Subcutaneous administration of BsAbs is a more convenient route and may mitigate the development of early, rapid, and uncontrollable immune responses and thus CRS. [12, 13] Subcutaneous elranatamab delayed Cmax and resulted in lower dose-normalized Cmax than IV elranatamab, leading to a lower grade ≥ 2 CRS incidence with the highest dose of SC elranatamab (1000 µg/kg; 33.3%) versus IV elranatamab (50 µg/kg; 83.3%), even though Cmax was higher with SC administration. The slower absorption with SC administration results in more gradual sensitization to initial immune stimulation versus IV administration [13]. The time to maximum observed serum concentration was delayed with the SC route (3–7 days) versus the IV route (~ 4 h, i.e., shortly after the end of infusion). The median time to onset of CRS tended to be later for SC cohorts and cohorts using priming regimens versus IV cohorts; CRS events occurring on the same day of elranatamab dosing were less frequent with SC versus IV administration. The CRS profile associated with SC dosing is consistent with the results of a recent meta-analysis of BCMA-targeting BsAbs [13]. The pharmacokinetic and CRS profile of IV and SC elranatamab during dose escalation in MagnetisMM-1 led to the selection of the SC route for subsequent studies.

Several elranatamab SC regimens for CRS mitigation were evaluated across four studies. The regimens spanned different approaches, including regimens without step-up doses, a 1-step-up dose priming regimen (44 mg), and two distinct 2-step-up dose priming regimens (12/32 mg and 4/20 mg). In the 1-step-up dose priming regimen cohorts, premedication reduced CRS incidence and was associated with a longer median time to onset and a shorter maximum CRS duration compared with the 1-step-up dose priming regimen without premedication. Although studies with elranatamab have not been performed, premedication is expected to have a minimal impact on the efficacy of elranatamab, given that corticosteroids are only administered with the initial doses. Therefore, premedication was implemented in subsequent cohorts/studies. However, the impact of premedication on grade ≥ 2 CRS appeared minimal, highlighting the need for an additional strategy, such as the 2-step-up dose priming regimen, to reduce the CRS risk.

Exposure–response modeling predicted that, compared with 44 mg, initial doses of 12 or 4 mg would result in lower CRS rates. These predications were confirmed via evaluation of the 12/32-mg and the 4/20-mg regimens. For the 12/32-mg regimen, the predicted rates agreed with the observed rates; however, the predicted rates for 4/20-mg regimen were underestimated, owing to the assumption that CRS events were driven by exposure after dose 1 (i.e., Cmax, 24h), which might not be valid if there is an inadequate stimulation of the immune system with dose 1. Alternative modeling strategies capturing CRS evolution and accounting for multiple events over time might better predict CRS with too low initial doses [22].

Cytokine release syndrome rates after dose 1 were higher with the 12/32-mg versus 4/20-mg priming regimen; however, CRS incidence and severity after subsequent doses were higher with the 4/20-mg regimen. Recurrent CRS events were also more frequent with the 4/20-mg regimen. Thus, an initial elranatamab dose of 4 mg may provide insufficient immune stimulation, leading to more CRS events with the second priming dose of 20 mg and the first full and later doses of 76 mg. The more predictable CRS profile observed with the 12/32 priming regimen led to the selection of this priming regimen for future use.

Currently approved BCMA-targeting T-cell redirecting therapies for RRMM include the chimeric antigen receptor T-cell therapies ciltacabtagene autoleucel (cilta-cel) and idecabtagene vicleucel (ide-cel) and the BsAbs elranatamab and teclistamab [4, 5, 23,24,25]. As with BsAbs, CRS is a common toxicity with chimeric antigen receptor T-cell therapies [10, 11]. Although the analysis is limited by cross-trial comparisons, CRS incidence (84–95%) and severity (grade ≥ 3, 5%; grade 5, ≤1%) with cilta-cel and ide-cel were higher and events lasted longer (median duration, 4.0–5.0 days) than those observed with elranatamab and teclistamab. [26,27,28,29] Differences in CRS premedication regimens (e.g., exclusion of corticosteroids for chimeric antigen receptor T-cell therapy) and a lack of fractionated dosing regimens may explain these differences [13].

