Valved allografts are used routinely in the repair of

Transcription

1 Prospective Randomized Trial of Azathioprine in Cryopreserved Valved Allografts in Children Robert E. Shaddy, MD, Linda M. Lambert, BSN, Thomas C. Fuller, PhD, Tracie Profaizer, BS, Dixie D. Thompson, RN, Shawna I. Baker, RN, Karen A. Osborne, RN, and John A. Hawkins, MD Departments of Pediatrics, Surgery, and Pathology, Primary Children s Medical Center and University of Utah, Salt Lake City, Utah Background. The purpose of this study was to prospectively assess the effects of azathioprine on the humoral immune response to HLA alloantigens and allograft function in children receiving cryopreserved valved allografts. Methods. We randomized 13 children to receive azathioprine or not to receive azathioprine (controls) after receiving a cryopreserved valved allograft. Azathioprine patients received intraoperatively 4 mg/kg of azathioprine and mg/kg once daily for 3 months after operation. Panel reactive antibodies against HLA class I and class II alloantigens were measured before, 1 month, and 3 months after operation. Results. Panel reactive antibodies were not significantly different between the azathioprine and control groups before (0.0% 0% versus 1.6% 1%), 1 month (59% 17% versus 71% 12%), or 3 months (84% 15% versus 96% 1.3%) after operation. There were no differences in degree of allograft valve stenosis between azathioprine ( mm Hg, months postoperatively) and control groups ( mm Hg, months postoperatively) or allograft valve insufficiency. Conclusions. Azathioprine does not significantly decrease the immune response to HLA alloantigens or affect the function of cryopreserved valved allografts used in children to repair congenital heart defects. (Ann Thorac Surg 2001;71:43 8) 2001 by The Society of Thoracic Surgeons Accepted for publication July 10, Address reprint requests to Dr Shaddy, Cardiology, Primary Children s Medical Center, Suite 1500, 100 North Medical Dr, Salt Lake City, Utah 84113; pcrshadd@ihc.com. Valved allografts are used routinely in the repair of congenital heart defects in children. We have previously shown that implantation of a cryopreserved valved allograft, but not open heart procedure alone, induces a significant HLA antibody response in children who receive valved allografts at the time of surgical repair for congenital heart defects [1]. Other investigators have demonstrated HLA class II alloantibodies [2] and HLA antibodies specific for valved allografts in adults and children receiving valved allografts [3]. Histologic examination in rats suggests that the immune response to valved allografts may be detrimental to valve function [4, 5]. Because of this concern of alloimmune-mediated allograft damage, animal studies have explored the role of immunouppression in altering this immune response. These studies have suggested that immunosuppressive therapy may reduce the alloimmune response to valved allograft implantation in rats and dogs [6 9]. However, data on the use of immunosuppressive therapy in humans receiving valved allografts is lacking. Previous studies have demonstrated that azathioprine suppresses the HLA alloantibody response to donor-specific blood transfusions in renal allograft recipients [10]. The purpose of this study was to prospectively study the effects of azathioprine on the humoral immune response and allograft function in children receiving valved allografts at the time of operation for repair of a congenital heart defect. Material and Methods Study Patients Between July 1996 and May 1999, we prospectively randomized 20 children to receive either azathioprine or no azathioprine (controls) after receiving a cryopreserved valved allograft in the pulmonary position as part of reparative operation for a congenital heart defect. The study was approved by the Institutional Review Boards at the University of Utah Health Sciences Center (July 1996) and Primary Children s Medical Center (August 1996). Informed consent was obtained from the parent or guardian of each patient, and informed assent was obtained from children 7 years of age or older. All procedures followed were in accordance with institutional guidelines. All patients who were scheduled to receive a cryopreserved valved allograft at the time of surgical repair of a congenital heart defect were approached for enrollment. A total of 32 patients were approached for enrollment. Twelve patients refused enrollment for a variety of reasons, including the parents or the patient just did not want to be involved in the study (n 6), concern about taking azathioprine (n 2), and concern about blood draws, taking daily medications, and infection in 1 patient each. One family refused to enroll their 2001 by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (00)

