Recurrence of Hepatitis B Infection in Liver Transplant Patients Receiving Long-Term Hepatitis B Immunoglobulin Prophylaxis

Background

Chronic hepatitis B virus (HBV) infection is a leading cause of cirrhosis and hepatocellular carcinoma (HCC) and is a major cause of mortality worldwide [1]. Although antiviral therapy with nucleos(t)ide analogues (NUCs) can prevent the need for transplantation in as many as 35% of patients with decompensated HBV-related cirrhosis, HBV infection remains a major indication for liver transplantation [2]. Due to the favorable impact of antiviral treatment on decompensated disease, the proportion of wait-listed HBV-positive patients who have HBV-related HCC (HBV-HCC) has increased, while decompensated HBV-disease has become less frequent [3]. Following transplantation, combined prophylactic treatment with hepatitis B immunoglobulin (HBIg) and NUC therapy has profoundly reduced the risk of HBV recurrence [4], substantially improving graft and patient survival rates [2], and is considered the standard of care in this setting [5]. HBIg therapy is usually started intravenously for at least 1 week post-transplant, after which it can be switched to subcutaneous or intramuscular preparations.

Although highly effective, even combined prophylactic therapy cannot entirely prevent recurrence of HBV infection after liver transplantation. A systematic review of 46 studies by Cholongitas and colleagues, involving over 2000 HBV-positive patients treated with HBIg and NUC therapy (lamivudine and/or adefovir), showed recurrence in 6.6% of cases [6]. Patients with HBV-HCC appear to be at higher risk for HCC recurrence than those with non-HBV-HCC [7], and a recurrence rate of 14.8% has been reported even under HBIg and NUC therapy [8]. The available evidence relating to HBV or HBV-HCC recurrence, however, is typically based on single-center cohorts of fewer than 100 patients [6]. Follow-up times are often short (mean 21 months in the 46 studies included in the analysis by Cholongitas et al) [6], but the average time to recurrence has been reported to be up to 44 months for HBV infection [9] and ~26 months for HBV-HCC [8]. Moreover, most estimates are derived from populations in which ongoing prophylaxis with both HBIg and NUC therapy was mandatory for inclusion. This does not necessarily reflect routine practice, in which HBIg is withdrawn in low-risk patients by some centers.

An international, multicenter retrospective analysis was undertaken to evaluate the recurrence of HBV infection and to assess other clinical and serological efficacy endpoints during long-term follow-up in a large cohort of patients who had undergone liver transplantation for HBV-related disease.

Material and Methods

STUDY DESIGN:

A retrospective analysis was performed at 20 liver transplant centers in Italy, Germany, Switzerland, the Netherlands, and the United Kingdom. The inclusion period was from January 2000 to May 2016. From a list of eligible patients, 50% of patients with HBV-HCC and 50% without HCC were to be selected randomly for documentation, stratified by year of transplant. Random selection was performed centrally. It was planned to recruit approximately 400 patients, with an average of 20–25 patients per center (approximately 200 from Italy and approximately 200 in total from Germany, Switzerland, the Netherlands and the United Kingdom).

For patients not alive at time of documentation, no demographic characteristics were collected except for the year of birth. Approval from the Institutional Review Board was obtained for all sites and all living patients signed informed consent before data were collected.

ELIGIBILITY CRITERIA:

Patients were eligible for inclusion if they were aged ≥18 years and had undergone primary transplantation for fulminant hepatitis B, hepatitis B cirrhosis, or HBV-HCC inside the Milan criteria, or liver re-transplantation except due to HBV recurrence, from 2000–2014. All patients were required to have received treatment with HBIg for at least 1 year during the post-transplant period, including a minimum of 6 months’ treatment with intravenous Hepatect^® CP (iv HBIgB, Biotest AG, Dreieich, Germany) or subcutaneous Zutectra^® (sc HBIg, Biotest AG, Dreieich, Germany), with or without concomitant NUC therapy, and to have a minimum of 1 year’s data available. Patients could receive more than 1 HBIg therapy.