Priming and full dosing regimens of approved BsAbs for cancer treatment have been previously summarized and discussed [14, 30]. Cytokine release syndrome with other bispecific antibodies approved for treating RRMM [25, 31], including teclistamab (any grade, 72%; grade 1, 50%; grade 2, 21%; grade 3, < 1%) and talquetamab, a GPRC5D-targeting BsAb (SC 400 µg/kg once weekly: any grade, 79%; grade 1, 62%; grade 2, 15%, grade 3, 2%), both utilizing 2 step-up dose priming regimens (teclistamab: 0.06 mg/kg and 0.3 mg/kg; talquetamab: 0.01 mg/kg and 0.06 mg/kg) and premedication, appear higher than those observed with elranatamab [28, 32]. After dose 1, CRS rates were similar between teclistamab and elranatamab (44% and 43%, respectively) and were lower with talquetamab (34%) [28, 32]. However, after dose 2, the CRS rate with elranatamab (19%) was lower than that with teclistamab (35%) and talquetamab (49%) [28, 32]. Similarly, the CRS incidence after the initial full dose of elranatamab [dose 3] (7%) was lower than that with teclistamab (24%) and talquetamab (27%) [28, 32]. The recurrent CRS rate was also lower with elranatamab (12.6%) than with teclistamab (33%) and talquetamab (32%) [28, 32]. Patients given the 12-mg priming dose of elranatamab initially received a higher percentage of the full dose (16%) and likely experienced a greater immune stimulation compared with teclistamab and talquetamab (4% and 2.5% of full dose, respectively), potentially explaining the reduced CRS incidence with later doses of elranatamab.

Fixed dosing was supported by similar CRS rates and profiles across a wide range of body weights and is preferred for drugs with a wide therapeutic window because of convenience, a reduced risk for medical errors, and cost effectiveness because of lower drug waste [33, 34]. Additionally, body weight was not a statistically significant covariate on elranatamab exposure and therefore, a fixed dosing approach is expected to result in less inter-individual variability in drug exposure and consequently the CRS profile [21].

Baseline sBCMA levels were negatively associated with the probability of any-grade CRS in a univariable analysis, possibly owing to sBCMA acting as a sink that reduces drug exposure [15], which is consistent with sBCMA not being a significant covariate in the multivariable analysis with exposure. Additional research is needed to establish the value of sBCMA as a biomarker for CRS.

Immune effector cell-associated neurotoxicity syndrome, another toxicity observed with T-cell redirecting therapies, is generally associated with CRS [9]. Immune effector cell-associated neurotoxicity syndrome was observed with the three highest doses of 360, 600, and 1000 µg/kg of SC elranatamab. The incidence of ICANS was similar in patients who received the 1-step-up dose priming regimen of 44 mg with or without premedication. Immune effector cell-associated neurotoxicity syndrome rates were lower with the 12/32 and 4/20 regimens, suggesting that lower starting doses are associated with a lower incidence of ICANS. However, the interpretation of the relationship between ICANS and the different priming regimens may be limited by the small number of events. The incidence and severity of ICANS were higher with cilta-cel (any grade, 17%; grade ≥ 3, 2%) [29] and ide-cel (reported as neurologic toxicity; any grade, 18%; grade ≥ 3, 3%) [27] versus the 12/32-mg priming regimen (any grade, 3.3%; grade 3, 1.1%), which could be related to the higher CRS rates and the inability/difficulty to implement fractionated dosing regimens with CAR T-cell therapies. Immune effector cell-associated neurotoxicity syndrome rates were similar between elranatamab (3.3%) and teclistamab (3%) but appeared higher with talquetamab (11%) [35, 36].

The limitations of this study include the small sample size of some of the cohorts and the low incidence of certain events per cohort, in particular ICANS. Additionally, the different priming regimens were not evaluated through a randomized comparison. However, the baseline characteristics for participants receiving theses regimens were generally comparable between groups. Despite these limitations, this analysis is the first describing multiple approaches to mitigate CRS and ICANS and may provide useful insights for the development of BsAbs.