Prognostic factors for survival of patients with newly diagnosed chronic GVHD according to NIH criteria.

Chronic graft versus host disease (cGvHD) is the most common cause of late morbidity and mortality after allogeneic hematopoietic stem cell transplantation (HSCT). We retrospectively evaluated the impact of NIH classification on outcome of patients at our center. Primary endpoint was overall survival at 5 years. Two hundred one patients with cGVHD according to NIH were included. Platelets <100,000/µl on day of diagnosis of cGvHD (HR 2.97, 95 % CI 1.7-5.3, p < 0.001), female donor (HR 1.78, 95 % CI 1.0-3.2, p = 0.05), and reduced intensity conditioning (HR 1.95, 95 % CI 1.0-3.8, p = 0.05) impacted overall survival. Non-relapse mortality (NRM) was higher for patients with low vs. high platelets: 26 % (95 % CI 14-40) vs. 6 % (95 % CI 2-10), p < 0.001, and tended to be higher for female vs. male donor: 14 % (95 % CI 7-23) vs. 7 % (95 % CI 3-13), p = 0.08. Relapse tended to be higher for recipients of reduced intensity conditioning (RIC) vs. myeloablative conditioning (MAC): 33 % (95 % CI 23-43) vs. 20 % (95 % CI 10-31), p = 0.06. After excluding patients with myeloma and lymphoma, IgG serum levels at diagnosis of cGvHD of 122 patients were correlated with survival. IgG levels above normal were associated with worse 2-year overall survival (OS), p = 0.04, compared to normal or low IgG levels. Platelet count at diagnosis remains the most valid prognostic factor for survival of patients with cGvHD even in the era of NIH grading. High IgG level at diagnosis of cGVHD represents a potential negative prognostic parameter that deserves further investigation.

Serum albumin level predicts survival of patients with gastrointestinal acute graft-versus-host disease after allogeneic stem cell transplantation.

In a retrospective single-centre study, we analysed the prognostic impact of factors identifiable at initial diagnosis of acute GVHD (aGVHD). We retrospectively analysed 495 adult patients of whom 308 (62 %) developed acute GVHD (I-IV) and were included in further analysis. Gut aGVHD was diagnosed in 163/308 cases (53 %). Conditioning was myeloablative conditioning (MAC) in 123 (39.9 %) and reduced intensity (RIC) in 185 (60.1 %) patients. Median serum albumin level at diagnosis of aGVHD was 34 g/l, which was used as cut-off for low vs. normal albumin levels. In patients with gut aGVHD, low albumin level at the time of diagnosis of aGVHD was associated with poorer overall survival (OS) which was 52 vs. 67 % at 1 year and 40 vs. 61 % at 3 years, p = 0.015. In patients with only skin aGVHD, 1- and 3-year OS of patients with low vs. normal albumin levels were 72 vs. 72 % and 59 vs. 57 %, respectively, p = 0.69. In multivariate analysis of patients with gut aGVHD, low serum albumin level ≤34 g/l (relative risk (RR) 2.13, p = 0.003), gut aGVHD grades 3-4 (RR 2.70, p = 0.001), RIC (RR 1.84, p = 0.024), matched unrelated donor (RR 1.86, p = 0.18) and mismatched unrelated donor (RR 2.76, p = 0.03) retained negative impact on OS. Subgroup analysis revealed that impact of albumin was restricted to patients with gut aGVHD after RIC. Low serum albumin levels are associated with poorer OS in patients with gut but not skin aGVHD after RIC but not MAC allogeneic stem cell transplantation.