FcRH5×CD3 bispecific antibody cevostamab in relapsed or refractory multiple myeloma: a phase 1 trial.

Main

Almost all patients with multiple myeloma eventually become triple-class refractory to proteasome inhibitors, immunomodulatory drugs and anti-CD38 antibodies 1 , 2 . Recently, chimeric antigen receptor (CAR) T cells and bispecific antibodies targeting B-cell maturation antigen (BCMA) or G protein-coupled receptor class C group 5 member D (GPRC5D) have shown activity in triple-class exposed and/or refractory patients 3 – 7 . However, some patients do not respond and almost all responders eventually progress 3 – 7 . In addition, the management of infection risk remains a key consideration for BCMA-targeted T cell-redirecting agents 3 , 5 – 7 , while GPRC5D-targeted agents may be compromised by on-target, off-tumor toxicities 4 . Continuous treatment with these agents may further exacerbate these toxicities and also lead to treatment resistance through target loss or mutation and/or acquired T cell dysfunction 8 – 10 . New treatments, especially those that target alternative antigens and can be given for a fixed duration, are needed.

Fc receptor-homolog 5 (FcRH5) is a membrane protein that is expressed exclusively in the B cell lineage and ubiquitously on myeloma cells, independent of BCMA expression, making it an attractive therapeutic target 11 – 16 . Cevostamab is a first-in-class, humanized, immunoglobulin G1-based FcRH5×CD3 bispecific antibody that induces FcRH5-targeted, T cell-mediated killing of myeloma cells 15 .

We present the results of a phase 1 study (GO39775; NCT03275103 ) evaluating the safety, pharmacokinetics, immunogenicity and anti-tumor activity of cevostamab in late-line relapsed or refractory multiple myeloma (RRMM) when administered for a fixed duration of treatment (~12 months).

Results

GO39775 is an open-label, multicenter, phase 1 dose-escalation and dose-expansion study in patients with late-line RRMM for whom no established therapy was available or considered appropriate or likely to be tolerable by the investigator. Prior T cell-redirecting therapy was allowed. Cevostamab (intravenous infusion) was initiated with step-up dosing for the mitigation of cytokine release syndrome (CRS) and continued at the target dose once every 3 weeks for 17 cycles (~12 months) unless disease progression or unacceptable toxicity occurred. In the dose-escalation stage, escalating step-up and target doses, and alternative step-up schedules, were evaluated in a standard 3 + 3 design. In the dose-expansion stage, dosing regimens showing promising safety and activity were selected for further evaluation. Primary objectives were to evaluate safety, including determination of the maximum tolerated dose (MTD), and to identify a recommended phase 2 dose and schedule (RP2D) for cevostamab monotherapy. Secondary objectives were to evaluate pharmacokinetics, immunogenicity and activity. Activity endpoints were the rate of response and duration of response (DOR) as determined by investigators according to International Myeloma Working Group (IMWG) criteria 17 . Progression-free survival (PFS) was assessed in a post hoc analysis. The safety and efficacy of tocilizumab as premedication for the mitigation of CRS after cevostamab dosing was evaluated in an additional expansion cohort, but these data were not included in the current analysis and will be reported elsewhere. Data on cevostamab retreatment were also not included in the current analysis and will be reported elsewhere.

Patients

Between 19 September 2017 and 18 July 2023, 324 patients were enrolled into the monotherapy cohorts at 17 centers in the United States, Canada, Spain and Australia (Supplementary Table 1 ). As of 24 February 2025, all patients had received at least one dose of cevostamab. Of these, 48 patients completed 17 cycles of treatment and 276 discontinued (Table 1 ). The most common reason for discontinuation was disease progression (64.8%). Twenty-two dose-escalation cohorts were opened (Extended Data Fig. 1 ). Six were expanded and one additional expansion cohort was opened. Dosing regimens included those assessing single, double and triple step-up dosing and 0.15–252 mg target doses. Table 1 Patient disposition n (%) All target-dose levels Inactive target-dose levels Active target-dose levels 0.15–252 mg 0.15–10.8 mg 20 mg 40 mg 60 mg 90 mg 132 mg 160 mg 198 mg 252 mg Allocated to intervention 324 17 3 6 12 63 7 167 16 33 Received allocated intervention 324 (100) 17 (100) 3 (100) 6 (100) 12 (100) 63 (100) 7 (100) 167 (100) 16 (100) 33 (100) Completed 17 cycles of treatment 48 (14.8) 0 0 1 (16.7) 0 7 (11.1) 2 (28.6) 28 (16.8) 3 (18.8) 7 (21.2) Completed <17 cycles of treatment 276 (85.2) 17 (100) 3 (100) 5 (83.3) 12 (100) 56 (88.9) 5 (71.4) 139 (83.2) 13 (81.3) 26 (78.8) Disease progression 210 (64.8) 15 (88.2) 1 (33.3) 4 (66.7) 12 (100) 42 (66.7) 4 (57.1) 104 (62.3) 10 (62.5) 18 (54.5) AEs 28 (8.6) 0 0 1 (16.7) 0 5 (7.9) 1 (14.3) 17 (10.2) 2 (12.5) 2 (6.1) Death 18 (5.6) 1 (5.9) 0 0 0 4 (6.3) 0 11 (6.6) 0 2 (6.1) Withdrawal by subject 8 (2.5) 0 0 0 0 2 (3.2) 0 2 (1.2) 0 4 (12.1) Symptomatic deterioration 4 (1.2) 1 (5.9) 0 0 0 1 (1.6) 0 2 (1.2) 0 0 Other 3 (0.9) 0 1 (33.3) 0 0 1 (1.6) 0 1 (0.6) 0 0 Physician decision 3 (0.9) 0 1 (33.3) 0 0 0 0 1 (0.6) 1 (6.3) 0 Lack of efficacy 2 (0.6) 0 0 0 0 1 (1.6) 0 1 (0.6) 0 0 Analyzed for primary objective 324 (100) 17 (100) 3 (100) 6 (100) 12 (100) 63 (100) 7 (100) 167 (100) 16 (100) 33 (100)

