Response of 864 Patients to Steroids as Primary Therapy for Acute Gvhd at a Single Institution: Day 28 Response Is the Best Endpoint to Predict Transplant-Related Mortality.

Abstract 1134

Poster Board I-156

The optimal primary endpoint for acute GVHD treatment trials has not been established. In a retrospective analysis, we examined the response of 864 patients who received prednisone 60 mg/m² for 14 days, followed by an 8 week taper, as initial therapy for acute GVHD from 1990-2007 at a single institution. Complete response (CR), partial response (PR), very good partial response (VGPR), and no response (NR) were scored at day 14, day 28 and day 56 after initiation of steroids. To determine the best endpoint, an index of concordance, the c-statistic (C), was performed to estimate the probability that patients with the better GVHD response will have lower transplant related mortality (TRM) than patients with a worse response. Median patient age was 32 (range, 0.2-69) years; 35% were <18 years old. Patients received grafts of HLA-matched sibling bone marrow (BM) or peripheral blood (PB, n=315), partially matched sibling BM or PB (n=24), unrelated donor (URD) BM or PB (n=313), single umbilical cord blood (sUCB, n=89) or double UCB (dUCB, n=123). Prior to initiation of steroid therapy, initial GVHD grades were grade I in 230 (27%), grade II in 504 (58%), grade III in 119 (14%), and grade IV in 11 (1%). Initial GVHD organ involvement was skin only (57%), gut only (17%), liver only (1%) or multiorgan (25%). Day 28 responses were similar to day 56 responses (p=0.14) and better than day 14 responses (p=0.03) in predicting TRM. In multiple regression analysis, patients with NR at day 28 were 2.77 times more likely to have TRM than patients with CR, VGPR or PR while any response favored lower TRM (table). Factors associated with significantly worse 2 year TRM in patients with acute GVHD include: NR to steroids, partially matched BM/PB, high risk disease, older age and skin only GVHD. Factors not associated with TRM include CMV serostatus, conditioning therapy, GVHD prophylaxis, days to steroid treatment, and initial grade of GVHD.

These data suggest that responses at day 28 or 56 are equally effective endpoints for acute GVHD trials. Day 14 responses cannot as accurately predict TRM. As patients with NR require further therapy in a timely manner, early progression or response by day 28 is the best endpoint to assess efficacy of initial therapy for acute GVHD. Prospective trials are still required to determine the best therapy for different subgroups of patients with acute GVHD, especially those identified to have predictably poor responses and high TRM.