 0.3 mg/kg/day at the end of the first week, tapered to 10-15 mg/day until the third month post-LT, and then gradually withdrawn over the next 3 months. MMF was given to patients with renal insufficiency or as additional immunosuppression for those who were experiencing acute rejection. Clinical examinations and liver biochemistry tests were performed daily during the first 3 weeks of the postoperative period, 3 times weekly during the next month, then once a week for 3 months, and once a month thereafter.
Anti-HIV Drug Administration
HAART therapy was reintroduced after the 14th day post-LT when liver function was stable and consisted of the same regimen as that administered prior to LT. After transplantation, the initial daily dosage of any HAART medication was given according to standard recommendations, taking account of graft function and body weight. Thereafter, drugs were administered as a function of plasma concentrations determined by regular drug monitoring. If HIV virological breakthrough occurred, the antiretroviral regimen was adjusted empirically and then readjusted subsequently as a function of genotyping results.17
Standard interferon-alfa 2b or pegylated interferon and ribavirin were prescribed in the event of an acute/and or chronic recurrence of HCV hepatitis, as assessed by the results of graft biopsies. In the coinfected group, interferon alfa and ribavirin were prescribed in the event of symptomatic acute hepatitis or because of a score of F2 on the liver biopsy. In the monoinfected group, the decision to administer anti-HCV therapy was made because of a fibrosis score of F2 or higher. Patients were either included in a prospective trial using pegylated peginterferon-alfa 2a with ribavirin (n = 11; peginterferon-alfa 2a was initiated at 90 μg/week and ribavirin was initiated at 600 mg/day, and then they were increased to 180 μg/week and 1000 mg/day or adjusted as a function of hematological tolerance) or treated with pegylated interferon-alfa 2a at different doses (50-180 μg/week) and ribavirin (400-600 mg/day) on the basis of hematological tolerance and body weight (n = 9).
Liver Pathology Monitoring
Graft liver biopsies were routinely planned for all patients at day 0 before perfusion and every year post-LT. In the coinfected group, an additional routine liver biopsy was scheduled at 6 months post-LT. Moreover, liver biopsies were performed when liver function test results were abnormal. Liver biopsy specimens were fixed and stained with paraffin-embedded (Picrosirius and Perls). Acute hepatitis was defined by the presence of necroinflammatory activity. The degree of severity of mononuclear cell infiltration and intralobular necrosis was scored as follows: mild (+), moderate (++), and severe (+++). Histological features consistent with recurrent hepatitis C, that is, activity and fibrosis, were assessed with the METAVIR scoring system. Fibrosis was scored from 0 to 4 (cirrhosis).18 Acute and/or chronic rejection was diagnosed according to the Banff classification.19
Quantitative variables were expressed as the mean ± standard deviation. For comparisons between categorical variables, the chi-square test or Fisher test was used. For continuous variables, the Student t test or Mann-Whitney U test was employed. Survival estimates at different time points were expressed as the cumulative proportion of survivors at the end of the follow-up period (that is, March 1, 2007). For survival analysis, data were censored at the date of death or when the last information on life status was obtained. Survival times were calculated from the date of LT. The univariate analysis of survival was performed with the Kaplan-Meier method. Curves were compared with the log-rank test. Independent prognostic factors for survival were identified by a stepwise forward Cox regression model, in which covariates were entered if a difference with P < 0.1 was detected under univariate analysis.
