Prevention of de novo hepatitis B with adefovir dipivoxil in recipients of liver grafts from hepatitis B core antibody–positive donors†
This investigator-initiated phase 4 study was funded in part by Gilead Sciences.
Lamivudine has been shown to prevent de novo hepatitis B virus (HBV) infections in liver transplantation (LT) patients receiving hepatitis B core antibody–positive (HBcAb+) grafts, but it may produce long-term resistance. Adefovir dipivoxil (ADV) might be effective in preventing de novo hepatitis and resistance. A single-center, prospective trial was conducted with 16 adults (10 men and 6 women, mean age = 54 ± 11 years) who underwent LT with HBcAb+ grafts between September 2007 and October 2009. After LT, patients were given ADV [10 mg daily (adjusted for renal function)]. No hepatitis B immune globulin was administered. At LT, all graft recipients were hepatitis B surface antigen–negative (HBsAg−), 38% were surface antibody–positive (HBsAb+), and 50% were HBcAb+. The median follow-up after LT was 1.8 years (range = 1.0-2.6 years). All recipients had undetectable HBV DNA (<40 IU/mL) after LT until the end of follow-up. One recipient (6%) who was HBsAb− and HBcAb− before LT became HBsAg+ after 52 weeks. One recipient was switched from ADV to entecavir for chronic renal insufficiency, and 19% of the patients had renal dose adjustments. There was a nonsignificant trend of increasing creatinine levels over time (1.2 mg/dL at LT, 1.3 mg/dL 1 year after LT, and 2.0 mg/dL 2 years after LT, P = 0.27). A comparison with a control cohort of LT recipients with hepatitis C virus who did not receive ADV showed no difference in the creatinine levels at LT or 1 year after LT. In conclusion, ADV prophylaxis prevents HBV replication in recipients of HBcAb+ livers but does not fully protect recipients from de novo HBV. Long-term follow-up is needed to better determine the risk of de novo infection. Liver Transpl, 2012. © 2012 AASLD.
Hepatitis B surface antigen–negative (HBsAg−) recipients who undergo liver transplantation (LT) with a hepatitis B core antibody–positive (HBcAb+) graft (a type of extended criteria donor organ) are at risk of developing de novo hepatitis, and they generally require lifelong antiviral prophylaxis.1 Various prophylaxis regimens consisting of antivirals, hepatitis B immune globulin (HBIG), and vaccination have been reported with variable degrees of effectiveness.2-5 The optimal regimen is still uncertain and must balance viral suppression, risks of adverse effects, and costs.
Lamivudine (LAM) has been demonstrated to prevent de novo hepatitis B virus (HBV) infections and is widely used,4 but it has a theoretical potential for producing resistance.6, 7 The actual risk of developing resistance is uncertain; most studies have been small and retrospective with limited follow-up. Adefovir dipivoxil (ADV) is used in patients with LAM-resistant chronic HBV and as a rescue treatment for patients with de novo HBV after LAM prophylaxis failure.8 Therefore, ADV might serve as effective primary prophylaxis against de novo HBV with a lower risk of resistance. Prophylaxis with ADV has not been described in a clinical trial.4, 9
We started a prospective, open-label trial of ADV monoprophylaxis in 2006 and reserved newer nucleos(t)ide analogues for cases of de novo HBV. No HBIG was used. We hypothesized that ADV would be effective in preventing de novo HBV in LT recipients of HbcAb+ grafts.
PATIENTS AND METHODS
We enrolled 16 adults without HBV who underwent deceased donor LT with HBcAb+ grafts between September 2007 and October 2009 at our institution (Columbia University Medical Center) in a single-arm, prospective trial. Patients were excluded if they were HBsAg+ before orthotopic LT, received an HBsAg+ graft, had a pre-LT creatinine level > 1.6 mg/dL, were pregnant or lactating, or were <21 years of age. The study was approved by the Columbia University Medical Center institutional review board and was registered at ClinicalTrials.gov (NCT01146808). Before LT, all patients signed a consent form to receive an HBcAb+ graft and provided written informed consent to participate in this protocol.