Among the 324 patients enrolled, the median age was 65 years (range: 33–90), 55.6% were male and 75.9% were white (Table 2 ). High-risk cytogenetics (t(4;14), t(14;16) or del(17p)) were present in 34.7% of patients with a conclusive assay result, while 1q21 gain or amplification was present in 52.4%. Extramedullary disease was present in 26.2%. The median number of prior lines of therapy was 6 (range: 2–21); 47.5% had received prior BCMA-targeted therapy (CAR T cell: 28.1%; antibody–drug conjugates (ADC): 17.6%; bispecific antibody: 10.8%) (Table 2 ). Triple-class refractory disease was present in 89.5% and penta-drug refractory disease in 72.5%; 87.0% had myeloma that was refractory to their last prior therapy. Table 2 Baseline characteristics of all patients who received cevostamab at the 160 mg target-dose level and of all patients n (%) unless stated otherwise 160 mg target-dose level ( n = 167) All patients ( n = 324) Age (years) Median (range) 66.0 (40–90) 65.0 (33–90) ≥65 years 95 (56.9) 174 (53.7) Sex Male 87 (52.1) 180 (55.6) Female 80 (47.9) 144 (44.4) Race White 125 (74.9) 246 (75.9) Black or African American 23 (13.8) 41 (12.7) Asian 5 (3.0) 15 (4.6) Other 4 (2.4) 7 (2.2) Unknown 10 (6.0) 15 (4.6) ECOG performance status 0 57 (35.0) 119 (37.3) 1 106 (65.0) 200 (62.7) IgG < 400 mg dl −1 53 (31.7) 90 (27.8) High-risk cytogenetics a t(4;14), t(14;16) and/or del(17p) 38/99 (38.4) 69/199 (34.7) t(4;14) 13/106 (12.3) 25/207 (12.1) t(14;16) 6/98 (6.1) 8/196 (4.1) del(17p) 24/124 (19.4) 45/238 (18.9) 1q21 gain or amplification a 43/93 (46.2) 100/191 (52.4) Extramedullary disease b 47 (28.1) 85 (26.2) Time since first multiple myeloma therapy (years; median (range)) 6.3 (0.3 to 21.8) 6.1 (0.3 to 22.8) Number of prior lines of multiple myeloma therapy Median (range) 6 (2.0 to 18.0) 6 (2.0 to 21.0) ≥4 lines of prior multiple myeloma therapy 146 (87.4) 281 (86.7) Triple-class refractory c 160 (95.8) 290 (89.5) Penta-drug refractory d 123 (73.7) 235 (72.5) Refractory to last prior therapy 142 (85.0) 282 (87.0) Any prior BCMA-targeted therapy 96 (57.5) 154 (47.5) Any prior BCMA-targeted CAR T cell therapy 58 (34.7) 91 (28.1) Any prior BCMA-targeted ADC therapy 32 (19.2) 57 (17.6) Any prior BCMA-targeted bispecific antibody therapy 29 (17.4) 35 (10.8) ECOG, Eastern Cooperative Oncology Group a n (%) of patients with a conclusive assay result. b As determined by investigator assessment. c ≥1 immunomodulatory drug, ≥1 proteasome inhibitor and ≥1 anti-CD38 antibody. d ≥2 immunomodulatory drugs, ≥2 proteasome inhibitors and ≥1 anti-CD38 antibody.

Outcomes in all patients

Dose-limiting toxicities (DLTs) occurred in one patient in the 3.6/90 mg single step-up cohort (grade 3 pneumonia) and one in the 3.6/198 mg cohort (grade 1 CRS, which was initially graded as grade 3 using criteria from a previously published grading system 18 ; Extended Data Fig. 1 ); grade 3 pneumonia occurred after the initial target-dose administration on cycle 1 day 8 (C1D8; onset of event: C1D19), and grade 1 CRS occurred after the initial target-dose administration on C1D9 (onset: C1D9). Both events resolved. The MTD was not reached per the protocol-defined criteria for MTD determination.

Among all safety-evaluable patients ( n = 324), the most common (≥20%) adverse events (AEs) of any grade were CRS (77.8%), neutropenia (33.3%), cough (31.5%), anemia (28.1%), diarrhea (24.4%), nausea (24.1%) and fatigue (21.0%) (Table 3 ). Grade 3 or 4 AEs and serious AEs (SAEs) of any grade occurred in 193 (59.6%) and 195 (60.2%) patients, respectively. Grade 5 (fatal) AEs excluding disease progression occurred in 15 patients (4.6%); events in 3 patients (0.9%) were considered treatment related. AEs leading to cevostamab discontinuation excluding myeloma progression occurred in 50 patients (15.4%); events in 20 patients (6.2%) were considered treatment related. Table 3 AEs regardless of attribution in ≥10% of all patients who received cevostamab at the 160 mg target-dose level and in all patients n (%) 160 mg target-dose level ( n = 167) All patients ( n = 324) Any Grade 3 or 4 Any Grade 3 or 4 AEs 167 (100) 96 (57.5) 323 (99.7) 193 (59.6) Hematologic AEs Neutropenia a 53 (31.7) 48 (28.7) 108 (33.3) 98 (30.2) Anemia 40 (24.0) 30 (18.0) 91 (28.1) 67 (20.7) Thrombocytopenia b 27 (16.2) 18 (10.8) 62 (19.1) 44 (13.6) Non-hematologic AEs CRS 124 (74.3) 3 (1.8) 252 (77.8) 4 (1.2) Cough 51 (30.5) 1 (0.6) 102 (31.5) 1 (0.3) Nausea 47 (28.1) 1 (0.6) 78 (24.1) 1 (0.3) Diarrhea 40 (24.0) 0 (0.0) 79 (24.4) 2 (0.6) Fatigue 36 (21.6) 2 (1.2) 68 (21.0) 5 (1.5) Headache 32 (19.2) 0 (0.0) 60 (18.5) 1 (0.3) Hypophosphatemia 32 (19.2) 16 (9.6) 55 (17.0) 26 (8.0) Hypokalemia 27 (16.2) 4 (2.4) 47 (14.5) 8 (2.5) Back pain 24 (14.4) 4 (2.4) 44 (13.6) 5 (1.5) Oropharyngeal pain 23 (13.8) 0 (0.0) 35 (10.8) 0 (0.0) Pyrexia 23 (13.8) 0 (0.0) 53 (16.4) 0 (0.0) Hypomagnesemia 22 (13.2) 0 (0.0) 46 (14.2) 1 (0.3) Constipation 20 (12.0) 0 (0.0) 37 (11.4) 0 (0.0) Pruritus 20 (12.0) 0 (0.0) 33 (10.2) 0 (0.0) Rash maculopapular 20 (12.0) 3 (1.8) 32 (9.9) 3 (0.9) ALT increased 18 (10.8) 9 (5.4) 38 (11.7) 15 (4.6) Nasal congestion 18 (10.8) 0 (0.0) 31 (9.6) 0 (0.0) Productive cough 18 (10.8) 0 (0.0) 25 (7.7) 0 (0.0) Dyspnea 17 (10.2) 2 (1.2) 33 (10.2) 6 (1.9) Edema peripheral 17 (10.2) 0 (0.0) 30 (9.3) 0 (0.0) Vomiting 17 (10.2) 1 (0.6) 32 (9.9) 1 (0.3) Grade 5 AE excluding disease progression 10 (6.0) 15 (4.6) SAE 96 (57.5) 195 (60.2) AE leading to treatment discontinuation 29 (17.4) 50 (15.4) AEs of interest Infections c 91 (54.5) 26 (15.6) 171 (52.8) 67 (20.7) Infusion-related reaction 12 (7.2) 0 (0.0) 45 (13.9) 0 (0.0) Tumor lysis syndrome 0 (0.0) 0 (0.0) 2 (0.6) 2 (0.6) Hemophagocytic lymphohistiocytosis 3 (1.8) 1 (0.6) 3 (0.9) 1 (0.3) Possible immune effector cell-associated neurotoxicity syndrome d 23 (13.8) 3 (1.8) 53 (16.4) 6 (1.9) a Group term including the preferred terms neutropenia, neutrophil count decreased and febrile neutropenia. b Group term including the preferred terms thrombocytopenia and platelet count decreased. c MedDRA System Organ Class level term. d Possible immune effector cell-associated neurotoxicity syndrome events were defined as neurologic or psychiatric adverse events that were consistent with immune effector cell-associated neurotoxicity syndrome as described by the American Society of Transplantation and Cellular Therapy 15 and satisfied either (1) a neurologic or psychiatric adverse event that presented as a symptom of CRS or (2) neurologic or psychiatric adverse events that were considered treatment related by the investigator.