Characteristics of Coinfected and Monoinfected Transplant Patients (Table 1)
|Men||29 (66%)||29 (83%)||0.09|
|Recipient age (years)||55.3 ± 8.3||43.2 ± 5.9||<0.0001|
|MELD score||14.8 ± 4.7||18.8 ± 7.4||0.008|
|Donor age (years)||48 ± 14.6||48.4 ± 14||0.9|
|Partial liver graft from living donor||6 (14%)||4 (11%)||1|
|Domino procedure||2 (5%)||13 (37%)||0.0004|
|Chemotherapy after LT||7 (16%)||0 (0%)||0.013|
|Viral load at the time of LT (Log10 IU/mL)||5.3 ± 1.5||5.6 ± 0.7||0.38|
|Hepatocellular carcinoma||15 (34%)||5 (15%)||0.052|
|Mean follow-up (months)||63.47 ± 24.7||43.6 ± 82.8||0.13|
Coinfected patients were younger (43 ± 6 versus 55 ± 8 years; P < 0.0001) and had a shorter (but not significantly) follow-up period (43.6 ± 82.8 versus 63.4 ± 24.7 months; P = 0.13) than monoinfected liver transplant patients. They were more numerous in receiving a domino liver graft (13/35 versus 5/44; P = 0.0004) but not in receiving a partial liver graft from a living donor (4/35 versus 5/44; P = 0.7). None of the coinfected patients received posttransplant antitumor chemotherapy, although 7/44 (16%) monoinfected patients received chemotherapy during the post-LT period. The Model for End-Stage Liver Disease (MELD) score was also higher in coinfected patients than in monoinfected patients (18.8 ± 7.4 versus 14.8 ± 4.7; P = 0.008) at the time of LT. The donor age, genders of the recipients and donors, distribution of genotype 1, and viral load at the time of LT were similar in the 2 groups. In coinfected patients, primary immunosuppression was based on tacrolimus (n = 25), tacrolimus and MMF (n = 8), cyclosporine (n = 2), and prednisone. Tacrolimus and prednisone (n = 42) or tacrolimus and MMF (n = 2) were administered to monoinfected patients. One coinfected patient underwent retransplantation a few days after receiving a partial liver graft from a living donor, and 2 monoinfected patients received a second liver graft 1 and 3 years post-LT.
The 2-year and 5-year survival rates were 73% and 51% and 91% and 81% in coinfected patients and monoinfected patients, respectively (log-rank P = 0.004; Fig. 1). Thirteen (37%) of the 35 coinfected patients died from cerebral hemorrhage (n = 2), a severe recurrence of HCV infection (n = 3), mitochondrial toxicity and a recurrence of HCV infection (n = 2), a recurrence of hepatocellular carcinoma (n = 1), pancreatic carcinoma (n = 1), cardiac arrest (n = 1), and sepsis (n = 3). Of the 44 monoinfected patients, 9 (20%) died from recurrent HCV cirrhosis (n = 2), veno-occlusive disease, chronic rejection and severe recurrent hepatitis C (n = 1), recurrent hepatocellular carcinoma (n = 2), sepsis (n = 2), and cardiovascular causes (n = 2). Under log-rank analysis, other qualitative variables (gender, partial graft, genotype 1 or not, and interferon therapy) were not significant in univariate analysis. Under the Cox proportional model, the MELD score was significantly correlated to survival (P = 0.03), but other variables (HIV status, recipient, age, donor age, genotype 1 or not, type of LT, viremia at the time of LT, and indication of LT) were not correlated to survival (Table 2).
|Factors||Univariate Analysis||Multivariate Analysis|
|Risk Ratio (95% CI)||P||Risk Ratio (95% CI)||P|
|HIV status||3.37 (1.39, 8.12)||0.007||1.91 (0.7, 5.18)||0.20|
|Recipient age||0.97 (0.93, 1.01)||0.15|
|MELD score||1.1 (1.04, 1.16)||0.0005||1.08 (1.01, 1.15)||0.03|
|Genotype 1||1.57 (0.45, 6.49)||0.48|
|Donor age||1.03 (1, 1.07)||0.06||1.04 (1, 1.07)||0.06|
|Living donor||1.14 (0.33, 3.86)||0.83|
|Domino procedure||1.87 (0.72, 4.81)||0.19|
|Chemotherapy post-LT||1.29 (0.38, 4.38)||0.68|
|Hepatocellular carcinoma||0.56 (0.19, 1.66)||0.29|
|Viral load at the time of LT||3.61 (0.93, 13.98)||0.063|
Recurrence of HCV Infection
Recurrent Hepatitis C.
