AUTHORS
Cindy Varga, MD,BSc, Myra Robinson2, Manisha Bhutani, MD3, Christopher J. Ferreri, MD4, Barry Paul, MD, MS5, Brandon Kale, MD6, Karun Neupane, MD7, Shebli Atrash, MD8, Reed Friend, MD9, Ciara Louise L Freeman, MD, PhD10, Peter M Voorhees, MD11, Ariel Grajales-Cruz, MD12, Doris K. Hansen, MD13, Rebecca Gonzalez, PharmD14
AFFILIATIONS
1 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
2 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
3 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
4 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
5 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
6 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
7 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
8 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
9 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
10 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
11 Atrium Health/Levine Cancer Institute, Charlotte, North Carolina, United States
12 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
13 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
14 H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida, United States
ABSTRACT
Background: Corticosteroids (CS) are the cornerstone of management for cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) after BCMA-directed CAR-T therapy, yet real-world data on how cumulative steroid exposure effects long term outcomes remain limited. We evaluated steroid utilization patterns and their dose dependent relationship with toxicity and survival in a large multicenter relapsed/refractory multiple myeloma (RRMM) CAR-T cohort.
Methods: We retrospectively analyzed 417 RRMM patients treated with commercial anti-BCMA CAR-T (230 ide-cel, 187 cilta-cel) at two academic centers (5/2021–12/2024). This analysis was performed prior to the adoption of suppressive dexamethasone for excessive lymphocytosis post CAR-T. Patients were classified as no corticosteroid exposure (NCS, n=263) or corticosteroid-exposed (CS, n=154) within 30 days post-infusion. The CS group was further stratified by cumulative dex-equivalent dosing using three dosing thresholds: low-dose (LD-CS, <60mg, n=106), intermediate-dose (ID-CS, 61–200mg, n=32), and high-dose (HD-CS, >200mg, n=16). PFS and OS were estimated by Kaplan-Meier; Cox proportional hazards models identified univariable and multivariable predictors.
Results: CS exposed patients were older (median 68 vs 66 years) and had more frequent IMWG high-risk cytogenetics (43% vs 29%), greater bone marrow plasmacytosis >50% (33.8% vs 11.8%), and higher median baseline ferritin (336 vs 180 ng/mL) than NCS. The HD-CS subgroup was most disadvantaged: 43.8% had extramedullary disease (EMD), 0% responded to bridging therapy and the median baseline/peak ferritin was 670/7500 ng/mL. ORR at Day 30 was similar between CS and NCS-exposed patients (63.6% vs 57.0%). At a median follow-up of 15.2 mos, dose-stratified survival analysis revealed a mPFS of 17.9 (NCS), 21.2 (LD), 9.4 (ID), and 1.6 mos (HD) (p<0.0001) and OS of 40.7 (NCS), NR (LD), 28.7 (ID) and 3.6 mos (HD) (P<0.0001). Higher toxicity within HD-CS subgroup was notable with grade ≥3 CRS (18.8%), grade ≥3 ICANS (50%), and IEC-HS (31.3%, all p<0.05 vs LD/NCS). Delayed non-ICANS neurotoxicity (NINT) occurred in 9 CS patients (2 IEC-parkinsonism (PKS), 7 IEC-cranial nerve palsy (CNP)) vs 7 in NCS patients (1 with both IEC-PKS and CNP, 1 CIDP, 5 CNP). Three patients experienced IEC-enterocolitis (IEC-EC) all within the NCS subgroup. Non-relapse mortality (NRM) was 5.5% (n=5) in NCS, 12.4% (n=19) in CS overall, and 56% (n=9/16) in HD-CS. HD-CS deaths (11 of 16 patients) were predominantly immune-mediated or infectious (3 infection, 3 high-grade ICANS, 1 IEC-HS, 1 high-grade CRS, 1 NINT-associated, 1 cardiac-related), with only 2 attributable to myeloma progression; infection-related death rates rose from 3.1% (NCS) and 3.8% (LD) to 12.5% (HD-CS). One patient experienced IEC-EC associated mortality. On multivariable analysis for OS, HD-CS was not independently associated with mortality after adjusting for EMD, baseline ferritin, and ICANS grade ≥2 (HR 1.86, 95% CI 0.83–4.15, p=0.098). Baseline ferritin (HR 1.00, p<0.001) and EMD (HR 2.13, p<0.001) were the strongest independent predictors of death. On multivariable analysis for PFS, LD-CS exposure remained protective (HR 0.77, p<0.001), while ID- and HD-CS were independently associated with progression/death (HR 1.39 and 3.51; overall p=0.002) after adjusting for EMD, CAR-T product, prior lines of therapy, baseline ferritin, and high-grade CRS/ICANS. Infection rates by day 100 were similar between CS and NCS (46.8% vs 43.3%, p=0.61) but highest in HD-CS (75%, mostly bacterial); steroid dose was not independently associated with infection risk after adjustment for baseline ferritin (p=0.36).
Conclusions: In this real-world analysis of early corticosteroid exposure after anti BCMA CAR-T, LD-CS use did not compromise disease response and was associated with favorable PFS while HD exposure correlated with markedly inferior survival and high NRM. Critically, multivariable analysis demonstrates that the poor outcomes observed in the HD group are largely attributed to adverse baseline disease features rather than a direct deleterious effect of CS on CAR-T function. HD-CS exposure appears to be a surrogate marker for high-risk disease biology. Corticosteroids should not be withheld when clinically indicated; early intervention with stepwise dose escalation may minimize cumulative exposure and reduce NRM.
DISCLOSURES
Cindy Varga, MD,BSc: Janssen - Honoraria; sanofi - Honoraria; mmrc - Research Funding; Janssen - Research Funding; K36 - Research Funding; kite - Research Funding