Prognostic Relevance of “Early-Onset” Graft-Versus-Host Disease Following Nonmyeloablative Hematopoietic Cell Transplantation.

We have shown that graft-versus-host disease (GVHD) after nonmyeloablative hematopoietic cell transplantation (HCT) occurs a median of 2 months later than after myeloablative HCT (Blood 102:756, 2003). Here, we asked whether there was an association between the time of onset of GVHD and survival after nonablative HCT. We retrospectively analyzed outcomes among 395 patients with hematologic diseases who had nonmyeloablative HCT from HLA-matched related (n=297) or unrelated donors (n=98). The nonablative regimen consisted of 2 Gy total body irradiation with or without fludarabine followed by postgrafting immunosuppression with mycophenolate mofetil and cyclosporine. The cumulative incidences of grades II-IV acute GVHD and extensive chronic GVHD were 45% and 47%, respectively, with grafts from related donors, and 68% and 68%, respectively, with those from unrelated donors. The median time to prednisone initiation, a marker for the onset of both acute and chronic GVHD, was 79 (range, 8–799) days among patients with sibling donors and 28 (range, 5–104) days among patients with unrelated donors. Among patients with GVHD following related grafts (n=187), the cumulative incidence of non-relapse mortality (NRM) at 4 years was 55% among the tertile of patients who initiated prednisone before day 50 (“early initiation”), and 29% when prednisone was initiated on or after day 50 (“late initiation”; p<0.001). After controlling for patient age, diagnosis, type of preparative regimen, number of previous chemotherapies and donor/recipient gender combinations in multivariate analysis, early initiation of prednisone was identified as an independent predictor of increased NRM (Table). The presence of comorbidid conditions at the time of transplant (Charlson Comorbidity Index >=1) was neither predictive for the time of onset nor the overall incidence of GVHD. Among patients who initiated prednisone for the treatment of GVHD, those with pretransplant comorbidities had a significantly greater hazard of NRM than those without comorbidities (hazard ratio, 2.6; 95% confidence interval, 1.2–5.4; p=0.01). There was no association between time to prednisone initiation and survival after nonmyeloablative HCT from unrelated donors. In conclusion, early-onset GVHD (prednisone initiation before day 50) was associated with increased NRM and poor survival after nonmyeloablative HCT from HLA-identical related donors and identified a subgroup of patients who might benefit from more aggressive primary therapy of GVHD.

Hazard of Non-Relapse Mortality According to Time to Prednisone Initiation (HLA-Matched Related Transplants)