2 44 SHADDY ET AL Ann Thorac Surg AZATHIOPRINE AND VALVED ALLOGRAFTS IN CHILDREN 2001;71:43 8 infant because they were only interested in receiving azathioprine. Twenty patients were enrolled in the study. Patient diagnoses were tetralogy of Fallot (n 9), transposition of the great arteries (n 4), truncus arteriosus (n 4), and aortic valve disease (n 3). Eleven patients had undergone previous surgical repair or palliation of a congenital heart defect, but none had previously received allograft material as part of that repair. Seven patients who were initially enrolled withdrew before completion of the study. Four patients in the azathioprine group withdrew: 1 patient with tetralogy of Fallot and atrioventricular septal defect died 12 hours postoperatively because of low cardiac output; 1 patient withdrew after the 1 week blood draw because he did not want any more blood draws; 1 patient was withdrawn by the attending physician because of fever and grampositive bacteremia that was successfully and uneventfully treated with antibiotics; 1 patient was withdrawn by the attending physician because of diarrhea and Clostridium difficile colitis. Three control patients withdrew: 1 infant with critical aortic stenosis died 5 weeks postoperatively after a successful pulmonary autograft procedure, but with persistent endocardial fibroelastosis and low cardiac output; 1 patient withdrew after the 1-week blood draw because he did not want any more blood draws; 1 patient was withdrawn because of fungal sepsis and mediastinitis. Thus, 13 patients completed the study. At the time of enrollment, patients were randomized to receive either once daily azathioprine in addition to their usual care (azathioprine group) or no additional treatment (control group) for 3 months after valved allograft implantation. A member of the research team randomized patients by drawing a blank envelope from a box with either azathioprine or control written on a paper in the envelope. All of the patient s healthcare providers in addition to the patient and the patient s family knew into which group the patient was randomized. The following preoperative blood studies were obtained on all patients: cell blood count, liver function tests (alanine aminotransferase, -glutamyl transferase, aspartate aminotransferase, protein, albumin, bilirubin), HLA type, and panel reactive antibody (PRA). Those who were randomized to receive azathioprine were administered intravenously 4 mg/kg of azathioprine in the operating room within 1 hour before operation. All patients received irradiated and leukocyte-filtered blood products to prevent sensitization to allogeneic blood cells. Blood products were filtered with Purecell leukocyte reduction filters (Pall Biomedical Products Co, East Hills, NY) and irradiated with Cs-137 at 30 Gy. Study Design and Measurements Starting on the day after operation, each patient received a single daily dose of azathioprine, 1 to 2 mg/kg, intravenously (if unable to take medications by mouth) or orally (when able to take medications by mouth) to maintain a white blood cell count (WBC) of 4,000/mm 2 to 7,000/mm 2. If the WBC decreased to less than 4,000, then the dose of azathioprine was either decreased or temporarily discontinued until the WBC was more than 4,000/ mm 2. If the WBC increased significantly above 7,000/ mm 2, the azathioprine dose was increased to a maximum of 2.5 mg/kg per day to attempt to maintain the WBC between 4,000/mm 2 and 7,000/mm 2. Panel reactive antibody, liver function tests, and blood cell counts were performed at the following times after operation: 1 week, 1 month, and 3 months. In those who received azathioprine, an additional blood draw was performed at 2 months postoperatively to monitor for adverse effects. Postoperatively, all patients were monitored for evidence of infection and were treated appropriately if concerns of infection arose. At the time of routine blood draws for the study, parents and patients were asked about any adverse effects or infections since the previous blood draw. Echocardiograms were obtained using standard views in all patients at the discretion of each patient s cardiologist. HLA-A, HLA-B, and HLA-C loci serotyping was performed on all patients using the standard complementdependent cytotoxicity (CDC) test and in-house serologic reagents. Panel Reactive Antibody Panel reactive antibody was measured in two ways: (1) using an antiglobulin cytotoxicity technique (AHG-CDC) against an HLA-select frozen T-lymphocyte panel and (2) by flow cytometry using a pool of HLA class I and class II purified antigens coupled to latex beads. The AHG-CDC technique increases the sensitivity of complementdependent cytotoxicity by incorporating an antihuman light-chain immunoglobulin reagent. We used a frozen T-lymphocyte panel composed of 40 individuals of diverse HLA type and racial background [11]. Panel reactive antibody is expressed as the percentage of lymphocyte panel members against which each patient s serum reacts and therefore, reflects the breadth of allosensitization against the potential donor population. Using flow cytometry, HLA-A, HLA-B, and HLA-C (class I) and HLA DR/DQ (class II) antibodies were determined. This technique uses affinity-purified, soluble class I and class II antigens from 30 different cell lines that are coupled individually to uniform latex beads and then pooled together to create a panel that represents the majority of serologically recognized HLA class I and class II alloantigens (Flow-PRA I and II Beads, respectively; One Lambda, Canoga Park, CA). After incubation of the beads with 0.02 ml of the patient s serum, the beads were washed and stained with saturating goat antihuman IgG conjugated with fluorescein isothiocyanate. The percent fluorescent-positive beads (% PRA) was calculated by analysis on a Becton Dickinson FACScan flow cytometer (Becton Dickinson, Fullerton, CA) [12]. Statistical Analysis Data are expressed as mean SD (parametric data) or as median and range (nonparametric data). Comparisons between groups were made using either an unpaired t test (parametric data) or Mann-Whitney rank sum test (nonparametric data).