DATA COLLECTION:

Data were captured from clinical documentation and patients’ hospital records at time of transplant, and after transplantation until the end of the documentation period. Documentation was obtained for a minimum of 1 year and a maximum of 10 years post-transplant. Demographic and clinical data at the point of transplantation were recorded (unless unavailable for deceased patients). Any HBV recurrence during the documentation period was captured, defined as detection of hepatitis B surface antigen (HBsAg), and/or HBV DNA in the serum based on licensed diagnostic test systems for HBV. Additionally, all pre- or post-transplant measurements of antibodies against hepatitis B surface antigen (anti-HBs) or core antigen (anti-HBc), HBsAg, and HBV DNA were recorded. All clinical signs of HBV recurrence (e.g., jaundice) were documented, as well as recurrence of HBV-HCC or development of de novo malignancies other than HCC. Liver and kidney function test results were recorded pre-transplant and, if applicable, at the time of HBV recurrence. Individual adverse drug reaction reporting was not performed other than routine spontaneous adverse drug documentation by the treating physician via standard pharmacovigilance procedures.

STUDY ENDPOINTS:

The primary efficacy variable was the proportion of patients free from HBV recurrence, as assessed by non-detectability of HBsAg and/or HBV DNA in serum during the documentation period. Secondary efficacy variables were the proportion of patients with HBV recurrence, the time to HBV recurrence, the proportion of patients with HBV-HCC recurrence, and serum levels of anti-HBs, HBsAg, and HBV DNA. Other efficacy variables included viral status, exposure time to different HBIg therapies, use of different antiviral treatments, immunosuppressive treatments, graft rejection (acute and chronic), and occurrence of de novo malignancies other than HCC. There were no safety variables.

STATISTICAL ANALYSIS:

No formal sample size calculation was performed. For the primary endpoint a 2-sided 95% confidence interval (CI) was calculated based on the Clopper-Pearson method. Two-sided 95% CIs were also calculated for the percentages of patients with HBV recurrence, HBV-HCC recurrence, or occurrence of any new non-HCC malignancy during the period of documentation. Times to first HBV recurrence, HBV-HCC recurrence, or onset of any de novo malignancy (non-HCC) were estimated by the Kaplan-Meier method. All analyses were performed using SAS^® version 9.3.

Results

STUDY POPULATION:

In total, 371 patients met all eligibility criteria and were included in the analysis (Figure 1). Of the 332 who were alive at the time of study initiation, 257 were male (77.4%) and the mean (±SD) age was 58.5 (±10.7) years (data were not available for deceased patients). Characteristics of the study population overall, and for the subgroup for whom the primary indication for liver transplantation was HBV-HCC (n=147), are shown in Table 1. Five patients with HBV-HCC as the primary indication for transplantation who underwent downstaging pre-transplant were formally outside the Milan criteria according to the documented radiological results. Since the histology results indicated successful downstaging and the participating physicians considered all patients to be inside the Milan criteria at the time of liver transplantation, the patients were not excluded from the analysis.

Among patients in whom serological data were available before liver transplantation, 183/227 (80.6%) were anti-HBc positive, 279/300 (93.0%) were HBsAg-positive, and 101/239 (42.3%) had detectable levels of HBV DNA. For donors in whom data were available, 36/131 (27.5%) were anti-HBs-positive, 53/264 (20.1%) were anti-HBc-positive, and 4/279 (1.4%) were HBsAg-positive.

The most frequent immunosuppressive agents in the initial post-transplant maintenance regimen were calcineurin inhibitors (93.8%), mycophenolate mofetil (37.7%), and corticosteroids (68.2%) (Table 1) and these percentages were 82.7%, 31.3%, and 28.6%, respectively, at the last documentation.

The mean (±SD) follow-up time was 6.8 (±3.5) years (median 7.0 years, range 1.0–15.1 years). The total observation period across all patients was 30 781 months, including time with or without HBIg treatment. The mean (±SD) survival time was 6.8 (±3.5) years.

HBIG THERAPY:

All patients received intravenous HBIg immediately after transplantation, except for 2 patients who were given sc HBIg. In total, iv HBIgB was given to 299/371 patients (80.6%). At least 1 change in HBIg therapy occurred in 250 patients (67.4%), with 179, 47, 17, 4, and 3 patients changing 1, 2, 3, 4, or 5 times, respectively. After initial intravenous therapy, a total of 236 patients (63.6%) were switched to sc HBIg (Table 2). One hundred and thirty patients (35.0%) received another HBIg product, including intramuscular Igantibe® (Instituto Grifols S.A., Barcelona, Spain) and intravenous Niuliva® (Instituto Grifols S.A., Barcelona, Spain). At the end of documentation, 270/371 patients (73.8%) were still receiving HBIg therapy (iv HBIgB n=37, sc HBIg n=205, other HBIgs n=28).