Three coinfected patients treated before LT were plasma HCV RNA–negative at LT and remained plasma and liver HCV RNA–negative for more than 2 years afterwards, with normal liver histology findings. During follow-up of the 33 coinfected patients who were HCV RNA–positive, 4 experienced severe symptomatic acute hepatitis C (D14, D16, M1, and D40) characterized by marked liver test abnormalities and severe histological features on pathological examination. In 3 patients, this acute hepatitis contributed to the development of fibrosing cholestasis hepatitis and to death in 2 patients (Table 3). Moreover, in 3 patients (6, 15, and 24), lobular necrosis was observed at M3 post-LT, and all these patients developed severe liver disease on the liver graft. Seven patients (8, 1, 31, 32, 12, 3, and 16; 20%) developed F3 fibrosis (M4, M6, M10, M13, M16, M20, and M21), and 2 patients (6%) developed cirrhosis at M18 and M24, respectively (Table 4). Three patients (10, 24, and 27) in the coinfected group developed fibrosing cholestasis hepatitis at M4, M6, and M24. No symptomatic acute hepatitis was observed in the monoinfected group; during the follow-up period, 8/44 (18%) monoinfected patients developed stage F3, and 1 (2%) progressed to cirrhosis 1 year post-LT.
|Patient||Time post-LT||AST (IU/L; n < 35)||ALT (IU/L; n < 39)||gGT (IU/L; n < 50)||Total Bilirubin (μM/L)||Genotype||HCV Viral Load (Log 10 IU/mL)||Histological Features|
|10||D40||550||550||2561||114||1a||6.5||Mononuclear lymphoid cells infiltrate portal tract, ++/ballooning of hepatocytes, +++|
|14||D16||718||1115||600||90||1a||6.84||Mononuclear lymphoid cells infiltrate portal tract, +++|
|24||M1||70||136||800||30||1a||6.4||Mononuclear lymphoid cells infiltrate portal tract, +++|
|28||D14||204||316||250||129||1||6.7||Mononuclear lymphoid cells infiltrate portal tract, +++/intralobular necrosis, +++|
|Patient||Genotype||Acute Hepatitis C||Chronic Hepatitis C||Modality of Anti-HCV Therapy||Biochemical Response||Virological Response||Outcome/Delay of Follow-Up (M)/Clinical Status/Karnovsky Score (%)/Cause of Death|
|1||1b||Mild (D12)||A3F3 (M5, M6); A2F3 (M12, M21); A1F3 (M34)||IFN 1.5 MU × 3/W from M4 to M12; PegIFN 50 μg + RBV 600 mg/D from M10 to M34||Complete||SVR||M94/well/80|
|2||4c/4d||Mild (M1)||A2F1 (M7, M12); A3F4 (M24)||PegIFN 50 μg/W + RBV 600 mg/D from M24 to M36||Complete||NR||Death/M66/cerebral hemorrhage|
|3||1a||Mild (M3, M6)||A1F2 (M9)||IFN 3.5 MU × 3/W + RBV 400 mg/D from M11 to M12||NR||NR||Death/M24/mitochondrial toxicity and recurrence of HCV infection|
|4||1a||Mild (M1)||A0F3 (M12); A1F2 (M18); (M24); A1F3/4 (M42)||No treatment||—||—||M63/well/80|
|5||1a||Mild (D10, M2)||A2F1 (M12); A2F4 (M18); A2F4 (M30)||IFN 1.5 MU × 3/W (M3); RBV 400 mg/D from M3 to M6; PegIFN 25 μg/W + RBV 400 mg/D from M11 to M13; PegIFN 40 μg + RBV 400 mg/D from M32 to M42||Partial||Relapse||M72/asthenia/50|
|6||1b||Moderate (M1, M5)||—||IFN 1.5 MU single dose + RBV 400 mg/D for 3 D at M3||NR||NR||Death/M6/mitochondrial toxicity and recurrence of HCV infection|
|7||1a||Moderate (M1)||A3F2 (M8); A2F1/2 (M12); A1F2 (M16); A2F2 (M23); A1F1 (M33); A3F1 (M51)||PegIFN 80 μg/W + RBV 800 mg/D from M10 to M36 stopped because of intolerance||Complete||NR||M60/well/80|
|8||3a||Moderate (M1)||A2F1 (M2); A3F1/F2 (M3); A3F3 (M4); A2F3 (M11)||PegIFN 80 μg/W + RBV 600 mg/D from M3 to M5||Death/M18/pancreatic carcinoma|
