The utility of the novel optimized HLH inflammatory (OHI) index for predicting the risk for mortality and causes of death in lymphoma

7570 Background: Hemophagocytic lymphohistiocytosis (HLH) is a life-threatening inflammatory syndrome that may complicate hematologic malignancies (HM). We recently developed a simplified diagnostic and prognostic index termed the ‘optimized HLH inflammatory’ (OHI) index comprising the combined elevation of sCD25 ( > 3,900 U/mL) and serum ferritin ( > 1,000 ng/mL), which in HM patients both identifies HLH and predicts mortality more accurately than conventional criteria for HLH. In this study, we examined whether mortality in our cohort is directly related to progressive malignancy vs. HLH-associated causes in OHI+ and OHI- patients. Methods: We performed a multicenter, retrospective study of patients with newly diagnosed lymphoma from Israel, the USA, and Japan for whom sCD25 and ferritin levels were measured either as routine surveillance or during investigation for HLH and classified patients by their OHI status. The International Prognostic Index, International Prognostic Score, and Follicular Lymphoma International Prognostic Index were used to estimate the predicted prognosis of T/B cell non-Hodgkin’s lymphoma (NHL), Hodgkin’s lymphoma, and follicular lymphoma, respectively. Predicted five-year overall survival was calculated based on the relevant prognostic index and was compared between OHI+ and OHI- patients using the unpaired t-test. The actual survival at five years/last follow-up was recorded, as was the cause of death. The odds ratios (ORs) for observed vs. predicted mortality, and for HLH- vs. malignancy-related death were calculated using the Chi-square test. Results: 100 lymphoma patients were studied: 65% with B cell NHL, 18% with natural killer/ T cell lymphoma, 17% with Hodgkin’s lymphoma; 37 were OHI+, and 63 were OHI-. The disease-relevant international prognostic index-predicted five-year survival did not differ between OHI + and OHI- patients (a mean of 58% n OHI+ and 57% in OHI- p = 0.62). However, the observed five-year survival in OHI+ patients was lower (12%) than predicted, reflecting a mortality incidence that was four times higher than predicted by the relevant prognostic score (OR 3.9; CI 1.3-12.1). By contrast, OHI- patients had better survival (79%) than predicted by their prognostic scores (OR 0.15; CI 0.07-0.34). More than half of the OHI+ patients died from non-malignant causes (39% multi-organ dysfunction or HLH, 18% infection), while most OHI- patients (92%) died from progressive malignancy. The likelihood of dying from multi-organ dysfunction or HLH was 26 times higher in OHI+ vs. OHI- patients (OR 26.2; CI 4.1-286.7). Conclusions: OHI index status strongly correlated with mortality in patients with lymphoma within our cohort, and death in OHI+ patients was largely due to causes other than progressive malignancy. The OHI index appears to identify a harmful inflammatory state and deserves further prospective study.