Winston DJ, Gale RP, Meyer DV, Young LS, the UCLA Bone Marrow ... Schneerson R, Robbins JB: Induction of serum Haemophilus influen- zae type B capsular ...
Biology of Blood and Marrow Transplantation 5:46–50 (1999) © 1999 American Society for Blood and Marrow Transplantation
Defective anticarbohydrate antibody responses to naturally occurring bacteria following bone marrow transplantation Neena Kapoor, Raymond Chan, Kenneth I. Weinberg, Felix Burotto, Robertson Parkman Division of Research Immunology/Bone Marrow Transplantation, Children’s Hospital–Los Angeles, Los Angeles, CA Offprint requests: Neena Kapoor, MD, Division of Research Immunology/Bone Marrow Transplantation, Children’s Hospital–Los Angeles, 4650 Sunset Boulevard, MS#62, Los Angeles, CA 90027 (Received 8 September 1998; accepted 23 December 1998)
ABSTRACT The long-term recipients of allogeneic bone marrow transplantation (BMT) are at an increased risk of death due to bacterial infections. We evaluated the anticarbohydrate antibody responses of BMT recipients to a naturally occurring bacterial carbohydrate, polyribose phosphate (PRP). The recipients of autologous BMT achieved protective anti-PRP levels (.100 ng/mL) by 3 years after transplantation, with a pattern consistent with a recapitulation of the ontogeny of anticarbohydrate antibody responses. None of the six recipients of unrelated BMT who were off immunosuppressive therapy had protective anti-PRP levels, though their response to a protein antigen (tetanus toxoid) was normal. Of 48 recipients of histocompatible BMT, 22 (46%) had protective anti-PRP antibody levels, whereas 13 (27%) recipients who were .3 years post-BMT did not have protective levels. Therefore, all unrelated recipients and a significant proportion of histocompatible recipients without clinical graft-vs.-host disease had persistent and prolonged defects in their capacity to produce antibodies to naturally occurring bacterial carbohydrate antigens. These results suggest that allogeneic BMT recipients should be longitudinally evaluated for their anticarbohydrate antibody responses and that patients with defective antibody responses should receive prophylactic antibiotics or replacement immunoglobulin therapy or both to reduce their risk of late bacterial infections.
KEY WORDS Polyribose phosphate • Allogeneic transplant • Immunocompetence • Posttransplant
INTRODUCTION Recipients of bone marrow transplants (BMTs) have an immunodeficiency of varying duration and severity [1–4]. BMT recipients have an increased incidence of viral and bacterial infections indicating defects in both cellular and humoral immunity. Recurrent infections with encapsulated respiratory bacteria are a particular problem in the recipients of unrelated BMT and the recipients of histocompatible transplants with chronic graft-vs.-host disease (cGVHD) [3,4]. BMT recipients may have deficiencies in antibody production following immunization with viral vaccines (measles, mumps, rubella) or bacterial protein antigens (tetanus toxoid, φX174) early after transplantation, whereas patients with cGVHD
This work was supported by the National Institutes of Health NCRR General Clinical Research Center Grant MO1 RR-43.
have long-term defects in antibody responses to vaccination with pneumococcal carbohydrate antigens . The responses of BMT recipients to endogenous wild-type bacterial carbohydrate antigens have not been systematically studied. Antibody to polyribose phosphate (PRP), the capsular polysaccharide of Haemophilus influenzae type b (Hib), can be induced by stimulation with Hib organisms or by cross-reactivity with the capsular polysaccharides of the K100 strains of Escherichia coli that are normally found in the gastrointestinal tract [5,6]. We report here the long-term evaluation of the anticarbohydrate antibody responses of BMT recipients to naturally occurring PRP and demonstrate that, while the anticarbohydrate antibody responses of the recipients of autologous transplants undergo a recapitulation of the normal ontogeny of anticarbohydrate antibody responses, significant and protracted defects in the antibody response to wild-type bacterial PRP exist in all recipients of unrelated BMT and a significant proportion of histocompatible recipi-
Defective Anticarbohydrate Antibody Responses
ents who do not have detectable cGVHD. The sustained persistence of defects in the anticarbohydrate antibody responses of BMT recipients suggest that the long-term administration of intravenous immunoglobulin or prophylactic administration of antibiotics or both may be required to protect BMT recipients with defective anticarbohydrate antibody responses from significant bacterial infections. The absence of clinical cGVHD does not guarantee that BMT recipients are capable of normal antibody responses to wildtype bacterial carbohydrate antigens.