3 Ann Thorac Surg SHADDY ET AL 2001;71:43 8 AZATHIOPRINE AND VALVED ALLOGRAFTS IN CHILDREN 45 Table 1. Clinical Characteristics of Patients Characteristics Azathioprine Group Control Group No. of patients 6 7 Age (y) (median, range) 1.1 ( ) 3.2 ( ) Weight (kg) a Allograft Position Pulmonary 6 7 Type Pulmonary 5 5 Aortic 1 2 Diameter (mm) Length (cm) Cryopreservation time (mo) Donor age (y) patients in both groups demonstrated a broad, multispecific antibody pattern. Thus, azathioprine did not alter the humoral immune response to valved allografts. Echocardiography Echocardiograms obtained in patients in the azathioprine group at months after operation were compared to echocardiograms obtained in patients in the control group at months after operation. When comparing patients who received azathioprine to control patients, echocardiograms demonstrated no significant difference in the Doppler gradient across the conduit from the right ventricle to the pulmonary artery ( mm Hg versus mm Hg) (Table 2). Similarly, there were no significant differences in the degree of a Mean standard deviation. p value is not significant. Results Clinical Characteristics Thirteen patients completed the study, 6 in the azathioprine group and 7 in the control group. There were no significant differences between groups with regard to patient age or weight, or with regard to allograft type, size, or cryopreservation time, or donor age (Table 1). Eight of the children in the study were less than 2 years old. ABO blood type was available in 5 allografts used in the azathioprine group and 3 allografts used in the control group. Purely by chance, in the 5 allografts used in the azathioprine group for which ABO blood type was available, allograft blood types were identical to patient blood types in those who received the allografts. In the 3 allografts used in the control group for which ABO blood type was available, 2 allograft blood types were identical and 1 was incompatible with patient blood type. HLA typing was not available in any of the allografts. Patients who were randomized to the azathioprine group received mg/kg per day at the time of discharge from the hospital after operation. HLA Antibody There were no significant differences in the measured class I PRA between those who received azathioprine and control patients at any point during the study: before operation, and 1 week, 1 month, and 3 months after operation (Fig 1A). The two methods of measuring PRA class I, the cytotoxicity technique and the flow cytometry technique, demonstrated nearly identical results between the two methods (Fig 1B). Similarly, there were no significant differences in the class II PRA measured by flow cytometry between patients who received azathioprine and control patients at any point during the study (Fig 1C). Because of the broad panel reactivity, it was only possible to discern antibody specificity in 2 patients who showed the following specificity: B17, B5 cross-reactive group, A2, and A26, B15, B62, respectively. All other Fig 1. Humoral immune response to surgically implanted valved allografts measured before operation (Pre), and 1 week, 1 month, and 3 months after operation (Post) in patients who received azathioprine (AZA) and control patients. Humoral immune response was measured as panel reactive antibody (PRA) against (A) HLA class I alloantigens using the AHG-CDC technique (AHG-CDC PRA); (B) HLA class I alloantigens using flow cytometry (Flow I PRA); and (C) HLA class II alloantigens using flow cytometry (Flow II PRA).

