303: Incidence and Outcome of Chronic Graft-Versus-host Disease (cgvhd) after Allogeneic Stem Cell Transplant (SCT) Using National Institute of Health (NIH) Consensus Criteria

M. Jagasia,W. Chinratanalab,J. Giglia,S. Dixon,H. Chen,H. Frangoul,B. Engelhardt,S. Goodman,J. Greer,A. Kassim,W. Morgan,K. Ruffner,F. Schuening
DOI: https://doi.org/10.1016/j.bbmt.2006.12.308
2007-01-01
Biology of Blood and Marrow Transplantation
Abstract:cGVHD, defined as GVHD after day 100, is common, with an impact on morbidity and survival. The limited/extensive cGVHD classification is not reproducible or prognostic for late non-relapse mortality (NRM). Recently the NIH consensus criteria were proposed, but the ability of these criteria to predict outcome of various types of cGVHD is unknown. Pts undergoing their 1st SCT from 1/01 to 12/03 were studied. 110 pts alive beyond day 100 met criteria for the study. GVHD after day 100 was classified using NIH criteria into: persistent acute GVHD (aGVHD) (assigned at day 100), recurrent aGVHD, delayed aGVHD, classic cGVHD, overlap GVHD (all assigned at time of onset). Severity scores were assigned to pts with classic and overlap GVHD at onset and clinical worsening. Overall survival (OS) both from time of transplant and time of GVHD onset were measured. 37 (34%) had no GVHD and 73 (66%) pts had GVHD. OS was 44% vs. 66% (no GVHD vs.GVHD, P=0.026). Of 73 pts with GVHD, 14 (19%) had limited and 59 (80%) had extensive cGVHD. Pts with limited GVHD were reclassified as persistent aGVHD (7%), recurrent aGVHD (29%), and classic CGVHD (64%). Pts with extensive cGVHD were reclassified as persistent (3%), delayed (3%), recurrent (31%), classic chronic (37%), and overlap GVHD (26%). 31 (42%) had no subsequent clinical worsening and 42 (58%) had subsequent clinical worsening of GVHD. 65% of pts with classic cGVHD (22/31) had worsening compared to other types (20/42, 47%) (P=0.046). Severity scores increased in 12/31 pts (39%) at time of subsequent clinical worsening. OS of pts with various types of GVHD were significantly different (P<0.0001). This was more apparent when pts with any acute features of GVHD were compared with classic cGVHD (3-yr OS 47% vs. 66%, P=0.0015). This effect persisted when survival was measured from onset of GVHD (P=0.0336). Severity at onset or clinical worsening in pts with classic or overlap GVHD did not impact survival. The 3-yr NRM (with relapse as a competing risk) for the cohort was 21% and was not affected by the presence or absence of GVHD, or subtypes of GVHD. Significant variables using Cox model with time dependent covariates were any aGVHD feature after day 100 (HHR 5.27, P=0.0004), and extensive cGVHD (HR 0.28, P=0.0041). The OS with different NIH subtypes after day 100 from SCT varies and is superior for pts with classic cGVHD. Global severity score, within the limits of our study had no prognostic value with respect to survival.
What problem does this paper attempt to address?