PATIENTS AND METHODS Study population and treatment Patients between the ages of 4 months and 21 years with both malignant leukemia and lymphoma, and non-malignant aplastic anemia, severe combined immune deficiency, and a variety of genetic diseases, and who had received transplants at Children’s Hospital–Los Angeles between December 1, 1983 and November 30, 1996 were included in the study. The median patient age was 7.5 years and the median donor age was 6 years. Patients received autologous, histocompatible, or unrelated bone marrow (BM). The majority of the patients with both malignant and nonmalignant diseases were prepared for transplantation with busulfan (1 mg/kg q. 6 hours 3 16 doses) and cyclophosphamide (16 mg/kg q.d. 3 4 days), except for patients with severe aplastic anemia who did not receive busulfan. Recipients of unrelated BM also received antithymocyte globulin (20 mg · kg–1 · day–1 3 4 days) before transplantation. BM for autologous transplants for acute leukemia was purged in vitro with 4-hydroperoxycyclophosphamide (4-HC) . GVHD prophylaxis consisted of methotrexate alone for histocompatible recipients ,10 years old and cyclosporine plus methotrexate or steroids for histocompatible recipients $10 years old and for all unrelated recipients. Immunosuppression was tapered by the attending physicians as tolerated by the patients. None of the patients were immunized with either conjugated or unconjugated Hib vaccine. Antibody assays Antibodies to both PRP and tetanus toxoid were measured by an enzyme-linked immunosorbent assay (ELISA) as previously described . Briefly, PRP-human serum albumin conjugate (Richard Insel, University of Rochester School of Medicine, NY) or tetanus toxoid (Wyeth Laboratories, Marettia, PA) were used to coat microtiter plates. The immunoglobulin (Ig)G antibody content of patient samples was determined with a goat antihuman antibody conjugated to horseradish peroxidase. Plates were cultured with o-phenylenediamine and the antibody content calculated from the optical density of triplicate samples compared with a standard pool of normal human Ig. The minimum antibody levels that can be routinely detected are 10 ng/mL for PRP and 0.01 IU/mL for tetanus toxoid.
RESULTS Autologous BMT Seven of 14 evaluated recipients of 4-HC–purged autologous BMT had protective levels (.100 ng/mL) of anti-PRP
Figure 1. Relationship between anti-PRP antibody levels and time from BMT (A) Recipients of autologous BMT; (B) recipients of unrelated BMT; (C) recipients of histocompatible BMT.
antibody. All patients who had non-protective levels of antibody were ,36 months posttransplantation when last evaluated (Fig. 1A). Patient age had no impact on the likelihood of patients having protective levels of anti-PRP antibody. Unrelated BMT Recipients of unrelated BM routinely received monthly Ig replacement therapy while receiving immunosuppressive therapy, thus their ability to produce antibodies could not be assessed. None of the six recipients of unrelated marrow followed for as long as 6 years had protective levels of anti-
N Kapoor et al.
Figure 2. Regression analysis of relationship between age and anti-PRP antibody levels (A) Relationship between histocompatible donor age and anti-PRP antibody levels and (B) histocompatible patient age and anti-PRP antibody levels.
PRP antibodies despite not having cGVHD and no longer receiving Ig replacement or immunosuppressive therapies (Fig. 1B). These patients were between 3 and 12 years old at the time of their most recent assessment. Histocompatible BMT Anti-PRP antibody levels from 48 histocompatible BMT recipients more than 1 year after transplantation are presented in Figure 1C. Twenty-two (46%) patients had protective levels of anti-PRP antibody, whereas 26 patients had nonprotective levels, 13 (27%) of whom were .3 years post-BMT. The impact of donor and recipient age on the production of anti-PRP antibodies was analyzed. A positive correlation existed between donor age and the levels of recipient anti-PRP antibodies (r 5 0.28, p 5 0.04; Fig. 2A) while no correlation was found between recipient age and anti-PRP antibody levels (r 5 0.10, p 5 0.21; Fig. 2B). Anti-PRP and antitetanus toxoid antibody levels To determine if the inability of unrelated BMT recipients to produce protective levels of anticarbohydrate antibodies was a generalized problem in antibody production or was limited to carbohydrate antigens, their antibody responses to a protein antigen (tetanus toxoid) were determined. In all cases recipients were able to produce protective levels (.0.1 IU/mL) of antibody to tetanus toxoid (Table 1).