ADV [10-mg tablets by mouth daily (adjusted for renal function)] was started on postoperative day 1. No HBIG was administered. The immunosuppression was comprised of mycophenolate mofetil (from which patients were weaned over the course of a year), a corticosteroid taper (which was diagnosis-dependent), and a calcineurin inhibitor [CNI; either tacrolimus or cyclosporine for patients infected with hepatitis C virus (HCV)]. Patients with renal insufficiency were treated with a CNI-sparing protocol using basiliximab to delay the initiation of CNI treatment, as described elsewhere.10 Laboratory monitoring, including HBV serology and HBV DNA tests, were performed every 3 months. The primary measured endpoint was the development of de novo HBV, which was defined as HBsAg positivity or detectable serum HBV DNA, in the first 2 years after LT. Secondary endpoints included renal function at LT and 1 year after LT; these results were compared with those for a separate control cohort of HCV+ LT recipients who did not receive ADV. HCV+ recipients were chosen as a representative control group because they are not typically exposed to ADV and receive the majority of HBcAb+ grafts at our center. A 1-way repeated measures analysis of variance was used to compare creatinine values before and after LT. A 2-tailed α value of 0.05 was considered significant. Data analysis was performed with SPSS 12 (SPSS, Inc., Chicago, IL). Additionally, as part of a second phase of the trial that is still underway, patients without de novo HBV 12 to 18 months after LT were given the option of following an accelerated HBV vaccination schedule (double-dose recombinant HBsAg monthly for 3 months) in an attempt to produce a protective hepatitis B surface antibody (HBsAb) titer and ultimately be weaned off ADV.
The patients had a mean age of 54 years, and 62% were male; no one was on dialysis (Table 1). Patient 2 underwent combined liver-kidney transplantation for concurrent chronic kidney disease. The median follow-up after LT was 1.8 years. Forty-four percent had an episode of acute cellular rejection (ACR) according to liver biopsy; only 2 of these 7 patients (29%) were treated with bolus steroids because of an institutional protocol for avoiding steroids for mild ACR in HCV+ patients. Patient 5 underwent repeat LT 2 years after LT for refractory ACR (Table 2). Patient 15 died 1 year after LT because of recurrent HCV without evidence of de novo HBV on allograft biopsy samples.
Table 1. Baseline Demographic and Clinical Information for the Patients
|Mean age (years)||54 ± 11|
|Sex [n (%)]|| |
| Male||10 (62)|
| Female||6 (38)|
|Race [n (%)]|| |
| Caucasian||9 (56)|
| Latino||3 (19)|
| African American||1 (6)|
| Other||3 (19)|
|Etiology of liver disease [n (%)]|| |
| HCV||9 (56)|
| Alcohol||2 (13)|
| Other||5 (31)|
|Presence of HCC [n (%)]||6 (38)|
|Pre-LT serological status [n (%)]|| |
| HBsAb+||6 (38)|
| HBcAb+||8 (50)|
| HBsAb−/HBcAb−||7 (44)|
|Post-LT serological status [n (%)]|| |
| HBsAb+||7 (44)|
| HBcAb+||10 (63)|
| HBsAb−/HBcAb−||5 (31)|
|Post-LT immunosuppression regimen [n (%)]|| |
| Tacrolimus||8 (50)|
| Cyclosporine||8 (50)|
| CNI-sparing||2 (12)|
Table 2. Demographic and Laboratory Data for the Patients Who Received HBcAb+ Grafts
|7/male||Latino||36.0||2.1||Secondary biliary cirrhosis||No||12||2.5||1.32||0.9||1.0||0.9|
|12/female*||Caucasian||48.1||1.3||Primary biliary cirrhosis||No||37||27.2||2.04||3.0||1.6||—|
|14/male||Other||24.3||1.0||Primary sclerosing cholangitis||No||18||14.8||1.42||0.8||0.6||—|
At LT, all graft recipients were HBsAg−. The serological profiles at LT and at the last follow-up after LT are described in Table 3. Fifty percent of the recipients (5/10) who were HBsAb− before LT became HBsAb+ after LT; 80% of these patients (4/5) were positive only temporarily (1 was vaccinated for HBV before LT, and 2 were vaccinated after LT), and 1 recipient (patient 5) continued to be positive and also underwent repeat LT. No patients who were HBsAb+ before LT lost their surface antibodies after LT. Among the recipients who were HBcAb− initially, 50% (4/8) became HBcAb+ transiently at some point after LT, and 25% (2/8) became HBcAb+ over the long term. One recipient (6%) who was HBsAb− and HBcAb− before LT (patient 3) developed HBsAg positivity after 52 weeks (with HBV DNA persistently undetectable), and he was switched to tenofovir. The patient had been adhering to his medications according to self-reports, the clinical team, and a review of the immunosuppressant drug levels. At the time of documented seroconversion, he had normal transaminase levels and no graft dysfunction. All graft recipients had undetectable HBV DNA (<40 IU/mL) after LT.