The mean duration of exposure was 30.8, 36.4, and 42.1 months for iv HBIgB, sc HBIg, and other HBIgs, respectively, with total treatment durations of 8993, 8379, and 5392 months. For a total of 8017 months, no HBIg treatment was given. The median time to starting therapy was 1 day for iv HBIgB compared to 847 days for sc HBIg and 208 days for other therapies. The median daily dose of iv HBIgB (238 IU, range 12–10 000) was higher than for sc HBIg (71 IU, range 16–143) or other HBIgs (71 IU, range 16–5000), reflecting use of higher dosing in the early post-transplant period (Figure 2A). From month 3 post-transplant onwards, the mean dose was stable for each category of HBIg therapy, but remained higher for iv HBIgB than for sc HBIg or other HBIg products (Figure 2B). Almost all patients received sc HBIg at home (96.6%), compared to no patients given iv HBIgB and 56.9% of patients given other HBIg therapies.

There were no reports of HBIg-associated adverse events.

ANTIVIRAL THERAPY:

Prior to transplantation, 217 patients (58.5%) received a NUC, most frequently lamivudine (n=127). After transplantation, 347 patients (93.5%) were given at least 1 antiviral medication, including a NUC in each case. Lamivudine was again the most frequently prescribed NUC (225/272, 60.6%) (Table 1), with lamivudine/adefovir the most common combination of NUCs (12.1% of patients). Twenty patients (5.4%) received antivirals for reasons other than HBV prophylaxis.

ANTI-HBS:

At the end of the documentation period, anti-HBs test results were available for 317 patients, of whom 275 (86.8%) had a positive result. Of the 273 patients with quantitative data, 231 (84.6%) had a serum level ≥100 IU/l.

In the subgroup who were receiving HBIg therapy at the end of documentation and for whom data were available, 239/243 (98.4%) were anti-HBs-positive and 204/239 (85.4%) had a serum level ≥100 IU/l (Table 3).

HBV RECURRENCE:

The proportion of patients free of HBV recurrence was 95.7% (95% CI [93.1%, 97.5%]). HBV recurred in 16 patients (4.3%; 95% CI [2.5%; 6.9%]) during the documentation period, at a mean (±SD) of 36.3 (±35.6) months post-transplant (median 22.3 months, range 0.2–103.3 months) (Figure 3). Characteristics of the 16 patients with recurrence are shown in Table 4. The immunosuppression regimen immediately after liver transplantation in this subgroup included calcineurin inhibition in 15/16 cases (93.8%), mycophenolate mofetil in 4/16 cases (25.0%), azathioprine in 4/16 cases (25.0%), and corticosteroids in 10/16 cases (62.5%), with anti-IL-2 receptor antibody induction given to 3/16 patients (18.8%).

The rate of HBV recurrence per year was 0.65% across the total observation period.

Among patients with HBV recurrence for whom pre-transplant HBV test results were available, 12/12 were HBsAg-positive and 8/12 were HBV DNA-positive. Hepatitis D virus (HDV) co-infection was present in 4/16 patients who experienced HBV recurrence (Table 4). Thus, HBV recurrence occurred in 4/114 (3.5%) of HDV-co-infected patients over the whole documentation period and in 0.5% patients per year.

The duration of exposure to sc HBIg or other HBIg therapies, but not iv HBIgB, was shorter in patients with HBV recurrence versus those without recurrence (Table 5). Eleven of the 16 patients with HBV recurrence were receiving HBIg therapy at time of recurrence (6 iv HBIgB, 1 sc HBIg, 4 other HBIg) (11/371, 3.0%) (Figure 4). The remaining 5 patients were not receiving HBIg (2 had stopped treatment <3 months previously). When HBV recurrence was analyzed according to the time during which patients were on HBIg treatment (22 764 months) or not on treatment (8017 months), there was 1 on-treatment recurrence per 2069 months of treatment (11/22 764) and 1 recurrence while not on HBIg therapy per 1603 months without treatment (5/8017). In patients who stopped HBIg treatment compared to those given sc HBIg, the risk of developing HBV recurrence was increased by a factor of 5.2 (1 per 1603 versus 1 per 8379 treatment months).

