Background/Purpose: Chimeric antigen receptor T-cell (CAR-T) therapy is a highly effective form of adoptive cell immunotherapy combining antigen specific targeting capabilities with T-cell based cytotoxicity. Particularly effective against B-cell antigens (CD19 or CD22), CAR-T cell activation leads to a systemic inflammatory response called cytokine release syndrome (CRS), that can subsequently evolve into hemophagocytic histiocytosis (HLH)-like symptomatology. This abstract details the presentation, incidence and management of HLH-like toxicities in children/young adults with relapse/refractory pre-B acute lymphoblastic leukemia (pre-B ALL) treated on a phase I study of anti-CD22 CAR-T cell therapy.

Methods: Using modified diagnostic criteria as developed by Neelapu et al to define CAR-T cell related HLH, HLH was established in those who had a ferritin level of > 100,000 ng/mL with at least one of the following: > grade 3* AST/ALT elevation or hyperbilirubinemia; > grade 3* oliguria or increase in serum creatinine; > grade 3* pulmonary edema; or hemophagocytosis in the bone marrow. Serial inflammatory markers and serum cytokines were prospectively monitored from CAR-T cell infusion through day +28 (+/-4 days) and retrospectively analyzed comparing peak values in HLH/Non-HLH. Treatments for HLH included a combination of supportive care, corticosteroids and/or high-dose anakinra.

Results: In 52 subjects, 46 experienced CRS, of whom 37 (80.4%) achieved complete remission and 18 (39.1%) developed HLH-like manifestations. The median ferritin in those with/without HLH was 206740 ng/mL vs 22758 ng/mL, respectively. Clinical manifestations included: > grade 3* creatinine (n=2); > grade 3* transaminase elevation or hyperbilirubinemia (n=23, including 6 without HLH), > grade 3* pulmonary edema (n=5, including 2 without HLH); and hemophagocytosis on bone marrow aspirate (n=9). Cytokine profiling demonstrated significantly higher levels of IFN-y, IL-1B, IL-6, IL-10, TNFa and MIP-1a. Limited paired samples of s-IL2R demonstrated statistically significant increase from baseline (median level 1254) to HLH presentation (median level 9310), with 6/9 having substantial increase (peak value 123700). All patients had full resolution of HLH symptoms, except 1 who died from complications of gram-negative rod sepsis prior to HLH resolution. Three patients had asymptomatic laboratory abnormalities which self-resolved in a median of 14 days without intervention. Anakinra monotherapy (median 5 mg/kg/day) was used in 3 and the remainder (n=3) received anakinra + steroids. Both regimens resolved HLH. There was no statistically significant difference in underlying leukemia burden or efficacy following CAR-T cell therapy in those who did/did not develop HLH. Notably, use of steroids +/- anakinra did not diminish the therapeutic efficacy of CAR-T cells.

Conclusion: Ferritin and cytokine profiling revealed that HLH-like patients had a different inflammatory response independent of disease burden. Use of steroids and/or anakinra was effective and did not impede CAR-T cell expansion. Further analysis to identify HLH-predictive parameters and optimizing interventions to treat/prevent these complications are ongoing.

« Back to 2020 Pediatric Rheumatology Symposium

ACR Meeting Abstracts - https://acrabstracts.org/abstract/experience-with-and-management-of-hlh-like-toxicities-following-chimeric-antigen-receptor-t-cell-therapy-for-treatment-of-relapsed-refractory-pre-b-all/