In all patients with ≥1 post-baseline result, anti-drug antibodies (ADAs) were detected in 63 of 296 (21.3%) and neutralizing antibodies in 23 of 290 (7.9%).

Among all efficacy-evaluable patients ( n = 323), the objective response rate was 42.1% and the very good partial response (VGPR) or better rate was 25.1% (Extended Data Table 1 ). Median time to first response was 1.4 months (minimum to maximum: 0.5–5.9) and to best response was 2.5 months (minimum to maximum: 0.5–27.3). The median DOR was 11.2 months (95% confidence interval (CI), 8.3, 15.6) among patients in partial response or better ( n = 136) and 25.1 months (95% CI, 19.4, 36.4) among those in VGPR or better ( n = 81) (Extended Data Table 1 ).

On the basis of a comprehensive assessment of cumulative efficacy and safety and a comprehensive quantitative clinical pharmacology evaluation that will be reported elsewhere, once every 3 weeks (Q3W) dosing at the 160 mg target-dose level and triple step-up dosing at the 0.3/1.2/3.6/160 mg dose level were found to have the optimal benefit–risk profile (RP2D for cevostamab monotherapy). Notably, across the range of Q3W target doses tested (0.15–252 mg), the clinical data, along with exposure–response analyses, showed that increasing cevostamab exposure was associated with a higher objective response and VGPR or better rate (Extended Data Table 2 ), with the attainment of a plateau at the 160 mg target-dose level, implying diminishing gains at target doses beyond 160 mg. While the majority of AEs of interest (including grade ≥1 and ≥2 CRS, grade ≥3 cytopenias and grade ≥3 AEs) did not show a strong correlation with the target dose or exposure, a significant exposure–response safety relationship was observed for SAEs and grade ≥3 infections, suggesting that, at target doses beyond 160 mg, there was an increased risk of these events compared with the 160 mg dose. In addition, triple step-up dosing at the 0.3/1.2/3.6/160 mg dose level showed the most favorable CRS profile among all of the tested priming regimens. Cumulative data for the 160 mg target-dose level and 0.3/1.2/3.6/160 mg triple step-up priming regimen are summarized below and in the tables.

Outcomes in patients who received the recommended phase 2 dose and schedule

A total of 167 patients were enrolled at the 160 mg target-dose level. Of these, 10 patients initiated treatment with single step-up dosing, 97 with double step-up dosing and 60 with triple step-up dosing (Extended Data Fig. 1 ). Baseline characteristics were generally comparable with those of all patients (Table 2 ). Median follow-up was 20.0 months (range: 0.5–54.9).

Safety

The most common (≥20%) AEs of any grade at the 160 mg target-dose level were CRS (74.3%), neutropenia (31.7%), cough (30.5%), nausea (28.1%), diarrhea (24.0%), anemia (24.0%) and fatigue (21.6%) (Table 3 ). Grade 3 or 4 AEs and SAEs of any grade occurred in 96 patients each (57.5%). Grade 5 (fatal) events excluding disease progression occurred in 10 patients (6.0%); events in 3 patients (1.8%) were considered treatment related (hemophagocytic lymphohistiocytosis (HLH) in 2 patients, disseminated intravascular coagulation in the context of pseudomonal sepsis in 1 patient) (Extended Data Table 3 ). AEs leading to cevostamab discontinuation excluding myeloma progression occurred in 29 patients (17.4%); events in 13 patients (7.8%) were considered treatment related (Extended Data Table 4 ).

Grade 3 or 4 neutropenia occurred in 48 patients (28.7%) and grade 3 or 4 febrile neutropenia in 8 (4.8%) at the 160 mg target-dose level. Thirty-seven (22.2%) patients received colony stimulating factors, and 19 (11.4%) received dose modifications or interruptions for any-grade neutropenia. Grade 3 or 4 thrombocytopenia occurred in 18 patients (10.8%). No neutropenia or thrombocytopenia leading to cevostamab discontinuation was reported.

Infections occurred in 91 patients (54.5%) at the 160 mg target-dose level (Table 3 and Extended Data Table 5 ). Infections in ≥5% were pneumonia (9.6%), upper respiratory tract infection (8.4%), urinary tract infection (7.2%) and rhinovirus infection (5.4%). Grade 3 or 4 infections occurred in 26 patients (15.6%) and grade 5 infections in 6 patients (3.6%). Infections leading to treatment discontinuation were uncommon (9 patients, 5.4%). Anti-infective prophylaxis for viral, fungal, bacterial or pneumocystis infections was not mandated but was recommended per institutional practice; one patient developed pneumocystis infection despite prophylaxis with sulfamethoxazole–trimethoprim. Forty-nine patients (29.3%) had a reported medical history of any hypogammaglobulinemia, while 13 (7.8%) had ≥1 treatment-emergent AE of hypogammaglobulinemia. Of the 79 patients who had an IgG level <400 mg dl −1 at any time between cycle 2 and the first post-treatment visit, 40 had an IgG level ≥400 mg dl −1 at baseline, with median time to first result below 400 mg dl −1 of 80 days. Fifty-five patients (32.9%) received intravenous immunoglobulin during treatment. Three of the 32 patients who had a grade 3+ infection had received intravenous immunoglobulin within the preceding 30 days.

Possible immune effector cell-associated neurotoxicity syndrome (see Table 3 for the definition) occurred in 23 patients (13.8%; 11 grade 1, 6.6%; 9 grade 2, 5.4%; 3 grade 3, 1.8%) at the 160 mg target-dose level (Table 3 ); events in ≥3 patients were confusional state (4.8%; 3 grade 1, 1.8%; 3 grade 2, 1.8%; 2 grade 3, 1.2%) and dysphonia (1.8%; 1 grade 1, 0.6%; 2 grade 2, 1.2%). Infusion-related reactions occurred in 12 patients (7.2%; all grade 1 or 2). No cases of tumor lysis syndrome were reported. Grade 3 HLH occurred in one additional patient who discontinued treatment.

CRS occurred in 19 of 30 patients (63.3%) who received the 0.3/1.2/3.6/160 mg triple step-up priming regimen (Table 4 ). CRS was grade 1 in 14 (46.7%) and grade 2 in 5 (16.7%). No grade 3 or 4 events or events leading to treatment discontinuation occurred. CRS occurred in 18 patients in cycle 1 and in 4 patients in cycle 2+ (Extended Data Fig. 2 ). Ten patients had ≥1 CRS event, including all 4 of the patients who had CRS in cycle 2+. All 4 of the patients who had CRS in cycle 2+ had previously experienced a CRS event. The median time to CRS onset from the last end of infusion was 13.1 h (minimum to maximum: 0–57.1) and median time to resolution was 1 day (minimum to maximum: 0–5). CRS was managed with tocilizumab in 9 patients (30.0%), steroids in 4 (13.3%) and both agents in 2 (6.7%). All events resolved. Table 4 AEs of CRS regardless of attribution in all patients who received cevostamab at the 160 mg target-dose level and initiated treatment with the cycle 1 0.3/1.2/3.6 mg triple step-up priming regimen 160 mg target-dose level + 0.3/1.2/3.6 mg triple step-up dosing ( n = 30) n (%) of patients with: CRS 19 (63.3) Grade 1 14 (46.7) Grade ≥2 5 (16.7) Grade 2 5 (16.7) Grade 3 0 (0.0) Grade 4 0 (0.0) CRS leading to discontinuation 0 (0.0) n (%) of patients receiving: Tocilizumab 9 (30.0) Steroids 4 (13.3) Tocilizumab and steroids 2 (6.7) n (%) of CRS events: n = 35 a Resolved at data cutoff 35 (100.0) a A total of 35 CRS events occurred in 19 patients; 6 patients had ≥1 CRS event.