4 46 SHADDY ET AL Ann Thorac Surg AZATHIOPRINE AND VALVED ALLOGRAFTS IN CHILDREN 2001;71:43 8 allograft valve insufficiency between patients who received azathioprine (3 moderate, 2 mild, 1 trace) and controls (3 moderate, 1 mild, 3 trace). One patient in the azathioprine group required replacement of his valved allograft 9 months after initial valved allograft because of significant conduit stenosis and calcification unresponsive to balloon valvuloplasty. Azathioprine Azathioprine was well tolerated. Three patients required reduction in azathioprine dosage because of mild neutropenia in 1 patient and mildly elevated transaminases in 2 patients, both of which resolved with reduction in dosage. There were no significant differences between azathioprine and control groups with regard to infection. In patients randomized to azathioprine, 1 patient had gram-positive bacteremia that required discontinuation of azathioprine and antibiotic treatment; 1 patient developed oral candidiasis and 1 patient had a streptococcal pharyngitis, both easily and successfully treated. In patients randomized to the control group, 1 patient developed fungal sepsis and mediastinitis, and 1 patient had an episode of otitis media requiring oral antibiotics. Comment This study demonstrates that azathioprine does not reduce the alloimmune HLA antibody response to valved allograft implantation in children, nor does it decrease the occurrence of allograft dysfunction. Although there was no beneficial effects from azathioprine, there also did not appear to be any significant increase in adverse effects from the use of azathioprine. The number of patients who did not complete the study was comparable between groups, as were the number of infections and deaths. It is possible that other, more potent immunosuppressants may more effectively reduce the humoral immune response to valved allografts and thus improve allograft function in these patients. However, more potent immunosuppressants have more adverse effects. Although experimental studies in animals have suggested that cyclosporine may reduce the T-cell-specific immune response to valved allografts [6 8, 9], we elected not to use cyclosporine because of its significant adverse effect profile, including susceptibility to infection, renal dysfunction, hirsuitism, and gingival hyperplasia. Instead, we elected to examine the effects of azathioprine because of its previous demonstrated efficacy in reducing allosensitization to donor-specific blood transfusions in renal transplant patients [10], the extensive decades-long experience in the prevention of rejection in solid organ transplantation in children, and its low adverse effect profile and ease of administration. We are confident that the patients received a therapeutic dose of azathioprine in this study for the following reasons. Although it is not possible to measure serum azathioprine levels, we used a dosage that has been widely used and accepted in pediatric solid organ transplantation. Furthermore, 3 patients required reduction in their azathioprine dosing during the study because of neutropenia (1 patient) and mildly Table 2. Echocardiographic Assessment of Stenosis and Insufficiency of Valved Allografts in Azathioprine Patients and Control Patients Patient Months After Operation Pulmonary Stenosis (mm Hg) Pulmonary Insufficiency Azathioprine Patients Moderate Mild Moderate Mild Trace Moderate Mean SD Control Patients Trace Mild Moderate Moderate Moderate Trace Trace Mean SD SD standard deviation. elevated serum transaminases (2 patients) that normalized after reduction in azathioprine dose. This suggests that the patients had adequate treatment with azathioprine. Thus, although reasonable doses of azathioprine were used and were well tolerated by this patient group, neither the humoral immune response to nor the function of valved allografts was significantly improved. This study confirms our previous finding that antibodies to HLA class I alloantigens are prevalent in children receiving cryopreserved valved allografts [1]. By using two independent methods for measuring HLA class I antibody, we demonstrated nearly identical class I PRA responses in both groups of patients. Furthermore, we also demonstrated the presence of HLA class II antibodies in these children using flow cytometry. Importantly, the relationship of the HLA antibody response to the histologic response of the valved allograft is unclear. Whereas some researchers have found no histologic evidence of immunologic injury to explanted valved allografts [13], other investigators have found evidence of a significant immunologic infiltration in explanted valved allografts from children [14]. Furthermore, the functional significance of any immunologic response to the allograft is even less clear. For example, Smith and colleagues [15] found no association between HLA mismatch and longterm valve function of homovital allografts. In older children, the freedom from reoperation in patients receiving valved allografts approaches 90% at 5 years in those who undergo pulmonary autograft for aortic valve disease [16 18]. However, valved allograft dysfunction continues to be a significant problem in pediatric heart surgery, particularly in small children [19, 20]. Thus, it is possible, although still speculative, that the HLA anti-