Table 3. Serological Profiles of the Patients Who Received HBcAb+ Livers
The mean creatinine level was 1.2 ± 0.6 mg/dL at LT (n = 16), 1.3 ± 0.5 mg/dL 1 year after LT (n = 15), and 2.0 ± 1.0 mg/dL 2 years after LT (n = 7). According to a comparison of the mean creatinine levels of patients at least 2 years after LT (n = 7), there was a nonsignificant trend of increasing creatinine levels over time (1.2 ± 0.3 mg/dL at LT, 1.3 ± 0.3 mg/dL 1 year after LT, and 2.0 ± 1.0 mg/dL 2 years after LT, P = 0.27). A cohort of HCV+ patients (n = 140) who underwent LT at our institution during the same time period but did not receive ADV was used as a control group for comparing renal function. There was also an increase in the mean creatinine level of the control cohort over time (1.1 ± 0.7 mg/dL at LT versus 1.4 ± 0.7 mg/dL 1 year after LT, P < 0.005), and there was no difference in the mean creatinine levels of patients who received ADV and patients in the control cohort (1.2 ± 0.6 mg/dL versus 1.1 ± 0.7 mg/dL at LT, P = 0.72; 1.3 ± 0.5 mg/dL versus 1.4 ± 0.7 mg/dL 1 year after LT, P = 0.97). Patient 6 was switched from ADV (10 mg every other day) to entecavir (0.5 mg every other day) for chronic renal insufficiency (pre-LT baseline creatinine clearance = 52 mL/minute), with no evidence of HBV seroconversion (HBsAb+, HBcAb+, and HBV DNA−) at the end of follow-up at 1-year (creatinine clearance = 29 mL/minute). Patients 1, 12, and 16 (19%) had ADV dose adjustments for renal dysfunction; 1 of these patients already suffered from renal insufficiency at LT (Table 2).
Our study is the first prospective trial of ADV prophylaxis for the prevention of de novo HBV in HBsAg− LT recipients of HBcAb+ donor grafts. The optimal utilization of HBcAb+ livers is becoming increasingly significant internationally11-13 and in the United States.14, 15 ADV was effective in suppressing HBV viral replication (undetectable DNA), but it was not able to prevent 1 patient (6%) from developing new HBsAg positivity. Although HBV DNA is generally considered to be more sensitive than serology, this patient likely had a transient breakthrough that was suppressed by ADV and that would have been missed if HBsAg had not been checked.
Our rate of de novo HBV was comparable to previously reported rates with LAM monoprophylaxis (2.6%-3.6%).3, 5 Interestingly, the recipient who developed de novo HBV had the highest risk profile: he was seronegative for HBsAb and HBcAb (HBV-naive). Our rate of de novo HBV (14% or 1/7) for HBV-naive patients on monoprophylaxis was similar to the reported rates (0%-16%) but was higher than the rates for patients on LAM and HBIG (0% for HBV-naive recipients and 2.8% overall); the reported rates were taken primarily from small series, which often had fewer than 10 to 20 patients and limited follow-up.2, 3 One of the largest cohorts had 108 patients, but this cohort was studied before the widespread use of prophylaxis.16 Our study is one of the few prospective studies that was conducted with regular serological testing and evaluated a sizeable cohort over a relatively long period of time. In particular, we examined a high proportion of HBV-naive recipients. It is important to note that the 1 patient who developed de novo HBV became HBsAg+ after 52 weeks, which is a relatively long time after LT. Additionally, several patients became HBsAb+ or HBcAb+ at some point after LT, but HBV DNA was undetectable; this suggests subclinical HBV and the risk of developing de novo HBV in the future. If HBIG were used as prophylaxis, our study indicates that it would have to be administered for at least 1 year and probably much longer; this would be both costly and inconvenient with likely little additional benefit.
One concern with ADV is the risk of nephrotoxicity.17 This requires close monitoring and dose adjustments for renal insufficiency, which was seen in 19% of our patients. Although no one developed frank renal failure, chronic kidney disease was very common (56% had stage III disease), and there was a nonsignificant trend of increased creatinine levels with increasing time on ADV. Whether this was primarily due to ADV or CNI use is unclear. We did not find a difference in creatinine levels at LT and 1 year after LT in comparison with a separate cohort that did not receive ADV; this suggests nephrotoxicity from CNI use rather than ADV itself, although this is difficult to determine conclusively.
One of the major limitations of our study was the limited sample size. Although our study did not include a comparison group using another intervention such as LAM, we were able to compare our findings to recent reviews2, 3, 5 and a control HCV group from our institution that had not been exposed to ADV. It is also unclear what effect vaccination may have had on patients' serological profiles.
In summary, ADV prophylaxis does not fully protect against de novo HBV. Whether chronic ADV use has an irreversible impact on renal function in this at-risk population is yet to be determined. Long-term follow-up and larger trials are needed to better determine the optimal prophylaxis regimen for reducing the risk of de novo infections.