Pharmacokinetics and immunogenicity

Cevostamab pharmacokinetics and immunogenicity were assessed with a data cutoff of August 2023. Systemic exposure increased in a dose-proportional manner over the tested dose range. Serum concentrations peaked at the end of infusion and declined in a multi-phasic fashion. At the 160 mg target-dose level, the estimated median half-life was ~13 days. In patients with ≥1 post-baseline result at the 160 mg target-dose level, ADAs were detected in 29 of 148 (19.6%) and neutralizing antibodies in 5 of 142 (3.5%). As the sample size for neutralizing antibodies was limited ( n = 5), no conclusive characterization could be carried out to assess their impact on pharmacokinetics, efficacy or safety.

Anti-tumor activity

At the 160 mg target-dose level, the objective response rate was 44.3% and the VGPR or better rate was 25.7% (Fig. 1a and Extended Data Table 1 ). Median time to first response was 1.4 months (minimum to maximum: 0.5–4.6) and to best response was 2.6 months (minimum to maximum: 0.5–13.4). The median DOR was 10.4 months (95% CI, 6.2, 15.0) among patients in partial response or better ( n = 74) and 22.7 months (95% CI, 15.5, not estimable) among those in VGPR or better ( n = 43) (Fig. 1b and Extended Data Table 1 ). Minimal residual disease negativity at the <10 −5 level was achieved in 11 of 18 patients who achieved a VGPR or better and were evaluated. Fig. 1 Rate and DOR. a , b , Best overall response rate in all patients and in those with and without prior BCMA-targeted therapies who received cevostamab at the 160 mg target-dose level ( a ) and Kaplan–Meier plot of the DOR among all patients in partial response (PR) or VGPR or better who received cevostamab at the 160 mg target-dose level ( b ). sCR, stringent complete response; CR, complete response.

At the 160 mg target-dose level, baseline characteristics were generally comparable between patients with and without prior exposure to BCMA-targeted therapies, although the median number of prior lines of therapy appeared lower and the percentage of patients with high-risk cytogenetics and who were refractory to their last prior line appeared higher in those without prior exposure (Extended Data Table 6 ). The objective response rate was 60.6% in those without prior BCMA-targeted therapy versus 36.2% with ≥1 BCMA-targeted CAR T cell, 17.2% with ≥1 BCMA-targeted bispecific antibody and 43.8% with ≥1 BCMA-targeted ADC (Fig. 1a ). The median DOR (partial or better) was 19.7 months (95% CI, 10.0, 36.4) in patients without prior BCMA-targeted therapy ( n = 71) versus 7.1 months (95% CI, 4.2, 15.0) with ≥1 BCMA-targeted CAR T cell ( n = 58), 3.8 months (95% CI, 1.6, not evaluable) with ≥1 BCMA-targeted bispecific antibody ( n = 29) and 5.5 months (95% CI, 3.5, 14.0) with ≥1 BCMA-targeted ADC ( n = 32).

At the 160 mg target-dose level, the objective response rate was 46.5% (20/43 patients) in patients with 1q21 gain or amplification at baseline and 36.2% (17/47 patients) in those with extramedullary disease.

At the 160 mg target-dose level, 26 of the 74 patients who achieved a partial response or better completed the planned 17 cycles of treatment in partial response or better (Fig. 2 ). Seventeen had ongoing responses off treatment at data cutoff, including 5 with responses lasting ≥30 months after treatment initiation. Sixteen of 19 patients with responses lasting ≥6 months after completion were in complete response or better at end of treatment. Fig. 2 Evolution of response over time. Swimlane plot showing the duration of follow-up and evolution of response over time in all patients who achieved a PR or better and completed 17 cycles of treatment at the 160 mg target dose. MR, minimal response; SD, stable disease.

FcRH5 expression on tumor samples was observed for all biomarker-evaluable patients who received cevostamab at the 160 mg target-dose level. Best overall response was independent of the baseline FcRH5 expression level (Extended Data Fig. 3a ). The proportion of myeloma cells expressing FcRH5 (%FcRH5+) was higher in patients with a 1q abnormal (gain or amplification) cytogenetic status compared with a 1q normal status (Extended Data Fig. 3b ). No difference in molecules of equivalent soluble fluorochrome was observed (Extended Data Fig. 3b ).

In a post hoc analysis of PFS, median PFS was 2.8 months (95% CI, 2.1, 4.2) and the 12-month PFS rate was 23.1% (95% CI, 16.6, 29.6) (Extended Data Fig. 4 ). Median PFS was 5.2 months (95% CI, 2.6, 11.1; n = 71), without prior BCMA-targeted therapy, and 2.1 months (95% CI, 1.4, 3.0; n = 96), with prior BCMA-targeted therapy.

Discussion

We evaluated the safety, pharmacokinetics, immunogenicity and activity of cevostamab, a first-in-class, FcRH5-targeted bispecific antibody, in late-line multiple myeloma. Dosing regimens included those assessing single, double and triple step-up dosing and 0.15–252 mg target doses. The MTD was not reached per the protocol-defined criteria. Q3W dosing at the 160 mg target-dose level and triple step-up dosing at the 0.3/1.2/3.6/160 mg dose level were found to have the optimal benefit–risk profile (RP2D for cevostamab monotherapy).

Notably, the population in this phase 1 study was heavily pretreated. At the 160 mg target-dose level, patients had received a median of 6 prior lines of therapy; 95.8% were triple-class refractory, and 73.7% were penta-drug refractory. In addition, 57.5% had received prior BCMA-targeted therapy, with the availability of these and other targeted agents evolving considerably during the enrollment period.

The most frequent grade 3 or 4 events at the 160 mg target-dose level were reversible neutropenia (28.7%), anemia (18.0%) and thrombocytopenia (10.8%). Grade 3 or 4 neutropenia (32–64%), anemia (23–37%) and thrombocytopenia (11–24%) were also reported in the pivotal studies of the GPRC5D-targeted bispecific antibody, talquetamab, and the BCMA-targeted bispecific antibodies, teclistamab and elranatamab 4 – 6 .

Grade 3 or 4 infections occurred in 15.6% of patients at the 160 mg target-dose level, which appears to differ from the rates reported in the pivotal studies of teclistamab (45%), elranatamab (40%) and talquetamab (7%), although cross-trial comparisons should be interpreted with caution 4 – 6 . Rates of grade 3 or 4 febrile neutropenia (4.8%) were also low in the current study. Grade 5 infections occurred with a low frequency in all studies 4 – 6 . Of note, anti-infective prophylaxis, including intravenous immunoglobulin replacement therapy, was not mandated in the current study and guidelines for patients receiving bispecific antibodies were evolving during the enrollment period, which may have influenced the infection rate. Cevostamab did not cause substantial dysgeusia, weight loss, nail changes, desquamation or ataxia, toxicities associated with GPRC5D-targeted therapies that can significantly impact quality of life 4 , 19 .

In the current study, extensive efforts were made to mitigate CRS. At the 160 mg target-dose level, treatment initiation with a 0.3/1.2/3.6 mg triple step-up priming regimen had a favorable CRS profile. Among 30 patients who received this schedule, CRS primarily occurred in cycle 1, and was all grade 1 or 2. No events leading to treatment discontinuation were reported, and all events resolved.

