Immunogenicity of recombinant hepatitis B vaccine (rHBV) in recipients of unrelated or related allogeneic hematopoietic cell (HC) transplants

Liver disease secondary to hepatitis B (HepB) virus is a vaccine-preventable disease.^1,2  It is currently estimated that 1 million to 1.3 million individuals living in the United States are chronically infected with HepB virus, 5000 of whom will die each year from liver failure or hepatocellular carcinoma.³  In 1991, the Immunization Practices Advisory Committee (ACIP) recommended that all infants be vaccinated with recombinant hepatitis B vaccine (rHBV) starting at birth.⁴  In 1999, this recommendation was expanded to include all children younger than 18 years of age.⁵  Immunization of infants, children, and young adults with 3 age-appropriate doses of rHBV is associated with a greater than 95% seroconversion rate.⁶  Approximately 85% of healthy adults older than 40 years of age will also develop positive titers following rHBV.⁷  The vaccine is less immunogenic in patients undergoing dialysis for end-stage renal disease, in patients with advanced liver disease due to hepatitis C, and in individuals immunocompromised by HIV (reviewed in Yu et al⁸ ).

To date, a limited number of studies have evaluated the ability of HC transplant recipients to develop and maintain anti–HepB surface antigen (anti-HBs) titers following rHBV.⁹  In 1997, Li Volti et al¹⁰  evaluated anti-HBs titers in 20 hepatitis B virus–seropositive patients with thalassemia who underwent an unmodified allogeneic transplantation from an HLA-matched hepatitis B virus–seronegative sibling. Fifty percent of patients, including 9 who were previously immunized, became seronegative after HCT. Following 2 doses of the recombinant hepatitis B virus vaccine (Engerix-B), all patients developed a rise in anti-HBs. Machado¹¹  and Machado et al¹²  reported a 100% seroconversion rate in 5 autologous and 45 HLA-matched related HC transplant recipients immunized at least 1 year after transplantation. Despite the excellent initial response, 60% of patients failed to sustain titers for more than 1 year following vaccination.^11,12 

To determine the ability of allogeneic transplant recipients to respond to rHBV, the immunization records and prevaccine and postvaccine titers of 292 patients vaccinated against HepB virus after an unrelated or related HCT were analyzed and compared with responses attained following vaccination with inactivated poliovirus (IPV). The effects of transplant type, patient age, stem-cell and donor type, use of T-cell depletion, history of prior acute or chronic graft-versus-host disease (GVHD), and the use of donor leukocyte infusion (DLI) and/or humanized monoclonal anti-CD20 antibody (rituximab¹³ ) on rates of seroconversion and duration of response were assessed.

This study was approved by the Memorial Sloan-Kettering Institutional Review Board. The medical records of all hepatitis B surface antigen (HBsAg)–negative patients who received an allogeneic HCT from an HBsAg-negative donor at Memorial Sloan-Kettering Cancer Center (MSKCC) from January 1, 1992, through April 15, 2004, and survived disease-free for longer than 1 year were reviewed. Three hundred seven patients were identified who received 3 recombinant DNA hepatitis B immunizations. Criteria for vaccination with rHBV and IPV at this center included a circulating CD4-cell count of at least 200 cells/μL, IgG level of 5 g/L (500 mg/dL) or greater at least 8 weeks following the last dose of intravenous immune globulin, and in vitro T-cell response to phytohemagglutinin at least 75% of the lower limit of normal. Patients with a prior history of GVHD (grade II-IV acute, limited, or extensive chronic GVHD) were eligible for revaccination if they were off systemic immunosuppressive therapy and had minimal or no signs or symptoms attributable to GVHD.^14,15  Patients who were hepatitis C antibody positive but did not have severe liver disease at the time of vaccination were included in the analysis (n = 7). Patients were vaccinated with an age-appropriate dose of rHBV from either Merck (Whitehouse Station, NJ) or SmithKline Beecham Pharmaceuticals (Philadelphia, PA), as recommended by the manufacturers. The majority of patients were vaccinated at times 0, 1, and 6 months. Seventy-six percent of patients were vaccinated at MSKCC. All patients were evaluated at MSKCC before and after vaccination, including assessment of acute and chronic GVHD using established criteria.¹⁴ 

Titers against hepatitis B surface antigen (anti-HBs) were obtained prior to vaccination and 2 to 12 months following vaccination (median 5.9 months). Anti-HBs titers were measured using the AUSAB enzyme-linked immunoassay kit manufactured by Abbott Laboratories (Abbott Park, IL) using controls provided by the manufacturer. Patients who tested seronegative prior to vaccination were considered responders if their postvaccine serum testing showed absorbance above or equal to the kit cutoff value.

