Evaluation of Immune Reconstitution in Patients with Relapsed/Refractory Multiple Myeloma (RRMM) Treated with Cevostamab in Phase I Study GO39775

Background: Immunoparesis (IP), the suppression of one or more uninvolved immunoglobulins (Ig) below the lower limit of normal (LLN), is frequently observed in patients with multiple myeloma (MM). IP is associated with a negative impact on survival in patients with relapsed MM (Chakraborty et al. Br J Haematol 2019) and may increase the risk of infection in patients with MM (Pratt et al. Br J Haematol 2007). Cevostamab is a FcRH5xCD3 T-cell engaging bispecific antibody that facilitates T-cell directed killing of myeloma cells. Cevostamab has demonstrated clinically meaningful activity and a favorable toxicity profile when given once every 3 weeks for up to 17 cycles (approximately 1 year) in an ongoing Phase I trial (GO39775; NCT03275103) involving patients with heavily pre-treated RRMM (Trudel et al. ASH 2021). Hypogammaglobulinemia is a potential side effect due to the targeting of FcRH5 on normal plasma cells (Hammons et al. JAMA Netw Open 2022). The purpose of the current study was to evaluate whether reducing monoclonal Ig burden while on therapy allows for polyclonal Ig recovery, despite the mechanism of action of cevostamab. Specifically, we evaluated whether responders in the GO39775 study demonstrated evidence of immune reconstitution (IR) while on cevostamab therapy.

Methods: All patients were aged ≥18 years and had RRMM for which no established therapies were available. Cevostamab was administered via intravenous infusion every 21 days for 17 cycles unless disease progression or unacceptable toxicity occurred. Patients who received concomitant IVIg therapy were excluded from the analysis. Patients were included if they were enrolled for ≥100 days, demonstrated a partial response or better, and had IP at baseline, defined as ≥1 Ig below the LLN. IR was defined as recovery of ≥1 polyclonal Ig within the normal range.

Results: At data cut-off (March 1, 2023), a total of 310 patients were enrolled in GO39775. 76 patients met the inclusion criteria and were eligible for analysis. Median age was 65 years (range: 43-82), with a median of 6 previous lines of therapy (range: 2-12). 46 patients (60.5%) had IgG isotype, 10 (13.2%) had IgA isotype, and 20 (26.3%) had light chain disease.

Of the 76 patients who met the inclusion criteria, 10 (13.2%) experienced IR (see Table); 5 patients had IgG isotype, 3 had IgA isotype, and 2 had light chain disease. The 10 patients with IR were treated for a median of 323 days (range: 64-364). Of these, 6 patients had IR at a median time of Day 156 (range: 123-260) while on cevostamab, while 4 patients had IR at a median time of Day 242 (range: 133-309) after stopping cevostamab. Of the 6 patients who reconstituted on cevostamab, 1 had IgG isotype, 3 had IgA isotype, and 2 had light chain disease. All 4 patients who reconstituted after stopping cevostamab had IgG isotype. Apart from disease isotype, no other baseline characteristic demonstrated a trend with IR.

Discussion: The majority of patients with a response to cevostamab in the GO39775 study had baseline IP, consistent with previous findings in RRMM. Among eligible responders, 10 patients (13.2%) reconstituted ≥1 polyclonal Ig, highlighting that disease control may allow some patients to reverse IP. Of the patients who had IR, 6 had immune recovery while on cevostamab therapy, which indicates that IR is possible on cevostamab once disease control is achieved. Our findings demonstrate that despite normal plasma cell targeting, IR can occur while receiving a FcRH5xCD3 bispecific antibody. Further analysis on a larger patient pool may provide more insight into the characteristics of patients with RRMM who experience IR on cevostamab therapy.