FLOW CROSS-MATCHING IDENTIFIES PATIENTS AT RISK FOR POSTOPERATIVE ELABORATION OF CYTOTOXIC ANTIBODIES

Abstract

Background. 

Cytotoxic IgG against class I antigens can contribute to renal dysfunction or failure after transplantation. However, the clinical relevance of IgG measured by flow cytometric cross-matching is controversial. This study correlated pre- and postoperative flow reactivity with clinical outcome among renal transplant patients with negative preoperative cytotoxic cross-matches.

Methods. 

Nonsensitized primary renal allograft patients (n=157) with negative preoperative cytotoxic cross-matches (complement-dependent lymphocytotoxicity assays) were stratified on the basis of IgG reactivity measured by flow cytometric cross-matching (FCXM) as FCXM negative (Neg) or positive against class I (T-pos FCXM) or class II (B-pos FCXM) antigens. The groups were compared in terms of frequency of early rejection and 1-year graft survival.

Results. 

Patient distribution was 67% Neg, 14% T-pos FCXM, 14% B-pos FCXM, and 5% IgM FCXM. The incidence of early rejection was 25±3% for Neg and 51±3% for T- and B-pos FCXM (P<0.05). One-year graft survival for Neg versus T-pos and B-pos FCXM was 97±3% versus 44±10% (P<0.05) and 77±5%(P=0.06), respectively. Rejections requiring plasmapheresis were found only among patients with T-pos FCXM. Among 29 patients, FCXM and complement-dependent lymphocytotoxicity assays were performed 10±2 and 28±4 days after transplantation. Pre- and posttransplant antibody levels were relatively unchanged among Neg and B-pos FCXM patient groups. In contrast, patients with T-pos FCXM produced cytotoxic IgG against class I after transplantation, which may have contributed to the severe graft dysfunction experienced by this group.

Conclusions. 

FCXM is a useful tool to stratify primary renal transplant candidates in terms of potential risk for severe rejection. Furthermore, demonstration of preoperative flow reactivity against class I may identify a subgroup of patients at risk for early elaboration of cytotoxic alloantibody.

Antibodies directed against HLA antigens can be measured by complement-dependent lymphocytotoxicity (CDC*) assays or by exquisitely sensitive flow cytometric techniques (FCXM). It is clear that CDC due to IgG against class I antigens can contribute to the pathogenesis of early graft dysfunction following renal transplantation (RTX)^(1-3). However, the clinical relevance of IgG measured by FCXM is controversial ^(4-6). Certainly allo-IgG identified by FCXM is associated with high rates of early graft failure among regraft patients. On the other hand, the clinical significance of preoperative IgG measured by FCXM among primary RTX patients has been variable and has led to the concern that FCXM sensitivity may inappropriately exclude patients from transplant consideration. This report summarizes our experience with nonsensitized (current and peak panel-reactive antibody <10%) primary RTX recipients stratified on the basis of preoperative FCXM reactivity. FCXM reactions partitioned patients in terms of relative risk for early rejection and rejection severity. A positive FCXM was associated with a greater incidence of early rejection than a negative FCXM. However, the most severe rejections were experienced only among patients with FCXM reactivity against class I targets. Demonstration of cytotoxic IgG after RTX among this same group of patients may be causally related to rejection severity and clinical outcome.

The study group comprised 157 RTX patients who received primary cadaveric transplants between 1989 and 1996. The mean follow-up time was 4.9 years. Immunosuppression consisted of induction prophylaxis with 5 mg of OKT3 (Ortho Pharmaceutical Corp., Raritan, NJ), methylprednisolone, and azathioprine. Cyclosporine and prednisone were started after the serum creatinine level fell to 2.5 mg/dl. Rejection was confirmed histologically. All patients showed negative preoperative CDC against donor T and B splenocytes using the Amos 1 and 3 wash. The average panel-reactive antibody was 8±2%. FCXM was performed using a 1024-channel fluorescence-activated cell sorter (FACScan, Becton Dickinson, San Jose, CA); positivity against donor T and B cells was defined as channel shifts exceeding 40 ⁽⁷⁾.

Patients were stratified on the basis of preoperative FCXM reactivity as FCXM negative (Neg) or FCXM positive due to IgG against T and B cells (T-pos FCXM) or only B cells (B-pos and T-neg FCXM). The majority of patients (67%) were FCXM Neg. In contrast, 28% elaborated IgG and 5% elaborated IgM. Patients with IgM FCXM were excluded from further consideration. Among the IgG FCXM-positive patients, 14% were T-pos FCXM and 14% were B-pos FCXM.

The incidence of early rejection and graft failure differed among the three groups of patients. The incidence of rejection within 3 months of RTX was 25±3% for Neg versus 51±2% (P<0.05) for T-pos FCXM and 51±3% (P<0.05) for B-pos FCXM. Graft survival was strikingly different among the groups. The 1-year graft survival rate was 97±3% for patients with a negative FCXM. In contrast, the 1-year graft survival rate was 44±10% (P<0.05) for patients with T-pos FCXM and 77±5% (P=0.06) for patients with B-pos FCXM.

