It has been shown that the effect of human leukocyte antigen (HLA) matching is potentiated in patients with highly reactive preformed lymphocytotoxic antibodies, whereas it is less pronounced in nonsensitized patients (1). We demonstrated recently that patients with a high pretransplant serum soluble CD30 (sCD30) have an impaired graft outcome, and that the sCD30 effect is independent of pretransplant lymphocytotoxic antibody status (2,3). In an analysis of recipients without lymphocytotoxic panel reactive antibodies (PRAs), we made the surprising observation that HLA compatibility has a strong effect in nonsensitized recipients with a high pretransplant sCD30.
Pretransplant sera of 3,980 nonsensitized first cadaveric kidney recipients (<10% PRA) who underwent transplantation between 1985 and 2002 were provided by 33 centers in 17 countries (see Appendix). Consecutivity was not a criterion because the centers were asked to send sera for all patients who underwent transplantation in whom a sufficient volume of the last serum obtained before transplantation was available. The sera were shipped frozen to the study center in Heidelberg and tested for serum sCD30 content using a commercially available enzyme-linked immunosorbent assay (ELISA) kit (Bender MedSystems, Vienna, Austria). For analysis, an sCD30 cut-off level of 100 U/mL was chosen on the basis of previous experience (3). In a subset of 2,628 patients, we obtained information on ELISA-detected immunoglobulin G anti-HLA antibodies using the Collaborative Transplant Study Anti-HLA Class I ELISA-Kit and the Class II-ID ELISA-Kit of GTI (Brookfield, WI) (4). An OD of 100 or more was considered positive for anti-HLA class I and 300 or more for anti-HLA class II.
The results were entered into the Collaborative Transplant Study database (5) and connected with previously registered information on the transplants in a blinded manner. Ninety-one percent of the patients studied received cyclosporine-based immunosuppression, 64% were male, 6% were aged less than 20 years, and 12% were aged more than 60 years. HLA typings and PRA determinations were performed at the tissue typing laboratories of participating centers. Information on graft function and patient survival was documented at 3 and 6 months, and at 1, 2, 3, 4, and 5 years. Actuarial survival rates were computed according to the Kaplan-Meier method (6) and expressed as percent±standard error. Log-rank test and weighted regression (7) were used for statistical analysis.
Although the influence of HLA compatibility on graft outcome was rather unimpressive in nonsensitized recipients with a low sCD30 of less than 100 U/mL (regression at 3 years:P =0.0095; at 5 years:P =0.1033), the effect was strong in nonsensitized patients with a high pretransplant sCD30 of 100 or more (at 3 years:P <0.0001; at 5 years:P =0.0001) (Fig. 1).
To further ascertain the absence of clinically relevant preformed antibodies in the nonsensitized group, we performed a separate analysis in the subset of patients who tested negative (in the ELISA) for antibodies against HLA classes I and II. These patients were previously found to have a particularly good graft survival rate (4). Because of the smaller patient numbers, this analysis was performed for transplants with three or less or more than three HLA-A+B+DR mismatches (logical cut-off because of the results shown in Fig. 1). In patients with a high pretransplant sCD30, the survival difference between relatively well- or poorly matched grafts was strong (71%±2% vs. 52%±4% at 5 years, log-rank P <0.0001), whereas it was weak in patients with an sCD30 of less than 100 (P =0.0776) (Fig. 2).
This study demonstrates an unexpectedly strong HLA effect in nonsensitized recipients with high pretransplant serum sCD30 content. Similar to presensitized patients with preformed PRA, these patients benefit greatly from HLA well-matched kidneys. As a consequence of this finding, we propose that patients on the transplant waiting list should be tested for serum sCD30, and that HLA matching algorithms should be adjusted. Because patients with a high pretransplant sCD30 benefit from a well-matched transplant even in the absence of PRA sensitization, special provisions should be made to allocate HLA well-matched kidneys to these patients.
We thank Tina Nonn, Martina Stolp, Nicole Schach, and Olga Wink for excellent technical assistance; Andrea Ruhenstroth, Bernd Döhler, and Gunter Laux for assistance for computer analysis; and staff members at the participating laboratories and clinical units for supplying us with sera and clinical follow-up data.
Sera for this multicenter study were provided by transplantation centers in the following cities: Barcelona, Spain (Dr. Jaume Martorell, n=41); Berlin, Germany (Dr. Constanze Schönemann, n=272); Budapest, Hungary (Dr. Agnes Padanyi, n=80); Cape Town, South Africa (Dr. Ernette Du Toit, n=227); Cardiff, United Kingdom (Dr. Tracy Rees, n=116); Dallas, Texas (Dr. Peter Stastny, n=264); Essen, Germany (Dr. Uwe Vögeler, n=78); Freiburg, Germany (Dr. Helmut Lang, n=8); Geneva, Switzerland (Dr. Michel Jeannet, n=73); Giessen, Germany (Dr. Rolf Weimer, n=53); Glasgow, United Kingdom (Dr. Alex M. Farrell, n=277); Goteborg, Sweden (Dr. Lennart Rydberg, n=50); Heidelberg, Germany (Dr. Manfred Wiesel, n=450); Helsinki, Finland (Dr. Saija Koskimies, n=614); Lausanne, Switzerland (Dr. Vincent Aubert, n=44); Leicester, United Kingdom (Dr. Terry Horsburgh, n=67); Lexington, Kentucky (Dr. John S. Thompson, n=57); Liege, Belgium (Dr. Claire Bouillenne, n=58); Ljubljana, Slovenia (Dr. Mateja Bohinjec, n=38); Mannheim, Germany (Dr. Peter Schnülle, n=70); Marburg, Germany (Dr. Harald Lange, n=44); Munich, Germany (Dr. Siegfried Scholz, n=10); New York, New York (Dr. Marilena Fotino, n=19); Phoenix, Arizona (Dr. Theona M. Vyvial, n=73); Portland, Oregon (Dr. Douglas J. Norman, n=135); Prague, Czech Republic (Dr. Eva Ivaskova, n=82); Quebec, Canada (Dr. Reynald Roy, n=315); Reims, France (Dr. Jacques H. Cohen, n=106); Rijeka, Croatia (Dr. Ksenija Vujaklija-Stipanovic, n=71); Rio de Janeiro, Brazil (Dr. Maria E. Goncalves De Freitas, n=4); Strasbourg, France (Dr. Marie-Marthe Tongio, n=52); Sydney, Australia (Dr. Trevor Doran, n=93); and Zagreb, Croatia (Dr. Adrija Kastelan, n=39).
1. Opelz G, Wujciak T, Döhler B, et al. HLA compatibility and organ transplant survival. Collaborative Transplant Study. Rev Immunogenet 1999; 1: 334.
2. Pelzl S, Opelz G, Wiesel M, et al. Soluble CD30 as a predictor of kidney graft outcome. Transplantation 2002; 73: 3.
3. Süsal C, Pelzl S, Döhler B, et al. Identification of highly responsive kidney transplant recipients using pretransplant soluble CD30. J Am Soc Nephrol 2002; 13: 1650.
4. Süsal C, Opelz G. Kidney graft failure and presensitization against HLA class I and class II antigens. Transplantation 2002; 73: 1269.
6. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457.
7. Dunn OJ, Clark VA. Applied statistics: analysis of variance and regression. 2nd ed. New York: John Wiley & Sons, 1987.