Fifteen (4.9%) of 307 vaccinated patients lacked follow-up titers and were excluded from analysis. The remaining 292 patients formed the basis of this report.

Patient and transplant demographics are shown in Table 1. The median patient age at transplantation was 24 years, with a range of 0.2 to 69.0 years. The stem-cell donor was an HLA-A, -B, -DRβ1–identical sibling; an HLA-mismatched family member; or unrelated donor in 65%, 9%, and 25% of cases, respectively. Sixty-five percent of patients received a T-cell–depleted (TCD) HC transplant. Bone marrow was T-cell depleted by soybean lectin agglutination followed by rosetting with sheep red blood cells (n = 149)¹⁶  or treatment with the monoclonal antibody T10B9 and complement (n = 6) as previously described.¹⁷  G-CSF–mobilized peripheral-blood stem cells underwent positive selection for CD34⁺ stem cells followed by T-cell depletion by rosetting with sheep erythrocytes (n = 37). The only patients who received GVHD prophylaxis following a TCD HCT were the 6 recipients of an unrelated T10B9-treated bone marrow (BM) transplant.

The majority of recipients of a T-cell replete (unmodified) HC transplant (n = 100) received cyclosporine A (CyA) and methotrexate (n = 57) for GVHD prophylaxis. Two infants with severe combined immunodeficiency disease and one adult recipient of a syngeneic transplant did not receive GVHD prophylaxis after an HLA-matched related unmodified transplantation. The remaining 40 patients received CyA (n = 6) or methotrexate alone (n = 3), CyA/steroids (n = 3), CyA/methotrexate/steroids (n = 7), CyA/alemtuzumab (n = 8), CyA/mycophenolic acid (n = 3), or tacrolimus and methotrexate (n = 10).

Twenty-five (8.6%) of the 292 patients received the monoclonal antibody rituximab for the treatment of a posttransplantation Epstein-Barr virus lymphoproliferative disorder (EBV-LPD; n = 8), EBV viremia (n = 12), autoimmune cytopenia (n = 2), or recurrent lymphoma (n = 3). Sixty-five patients received adoptive immunotherapy in the form of unfractionated donor lymphocytes for the treatment or prevention of an EBV-LPD (n = 13); to prevent or treat recurrent malignancy (n = 43), an autoimmune cytopenia (n = 1), or mixed T-cell chimerism (n = 4); or to promote immune reconstitution (n = 4).

Frequency distributions of characteristics of patients who responded to rHBV or IPV vaccination and of the study cohort were summarized. Univariate and multivariate logistic regression analysis was used to examine the odds ratio for differences between vaccine responders and nonresponders. The Cuzick¹⁸  test for nonparametric trends across ordered groups was used to determine the effect of age categories on seroconversion rates. All statistical tests were considered significant at a 2-sided α-level of 0.05. Statistical analyses were performed using Stata 7 (release 7, 2000; Statacorp, College Station, TX).

The median time to initiate vaccination against hepatitis B was 23.4 months (range, 5.3-101.6 months) after HCT. There was no significant difference in the time to vaccination in T-cell–depleted versus T-cell–replete transplant recipients (23.5 months versus 22.5 months). In patients younger than or older than 18 years of age, the median time to vaccination was 20.0 and 24.5 months after HCT, respectively. This reflected in part the time to achieve immune milestones required for revaccination in children and adults following HCT.

Two hundred sixty-seven of 292 patients were evaluable for seroconversion. This included 58 of 83 patients who were anti-HBs positive before HCT but lacked detectable titers when re-evaluated at a median 22.9 months (range, 5.6-73.6 months) after transplantation. At the time of vaccination, none of the 30 anti-HBs–negative patients who received a transplant from an anti-HBs–positive donor had passively transferred antibody detected.