The serum anti-HBs level at time of recurrence was available in 10/16 patients, 7 of whom had a level <100 IU/l (with 6 patients <50 IU/l) (Table 6). Among these 7 patients, the HBIg treatment was iv HBIgB in 1 case, sc HBIg in 1 case, other HBIgs in 2 cases, and no HBIg in 3 cases. Seven of the remaining 9 patients were receiving HBIg at the time of recurrence, 2 of whom were known to have serum anti-HBs ≥100 IU/l. Assuming conservatively that the other 5 patients also had serum anti-HBs ≥100 IU/l, the rate of recurrence in adequately-treated patients was 1.9% (7/371) (Figure 4). No patient switched to adequate treatment with sc HBIg had recurrent HBV.

At time of recurrence, 13/16 patients were under NUC treatment. Recurrences were observed in patients receiving NUC alone, HBIg with lamivudine, and HBIg with adefovir, but also when HBIg was given in combination with more potent NUCs (Table 4).

GRAFT REJECTION:

Acute rejection occurred in 37/371 (10.0%) of patients and chronic rejection occurred in 7/371 (1.9%) of patients.

HBV-HCC RECURRENCE:

HBV-HCC recurred in 14/147 (9.5%; 95% CI [5.3%, 15.5%]) of patients transplanted due to HBV-HCC. The annual rate of recurrence was 1.7% and the mean (±SD) time to recurrence was 17.6 (±10.8) months (median 14.1 months, range 5.5–41.9 months) (Figure 5A).

Characteristics of the patients with HBV-HCC recurrence are shown in Table 6. Each of these 14 patients had HBV-HCC within Milan criteria confirmed by histopathology at the time of transplantation. Two of the patients also experienced HBV recurrence. Two patients had concomitant alcoholic liver disease and 3 patients had HDV co-infection. HCC recurrence occurred in 3/37 (8.1%) patients co-infected with HDV. The diagnosis of recurrence was based on clinical/radiological findings in 8 cases, with histopathological evidence in 7 cases (confirmation was by both methods in 1 case).

For all 3 categories of HBIg, the mean duration of HBIg therapy was shorter among patients with HBV-HCC recurrence (9, 23, and 16 months for iv HBIgB, sc HBIg, and other HBIgs, respectively) versus patients without HBV-HCC recurrence (32, 37, and 44 months, respectively) (Table 5). Documentation in patients with HBV or HCC recurrence was stopped at the time of determination, which led to a shorter documentation time in these patients. At the time of recurrence, 8 patients were receiving either iv HBIgB or sc HBIg, 5 were being treated with another HBIg, and 1 was not receiving HBIg therapy. All 14 patients were being treated with antivirals, most frequently lamivudine (9/14) (Table 6).

:

In total, 26/371 patients (7.0%; 95% CI [4.6%, 10.1%]) were diagnosed with de novo malignancies during the documentation period, representing an incidence of 1.05% per year. The most frequent of these were skin cancers (squamous cell carcinoma n=3, basal cell carcinoma n=2, cutaneous carcinoma n=2). Kaposi’s sarcoma and lung cancer occurred in 2 patients each; no other cancer occurred in more than 1 patient. The mean (±SD) time to onset was 59.4 (±40.7) months (median 61.4; range 0.8–141.7 months) (Figure 5B). At the time of diagnosis, 16 patients were receiving HBIg therapy and 10 patients had stopped HBIg treatment. The incidence of de novo malignancies in the cohort of patients with HCC recurrence was 12.5% (2/16 patients).

Discussion

This retrospective analysis represents one of the largest data sets available to determine HBV recurrence rates in HBV-infected patients after liver transplantation and explore HCC recurrence in this population. In this European multicenter study, combination prophylaxis with HBIg and NUC therapy was associated with low rates of HBV recurrence (4.3%) over a mean follow-up of almost 7 years. The HBV-HCC recurrence rate was 9.5%.