|9||1a||Mild (D16, M2)||A2F1 (M12); A2F2 (M24); A2F2 (M32); A1F2 (M37)||PegIFN 60 μg/W + RBV 600 mg/D from M13 to M24||Complete||NR||Death/M47/Recurrence of hepatocellular carcinoma|
|10||1a||Mild (D6); severe (D40)||A3F1 (M5); FCH (M10)||PegIFN 65 μg/W + RBV 600 mg/D from M1 to M8, stopped because of weight loss and severe asthenia||Complete||NR||Death/M27/recurrence of HCV infection|
|11||1a||Mild (M1)||A0F0 (M6, M12, M24)||No treatment||M30/well/80|
|12||3a||Mild (D25, M3)||A2F1 (M6); A2/3F3 (M14); A2F3/4 (M24)||PegIFN 50 μg/W + RBV 600 mg/D from M1 to M5 (intolerance = major asthenia)||Complete||SVR||Death/M42/heart arrest|
|13||1a||Moderate (D15)||A3F1 (M3); A1F1 (M12, M20); A2F1 (M26); A2/3F2 (M37)||PegIFN 50 μg/W + RBV 600 mg/D from M4 to M18 because of myocarditis due to interferon||Complete||NR||M45/well/80|
|14||1a||Severe (D16); moderate (M2)||A2F1 (M6); A1F1 (M12); AR Banff 3 and A2/3F2 (M29)||No treatment||—||—||M40/well/80|
|15||3a||Moderate (M3)||A3F2 (M5); A1F3 (M17); A1F2 (M24); A1F3 (M38)||PegIFN 40 μg/W + RBV 400 mg from M4 to M20 (thrombopenia)||Partial||SVR||M40/well/80|
|16||3a||Mild (D23); moderate (D37)||A3F1 (M2); A3F2 (M12); A2/A3F2 (M14); A2/A3F3 (M21)||PegIFN 50 μg/W + RBV 100 mg/D from M2 to M10 (acute pancreatitis)||Complete||NR||M39/refractory ascites/50|
|17||—||Mild (M1, M3, M5)||A2F1 (M18); A0F1 (M25)||PegIFN 100 μg/W + RBV 1000 mg/D from M16 to M18||NR||M27/well/80|
|18||4c/4d||Mild (M1)||A2F1 (M3); A1F1 (M6); A1F1 (M13); A2F1 (M16); A1F2 (M22)||PegIFN 60 μg/W + RBV 400 mg/D from M3 to M15; PegIFN 180 μg/W + RBV 800 mg/D from M22 in course||Partial||NR||M36/well/80|
|19||3a||Mild (D17, M1, M3)||—||PegIFN 80 μg/W + RBV 600 mg/D from M2 to M14 (asthenia)||Complete||Partial||M36/well/80|
|20||1a||Moderate (M1, M4)||A2F2 (M21)||PegIFN 80 μg/W + RBV 600 mg/D from M26 in course||—||—||M33/well/80|
|21||2||Moderate (D13, D36, M3)||A1F1 (M6, M12)||No treatment||—||—||M33/well/80|
|22||4||Moderate (D24)||No treatment||—||—||Death/M6/sepsis|
|23||1a||Mild (M1)||—||No treatment||—||—||Death/M3/cerebral hemorrhage|
|24||1a||Severe (M1)||A2F1 (M3); FCH (M6); A1F1/F2 (M12); A2F3 (M19)||PegIFN 80 μg/W + RBV 600-800 mg/D from M4; PegIFN 60μg/W from M8 to M13||Partial||NR||M25/well/80|
|25||3a||Mild (M1)||A1F0 (M6)||No treatment||M22/well/80|
|26||1a||Mild (M4)||A1F1 (M12)||No treatment||Death/M19/recurrence of HCV infection|
|27||1b||RAA Banff 2 (D13)||FCH (M4, M5, M9)||PegIFN 180 μg/W + RBV 800 mg/D from M10 in course (to M17 RBV 600 mg/D)||Partial||Partial||M24/well/80|
|28||1||Severe (D14)||FCH (M2)||No treatment||Death/M3/sepsis|
|29||1b||Cholangiolitis (D5)||A2F2 (M2, M5)||PegIFN 80 μg/W + RBV 800 mg/D M4-M16 (dosage was discontinued because of severe anemia)||Partial||NR||M19/well/80|
|31||3a||—||A3F2 (M4, M8); A3F3 (M10); A3F3 (M15)||PegIFN 180 μg/W + RBV 800 mg/D from M4 to M8; PegIFN 70 μg/W + RBV 800 mg/D from M10 to M15||Complete||Virological response from M7||M18/well/80|
|32||4a||—||A1F0 (M6)||No treatment||M18/well/80|
|33||3a||Mild (M1)||FCH (M6); A2F3 (M12)||PegIFN 180 μg + RBV 800 mg/D from M4; PegIFN 135 μg from M5, 90 μg from M6 135 μg from M6; stopped RBV M9; stopped PegIFN M13||Partial||Virological response from M2, relapse from M3, response from M5 post-therapy||Death/M14/recurrence of HCV infection|