Overall, 171 (64%) of 267 patients seroconverted after vaccination (Table 2), including 72 (73%) of 99 children and 99 (59%) of 168 adults (P = .02). There were no significant differences in the proportion of responders on the basis of pretransplantation diagnosis, stem-cell source (bone marrow or peripheral blood), donor type, use of T-cell depletion, inclusion of total body irradiation in the conditioning regimen, or seropositive donor (Table 2). In univariate analysis, younger age at transplantation (P < .001) and at initial immunization was associated with significantly higher rates of seroconversion (P < .05). Prior acute grade II-IV GVHD (P = .02) or chronic GVHD (P < .001) was associated with poorer response. In multivariate analysis, younger age (P = .001) remained associated with vaccine response. Chronic GVHD (P < .001) remained associated with nonresponse.

The results of vaccination with rHBV contrast sharply with results of IPV vaccination. At the time of revaccination (median, 19.2; range, 4.6-113.3 months after HCT), only 46 patients lacked residual polio titers. Ninety-six percent of 219 patients developed a 3-fold or greater rise in titer against all 3 polio serotypes after receipt of 3 IPV vaccinations given at 2-month intervals, including 70 (89%) of the 79 non-rHBV responders. Only younger age at transplantation (P = .01) was associated with improved response to IPV vaccination (Table 3).

Although the number of patients is small, the poorest response to rHBV was observed in recipients of an HLA-mismatched related HC transplant, irrespective of prior GVHD or use of T-cell depletion. Following an HLA-mismatched related HCT, 56% of patients developed positive anti-HBs titers compared with 63.8% and 64.8% of recipients of an HLA-matched related or unrelated HC transplant, respectively. Within all HCT groups, there was a trend for decreasing seroconversion with increasing patient age (P = .04; Table 4). There were no significant differences in responses among recipients of a TCD or unmodified related or unrelated transplant (P = .2). Within the pediatric age group, 18 (85.7%) of 21 patients who received a TCD unrelated transplant responded to rHBV compared with 8 (72.7%) of 11 unmodified unrelated transplant recipients. Of the 12 children vaccinated against HepB less than 1 year after transplantation, 5 of 5 unrelated transplant recipients and 5 of 7 HLA-matched related transplant recipients developed positive titers. Fifty percent of adults seroconverted following an unrelated unmodified HCT (n = 8) compared with 58% of adults following a unrelated TCD HCT (n = 24).

Twenty-five patients vaccinated against the HepB virus received rituximab after HCT. The median age of these patients was 31 (range, 4-63) years. The median time to vaccination was 23.6 (range, 8.8-113.3) months after HCT and a median of 16.6 months after the last dose of rituximab. Sixteen of 25 patients lacked anti-HBs at transplantation, and 23 lacked detectable anti-HBs titers at the time of vaccination. Twelve patients seroconverted, 11 did not respond, and 2 patients retained immunity following rHBV.

Sixty-five vaccinated patients received DLIs after HCT. Fifty-two of the 65 recipients of DLIs were anti-HBs negative before HCT. DLI was derived from a HepB virus–seronegative donor in 36 cases, a seropositive donor in 6 cases, or HepB core antibody–negative donor whose anti-HBs status was unknown in 10 cases. Passive transfer of anti-HBs was not observed in any patient. Thirty (57.6%) of 52 HepB virus–seronegative patients seroconverted. When compared with age-matched patients who did not receive DLI or rituximab, there was no significant difference in the rate of seroconversion.

The proportion of patients who maintained positive anti-HBs titers following immunization was evaluated. Of the 149 patients who were more than 24 months out from their last rHBV, 98 had serial titers obtained. Eighteen patients became seronegative by 5 years after HCT and an additional 8 patients by 10 years following transplantation. During the 10-year observation period, only one patient lost protective polio titers. Loss of titers against hepatitis B did not correlate with age or use of T-cell depletion but did correlate with transplantation from an unrelated or HLA-mismatched related donor (P < .05).

Ninety-six patients failed to respond to immunization with 3 rHBVs. After failing a primary series, 31 patients were revaccinated with rHBV. Although no patient responded to a single booster vaccine, 17 (55%) patients responded after receiving an additional 3 rHBVs.

Over the last decade, improved disease-free survival following HCT has resulted in increasing numbers of HCT survivors returning to school and jobs that require vaccination against hepatitis B prior to re-entry. Whereas vaccination of healthy children and young adults is considered protective against vaccine-preventable diseases due to the greater than 90% seroconversion rates observed, the ability of HCT patients to respond to childhood vaccines is less well studied. Moreover, the immunogenicity of vaccinations in recipients of an HLA-unrelated HC transplant, HLA-mismatched related HC transplant, or T-cell–depleted HC transplant remains largely unknown (reviewed in Machado⁹ ).

