Supportive care alone is effective for managing low-grade CRS in multiple myeloma patients treated with bispecific antibodies

Abstract

Abstract Introduction: Patients with relapsed/refractory multiple myeloma (RRMM) who receive BCMA or GPRC5D-directed bispecific antibodies (BsAb) commonly experience low-grade cytokine release syndrome (CRS). Based on limited data, IMWG and mSMART guidelines recommend CRS management with tocilizumab and/or dexamethasone. However, as most CRS is low-grade, we evaluated the safety and efficacy of CRS management with supportive care alone compared to alternative management strategies. Methods: Data on 555 patients with RRMM who initiated elranatamab, teclistamab, and/or talquetamab as of March 1st, 2025, were collected by 7 US academic medical centers. All patients received standard pre-medications and BsAb step-up dosing (SUD) per the package insert. Toxicity management and supportive care were based on institutional standards. Supportive care consisted of antipyretics, hydration, and/or oxygen supplementation. Tocilizumab was dosed at 8mg/kg (maximum 800mg) and dexamethasone ranged from 4-20mg per dose. CRS and ICANS were graded per ASTCT criteria. Outcomes included CRS incidence, severity, recurrence, and duration of hospitalization. Results: Of 555 patients, 36 received elranatamab, 307 received teclistamab, and 212 received talquetamab. Eighteen patients received both a BCMA BsAb and talquetamab at separate time. Median age was 68 years, 52% were male, and 22% were African American. Median prior lines of therapy was 5 (range 2-18), 30% had extramedullary disease and 65% of patients would have been ineligible for registrational trials. Forty-two (7.5%) patients received SUD in the outpatient setting. Overall incidence of CRS was 55% (n=306) (grade 1: 38%, grade 2: 15%). Eight patients experienced grade 3 CRS (1.4%) and two patients experienced grade 4 CRS. Median duration of CRS was 1 day (range 0.5-8 days). Of 29 patients who received prophylactic tocilizumab, 52% (n=15) experienced CRS (grade 1: 13, grade 2: 1, grade 4: 1). Incidence of ICANS was 15% (grade 1: 7%, grade 2: 4%, grade 3: 2.2%, grade 4: 1.3%). There were no CRS-related deaths. Three deaths were possibly related to neurotoxicity and/or progressive disease. Median PFS and OS were 8.3 months and 19 months, respectively. Among patients with CRS, 87 (28%) were managed with supportive care alone. The remaining 219 patients received pharmacologic intervention: dexamethasone (n=68, 22%), tocilizumab (n=85, 28%), or both dexamethasone and tocilizumab (n=66, 22%). Patients received a median of 5 dexamethasone doses (range 1-34). In those receiving both dexamethasone and tocilizumab, 30 patients (45%) had concomitant ICANS. Supportive care consisted of acetaminophen (90%), hydration (30%), and supplemental oxygen (5%). Thirteen patients in the supportive care group (15%) received prophylactic tocilizumab. There was no difference in age, baseline LDH, presence of extramedullary disease, or performance status between those managed with supportive care versus those managed with pharmacologic intervention. Ten patients in the supportive care group and 14 patients in the pharmacologic intervention group initiated SUD outpatient (p=0.134); of them, 3 and 9 patients were admitted for CRS (p=0.114). Patients treated with supportive care experienced more grade 1 CRS (92% vs 61%, p<0.001) and less grade 2 CRS (8% vs 35%, p<0.001) compared to those receiving pharmacologic intervention. Duration of CRS was 1.3 days for supportive care vs 1.8 days for pharmacologic intervention (p=0.003). There was no difference in incidence of dose delay due to CRS (33% vs 36%, p=0.304) or recurrent CRS (26% vs 32%, p=0.306) in the supportive care and pharmacologic intervention groups, respectively. Doses were delayed by a median of 2 days in those treated with supportive care and 1 day in those treated with pharmacologic intervention (p=0.458). Median duration of hospitalization for SUD was 9 for both groups (p=0.809). Conclusion: In this large real-world cohort of patients with RRMM receiving bispecific antibodies, most CRS events were low grade and short duration. Supportive care alone was associated with comparable outcomes to pharmacologic intervention with dexamethasone and/or tocilizumab for low grade CRS (particularly grade 1), including similar rates of dose delay, recurrent CRS, and duration of hospitalization. These findings suggest that supportive care may be an appropriate management strategy for low-grade CRS, reducing the need for immunosuppressive therapy and cost avoidance.