|34||3a||—||A1F0 (M3); A2F2 (M24)||No treatment||M17/well/80|
The 2-year and 5-year actuarial survival rates with lower than stage F3 at 12 months reached 65% and 43% in coinfected patients, 77% and 77% in monoinfected patients, respectively (P = 0.07). The 2-year and 5-year actuarial survival rates with higher than stage F3 at 12 months were 72% and 53% in coinfected patients and 90% and 82% in monoinfected patients, respectively (P = 0.03). Finally, histology findings assessed 12 months post-transplant showed fibrosis scores of 1.1 ± 1.03 in monoinfected patients and 1.7 ± 1.2 in coinfected patients (P = 0.06); at 24 months, these scores reached 1.4 ±1.1 in monoinfected patients and 2.4 ± 1.3 in coinfected patients (P = 0.01). Figure 2 shows the number of patients with the different fibrosis scores (F0 to F4) at M12 (Fig. 2A) and M24 (Fig. 2B). With the Kaplan-Meier method, progression to a fibrosis score ≥ F2 was significantly higher in coinfected patients than in monoinfected patients (P < 0.0001; Fig. 3). With respect to the HCV viral load, in nontreated patients at 6 months post-LT, it did not differ significantly between coinfected and monoinfected patients (6.28 log10 ± 0.9 versus 6.4 log10 ± 0.37). Viremia results were not interpretable thereafter because of the effects of antiviral therapies.
Various antiviral therapies based on standard or pegylated interferon alfa 2b were attempted in 19/35 (54%) coinfected patients and 20/44 (45%) monoinfected patients. Coinfected patients received either standard interferon alfa 2b thrice weekly (which was switched to pegylated interferon 2b in 4 cases or pegylated interferon from the start in 15 cases) and ribavirin. Ribavirin doses differed because of hematological intolerance and were 400 mg/day in 5 patients, 600 mg/day in 8 patients, and 800 mg/day in 6 patients. A virological response during therapy was observed in 15/20 (75%) monoinfected patients, and 3 of them experienced a sustained virological response. A virological response was also obtained in 4/19 (21%) coinfected treated patients, which was sustained in 3 of them (Table 4). Interestingly, the virological response was observed at a late stage: in 1 patient after 6 months of antiviral therapy and in the other 18 months after the withdrawal of antiviral therapy. In coinfected patients, asthenia was a major limiting factor requiring the cessation of antiviral therapy (n = 4), despite a low dose of pegylated interferon. In 1 patient, lethal lactic acidosis was observed at M4, whereas another patient suffered from acute pancreatitis a few weeks after the initiation of therapy. This type of adverse effect was not observed in the monoinfected group.
HIV Status During the Post-LT Period
Plasma HIV RNA reappeared in 5 patients because of the temporary withdrawal of HAART at D3 (patient 11, 2.3 log10), D10 (patient 14, 2.2 log10), M6 (patient 21, 5 log10), M7 (patient 3, 4.9 log10), and M12 (patient 17, 4 log10). Drug resistance was observed in 2 patients at M2 (patient 8, 2.7 log10) and M3 (patient 2, 3.9 log10). At the time of LT, 4 patients (patients 2, 3, 14, and 15) had a CD4 cell count < 150 cells/mL. CD4 counts were maintained within a range of 80-630 cells/mL during the post-LT period. In 3 patients (patients 14, 16, and 19), CD4 levels fell below 100 cells/mL during the post-LT period at M1, M24, and M24, respectively.