At the 160 mg target-dose level, cevostamab was associated with an objective response rate of 44.3% and a VGPR or better rate of 25.7%. Responses were durable, especially in patients in VGPR or better. These data are compelling, given that FcRH5 is a new therapeutic target for myeloma, the population was highly refractory and >50% had received prior BCMA-targeted therapy. Notably, the objective response rate in patients without prior BCMA-targeted therapy (60.6%) was higher than in those with prior exposure (17.2–43.8%), and appeared to be generally comparable with those reported in the pivotal studies of teclistamab (63%), elranatamab (61%) and talquetamab (64–70%) in BCMA therapy-naive populations, while again acknowledging the limitations of cross-trial comparison 4 – 6 . Interestingly, the VGPR or better rate in patients without prior BCMA-targeted therapy (39.4%) appeared to be lower than in the abovementioned pivotal studies (58.8%, 56.1% and 52–57%, respectively), potentially reflecting the higher rate of penta-drug-refractory disease in this study 4 – 6 . Response rates in patients with prior BCMA-targeted CAR T cell (36.2%) and ADC (43.8%) exposure were also clinically meaningful, supporting exploration of cevostamab as a non-BCMA-targeted option for patients relapsing after these therapies.

The short median PFS observed at the 160 mg target-dose level was largely driven by the high proportion of BCMA therapy-exposed, penta-drug-refractory patients in this cohort, and was relatively similar to that reported with other T cell-redirecting therapies when used after prior BCMA-targeted therapies 5 , 20 – 22 . The relatively short washout period in this trial (≥12 weeks for CAR T cells, ≥4 weeks for bispecific antibodies) may also have had a role, as recent reports suggest that a shorter interval between T cell-redirecting therapies may be associated with poorer T cell function and worse outcomes 20 , 23 , 24 . Further studies exploring the optimal timing and sequencing of T cell-redirecting therapies are needed. Notwithstanding this observation, it is notable that durable responses were observed in this cohort of patients, especially among those who achieved a VGPR or better.

Most targeted therapies for RRMM, including approved bispecific antibodies, are given until progression. However, indefinite T cell engagement can produce unintended negative consequences, including persistent hypogammaglobulinemia, progressive T cell dysfunction and high rates of tumor antigen loss or mutation 8 – 10 , 25 . This can not only lead to high infection rates and loss of efficacy, but may also compromise future therapies against the same target. In this study, a fixed duration of treatment of ~12 months was evaluated. At the 160 mg target-dose level, 26 patients completed treatment in partial response or better, with 17 having ongoing responses off treatment at data cutoff. Most durable responders were in complete response or better at end of treatment, suggesting that, in patients who achieve a complete response to cevostamab, it may be possible to achieve an extended treatment-free interval that could help to alleviate cumulative toxicities and improve quality of life.

Limitations of this study include small patient numbers in the step-up dosing cohorts; non-uniform follow-up, with relatively short follow-up in the latest cohorts; heterogeneity in the step-up dosing schedules used in the 160 mg target-dose level group; and the evolution in CRS management and availability of targeted myeloma therapies during the enrollment period. A standardized approach for anti-infective prophylaxis was also lacking, with guidance for patients receiving bispecific antibodies evolving over the duration of study. Risk status according to the most recent IMWG criteria could not be determined post hoc because key genetic alterations were not evaluated.

Collectively, these data validate FcRH5 as a therapeutic target for myeloma and show that fixed-duration cevostamab has manageable safety and induces durable remissions in patients with late-line disease. These data support the continued development of cevostamab alone or in combination with standard of care or investigational agents 26 , 27 . Given its distinct target, lack of GPRC5D-related skin or mucosal toxicities, and manageable infection risk profile, cevostamab may offer a complementary therapeutic option within the expanding landscape of T cell-redirecting therapies for multiple myeloma.

Methods

The GO39775 study was conducted in accordance with the International Conference on Harmonization guidelines for Good Clinical Practice. The protocol was approved by the institutional review board and ethics committee at each participating center (Supplementary Data Table 1 ). Patients provided written informed consent and were not compensated.

Patients

In summary, eligible patients were those with measurable disease (defined by at least one of serum monoclonal protein ≥0.5 g dl −1 (≥5 g l −1 ), urine monoclonal protein ≥200 mg per 24 h or involved serum free light chains (SFLCs) ≥10 mg dl −1 (≥100 mg l −1 ) and an abnormal SFLC ratio (<0.26 or >1.65)) for whom no established therapy for RRMM was available or considered appropriate or likely to be tolerable by the investigator. Prior CAR T cells ≥12 weeks and ADCs or bispecific antibodies ≥4 weeks before cevostamab initiation were allowed.

Detailed inclusion criteria

Signed informed consent form(s) Age ≥18 years at the time of signing of the informed consent form Ability to comply with the study protocol in the investigator’s judgment ECOG performance status of 0 or 1 Life expectancy of ≥12 weeks RRMM for which no established therapy for MM is considered by the investigator to be appropriate or tolerable, or is available to the investigator for clinical use Agreement to provide bone marrow biopsy and aspirate samples ∘ AEs from prior anti-cancer therapy resolved to grade ≤1, with the following exceptions: Any-grade alopecia Peripheral sensory or motor neuropathy resolved to grade ≤2 Measurable disease defined as ≥1 of the following: ∘ Serum monoclonal protein (M-protein) ≥ 0.5 g dl −1 (≥5 g l −1 ) ∘ Urine M-protein ≥ 200 mg per 24 h ∘ SFLC assay: involved SFLCs ≥10 mg dl −1 (≥100 mg l −1 ) and an abnormal SFLC ratio (<0.26 or >1.65) Laboratory values as follows: ∘ Hepatic function Aspartate aminotransferase (AST) and alanine transferase (ALT) ≤ 2.5 × upper limit of normal (ULN) Total bilirubin ≤ 1.5 × ULN; patients with a documented history of Gilbert syndrome and in whom total bilirubin elevations are accompanied by elevated indirect bilirubin are eligible. ∘ Hematologic function (requirement before the first dose of cevostamab) Platelet count ≥ 50,000 mm − 3 without transfusion within 7 days before the first dose ANC ≥ 1,000 mm − 3 Total hemoglobin ≥ 8 g dl −1 Note: Patients may receive supportive care to meet hematologic function eligibility criteria (for example, transfusions, G-CSF and so on) ∘ Creatinine ≤ 2.0 ml dl −1 and creatinine clearance ≥ 30 ml min −1 (either calculated or per 24-h urine collection) ∘ Serum calcium (corrected for albumin) level at or below grade 1 hypercalcemia (patient may receive treatment for hypercalcemia to meet eligibility criteria) For women of childbearing potential: agreement to remain abstinent (refrain from heterosexual intercourse) or use contraceptive measures, as defined below: ∘ Patients treated with cevostamab: women must remain abstinent or use contraceptive methods with a failure rate of <1% per year during the treatment period (including treatment interruptions) and for ≥5 months after the last dose of cevostamab was administered; women must refrain from breastfeeding during the same period ∘ Patients treated with tocilizumab (if applicable): women must remain abstinent or use contraceptive methods with a failure rate of <1% per year during the treatment period and for ≥3 months after the last dose of tocilizumab was administered. Women must refrain from breastfeeding during the same period ∘ A woman is considered to be of childbearing potential if she is postmenarcheal, has not reached a post-menopausal state (≥12 continuous months of amenorrhea with no identified cause other than menopause) and has not undergone surgical sterilization (removal of ovaries and/or uterus) ∘ Examples of contraceptive methods with a failure rate of <1% per year include bilateral tubal ligation, male sterilization, hormonal contraceptives that inhibit ovulation, hormone-releasing intrauterine devices and copper intrauterine devices. For men: agreement to remain abstinent (refrain from heterosexual intercourse) or use a condom, and agreement to refrain from donating sperm, as defined below: ∘ With female partners of childbearing potential or pregnant female partners, men must remain abstinent or use a condom during the treatment period and for ≥2 months after the last dose of tocilizumab (if applicable) to avoid exposing the embryo. Men must refrain from donating sperm during this same period.