A recent systematic review which pooled data from 17 randomized or observational studies found the HBV recurrence rate to be 7.1% for patients treated with HBIg and lamivudine, or 1.3% with HBIg and either entecavir or tenofovir (there was no restriction of HBIg duration in either group) [10]. The latter rate is strikingly lower, but comparison with the recurrence rate in the present study (4.3%) is difficult because the follow-up periods differ markedly, with a mean of ~3 years in the pooled analysis compared to 6.8 years in the present study. HBV recurrence has been reported to occur most frequently at approximately 2–4 years post-transplant [8,11], consistent with our observation (mean 36±36 months).

NUC therapy alone, however, has also been explored. Recently, an observational study of 265 Asian liver transplant patients with chronic HBV infection was published, in which patients received entecavir without HBIg therapy [12]. During a maximum follow-up of 8 years, 14/265 (5.3%) patients remained persistently positive for HBsAg. Among 242 patients who experienced HBsAg seroclearance after transplantation, 36 developed HBsAg and none had HBV DNA reappearance. This corresponds to a recurrence rate of 14.8% (36/242), considerably higher than in our study. However, in the Asian population, no HDV-co-infected patient was included, in contrast to our European cohort in which the proportion with HDV infection was relatively high. In our study, the HBV recurrence rate was 3.5% for HDV-co-infected patients, a subgroup in whom life-long combination therapy with HBIg and NUC is recommended [5].

In the present series, 5 of the 16 patients with HBV reinfection had stopped HBIg therapy by the time of recurrence. Notably, 7 patients had documented inadequate HBIg therapy (6 of those had serum anti-HBs <50 IU/l and 1 had <100 IU/l). The threshold of ≥100 IU/l is defined by the European Medicines Agency Committee for Medicinal Products for Human Use as the minimum threshold for effective protection against HBV reinfection using HBIg therapy [13]. Excluding the 7 patients in whom an inadequate serum anti-HBs level (<100 IU/l) was confirmed, the reinfection rate was only 2.4% of all 371 patients at risk. Higher HBIg dosage is associated with a lower frequency of HBV recurrence even in patients receiving combination prophylaxis [7]. When dosed correctly, HBIg is highly effective in maintaining adequate titers [14,15]. In the present cohort, 14.6% of patients had a serum anti-HBs level <100 IU/l, suggesting failure to maintain protective titers in this subpopulation. Withdrawal of HBIg in low-risk patients after a defined course of combined therapy, or use of low-dose or on-demand HBIg therapy, are valid strategies to reduce treatment costs [16,17]. Typically, transplant centers administer higher HBIg doses during the anhepatic phase and the very early phase post-transplant. Various HBIg minimization strategies are then pursued [18]. However, patients must be evaluated carefully before HBIg reduction, based on risk factors such as high pre-transplant HBV DNA, co-infections, and transplantation for concomitant HBV-HCC [16,17]. According to EASL 2017 guidelines, discontinuation should be restricted to patients who were HBV DNA-negative pre-transplant [5]. Here, the shorter duration of exposure to HBIg in patients with HBV recurrence compared to recurrence-free patients is of interest. The observational study design, however, does not permit firm conclusions from this finding; for example, HBIg therapy may have been discontinued after diagnosis of HBV recurrence. Another interesting finding is that, despite some missing values, patients with HBV recurrence had higher HBV DNA levels prior to liver transplantation (median HBV DNA levels were 2 log copies/mL, and only 1 of the patients with recurrence had DNA levels below 2 log copies/mL). This is a risk factor influencing recurrence in some patients.

A prospective study showed that early switch to subcutaneous treatment with sc HBIg (days 8–18 post-transplant), dosed according to serum anti-HBs level in patients who were HBV-DNA-negative at time of transplant, maintained serum anti-HBs levels ≥100 IU/l and no patient developed reinfection [19]. It is striking in the present series that there was only 1 patient with HBV recurrence over a total of 8379 months of sc HBIg treatment. Sub-optimal dosing with sc HBIg may be responsible for this single occurrence, as the anti-HBs level of 18 IU/l was below the minimum threshold for effective protection. In terms of dosing, the median daily dose of sc HBIg over the entire study duration was lower than with iv HBIgB (71 versus 238 IU). This partly reflects the delayed starting time for sc HBIg compared to iv HBIgB, but a dosing difference was sustained even after the early post-transplant period (Figure 2B). sc HBIg was not licensed until 2009, when initiation of therapy was approved at month 6 after transplantation. It was not until late 2015 that a change in the labeling allowed initiation of therapy as early as 1 week post-transplantation. By this time, the last patient included in this study had already been transplanted.