Our data demonstrate that 64% of patients vaccinated against hepatitis B after HCT develop protective titers. The immunogenicity of the recombinant vaccine is decreased, however, in patients older than 18 years of age when compared with age-matched controls and in patients with a history of grade II-IV acute GVHD or chronic GVHD. In a smaller study of allogeneic or autologous HC transplant recipients, Machado et al¹²  also found an association of older age and prior chronic GVHD with poorer response to the rHBV. Similar to reports in healthy controls, approximately 50% of allogeneic transplant recipients who fail an initial series of rHBV will respond to a second series.¹  In contrast to the results of rHBV, response to polio virus was excellent in all age groups and not adversely affected by prior acute or chronic GVHD. Ninety-six percent of patients responded to IPV vaccination. The greater response to IPV vaccination likely reflects in part the multiple antigenic peptides included in the Salk vaccine compared with the single small envelope (S protein) contained in recombinant hepatitis B vaccines currently licensed in the United States (reviewed in Poland and Jacobson¹ ). It is also possible that responses reflect expansion of memory T and B cells persisting after childhood immunization, which were transferred in the unmodified or even the T-cell–depleted grafts.

The impact of age and prior chronic GVHD on hepatitis B responses observed in this study occurred despite our requirement that patients be off all immunosuppressive medications and that they had achieved specific minimal milestones of T- and B-cell immune competence. Our rationale for this approach was based on concerns that vaccination prior to the development of these milestones would be less effective in inducing immediate and long-term T- and B-cell responses. Indeed, qualitative and quantitative differences in effector and memory helper T and/or B cells may differentiate long-term responders and nonresponders.^19,20  Although all HC transplant recipients are subject to a varying period of T and B lymphocytopenia and subsequent functional incompetence despite return of normal numbers of T and B cells, the severity and duration of this immunoincompetence is not uniform across transplant types and/or patients of all ages.^21,22  For these reasons, we hypothesized that immunization guidelines based on fixed times after HCT^23,24  are unlikely to result in adequate protection against vaccine-preventable diseases in all patient groups. Indeed, based on studies from our own series and that of others,^11,12  vaccination with rHBV starting at 6 or 12 months following HCT, particularly in older adults, will likely precede reconstitution of required memory and effector lymphoid populations necessary for long-term immunity.^19,20  In the study of Machado et al,¹²  60% of patients vaccinated lost protective anti-HBs titers by one year following immunization. In contrast, in our own study in which patients were immunized only after recovering a specific level of immune reconstitution (which varied in time from 6 months to > 2 years), only 20% of patients lost protective levels of antibody by 5 years after immunization. Clearly, differences in several variables such as age, incidence and severity of acute and/or chronic GVHD, and/or the type of transplant might contribute to these differences. However, differences in the durability of the response do underscore the need for prospective studies to examine and compare the effectiveness of current vaccination protocols.

Studies in healthy individuals have demonstrated that at least 60% of children vaccinated as infants and 98% of young adults vaccinated in adolescence retain protective titers 10 years following HepB vaccination.²⁵  The 1990 measles epidemic in the United States (reviewed in Ada²⁶ ), the spread of measles in a bone marrow transplantation (BMT) unit in São Paolo,²⁷  the outbreak of varicella in a daycare center²⁸  in which the index case and the majority of affected children had previously received varicella virus vaccine live, and the recent increase in pertussis in healthy²⁹  individuals and oncology patients³⁰  underscore the importance of systematic evaluation of the duration of positive antibody titers following vaccination, even in immunocompetent individuals. Despite the retrospective nature of this study, it underscores the need to obtain prevaccine and postvaccine titers in order to ensure seroconversion and long-term immunity against hepatitis B virus and suggests that studies using higher doses of rHBV in patients older than 40 years or with a prior history of GVHD, a strategy efficacious in dialysis patients, should be undertaken (reviewed in Poland and Jacobson¹  and Yu et al⁸ ). Prospective trials that compare initiation of immunization based on the time interval from transplantation versus schedules based upon achievement of specific benchmarks of functional immune reconstitution, irrespective of the time elapsed from transplantation, are needed.

We wish to thank the nurses and nurse practitioners, including Bernadette Cuello, Ruth Ford, Joanne Taylor, Anne Casson, Joanne Torok-Castanza, Catherine Copeland, and Heidi Abendroth, of the bone marrow transplant service who helped vaccinate and obtain titers on our patients.