Detailed exclusion criteria

Inability to comply with protocol-mandated hospitalization and activity restrictions Pregnant or breastfeeding, or intending to become pregnant during the study or within 5 months after the last dose of cevostamab or within 3 months after the last dose of tocilizumab (if applicable) ∘ Women of childbearing potential must have a negative serum pregnancy test result within 7 days before the initiation of the study drug Prior use of any monoclonal antibody, radioimmunoconjugate or ADC as anti-cancer therapy within 4 weeks before first cevostamab infusion Prior treatment with systemic immunotherapeutic agents, including, but not limited to, cytokine therapy and anti-CTLA4, anti-PD-1 and anti-PD-L1 therapeutic antibodies, within 12 weeks or 5 half-lives of the drug, whichever is shorter, before first cevostamab infusion Prior treatment with CAR T cell therapy within 12 weeks before first cevostamab infusion Known treatment-related, immune-mediated AEs associated with prior immunotherapeutic agents as follows: ∘ Prior PD-L1/PD-1 or CTLA4 inhibitor: grade ≥3 AEs with the exception of grade 3 endocrinopathy managed with replacement therapy ∘ Grade 1 or 2 AEs that did not resolve to baseline after treatment discontinuation Treatment with radiotherapy, any chemotherapeutic agent or treatment with any other anti-cancer agent (investigational or otherwise) within 4 weeks or 5 half-lives of the drug, whichever is shorter, before the first cevostamab infusion Autologous stem cell transplantation within 100 days before the first cevostamab infusion Prior allogeneic stem cell transplantation Absolute plasma cell count exceeding 500 µl −1 or 5% of the peripheral blood white cells Prior solid organ transplantation History of autoimmune disease, including, but not limited to, myasthenia gravis, myositis, autoimmune hepatitis, systemic lupus erythematosus, rheumatoid arthritis, inflammatory bowel disease, vascular thrombosis associated with antiphospholipid syndrome, Wegener’s granulomatosis, Sjögren’s syndrome, Guillain–Barré syndrome, multiple sclerosis, vasculitis or glomerulonephritis ∘ Patients with a history of autoimmune-related hypothyroidism on a stable dose of thyroid replacement hormone may be eligible for this study Patients with history of confirmed progressive multifocal leukoencephalopathy. Known history of HLH or macrophage activation syndrome History of severe allergic or anaphylactic reactions to monoclonal antibody therapy (or recombinant antibody-related fusion proteins) Patients with known history of amyloidosis (for example, positive Congo Red stain or equivalent in tissue biopsy) Patients with lesions in the proximity of vital organs that may develop sudden decompensation or deterioration in the setting of a tumor flare History of other malignancy that could affect compliance with the protocol or interpretation of results ∘ Patients with a history of curatively treated basal or squamous cell carcinoma of the skin or in situ carcinoma of the cervix are allowed ∘ Patients with a malignancy that has been treated with curative intent will also be allowed if the malignancy has been in remission without treatment for ≥2 years before the first cevostamab infusion. Current or past history of CNS disease, such as stroke, epilepsy, CNS vasculitis, neurodegenerative disease or CNS involvement by MM ∘ Patients with a history of stroke who have not experienced a stroke or transient ischemic attack in the past 2 years and have no residual neurologic deficits as judged by the investigator are allowed ∘ Patients with a history of epilepsy who have had no seizures in the past 2 years while not receiving any anti-epileptic medications are allowed Significant cardiovascular disease (such as, but not limited to, New York Heart Association Class III or intravenous cardiac disease, myocardial infarction within the last 6 months, uncontrolled arrhythmias or unstable angina) that may limit a patient’s ability to adequately respond to a CRS event Symptomatic active pulmonary disease or requiring supplemental oxygen Known active bacterial, viral, fungal, mycobacterial, parasitic or other infection (excluding fungal infections of nail beds) at study enrollment, or any major episode of infection requiring treatment with intravenous antibiotics within 14 days before the first cevostamab infusion. Positive and quantifiable Epstein–Barr virus (EBV) PCR or cytomegalovirus PCR before the first study treatment Known or suspected chronic active EBV infection ∘ Guidelines for diagnosing chronic active EBV infection provided by Okano et al. 28 Recent major surgery within 4 weeks before the first cevostamab infusion ∘ Protocol-mandated procedures (for example, bone marrow biopsies) are permitted. Positive serologic or PCR test results for acute or chronic HBV infection ∘ Patients whose HBV infection status cannot be determined by serologic test results ( www.cdc.gov/hepatitis/hbv/pdfs/serologicchartv8.pdf ) must be negative for HBV by PCR to be eligible for study participation. Acute or chronic hepatitis C virus (HCV) infection ∘ Patients who are positive for HCV antibody must be negative for HCV by PCR to be eligible for study participation. Known history of HIV seropositivity Any episode of active, symptomatic COVID-19 infection, or requiring treatment with intravenous antivirals for COVID-19 (not including COVID-19 primary prophylaxis) within 14 days, before the first study treatment Patients with symptomatic COVID-19 infection must have two negative antigen tests at least 1 day apart before the first study treatment Administration of a live, attenuated vaccine within 4 weeks before the first cevostamab infusion or anticipation that such a live attenuated vaccine will be required during the study ∘ Influenza vaccination should be given during influenza season (approximately October to May in the Northern Hemisphere and approximately May to October in the Southern Hemisphere). Patients must not receive live, attenuated influenza vaccine (for example, FluMist) at any time during the study treatment period. ∘ SARS-CoV-2 vaccines may be given in accordance with the approved or authorized vaccine label and official or local immunization guidance. SARS-CoV-2 vaccines must not be administered within 1 week before the first study treatment or during cycle 1. ∘ Investigators should review the vaccination status of potential study patients being considered for this study and follow the U.S. Centers for Disease Control and Prevention guidelines for adult vaccination with any other non-live vaccines intended to prevent infectious diseases before the study. Received systemic immunosuppressive medications (including, but not limited to, cyclophosphamide, azathioprine, methotrexate, thalidomide and anti-tumor necrosis factor agents), with the exception of corticosteroid treatment ≤10 mg d −1 prednisone or equivalent within 2 weeks before the first dose of cevostamab and, if applicable, tocilizumab premedication before the first dose of cevostamab ∘ Patients who received acute, low-dose, systemic immunosuppressant medications (for example, a single dose of dexamethasone for nausea) may be enrolled in the study after discussion with and approval of the medical monitor. ∘ The use of inhaled corticosteroids is permitted. ∘ The use of mineralocorticoids for management of orthostatic hypotension is permitted. ∘ The use of physiologic doses of corticosteroids for management of adrenal insufficiency is permitted. History of illicit drug or alcohol abuse within 12 months before screening, in the investigator’s judgment Any medical condition or abnormality in clinical laboratory tests that, in the investigator’s or medical monitor’s judgment, precludes the patient’s safe participation in and completion of the study, or which could affect compliance with the protocol or interpretation of results