5 Ann Thorac Surg SHADDY ET AL 2001;71:43 8 AZATHIOPRINE AND VALVED ALLOGRAFTS IN CHILDREN 47 body response to valved allografts may lead to accelerated allograft dysfunction, particularly in small children. Because the majority of the children in the current study were in the high-risk age group for early allograft failure (less than 2 years of age), we can conclude that immunosuppressive therapy with azathioprine has no effect on humoral immune response to or function of valved allografts in these younger children. Continued investigation into methods of optimizing allograft function in children is warranted. To totally remove the immune response to valved allografts, it would probably require significantly stronger immunosuppressive therapy or ABO and HLA matching of allografts to recipients. Although it may be ideal to match patients and allografts with regard to ABO blood type and HLA type, it is currently impractical to do this because of the limited availability of valved allografts. The effect of ABO incompatibility on valved allograft function is unknown. Retrospective analyses of ABO incompatibility have shown this to be a risk factor for graft failure in some studies [20], but not in others [18]. In the current study, the majority of patients received an ABO-compatible valved allograft, thus lessening the impact of ABO incompatibility. Where possible, matching of ABO blood type may be worthwhile because of the presence of preexisting humoral immunity against incompatible blood group A and B antigens. It is much less clear whether attempts at more potent immunosuppressive therapy in this setting is warranted. Thus, immunosuppressive therapy with azathioprine during the first 3 months after valved allograft implantation has no effect on the humoral immune response or function of the allograft valve in the short term after implantation. This investigation was supported by Public Health Service research grant No. M01-RR00064 from the National Center for Research Resources. References 1. Shaddy RE, Hunter DD, Osborn KA, et al. Prospective analysis of HLA immunogenicity of cryopreserved valved allografts used in pediatric heart surgery. Circulation 1996; 94: Hoekstra FM, Witvliet M, Knoop CY, et al. Immunogenic human leukocyte antigen class II antigens on human cardiac valves induce specific alloantibodies. Ann Thorac Surg 1998; 66: den Hamer I, Hepkema B, Prop J, Elzenga N, Ebels T. HLA antibodies specific for cryopreserved heart valve homografts in children. J Thorac Cardiovasc Surg 1997;113: Motomura N, Imakita M, Yutani C, Kitoh Y, Kawashima Y, Oka T. Histological change in cryopreserved rat aortic allograft. J Cardiovasc Surg (Torino) 1995;36: Moustapha A, Ross DB, Bittira B, et al. Aortic valve grafts in the rat: evidence for rejection. J Thorac Cardiovasc Surg 1997;114: Schmitz-Rixen T, Megerman J, Colvin RB, Williams AM, Abbott WM. Immunosuppressive treatment of aortic allografts. J Vasc Surg 1988;7: Wagner E, Roy R, Marois Y, Douville Y, Guidoin R. Fresh venous allografts in peripheral arterial reconstruction in dogs. Effects of histocompatibility and of short-term immunosuppression with cyclosporine A and mycophenolate mofetil. J Thorac Cardiovasc Surg 1995;110: Vischjager M, Van Gulik TM, De Kleine RH, et al. Experimental arterial allografting under low and therapeutic dosages of cyclosporine for immunosuppression. Transplantation 1996;61: Yankah AC, Wottge HU, Muller-Ruchholtz W. Short-course cyclosporin A therapy for definite allograft valve survival immunosuppression in allograft valve operations. Ann Thorac Surg 1995;60:S Anderson CB, Tyler JD, Sicard GA, Anderman CK, Rodey GE, Etheredge EE. Pretreatment of renal allograft recipients with immunosuppression and donor-specific blood. Transplantation 1984;38: Fuller TC, Phelan D, Gebel HM, Rodey GE. Antigenic specificity of antibody reactive in the antiglobulinaugmented lymphocytotoxicity test. Transplantation 1982;34: Pei R, Wang G, Tarsitani C, et al. Simultaneous HLA Class I, and Class II antibodies screening with flow cytometry. Hum Immunol 1998;59: Mitchell RN, Jonas RA, Schoen FJ. Structure function correlations in cryopreserved allograft cardiac valves. Ann Thorac Surg 1995;60:S Vogt PR, Stallmach T, Niederhauser U, et al. Explanted cryopreserved allografts: a morphological and immunohistochemical comparison between arterial allografts and allograft heart valves from infants and adults. Eur J Cardiothorac Surg 1999;15: Smith JD, Hornick PI, Rasmi N, Rose ML, Yacoub MH. Effect of HLA mismatching and antibody status on homovital aortic valve homograft performance. Ann Thorac Surg 1998; 66:S Elkins RC. Pulmonary autograft: expanding indications and increasing utilizations [editorial]. J Heart Valve Dis 1994;3: Niwaya K, Knott-Craig CJ, Lane MM, Chandrasekaren K, Overholt ED, Elkins RC. Cryopreserved homograft valves in the pulmonary position: risk analysis for intermediate-term failure. J Thorac Cardiovasc Surg 1999;117: Shaddy RE, Tani LY, Sturtevant JE, Lambert LM, McGough EC. Effects of homograft blood type and anatomic type on stenosis, regurgitation and calcium in homografts in the pulmonary position. Am J Cardiol 1992;70: Clarke DR, Campbell DN, Hayward AR, Bishop DA. Degeneration of aortic valve allografts in young recipients. J Thorac Cardiovasc Surg 1993;105: Yankah AC, Alexi-Meskhishvili V, Weng Y, Schorn K, Lange PE, Hetzer R. Accelerated degeneration of allografts in the first two years of life. Ann Thorac Surg 1995;60:S71 6. INVITED COMMENTARY This study by Shaddy and colleagues from Primary Children s Medical Center in Salt Lake City is extraordinarily interesting and despite the low number of patients involved is a significant contribution to the evolving literature on immunology of cryopreserved homograft valve conduit transplants. The first point of interest is 2001 by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (00)