Overlap of post-CAR T HLH toxicity definitions and their prognostic significance in children and young adults with B-acute lymphoblastic leukemia Free

Introduction: Hemophagocytic lymphohistiocytosis (HLH) toxicities are increasingly recognized following chimeric antigen receptor (CAR) T-cell therapies and are included on the FDA label of each commercial CAR product. Prior to the ASTCT immune effector cell-associated HLH-like toxicity (IEC-HS) consensus definition, various criteria were used. Post-CAR HLH toxicities have been linked to poor survival in B-acute lymphoblastic leukemia (B-ALL), lymphoma, and multiple myeloma. However, non-harmonized criteria have limited prior analyses. We retrospectively apply three CAR T-cell-associated HLH definitions (ASTCT IEC-HS, PRWCC HLH-like toxicity [HLH-LT] classification, institutional designation) to characterize classification overlap, identify core features of HLH, and assess associated outcomes in children and young adults receiving tisagenlecleucel (Tisa-cel).

Methods: This retrospective multi-institutional study included all patients infused with commercial Tisa-cel at five pediatric centers between January 3, 2018, and December 18, 2023. IEC-HS was defined per ASTCT as the development of a pathological and biochemical hyperinflammatory syndrome independent from CRS and ICANS that (1) manifests with features of macrophage activation/HLH, (2) is attributable to IEC therapy, and (3) is associated with progression or new onset of cytopenias, hyperferritinemia, coagulopathy with hypofibrinogenemia, and/or transaminitis. PRWCC HLH-LT was defined as a peak ferritin ≥10,000 ng/ml within 30 days of Tisa-cel with >=2 of: grade >=3 organ toxicities (renal, pulmonary, hepatic), hypofibrinogenemia (<150mg/dL), or hemophagocytosis. Institutionally-defined HLH was based on clinician determination within 30 days of Tisa-cel infusion, as documented in the electronic medical record. High disease burden was defined as described previously (Schultz L et al. JCO. 2022). Analyses for overall survival (OS) and event free survival (EFS) were conducted stratified by HLH definitions and baseline disease burden. Cox proportional hazards models were adjusted for baseline disease.

Results: 154 patients with a median age of 12 years were included. The median baseline bone marrow blast percentage was 1.36% (25^th-75^th interquartile range [IQR] 0%-30%), measured a median of 14 days (25^th-75^th IQR 10-20 days) prior to Tisa-cel. Following Tisa-cel, 23 (14.9%) met criteria for IEC-HS, 33 (21.4%) met criteria for HLH-LT, and 20 (13.0%) had HLH by institution determination. Median times from Tisa-cel to IEC-HS and HLH-LT onset were both 8 days. HLH definitions showed partial overlap: HLH-LT criteria captured the most patients, including 8 unique cases; IEC-HS captured one unique case; and institutional HLH captured no unique cases. Twenty-one (13.6%) met criteria for HLH-LT and IEC-HS, and 15 (9.7%) met all three definitions. Clinical features present in >=50% of patients with each respective HLH definition and in <25% of those without each respective HLH definition included: peak ferritin >=10,000ng/ml, AST >5x upper limit of normal (ULN), ALT >5x ULN, hypoxia or pulmonary edema, cytopenias (new onset/worsening/refractory), renal insufficiency, hypertriglyceridemia (>265mg/dL), and fibrinogen <150 mg/dL. LDH and PT/INR were elevated >50% of those with and without HLH definitions. Hemophagocytosis and splenomegaly were identified in <25% with any HLH definition.

Each HLH definition was associated with inferior OS and EFS compared with groups without respective definitions of HLH (p<0.0001). When stratified by high vs. no/low disease burden, those with each HLH definition experienced significantly inferior OS and EFS compared with those without respective HLH definitions (p<0.0001). Severe CRS without HLH was not associated with inferior outcomes compared with no or low-grade CRS without HLH. In Cox models, disease burden and HLH definitions were independently predictive of OS and EFS for each HLH definition. Of the three HLH definitions, IEC-HS carried the highest hazard ratios (HR) for EFS and OS of 6.78 and 5.98, respectively.

Conclusions: Post-CAR T HLH toxicity definitions have substantial overlap, occurred in 13-21% of those treated with Tisa-cel, and were independently associated with inferior OS and EFS. IEC-HS carried the highest HRs for OS and EFS events. A mechanistic understanding of the cause of poor outcomes and strategies to predict and prevent this severe post-CAR T sequela are critical for improving outcomes in CAR T-recipients.