Treatment

Cevostamab (intravenous infusion) was initiated with step-up dosing for CRS mitigation and continued at the target dose Q3W for 17 cycles unless disease progression or unacceptable toxicity occurred (Extended Data Fig. 1 ). In the dose-escalation stage, escalating step-up and target doses, and alternative step-up schedules, were evaluated in a standard 3 + 3 design (see the protocol for full details). In the dose-expansion stage, dosing regimens showing promising safety and activity were selected for further evaluation. Corticosteroid premedication (20 mg dexamethasone intravenous or 80 mg methylprednisolone intravenous) was initially mandated before each cevostamab administration for the mitigation of CRS. This requirement was subsequently updated to corticosteroid premedication before each administration in cycle 1 and cycle 2, with the option to discontinue corticosteroid premedication from cycle 3 onwards in patients who do not experience CRS with their previous administration. Premedication with acetaminophen or paracetamol (500–1,000 mg, oral) and diphenhydramine (25–50 mg, oral) was also required before each administration. Patients were hospitalized for ≥48 h after each cevostamab infusion in cycle 1. Anti-infective prophylaxis, including intravenous immunoglobulin, was per institutional practice and was not protocol mandated.

Objectives and endpoints

Primary objectives were to evaluate safety, including characterization of DLTs (see below for information regarding the DLT assessment windows and DLT definition) and determination of the MTD (see the protocol for the dose-escalation rules, including the criteria for MTD determination), and to identify an RP2D for cevostamab monotherapy. Secondary objectives were to evaluate pharmacokinetics, immunogenicity and activity.

Safety was assessed by evaluating the incidence and severity of AEs, graded by National Cancer Institute criteria 29 . CRS was initially graded by Lee criteria 18 and was subsequently graded by American Society of Transplantation and Cellular Therapy criteria 30 , and is reported according to the latter. Re-grading was carried out algorithmically using statistical programming based on data on signs and symptoms and interventions collected in the electronic case report form.

Pharmacokinetics were assessed by measuring the concentration of cevostamab in serum samples. Immunogenicity was assessed by detecting ADAs in serum samples using validated homogeneous bridging enzyme-linked immunosorbent assays following a tiered approach (screening, confirmatory, titer). Confirmed positive samples were also analyzed in a cell-based neutralizing antibody assay using bead extraction with acid dissociation pretreatment. Immunogenicity samples were obtained at baseline; before dosing in cycles 2, 3 and 4; and subsequently at every even-numbered cycle (cycles 6–16). Additional samples were collected at treatment discontinuation or early termination, as well as during follow-up visits.

Activity endpoints were the rate of response and DOR, defined as time from first partial response or better to disease progression or death from any cause (whichever occurs first), as determined by investigators according to IMWG criteria 30 .

Minimal residual disease negativity was an exploratory endpoint that was evaluated at the <10 −5 level by next-generation sequencing in study samples collected at the time of suspected complete response in patients who achieved a VGPR or better. Incomplete data are not reported. FcRH5 expression level on myeloma cells collected at baseline by bone marrow aspiration was an exploratory endpoint that was assessed by flow cytometry. The relationship between FcRH5 expression and 1q abnormal status, and FcRH5 expression and response, was assessed in an analysis that was not prespecified.

PFS, defined as time from first exposure to treatment to disease progression or death, was assessed in a post hoc analysis.

DLT assessment windows

Single step-up dose-escalation cohorts

Two DLT assessment windows will be used as follows: The first DLT assessment window (window 1, step-up DLT window) will consist of the period of time between C1D1 and the initiation of the cevostamab infusion on C1D8. The second DLT assessment window (window 2, target-dose DLT window) will be defined as a period of 14 days following the initiation of the C1D8 infusion.

Double step-up dose-escalation cohorts

The DLT assessment windows will be used as follows: Each step-up dose will have a DLT assessment window defined as a period of 7 days following the initiation of the step-up dose. If the C1D1 or D8 step dose is less than or equal to a previously cleared C1D1 or D8 dose, respectively, in either the single step-up or double step-up dose-escalation cohorts, the DLT assessment window is not required. The target dose DLT assessment window will be defined as a period of 7 days following the initiation of the target dose cevostamab infusion.

Triple step-up dose-escalation cohorts

The DLT assessment windows will be used as follows: DLT window 1 will be from C1D1 to C1D7 and includes the first two step doses. DLT window 2 will be from C1D8 to the end of C1 and includes the third step dose and the first target dose.

DLT definition

A DLT will be defined as any of the following AEs occurring during the DLT assessment windows: Any grade 4 or 5 AE not considered by the investigator to be attributable to another clearly identifiable cause, with the following exceptions: Grade 4 lymphopenia, which is an expected outcome of therapy Grade 4 neutropenia that is not accompanied by temperature elevation (oral or tympanic temperature of ≥100.4 °F (38 °C)) and improves to grade ≤2 (or to ≥80% of the baseline ANC, whichever is lower) within 1 week with or without G-CSF Grade 4 thrombocytopenia that improves to grade ≤2 (or to ≥80% of the baseline platelet count, whichever is lower) within 1 week, does not require platelet transfusion or is not associated with bleeding. Any grade 3 hematologic AE not considered by the investigator to be attributable to another clearly identifiable cause, with the following exceptions: Grade 3 lymphopenia, which is an expected outcome of therapy Grade 3 neutropenia that is not accompanied by temperature elevation (oral or tympanic temperature of ≥100.4 °F (38 °C)) and improves to grade ≤2 (or to ≥80% of the baseline ANC, whichever is lower) within 1 week with or without G-CSF Grade 3 thrombocytopenia that improves to grade ≤2 (or to ≥80% of the baseline platelet count, whichever is lower) within 1 week without platelet transfusion and is not associated with bleeding that is considered clinically significant by the investigator. Any grade 3 non-hematologic AE not considered by the investigator to be attributable to another clearly identifiable cause, with the following exceptions: Grade 3 nausea or vomiting in the absence of premedication or that can be managed with resulting resolution to grade ≤2 with oral or intravenous anti-emetics within 24 h Grade 3 nausea or vomiting that requires total parenteral nutrition or hospitalization are not excluded and should be considered a DLT Grade 3 fatigue lasting ≤3 days Grade 3 laboratory abnormalities that are asymptomatic and resolve to grade ≤1 or baseline within 7 days Any hepatic function abnormality as defined by the following: AST or ALT > 3 × the ULN and total bilirubin >2 × ULN, with the following exception: Any AST or ALT > 3 × the ULN and total bilirubin >2 × ULN in which no individual laboratory value exceeds grade 3 that occurs in the context of grade ≤2 CRS (as defined by previously established criteria 30 ) and resolves to grade ≤1 within <3 days will not be considered a DLT Any grade 3 AST or ALT elevation with the following exception: Any grade 3 AST or ALT elevation that occurs in the context of grade ≤2 CRS (as defined by previously established criteria 30 ) and resolves to grade ≤1 within <3 days will not be considered a DLT. Any grade 2 neurologic toxicity mapping to a MedDRA High-Level Group Term listed in Appendix 11 that is not considered by the investigator to be attributable to another clearly identifiable cause and that does not resolve to baseline within 72 h will be considered a DLT. Grade 1 depressed level of consciousness or grade 1 dysarthria that is not considered by the investigator to be attributable to another clearly identifiable cause and that does not resolve to baseline within 72 h will be considered a DLT. Any-grade seizure that is not considered by the investigator to be attributable to another clearly identifiable cause will be considered a DLT. Decreases in B cells, lymphopenia and/or leukopenia due to decreases in B cells or T cells will not be considered DLTs as they are expected pharmacodynamic (PD) outcomes of cevostamab treatment based on nonclinical testing of this molecule.