DISCUSSION Antibody-mediated protection against infection with encapsulated respiratory bacteria (e.g., Haemophilus influenzae, Staphylococci, Streptococci pneumococci) is mediated by antibodies directed against carbohydrate antigens present in the bacterial cell wall or capsule. Ontological delays in the capacity of humans to produce anticarbohydrate antibodies result in an increased incidence of invasive bacterial infections once the protection of transplacentally-derived maternal IgG is lost . Following BMT, a similar ontogeny in the antibody response to bacterial carbohydrate antigens occurs.
In the present study, the recipients of autologous 4-HC–treated BMT recovered protective levels of anticarbohydrate antibodies in a time course similar to that seen in normal human ontogeny (i.e., protective antibody levels were present in all recipients by 2–3 years following BMT, though many recipients did not have protective levels before 2–3 years). Previous studies evaluating the transfer of mature antibody-producing B cells has demonstrated a lack of transfer of antigen-specific B lymphocytes unless the donor is reimmunized before transplantation and the recipient is immunized following transplantation [9–11]. Thus, the development of antibody responses to both protein and carbohydrate antigens after transplantation is due to either the clonal expansion of the limited number of antigenspecific B lymphocytes transferred at the time of transplantation or the generation of new antigen-specific B lymphocytes following hematopoietic stem cell engraftment. Although delays in the production of anticarbohydrate antibodies may be present during the first 3 years after autologous BMT, all patients did develop antibody responses to a protein antigen (tetanus toxoid) by 1 year after immunization (data not presented). Thus, delays in the development of anticarbohydrate antibodies are not part of a generalized defect in antibody production. When recipients of histocompatible BMT were evaluated, approximately one-half of the patients had protective levels of anticarbohydrate antibodies. The patients who did not have protective levels of antibodies could be divided into two groups. Recipients who were ,3 years post-BMT might be expected to have an ontological inability to produce anticarbohydrate antibodies, whereas those recipients who were more than 36 months post-BMT would be expected to have protective levels of anti-PRP antibody. Thirteen patients who were .36 months post-BMT and not on immunosuppressive therapy had a lack of protective levels of anticarbohydrate antibodies for as long as 6 years after BMT, although they were capable of normal antitetanus toxoid antibody responses (data not shown). Five of the patients had no detectable anticarbohydrate antibodies
Defective Anticarbohydrate Antibody Responses
(,10 ng/mL) for up to 78 months posttransplant. Thus, a subset of histocompatible BMT recipients without clinical evidence of cGVHD had persistent defects in the production of anticarbohydrate antibodies despite normal antibody responses to protein antigens. Our results confirm those of other studies that demonstrated that BMT recipients could respond to immunization with protein antigens by 1 year, whereas the majority of the patients (84%) were unable to respond to immunization with purified pneumococcal polysaccharide antigens for up to 2 years [11–14]. As in human infants, BMT recipients are capable of responding to immunization with conjugated PRP vaccines [11,14–17]. The ability to respond to a conjugated carbohydrate vaccine, however, does not predict the capacity of a BMT recipient to respond to wild-type PRP, as demonstrated in the present study. Many recipients of unrelated BMT remain on immunosuppressive therapy because of cGVHD, making the evaluation of their ability to make anticarbohydrate antibodies difficult. However, six patients who were no longer receiving immunosuppressive therapy had nonprotective levels of anticarbohydrate antibodies for as long as 6 years following BMT and 3–4 years following the cessation of their immunosuppressive therapy. Thus, sustained and potentially permanent defects in the ability of BMT recipients to produce protective levels of anticarbohydrate antibodies may exist in both recipients of unrelated BMT as well as some histocompatible recipients. Unrelated BMT recipients have an increased incidence of late-infection mortality unrelated to their GVHD status . The basis for the protracted inability of some BMT recipients to produce antibodies to wild-type carbohydrate antigens is unknown. Because of the inability of some BMT recipients to produce protective levels of anticarbohydrate antibodies, our standard practice is to maintain such patients on daily antibiotic prophylaxis (trimethoprim-sulfamethoxazole) against bacterial infections until they can produce protective levels of anticarbohydrate antibodies. Many transplant programs routinely administer prophylactic antibiotics for only 1 year after BMT and do not assess the ability of their patients to produce anticarbohydrate antibodies. The lack of protective levels of anticarbohydrate antibodies in conjunction with the absence of Ig therapy or antibiotic prophylaxis may contribute to the increased incidence of invasive bacterial infections seen in long-term survivors of BMT. Studies should be undertaken to determine if long-term recipients, especially of unrelated BM, can respond to immunization with conjugated polysaccharide vaccines. The development of protective antibody levels after immunization with conjugated vaccines (Hib, pneumococci), however, will not indicate that such recipients can develop protective immunity to wild-type encapsulated bacteria (e.g., Staphylococci, Streptococci, and Meningococci). We have previously reported that the capacity of BMT recipients to produce new CD41 T lymphocytes (CD41, CD45RA1 cells) is predictive of their capacity to develop Tlymphocyte proliferative responses to tetanus toxoid following transplantation . We anticipated, therefore, that an inverse correlation might exist between patient age and anti-PRP antibody levels since in humans, as opposed to mice, anticarbohydrate antibody responses are T lymphocyte–dependent.