The incidence of bacteremia with fever or shock was higher in monoinfected patients than in coinfected patients (16/44, 36% versus 3/35, 8.5%; P = 0.01). Symptomatic cytomegalovirus pneumonia was observed in 1 of the coinfected patients, and cytomegalovirus viremia with fever was observed in 3 of the monoinfected patients (not significant). One case of Pneumocystis carinii infection and 2 cases of systemic candidiasis were observed in a monoinfected subject. Moderate esophageal candidiasis was observed in 1 coinfected patient
Histologically proven acute rejection was observed in 11/35 (31%) coinfected patients and in 12/44 (27%) monoinfected patients (not significant). The mean interval between acute rejection and LT was similar in coinfected and monoinfected patients (8 days ± 2 versus 7 days ± 3). Early chronic rejection defined by the disappearance of more than 20% of bile ducts was similar in coinfected (6/35, 17%) and monoinfected (8/44, 18%) patients. Late chronic rejection (more than 50% of bile duct vanishing) was observed in 1/35 coinfected patients and 3/44 monoinfected patients (not significant). No correlation was demonstrated between initial immunosuppression (tacrolimus, cyclosporine, and interleukin 2 receptor monoclonal antibody) or interferon-ribavirin therapy and the occurrence of acute, early, or late chronic rejection.
Compliance problems were observed in 4/35 coinfected patients and 1/44 monoinfected patients. In 1 patient, HAART withdrawal led to a 4-log rebound of the HIV viral load with the appearance of esophageal candidiasis. In 2 other patients, follow-up was complicated by irregular attendance that delayed the introduction of anti-HCV therapy. In the last coinfected patient, a recurrence of alcohol consumption was observed 3 months after LT.
Our results clearly indicate that after LT, the recurrence of HCV is more severe in coinfected patients than in monoinfected patients. By comparing HIV-HCV–coinfected and HCV-monoinfected patients receiving transplants in a single center and followed by the same team, we showed that fibrosis-free survival rates were extremely low in coinfected patients after the second year post-transplant, whereas a majority of monoinfected patients experienced mild recurrent hepatitis.
This study also shows that the survival of coinfected patients tended to be poorer than that of monoinfected patients, but this was mainly due to the severity of their condition at the time of LT. This fact was clearly demonstrated in this series by the significantly higher MELD score at the time of LT in the coinfected group. The poor prognosis of HIV-HCV–coinfected patients with end-stage liver disease has already been described elsewhere.4, 19, 20, 21 We could transplant only a small percentage of such coinfected patients referred to our center for LT. The fact that MELD was the principal survival criterion in both coinfected and monoinfected patients clearly reflected the particularly serious condition of such patients. Moreover, HIV status added a further cause for death after transplantation, that is, mitochondrial toxicity due to the use of nucleoside analogs. Two patients died from mixed liver disease including a severe recurrence of HCV infection and HAART therapy toxicity, and another died from acute toxic pancreatitis. Following our experience with the first 7 patients,9 and in light of the fact that of the 9 patients who underwent transplantation between December 1999 and February 2003, 7 died and 3 were treated with DDI and/or d4T, whereas 1 died because of mitochondrial toxicity and a recurrence of HCV, we decided to avoid using nucleosides such as DDI and d4T analogs after LT and subsequently did not observe any lethal complications of HAART therapy during the post-LT period. The avoidance of such complications, together with the transplantation of patients with lower MELD scores, will probably improve the survival of HCV-HIV–coinfected patients.22–25
The main problem encountered was that the liver grafts given to these coinfected patients were clearly a target for a severe recurrence of hepatitis C. Our series showed that under both cross-sectional and actuarial analyses, HIV status was independently related to the degree and rapidity of the onset of fibrosis and the HCV viral load. The precise reasons for this severe recurrence remain unclear; a host factor such as a low CD4 count, as suggested by Ragni et al.,7 could not be determined clearly because the HIV status and CD4 count were confusing variables; in our series, the CD4 cell count during the pre-LT period did not constitute a factor affecting post-LT survival. However, in 7 patients, we observed a CD4 cell count < 150 cells/mL at least at one time point (at M1, M3, M6, M12, M18, M24, M36, and M48 during the post-LT period). Of these 7 patients, 6 developed severe chronic hepatitis C; moreover, 3 of these 6 patients had a CD4 cell count > 150 cells/mL before starting pegylated interferon therapy. The impact of a low CD4 cell count on survival and the degree of severity of HCV recurrence on the liver graft needs to be demonstrated in a larger series of patients.
