Validation of NIH consensus criteria for diagnosis and severity-grading of chronic graft-versus-host disease.

To validate the National Institutes of Health (NIH) consensus criteria for chronic GVHD, we retrospectively reviewed 143 patients who developed GVHD later than 100 days after allogeneic hematopoietic stem cell transplantation. Their GVHD was reclassified and the severity was graded according to the criteria. Only four patients (2.8 %) could not be reclassified into any type of GVHD. In the remaining 139 patients, reclassified subtypes were late acute GVHD in 52 patients (37.4 %), classic chronic GVHD in 33 (23.7 %), and overlap syndrome in 54 (38.8 %). Of 87 patients with classic chronic GVHD or overlap syndrome, the severity was graded as mild in 21 patients (24 %), moderate in 53 (61 %), and severe in 13 (15 %). The proportions of moderate (70 %) and severe (20 %) disease were significantly higher in patients with overlap syndrome than those with classic chronic GVHD (46 and 6 %, respectively; P < 0.001). Univariate and multivariate analyses of subtypes and severity did not identify any significant prognostic values in any of the transplant outcomes, such as transplant-related mortality, overall survival, GVHD-specific survival, or discontinuation of systemic immunosuppressants. These findings suggest that the NIH consensus criteria are useful for classification of chronic GVHD, but have limited significance in predicting clinical outcomes. The validity of these criteria remains inconclusive, and future prospective studies will be required to refine them.