Prognostic Factors Identifiable at the Time of Onset of Acute Graft-Versus-Host Disease after Allogeneic Hematopoietic Cell Transplantation.

Existing grading systems of acute GVHD, based on the overall degrees of skin, liver, and gut involvement, may not identify patients who are likely to have poor outcome at the time of onset of acute GVHD due to the subjective nature of determining the extent of skin involvement and wide variation of daily amount of diarrhea. This retrospective analysis was to identify prognostic factors for acute GVHD, with an emphasis given to the factors available at the time of onset. Among 303 adult patients who were given allogeneic HCT between Jan 1996 and Aug 2003, 83 patients (27.4%, 50 males and 33 females) developed ≥ grade II acute GVHD, received systemic immunosuppressive treatment, and were included in the analysis. The median age of the patients was 32 years (range, 15–59). The indications for HCT were acute leukemia in 34 patients, CML in 20, MDS in 13, SAA/PNH in 8, and others in 8. Initial treatment for acute GVHD was corticosteroids in 81 and mycophenolate mofetil in 2. Of those, 40 (48.2%) required additional immunosuppressive agents on median day 10 (range, 3–48) of onset of acute GVHD (time to initial treatment failure, TITF). Overall, 52 patients (62.7%) discontinued immunosuppressive treatment successfully except for the prophylactic dose of cyclosporine or tacrolimus on median day 44 (range, 7–800, time to completion of treatment, TCT). After median follow-up of surviving patients of 805 days (range, 302–2290), 42 patients died. Of 42 deaths, 27 died due to complications related uncontrolled acute or chronic GVHD within 1 year of onset of acute GVHD and these deaths were considered as GVHD-related death (GVHD-specific survival). Two died due to complications of chronic GVHD beyond 1 year and were not considered as an event related to acute GVHD. Thirteen died due to relapse of malignancie (6) or other causes (7). Multivariate analysis with variables of P < 0.05 on univariate analysis showed that GVHD prophylaxis (cyclosporine plus methotrexate vs. cyclosporine alone; P=0.005; odd ratio, 0.282; 95% CI, 0.117–0. 677), HLA match (full vs. 1 ag mismatch; P=0.023; odd ratio, 2.591, 95% CI, 1.142–5.878), lower G-I involvement (no vs. yes; P=0.000; odd ratio, 4.868; 95% CI, 2.501–9.475), and absolute lymphocyte count (> 100 vs. < 100/ul; P=0.003; odd ratio, 2.786; 95% CI, 1.418–5.473) were significant independent variables predicting TITF. HLA match (P=0.037; odd ratio, 0.218; 95% CI, 0.052–0.909), lower G-I involvement (P=0.020; odd ratio, 0.420; 95% CI, 0.202–0.874), and absolute lymphocyte count (P=0.040; odd ratio, 0.448; 95% CI, 0.208–0.963) were significant independent variables predicting TCT. For the prediction of GVHD-specific survival, type of cell donor (sibling vs. unrelated; P=0.042; odd ratio, 0.420; 95% CI, 0.182–0.969), HLA match (P=0.014; odd ratio, 0.286; 95% CI, 0.105–0.779), skin vs. visceral initiation (P=0.033; odd ratio, 0.287, 95% CI, 0.091–0.903), upper G-I involvement (P=0.015; odd ratio, 0.213; 95% CI, 0.061–0.739), and absolute lymphocyte count (P=0.003; odd ratio, 0.260; 95% CI, 0.106–0.641) were significant independent variables. Our study showed that, at the time of onset of acute GVHD, peripheral blood lymphocytopenia was an independent variable that predicted shorter TITF, longer TCT, and shorter GVHD-specific survival, in addition to visceral organ involvement. Early aggressive treatment for acute GVHD in this population of patients may improve overall outcome of acute GVHD after allogeneic HCT.