The issue of HCV recurrence is problematic in view of the low rates of sustained virological response induced by interferon-based therapies in both the coinfected and monoinfected populations. New antiprotease or antipolymerase drugs currently under phase IIB development are likely to be the best answer. The apparently poor results achieved with interferon in this population should be interpreted with caution. The adverse effects of anti-HCV therapies were particularly marked in coinfected patients in whom they were frequently introduced sooner after LT than in monoinfected patients. The durations of treatments and doses used were clearly lower in coinfected patients. Interestingly, sustained virological responses were possible and were observed in 3 coinfected LT patients who could afford prolonged treatment (patients 1, 12, and 16; Table 4). These patients subsequently remained HCV RNA–negative, and 1 improved his liver graft histology findings (patient 1). Another encouraging result concerned the 3 coinfected patients who had been treated successfully before LT (patients 12, 17, and 25; Table 4). Two of them remained HCV RNA–negative (patients 12 and 25), whereas the other developed mild recurrent liver disease (patient 17). Finally, although we observed impressive rates of virological response in monoinfected patients, sustained virological response was relatively infrequent during prolonged periods of treatment in a setting of hard-to-treat patients. In summary, we feel that active therapies should be attempted, not only soon after LT but also (and whenever possible) before LT, in order to improve the results of this new indication for LT.
Because the early case reports on LT in coinfected patients reported severe recurrence and high mortality, at the beginning of the pilot program, we felt it reasonable to encourage the use of all types of LT (domino LT, split cadaveric LT, and partial LT from living donors) rather than solely cadaveric LT in these patients. This was achieved because 16/23 (69%) coinfected patients (as opposed to 20% of monoinfected patients) received such liver grafts. What was interesting is that we did not observe any deleterious effects of domino LT or partial LT from living donors on either survival or histological prognosis. Strikingly, we found that the only other variable related to prognosis was donor age. Domino and living donor transplants favor young donors and minimize the recurrence of HCV.26–29 In fact, although the effect of donor age was independent of HIV status, this was found only in the subgroup of monoinfected patients; this suggests that the effect might be a far less powerful predictor than HIV status. In the domino population, objective tests were performed on a regular basis to determine the occurrence of neuropathies in these patients and in the larger group of those receiving domino, and to date we have not detected any amyloidosis, despite a long follow-up period. Domino strategies are available only in a few LT centers. On the other hand, living donor and split cadaveric livers constitute a growing source of liver grafts worldwide. Nevertheless, some studies have demonstrated that HCV recurrence is more severe after a living donor transplant than after cadaveric LT.30, 31
Another means of improving the results of LT is to modify immunosuppression. Both chronic rejection and infectious complications (other than HCV) were uncommon in coinfected patients. This provided an opportunity to modulate immunosuppressant therapy in this population. It should be pointed out that all coinfected patients were receiving relatively standard immunosuppressant therapy with tacrolimus and gradually tapered corticosteroids. Almost all these patients experienced prolonged serum tacrolimus peaks due to interactions with antiprotease drugs, but none developed rejection under interferon. To date, in non-HIV LT recipients, the links between immunosuppression and the severity of recurrent hepatitis C do not appear to have been very strong, especially during prospective trials. However, this situation may differ in HIV-HCV–coinfected patients with relatively low CD4 levels who experience more severe hepatitis C both before and after LT. Our findings and those of others3–8 clearly suggest that host control of HCV is weaker in HIV patients, and this probably contributes to more severe liver disease. A reduction in immunosuppression may be of considerable value in this population if it enables partial restoration of the immune control of HCV replication. In the light of recent data indicating that cyclosporine exerts an anti-HCV activity, the usefulness of cyclosporine as an immunosuppressive agent of choice in HIV-HCV–coinfected patients needs to be demonstrated.32, 33 Of course, viral factors that are currently under study (such as the presence of pathogenic quasispecies in the serum and lymphocytes of coinfected patients) may constitute other arguments suggesting more severe disease on the liver graft.34, 35
In conclusion, the results of LT in HIV-HCV–coinfected patients were satisfactory in terms of survival benefit because these patients had high MELD scores. It will probably be possible to improve upon these results by an earlier referral of these patients to a liver transplant unit, by reduction of drug toxicity, and by the use of new anti-HCV specific inhibitors.