Table 1. Comparison of anti-PRP (PRP) and tetanus toxoid (TT) antibody levels in recipients of unrelated BMT Patient 1 2 3 4 5 6
Time post-BMT (months)
74 57 54 43 31 26
572 510 ,10 ,10 517 ,10
3.0 8.0 3.0 2.1 4.4 3.0
Protective level, .100 ng/mL. Protective level, .0.1 IU/mL.
Thus, defects in the capacity of the thymus to produce new T lymphocytes might be manifested by a decreased capacity to produce anticarbohydrate antibodies. No correlation between patient age and anti-PRP levels was found. However, the maximum age of the patients in the present study was only 20 years, whereas in our previous study, patients as old as 45 years were evaluated. Thus, the negative impact of decreased thymic function on anticarbohydrate antibody production may only be observed in older patients. The molecular basis for the normal ontological delay in production of antibodies to carbohydrates compared with protein antigens is unknown. Immunization with carbohydrate antigens conjugated to immunogenic proteins (conjugated vaccines) results in protective immunization, suggesting that defects in the processing or presentation of unconjugated carbohydrate antigens to T lymphocytes may exist. When normal infants or BMT recipients are immunized with conjugated PRP antigen, protective antibody responses are obtained demonstrating that the T and B lymphocyte interactions necessary for the production of anticarbohydrate antibodies exist following transplantation if the carbohydrate antigen is presented in an immunogenic manner [11–17]. In a previous study, a correlation between both donor and recipient age and the ability of BMT recipients to respond to conjugated PRP immunization was seen . In the study, we noted a correlation between donor and recipient age, therefore, the relative contribution of the donor and recipient could not be evaluated. In the present study, no correlation between donor and recipient age existed (r 5 –0.04); however, a correlation between donor age, but not recipient age, and the development of protective anti-PRP antibody levels was found. The fact that older donors produced higher levels of posttransplant anticarbohydrate antibodies suggests that either there is the transfer of more primed donor B lymphocytes at the time of transplantation or that B lymphocytes derived from older hematopoietic stem cells have a greater capacity to respond to carbohydrate antigens. This study demonstrates that significant and protracted defects in the production of protective levels of anticarbohydrate antibodies to naturally occurring bacteria exist in the recipients of allogeneic BMT despite their documented ability to produce antibodies to protein antigens and conjugated polysaccharide vaccines. The defects in the production of
N Kapoor et al.
anticarbohydrate antibodies may, therefore, contribute to the increased incidence of severe and potentially fatal invasive bacterial infections seen in long-term survivors of unrelated BMT and histocompatible recipients with cGVHD. These results suggest that the routine assessment of the capacity of BMT recipients to produce anticarbohydrate antibodies to wild-type bacteria is indicated. Patients who have defects in anticarbohydrate antibody production might be treated in a fashion analogous to patients with congenital or acquired agammaglobulinemia, i.e., the administration of prophylactic antibodies or the routine administration of replacement immunoglobulin therapy or both.
