Chronic Gvhd Global Severity According to NIH Consensus Criteria: Results From the Chronic Gvhd Consortium

The classification of chronic graft-versus-host disease (cGVHD) proposed by the 2005 National Institutes of Health (NIH) Consensus Criteria established a new scoring system for individual organs (0-3) and a global severity scale (mild, moderate, and severe) according to the number and severity of organs involved. The Chronic GVHD Consortium was established to validate these criteria, since the scoring system was based on consensus opinion not yet supported by empiric data. This report includes results from the first 298 adult patients (pts) enrolled at 5 centers of this Consortium. Pts were assessed using standardized clinical data collection forms every 3–6 months (741 total visits). At the time of study enrollment, global cGVHD severity according to NIH consensus criteria was calculated from reported data as mild in 10% (n=31, no more than 2 organs with score 1), moderate in 59% (n=175, 3 or more organs with score 1, lung score 1 or any other organ score 2) and severe in 31% (n=92, lung score=2 or any organ score=3). Severity distribution was similar across incident (54%, cGVHD diagnosis within 3 months of enrollment) vs. prevalent (46%) cases (p=0.35), and between pts with overlap acute and chronic GVHD (40%) vs. classic chronic (60%) cGVHD (p=0.61). Skin, mouth, eye and liver were most commonly involved. (Figure) Score 3 skin (> 50% BSA, or deep sclerotic features, or impaired mobility) and score 2 lung involvement (FEV1 40–59% or lung function score 6–9 or shortness of breath after walking on flat ground) accounted for most of the cases in the severe category, while scores of 3 in the mouth, eye, GI tract, joints, genital tract (females only) occurred in only 3–11% of patients in this category. Higher global severity scores (n=741) were attributable to skin (32%), lung (21%), eye (18%), mouth (9%), liver (6%), joint (6%), GI tract (4%) and genital involvement (3%). Previously reported risk factors associated with onset of cGVHD or treatment-related mortality in pts with cGVHD were analyzed for their association with severity of cGVHD using generalized estimating equations (GEE) to adjust for baseline characteristics, repeated observations per pt, variable time to study enrollment and center effect (n=296 adults, 731 assessments). Global severity of cGVHD was not associated with age, gender-match, donor type, conditioning intensity, stem cell source, prior CMV infection, underlying disease, disease status, prior acute GVHD, time intervals from transplant to onset of cGVHD and enrollment, or platelet count at onset of cGVHD (< or ≥ 100 × 10⁹/L). Moderate or severe cGVHD at onset was associated with Karnofsky performance status ≤ 70 (odds ratio [OR] 2.41, 95% CI (1.5-4.0), p=0.0004). In conclusion, skin, lung and eye involvement accounted for 71% of the global severity score, although all organs contributed and should be scored. Conventional risk factors for development of cGVHD were not associated with NIH cGVHD global severity categories. Most pts fit into the moderate category, suggesting that additional refinements to the global severity scoring may be able to distinguish prognostically different subgroups within this category. Funded by NCI CA118953. Registered as NCT00637689.