The authors thank Dr. Alloua Smail and Dr. Sonia Benhamida for their assistance and Isabelle Ogier, Noël Boujema, Virginie Lambert, Claire Mony, Frédérique Blandin, Vincent Karam, and Vicky Hawken for their technical help.
The Agence Nationale de Recherches sur le Sida et les Hépatites Virales HC08 THEVIC study group has the following members: Michelle Gigou (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Mariagrazia Tateo (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Valerie Cailliez (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Elisabeth Dussaix (Laboratoire de Virologie, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France) Catherine Guettier (Laboratoire d'Anatomopathologie, AP-HP Hôpital Paul Brousse, UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Marie-Pierre Bralet (Laboratoire d'Anatomopathologie, AP-HP Hôpital Paul Brousse, UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Philippe Ichaï (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Faouzi Saliba (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Theresa Antonini (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Eric Vibert (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Luc Antoine Veilhan (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Paola Andreani (Centre Hépato-Biliaire, AP-HP Hôpital Paul Brousse, Villejuif, France; UMR-S 785, Université Paris Sud, Villejuif, France; and Institut National de la Santé et de la Recherche Médicale Unité 785, Villejuif, France), Marie-Christine Gillon, Luce Kullman, Catherine Josse, Alain Mirand (Département d'Anesthésie, AP-HP Hôpital Paul Brousse, Villejuif, France), Marie-Gisèle Lebrette (Service des Maladies Infectieuses, AP-HP Hôpital Tenon, Paris, France), Gilles Pialoux, Service des Maladies Infectieuses, AP-HP, Hôpital Tenon, Paris, France, Yves Benhamou (Service d'Hépato-Gastroentérologie, AP-HP Groupe Hospitalier Pitié Salpêtrière, Paris, France), Danièle Botta, Service d'Hépato-Gastroentérologie, Hôpital de la Conception, Marseille, France, Marc Bourlière, Service d'Hépato-Gastroentérologie, Hôpital Saint-Joseph, Marseile, France, Odile Goria (Service des Maladies de l'Appareil Digestif, CHU de Rouen, Rouen, France), Gilles Cessot (Institut Alfred Fournier, Paris, France), Sylvie Grimbert (Institut Alfred Fournier, Paris, France), Hervé Zylberberg (Paris, France), Jean-Francois Cadranel (Unité d'Hépatologie, Hôpital de Creil, Creil, France), Francois Cordier (Service des Maladies Infectieuses, Hôpital de Creil, Creil, France), Denis Ouzan (Institut Arnault Tzanck, Saint Laurent du Var, France), Christophe Duvoux (Service d'Hépatologie, AP-HP Hôpital Henri Mondor, Créteil, France), Daniel Dhumeaux (Service d'Hépatologie, AP-HP Hôpital Henri Mondor, Créteil, France), Patrice Cassuto (Service de Médecine Interne, Nice, France), Michèle Bentata (Service de Médecine Interne, AP-HP Hôpital Avicenne, Bobigny, France), Michel Beaugrand (Service d'Hépato-Gastroentérologie, AP-HP Hôpital Jean Verdier, Bondy, France), Nathalie Ganne-Carrié (Service d'Hépato-Gastroentérologie, AP-HP Hôpital Jean Verdier, Bondy, France), Vincent Jeantils (Service de Médecine Interne, AP-HP Hôpital Jean Verdier, Bondy, France), Pascal Longuet (Service des Maladies Infectieuses et Tropicales, AP-HP Hôpital Bichat Claude Bernard, Paris, France), Jean-Didier Grangé (Service d' Hépato-Gastroentérologie, AP-HP Hôpital Tenon, Paris, France), Stanislas Pol (Service d'Hépatologie, AP-HP Hôpital Cochin, Paris, France), and Anne Gervais (Service des Maladies Infectieuses et Tropicales, Hôpital Bichat Claude Bernard, Paris, France).
- 17the CPCRA 046 Study Team for the Terry Beirn Community Programs for Clinical Research on AIDS. A randomized study of antiretroviral management based on plasma genotype antiretroviral resistance testing in patients failing therapy. AIDS 2000; 14: F83–F92., , , , , , et al., and