Pancreas transplantation in crossmatch-positive recipients

Background: Prolonged cold preservation time can unfavorably affect outcome in pancreas transplantation. To reduce this ischemic time, cadaver pancreas grafts, in selected cases, are sometimes transplanted before crossmatch results are known. We report our experience with pancreas transplants in recipients with either current or historically positive T- or B-cell crossmatches. Methods: Crossmatch-positive pancreas transplants were identified using a computerized database. T-cell crossmatches were performed using an antihuman-globulin-augmented complement-dependent cytotoxicity (CDC) test; B-cell crossmatches were performed using an extended incubation CDC test. All patients received anti-T-cell induction therapy and either cyclosporine (1987-1993) or tacrolimus-based (1994-2001) immunosuppression. More recent recipients (2000-2001) also received intravenous gamma globulin and postoperative plasmapheresis. Results: Between October 1, 1987 and March 31, 2001, of a total of 1076 pancreas transplants performed, 59 (5.48%) were crossmatch-positive. Of these, 8 had a current T-cell-positive crossmatch and 15 had a current B-cell-positive crossmatch. One recipient was both current B- and T-positive, and the rest were past B- and/or T-cell positive. One-year pancreas graft survival for current T- and B-cell crossmatch-positive transplants was 63% and 67%, respectively. T- or B-cell crossmatch-negative transplants had a 1-yr survival of 70%. In the T-cell crossmatch-positive group, four grafts are still functioning (follow-up range, 2-12 yr), one patient died with a functioning graft at 4 months, and four grafts failed (one each from pancreatitis, infection, primary nonfunction, and vascular thrombosis). No grafts were lost to rejection. In the B-cell crossmatch-positive group, six grafts are still functioning (follow-up range, 2-11 yr) and nine have failed (four from chronic rejection, three from vascular thromboses, and two from pancreatitis). Crossmatch-positive cases were significantly more likely to be retransplants (70.8%) than crossmatch-negative cases (14.8%, p < 0.0001). In a multivariate analysis, crossmatch positivity did not affect pancreas graft outcome, whereas retransplants had a significant impact on outcome (relative risk 1.84, p < 0.0001). Conclusions: (i) Pancreas transplants performed in the setting of a positive current crossmatch may have long-term function. (ii) With current immunosuppressive protocols, graft loss from hyperacute and acute rejection may be prevented in current crossmatch-positive pancreas transplants. Chronic rejection was only seen in B-cell crossmatch-positive cases. (iii) High rates of technical graft loss in crossmatch-positive cases may reflect a high frequency of retransplants in this group.