The annual incidence of HCC recurrence in the present study was very low, at 1.7%. This compares favorably with rates of 3.2–6.7% reported elsewhere [8,12,20,21]. It is possible that factors such as tumor history (which was not captured by the present study), tumor stage at transplantation, differences in malignancy surveillance, and the period of transplantation may account for this difference. HBIg treatment may positively influence the risk of HBV-HCC recurrence after liver transplant, in particular for HCC patients within the Milan criteria [11]. Observational studies have shown an association between HBV recurrence and HBV-HCC recurrence after transplantation [20,22,23]. In addition, high HBV viral load is an independent risk factor for HBV-HCC recurrence [24]. These factors may explain why 2 patients with HBV recurrence also experienced HCC recurrence in our study; however, this remains speculative. High-dose HBIg may reduce the risk for HBV-HCC recurrence in patients who meet the Milan criteria [12]. In HDV-co-infected patients, the HCC recurrence rate was 8.1%, suggesting that HCC recurrence does not occur at any higher frequency in this high-risk group using HBIg and NUC combination prophylaxis.

The incidence of skin cancer and lymphoma in liver transplant patients is substantially increased compared to in healthy individuals [25,26]. Published reports of the incidence of de novo malignancy after liver transplantation vary widely, from 2.6% to 15.7% [27], influenced by risk factors such as older age [28], smoking [28,29], transplantation for alcoholic cirrhosis [27], and more intensive immunosuppression [30,31]. Against this background, the low rate of de novo malignancies observed here over this long-term follow-up, at an annual incidence of 1.05%, is encouraging but might be an underestimation due to the study design. Although numbers were small, results in our population did not appear to indicate any marked increase in risk among patients with HCC recurrence compared to the overall population.

In this series, there was a 10% incidence of acute rejection, which is relatively low for a mean follow-up period of almost 7 years. Long-term follow-up data from randomized trials have reported 5-year rejection rates of 11–18% [32,33]. Patients transplanted for chronic HBV-related disease have been reported to be at lower risk for acute rejection versus non-HBV indications [34–36]. There are limited in vitro data suggesting that HBIg may inhibit alloantigen specific T cell responses, dendritic cell maturation, and cytokine production [35]. A recent study [37], albeit with lower HBIg concentrations as in prior in vitro studies [38,39], found no enhancing role on regulatory T cell generation. However, more in vivo and in vitro data are needed to determine the possible immunosuppressive effect of HBIg.

Administration of HBIg in combination with NUC was well tolerated and no HBIg-related adverse events were reported. These findings are consistent with prospective and retrospective studies of combined post-transplant prophylaxis with HBIg and NUC therapy [40].

Some characteristics of the present study should be considered. The analysis was restricted to patients with a minimum of 1 year’s HBIg therapy (and at least 6 months under either iv HBIgB or sc HBIg), but this applied to nearly all cases transplanted at the study centers over the period of documentation. The findings cannot be extrapolated to shorter periods of HBIg treatment. As a retrospective analysis, serological tests, notably anti-HBs levels, were missing or incomplete in some patients and differences in treatment protocols made comparisons between specific therapies impractical. It would have been relevant to document levels of anti-HBs routinely, but regular monitoring was not standard practice in this non-interventional study; levels would be expected to be low at the point of hepatitis B recurrence diagnosis. Patient monitoring, such as for skin cancers, may have been less systematic than in a controlled trial. Inevitably, in a long-term observational study of this type, the dataset is not complete, thus introducing a potential risk of bias. However, the analysis benefitted from a large ‘real-world’ population transplanted since 2000, and a consistent definition for HBV recurrence was applied across all patients. Importantly, it offers a long-term follow-up period of up to 15 years.

Conclusions

Results from this multicenter European retrospective analysis in a large series of patients undergoing liver transplantation for HBV-related disease indicate that routine clinical use of HBIg prophylaxis for a minimum of 1 year, combined with NUC therapy, is associated with a low rate of HBV recurrence and HBV-HCC recurrence over the long-term. Low recurrence rates were also seen in patients with HDV co-infection. These data support the current EASL guidelines [5] that combination therapy with HBIg and NUC is an effective prophylactic strategy for the management of HBV infection in liver transplant recipients.