Preoperative T-pos FCXM was clearly associated with a greater risk for graft failure among this cohort of nonsensitized primary RTX patients. We speculated that if FCXM antibodies contributed to the pathology of graft dysfunction, then antibody levels might change quantitatively or qualitatively in the period immediately after RTX. Consequently, pre- and postoperative antibody content was evaluated among 29 consecutive patients (i.e., first 10 Neg, first 10 B-pos FCXM, and first 9 T-pos FCXM). Patient demographics were identical in retrospective and prospective studies. As before, patients were stratified by preoperative FCXM reactivity. After transplantation, FCXM and CDC using cryopreserved donor splenocytes were evaluated on days 10±2 and 28±4. Patients with preoperative Neg showed minimal changes in antibody levels after RTX (Table 1). Only three patients(3/10) had measurable IgG FCXM after RTX, but it was low titer (channel shifts of 47, 72, and 98) and exclusively directed against B-cell targets. Cytotoxic antibodies were not produced by any patient in this group. Among the B-pos FCXM group, pre- and post-RTX antibody levels were similar(Table 2). B-cell reactivity was unchanged in four patients and decreased in four patients. Reactivity increased in two patients for whom FCXM IgG against T cells became measurable. Cytotoxic antibody was not found in any patient in this group. In contrast, patients with preoperative T-pos FCXM showed strong FCXM reactivity against T and B targets before and after RTX (Table 3). Similar to all of our patients, this cohort was poorly matched for HLA antigens (six were six-antigen mismatched and three were one-antigen matched at the A or B locus). Postoperative T-pos FCXM reactivity was unchanged for three patients and decreased for one patient and became highly elevated among the remaining five patients. All five patients simultaneously elaborated cytotoxic IgG, which showed individual end-point titrations of 8, 8, 64, 64, and 200. Unlike the other patients in the study, the patients producing CDC also experienced steroid-resistant rejections and underwent therapeutic plasma exchange (one exchange per day for 7 days with albumin as replacement)⁽⁷⁾. Two patients sustained high levels of cytotoxic IgG and experienced graft failure within 30 days of RTX. In contrast, CDC was eliminated and FCXM reactivity returned to pretransplant levels among the other three patients who currently demonstrate functioning grafts.

Allograft integrity can be severely compromised in the presence of allospecific complement-fixing IgG. When preoperative IgG CDC is directed against class I antigens, the impact upon RTX survival is so profoundly deleterious that it generally contraindicates transplantation^{(1, 2)}. Ten Hoor et al. ⁽⁸⁾ showed anamnestic responses may be equally problematic to renal allograft survival when IgG CDC is detected in noncurrent sera. The critical feature of these studies is that elaboration of IgG CDC against class I antigens is pivotal to the pathology of vascular destruction in the allograft. In contrast, the relevance of IgG measured by FCXM is less clear. The ability to measure antigen-antibody binding by FCXM is a boon to sensitivity but a bane to interpretation. FCXM detects but does not discriminate between low levels of Ig antibodies with variable degrees of complement fixation. Thus, there is a real concern that use of FCXM positivity as a criterion for patient selection may inappropriately exclude patients from RTX consideration.

Our results support a role for FCXM as a means to stratify RTX candidates in terms of relative risk for early rejection, rejection severity, and/or graft failure. Similar to the findings of Karuppan et al.⁽⁵⁾ and Lazda ⁽⁹⁾, we found that patients with no evidence of sensitization (Neg) enjoyed the lowest rates of rejection and the highest 1-year graft survival. In contrast, the risk for early rejection was twofold greater among patients with IgG FCXM. Although the incidence of rejection was equal among patients with IgG FCXM against class I or II antigens, patients with B-pos FCXM experienced easily reversed rejections and did not manifest cytotoxic antibody. In contrast, the patients with IgG FCXM against class I antigens experienced severe rejections as well as the shortest allograft life expectancy. We believe that the rejection severity experienced by these patients is linked to their proclivity to produce cytotoxic IgG against class I in the early postoperative period, a phenomenon that did not occur in the other groups. Conversely, normalization of graft function was associated with CDC IgG depletion. We speculate that postoperative IgG CDC may be as destructive to allograft survival as is preoperative IgG CDC. However, removal of these antibodies by pharmacological or mechanical means may permit restoration of graft function, as happened with three of five patients in this study.

FCXM may be useful to stratify patients on a risk management basis and predict the likelihood of rejection and rejection severity. Demonstration of preoperative T-pos FCXM appears to identify a subgroup of patients who are particularly disposed to making cytotoxic IgG. Even among this high-risk group, it is clear that not all patients will experience this complication. However, identification of a high-risk group allows the clinician to avoid transplantation or undertake the procedure with the knowledge that more aggressive therapeutic management may be needed. In fact, we now use this information to select the intensity of immunosuppression during the immediate postoperative period.