REFERENCES 1 Winston DJ, Gale RP, Meyer DV, Young LS, the UCLA Bone Marrow Transplantation Group: Infectious complications of human bone marrow transplantation. Medicine 58:1, 1979. 2 Witherspoon RP, Kopecky K, Storb RF, Flournoy N, Sullivan KM, Sosa R, Deeg HJ, Ochs HD, Cheever MA, Fefer A, Thomas ED: Immunological recovery in 48 patients following syngeneic marrow transplantation for hematological malignancy. Transplantation 33:143, 1982. 3 Atkinson K, Storb R, Prentice RL, Weiden PL, Witherspoon RP, Sullivan K, Noel D, Thomas ED: Analysis of late infections in 89 long-term survivors of bone marrow transplantation. Blood 53:720, 1979. 4 Ochs L, Shu XO, Miller J, Enright H, Wagner J, Flipovich A, Miller W, Weisdorf D: Late infections after allogeneic bone marrow transplantations: comparison of incidence in related and unrelated donor transplant recipients. Blood 86:3979, 1995. 5 Schneerson R, Robbins JB: Induction of serum Haemophilus influenzae type B capsular antibodies in adult volunteers fed cross-reacting Escherichia coli 075:K100:H5. N Engl J Med 292:1093, 1975. 6 Käyhty H, Peltola H, Karanko V, Mäkelä PH: The protective level of serum antibodies to the capsular polysaccharide of Haemophilus influenzae type b. J Infect Dis 147:1100, 1983. 7 Lenarsky C, Weinberg K, Petersen J, Nolta J, Brooks J, Annett G, Kohn D, Parkman R: Autologous bone marrow transplantation with 4hydroperoxycyclophosphamide purged marrows for children with acute non-lymphoblastic leukemia in second remission. Bone Marrow Transplant 6:425, 1990. 8 Peterson J, Church J, Gomperts E, Parkman R: Lymphocyte phenotype does not predict immune function in pediatric patients infected with human immunodeficiency virus type 1. J Pediatr 115:944, 1989.
9 Wahren B, Gahrton G, Linde A, Ljungman P, Lonnqvist B, Ringden O, Sundqvist VA: Transfer and persistence of viral antibody-producing cells in bone marrow transplantation. J Infect Dis 150:358, 1984. 10 Wimperis JZ, Brenner MK, Prentice HG, Rettie JE, Karayiannis P, Griffiths PD, Hoffbrand AV: Transfer of a functioning humoral immune system in transplantation of T-lymphocyte-depleted bone marrow. Lancet 1:339, 1986. 11 Molrine DC, Guinan EC, Antin JH, Parson SK, Weinstein HJ, Wheeler C, McGarigle C, Blanding P, Phillips NR, Kinsella K, Deans K, Ciamarra A, Goorin A, George S, Ambrosino DM: Donor immunization with Haemophilus influenzae type b (HIB)-conjugate vaccine in allogeneic bone marrow transplantation. Blood 87:3012, 1996. 12 Shiobara S, Lum LG, Witherspoon RP, Storb R: Antigen-specific antibody responses of lymphocytes to tetanus toxoid after human marrow transplantation. Transplantation 41:587, 1985. 13 Parkkali T, Käyhty H, Ruutu T, Volin L, Eskola J, Ruutu P: A comparison of early and late vaccination with Haemophilus influenzae type b conjugate and pneumococcal polysaccharide vaccines after allogeneic BMT. Bone Marrow Transplantation 18:961, 1996 14 Ambrosino DM, Sood SK, Lee MC, Chen D, Collard HR, Bolon DL, Johnson C, Daum, RS: IgG1, IgG2 and IgM responses to two Haemophilus influenzae type b conjugate vaccines in young infants. Pediatr Infect Dis J 11:855, 1992. 15 Guinan EC, Molrine DC, Antin JH, Lee MC, Weinstein HJ, Sallan SE, Parsons SK, Wheeler C, Gross W, McGarigle C, Blanding P, Schiffman G, Finberg RW, Siber GR, Bolon D, Wang M, Cariati S, Ambrosino DM: Polysaccharide conjugate vaccine responses in bone marrow transplant patients. Transplantation 57:677, 1994. 16 Black SB, Shinefield HR, Fireman B, Hiatt R, Polen M, Vittinghoff E: Efficacy in infancy of oligosaccharide conjugate Haemophilus influenzae type b (HbOC) vaccine in a United States population of 61,080 children. Pediatr Infect Dis J 10:97, 1991. 17 Santosham M, Wolff M, Reid R, Hohenboken M, Bateman M, Goepp J, Cortese M, Salck D, Hill J, Newcomer W, Capriotti L, Smith J, Owen M, Gahagan S, Hu D, Kling R, Lukacs L, Ellis RW, Vella PP, Calandra G, Matthews H, Ahonkhai V: The efficacy in Navajo infants of a conjugate vaccine consisting of Haemophilus influenzae type b polysaccharide and Neisseria meningitidis outer membrane protein complex. N Engl J Med 324:1767, 1991. 18 Weinberg K, Annett GM, Kashyap A, Lenarsky C, Forman SJ, Parkman R: The effect of thymic function on immunocompetence following bone marrow transplantation. Biol Blood Marrow Transplant 1:18, 1995.