Treatments and Outcomes of Lymphoma-Associated Hyperinflammation- a Report from the HLH in Adults Working Group for the Histiocyte Society

Abstract

Purpose: Hemophagocytic lymphohistiocytosis (HLH) is a life-threatening inflammatory syndrome. Lymphoma is the most common cause of HLH in adults. Previous studies have indicated poor patient outcomes in lymphoma-associated HLH. The recently introduced optimized HLH inflammatory (OHI) index has improved prognostication and identification of the hyper-inflammatory phenotype of lymphoma; however, more comprehensive information on treatments and outcomes in a large cohort in the contemporary era is necessary. Our study aimed to investigate the clinical features, treatments, and outcomes of lymphoma-associated HLH through a multi-national effort by the HLH in Adults Working Group of the Histiocyte Society (NCT05898477). Methods: We conducted a multicenter, retrospective study evaluating patients diagnosed with lymphoma and hyperinflammation, defined by fulfilling 5/8 HLH diagnostic criteria or the OHI index from 2003-2022. Clinical characteristics were summarized using median and interquartile ranges (IQR). Survival analysis employed Kaplan-Meier curves and the log-rank test, considering HLH diagnosis as the zero time-point. Results: The analysis encompassed 181 patients with lymphoma: 55% had B-cell non-Hodgkin lymphoma (NHL), 35% had natural killer (NK) cell or T-cell lymphoma, and 10% had classical Hodgkin lymphoma. 35% of patients were from China, 21% from the USA, 17% from Israel, 16% from Europe, and 11% from Japan. Of the entire cohort, 157 (87%) patients met the HLH-2004 criteria, and 153 (83%) met the OHI index; The median age at HLH diagnosis was 60 years (IQR, 45-68), and 59% were males. 50% of patients presented with HLH before/at the time of lymphoma diagnosis (median time to lymphoma diagnosis 9 days), while 50% developed HLH after the initial lymphoma diagnosis (median time to HLH diagnosis 181 days). The median values of hemoglobin, absolute neutrophil count (ANC), and platelet count at HLH diagnosis were 8.6 g/dL, 2.1 K/microL, and 54.7 K/microL, respectively. The median values of triglycerides and fibrinogen were 241 mg/dL and 250 mg/dL, respectively. Interestingly, the median values for ANC, triglycerides, and fibrinogen fell outside the HLH-2004 diagnostic threshold. The median and interquartile ranges for ferritin and sCD25 (sIL-2r) were 3017 ng/mL (1337-8958) and 14098 U/mL (7894-39916), respectively. 48% of physicians altered their initial treatment approach due to the HLH diagnosis. Of these instances, 48% were treated with dexamethasone and etoposide outside of a formal treatment protocol, 11% with the HLH-94/2004 protocol, 24% with DEP regimen (liposomal doxorubicin, etoposide, and methylprednisolone), and 4% with DEP and ruxolitinib. Dexamethasone alone was used in 15% of the patients, while another 4% were treated with anakinra +/- steroids. After a median follow-up duration of 2.1y (range, 0.005-6.1), the median overall survival was 0.58 years. 46% of the patients died from multi-organ dysfunction/HLH, 17% from single-organ dysfunction, 16% from infection or septic shock, and 8% from bleeding. The risk of mortality among T-cell or NK cell lymphoma was higher than B-cell lymphoma (HR 1.7, 95% CI 1.16-2.61, Figure 1A). Finally, patients positive for the OHI index alone at diagnosis had a higher mortality risk (Figure 1B) compared to those positive for both the OHI index and HLH-2004 (HR 2.5; CI 1.26 - 4.8) or those positive for HLH-2004 alone (HR 3.72; CI- 1.62-8.58). Conclusions: Preliminary findings from this large multi-institutional study show that the overall survival for patients with lymphoma and HLH remains poor despite the institution of HLH-specific therapies. In nearly half of the cases, the physicians altered lymphoma treatment to control HLH, primarily through dexamethasone and etoposide-based or DEP regimens. Although B-cell lymphoma was our cohort's most common cause of HLH, overall survival was worse among those with T-cell or NK cell lymphoma, suggesting that the underlying lymphoma type might drive the prognosis even in cases with hyperinflammation. The most common causes of death included multi-organ dysfunction and infection. The OHI index-positive patients experienced higher mortality risk than those meeting the HLH-2004 criteria. The study is ongoing, and we plan to present data on the differential effects of treatment regimens at the meeting.