Overall Survival of HIV-Positive Patients with HLH

Timothy J Brown, Bonnie C Prokesch, Srikanth Nagalla and Christian Wysocki


Introduction: Hemophagocytic lymphohistiocytosis (HLH) is a rare but life-threatening disorder affecting the interplay between natural killer cells, macrophages, and cytotoxic lymphocytes. In HLH, persistent macrophage activation results from dysregulated cytokines produced from T-lymphocytes. The disease can be primary (due to a genetic defect) or secondary (triggered by another disease process). The diagnosis of HLH requires five of eight physical exam, biochemical, and histologic abnormalities: fever, splenomegaly, cytopenias, hypertriglyceridemia and/or hypofibrinogenemia, hyperferritinemia, hemophagocytosis on biopsy, low or absent NK cell activity, or elevated soluble CD25. The body of literature describing the link between HIV and HLH is limited to case reports and small case series, thus we sought to further define the characteristics and outcomes of this patient population. Further, we sought to characterize the interaction between HIV, malignancies, and HLH.

Methods: We retrospectively extracted data with regards to human immunodeficiency virus (HIV) status, treatment, and associated malignancies of all patients with confirmed HLH from 2008-2018. The outcome of interest was overall survival time from date of admission culminating in the diagnosis of HLH to death. Survivors were censored at the last clinical encounter. Survival was analyzed using Kaplan-Meier curves and log-rank test with significance set at P≤0.05. Hazard ratios were calculated using the logrank test.

Results: Overall, 43 patients fulfilled diagnostic criteria for HLH, with the first case registered in 2008. Of these 43, 11 patients had a concomitant HIV infection. All of those with HIV met criteria for acquired immune deficiency syndrome (AIDS). Associated illnesses included lymphoma (n=4), disseminated histoplasmosis (n=3), multi-centric Castleman's disease (N=1), Kaposi sarcoma (n=1), and autoimmune hepatitis (n=1). One patient had no identifiable associated illness. Six of the HIV patients were on anti-retroviral therapy prior to the diagnosis of HLH. Five patients with HIV received etoposide-based therapy and seven received steroids. Of the 32 patients without HIV, 15 had a malignancy. Compared to those without HIV infection, patients with HIV did not have worse survival (median survival for those with HIV = 8.2 months vs 3.6 months in those without HIV p=0.75, Hazard ratio for death (HR) 0.87 [95% confidence interval (CI) 0.37-2.904]). All patients with malignancy had a worse survival than those who did not (median survival with malignancy 1.9 months vs 37 months without malignancy p=0.0009, HR for death 3.648 [95% CI 1.804-9.169]), regardless of HIV status. Among those with HIV infection, presence of malignancy resulted in a trend towards worse survival (median survival HIV positive with malignancy = 2.03 months, without malignancy = 37.03 months, p=0.0578). In HIV-negative patients, the presence of malignancy is associated with worse survival (median survival with malignancy 1.9 months vs unreached in patients without malignancy p=0.0063 HR 3.56 [95% CI 1.475-10.11]).

Conclusions: While malignancy-associated HLH has been found in this patient population and others to confer poorer survival, HIV infection was not associated with worse overall survival regardless of the presence of a concomitant malignancy in our patient cohort. This analysis is limited by the small number of patients with HIV and the varied therapies used in treatment. Prospective databases and collaborative trials are needed to optimize therapy for HLH in patients with HIV with and without malignancies.

Disclosures No relevant conflicts of interest to declare.

  • * Asterisk with author names denotes non-ASH members.