Treatment assignment and statistical methods

Patients were identified for potential recruitment by the investigators using prescreening enrollment logs and institutional databases. Patients were assigned to open study cohorts in the order they were enrolled. Planned enrollment was ~180 patients in the dose-escalation stage and ≥30 patients per dosing regimen in the dose-expansion stage (see the protocol for full details). Notably, 30 patients is a reasonable size to obtain an estimate of the response rate; with a sample size of 30 and an observed response rate of 20%, the 90% CI for the true response rate is (8%, 32%) and the expected number of responses is six. Thirty patients is also a reasonable size to obtain an estimate of the rate of grade ≥2 CRS events; with a sample size of 30 and an observed grade ≥2 CRS rate of 27%, the 90% CI for the true grade ≥2 CRS rate is (13%, 40%) and the expected number of grade ≥2 CRS events is eight.

The Medidata RAVE electronic data capture system was used to capture, manage, clean and report the trial data. Descriptive statistics were used to describe and summarize the data in patients who received at least one dose of cevostamab. Tables were generated using SAS version 9.4. Figures were generated using R version 4.4.3. Patients with missing or no response assessment were classified as nonresponders. No formal hypothesis testing was planned. Sex was based on self-reported demographics obtained at screening.

An internal safety committee was used during the study to make recommendations regarding study conduct on the basis of trial safety data to ensure enhanced patient safety monitoring while receiving study treatment. The internal safety committee consisted of the study medical monitor, safety scientist, biomarker scientist and biostatistician, and a medical monitor who was not associated with the study.

Protocol deviations

Investigators were asked to document and promptly report to the sponsor, institutional review board and ethics committee any protocol deviations that might have had an impact on patient safety and/or data integrity. Protocol deviations were reviewed by the sponsor to assess whether any represented a serious breach of good clinical practice guidelines and required reporting to health authorities. Prospective requests to deviate from the protocol were not allowed. Corrective measures are identified in the summary of protocol amendments, which has been included in Supplementary Information .

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Online content

Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at 10.1038/s41591-026-04522-3.

Supplementary information

Supplementary Information Supplementary Table 1, supplementary tables listing all adverse events, study protocol and amendment history. Reporting Summary

Extended data

Extended Data Fig. 1 Cevostamab dose-escalation and dose-expansion cohorts. Summary overview of the cevostamab doses and dosing schedules evaluated in the dose-escalation and dose-expansion cohorts. Cevostamab was administered by intravenous infusion in all cohorts (initial step-up doses and target dose in Cycle 1: 4 hours ± 15 minutes; subsequent target doses in the absence of signs or symptoms of infusion-related reactions or CRS: 2 hours). * Or Day 3 or Day 4 depending on the emergence and resolution of CRS with the previous administration. † Or Day 10 or Day 11 depending on the emergence and resolution of CRS with the previous administration. ‡ All doses in mg. § Dose-expansion cohorts that were expanded from dose-escalation cohorts. ¶ Dose-expansion cohort.

Extended Data Fig. 2 CRS in the Cycle 1 0.3/1.2/3.6 mg triple step-up dosing cohort. Patients (%) with CRS in the Cycle 1 0.3/1.2/3.6 mg triple step-up dosing cohort at the 160 mg target-dose level by dose and grade. * CRS occurred in 3 patients in C2, 3 patients in C3, 1 patient in C5 and 1 patient in C6. Gr 2 CRS occurred in 1 patient in C2 only. † Or D3 or D4 depending on the emergence of CRS with the previous administration. ‡ Or D10 or D11; data cut-off: August 22, 2024.

Extended Data Fig. 3 Relationship between baseline FcRH5 expression and best overall response and 1q cytogenetic status. Relationship between baseline FcRH5 expression on myeloma cells in bone marrow aspirate assessed centrally by flow cytometry among patients who received treatment at the 160 mg dose level and ( A ) best overall response as determined by investigators according to International Myeloma Working Group criteria and ( B ) 1q cytogenetic status as assessed centrally by fluorescence in situ hybridization (FISH). Bone marrow aspirates that were considered hemodiluted by the study pathologists or did not have enough tumor events ( < 100 events) were not included in the analysis. Boxplots represent the median, with lower and upper hinges corresponding to the first and third quartiles (the 25th and 75th percentiles). In the analysis of the relationship between baseline FcRH5 expression and 1q cytogenetic status (Extended Data Fig. 3b), statistical significance was evaluated using the two-sided Mann-Whitney U test. No adjustments for multiple comparisons were performed. amp, amplification; MESF, molecules of equivalent soluble fluorochrome; MMPCs, multiple myeloma plasma cells.

Extended Data Fig. 4 Progression-free survival. Progression-free survival among patients who received cevostamab at the 160 mg target-dose level along with 95% confidence bands.

Extended Data Table 1 Anti-tumor activity in all patients who received cevostamab at the 160 mg target-dose level and in all patients

Extended Data Table 2 Rates of objective response and very good partial response or better in all patients and by target-dose level

Extended Data Table 3 Grade 5 adverse events excluding disease progression in all patients who received cevostamab at the 160 mg target-dose level

Extended Data Table 4 Adverse events leading to cevostamab discontinuation excluding plasma cell myeloma progression in all patients who received cevostamab at the 160 mg target-dose level

Extended Data Table 5 Adverse events of infection regardless of attribution in all patients who received cevostamab at the 160 mg target-dose level and in all patients

Extended Data Table 6 Baseline characteristics in patients with and without prior BCMA-targeted therapy who received cevostamab at the 160 mg target-dose level

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Adam D. Cohen, Joshua Richter, Suzanne Trudel, Alexander Lesokhin.

Extended data

is available for this paper at 10.1038/s41591-026-04522-3.

Supplementary information

The online version contains supplementary material available at 10.1038/s41591-026-04522-3.

Acknowledgements

Third-party medical writing assistance, under the direction of all authors, was provided by S. Malkin of Ashfield MedComms, an Inizio company, and was funded by F. Hoffmann-La Roche Ltd.

Author contributions

Conception and design: A.D.C., J.R., S.T., T.T., D.S., W.L., R.N., V.C., C.W., J.C. and A.K. Provision of study materials or patients: all authors. Collection and assembly of data: T.T., D.S., W.L., R.N., V.C., C.W. and J.C. Data analysis and interpretation: all authors. Writing and final approval of the paper: all authors. Accountable for all aspects of the work: all authors.

Peer review

Peer review information

Nature Medicine thanks Benjamin Diamond, Carsten Müller-Tidow and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Saheli Sadanand, in collaboration with the Nature Medicine team.

Funding

The GO39775 study was sponsored by F. Hoffmann-La Roche Ltd/Genentech, Inc. Employees of the sponsor were involved in the design and conduct of the study; the collection, management, analysis and interpretation of the data; and the preparation of the paper.

Data availability

For eligible studies, qualified researchers may request access to individual patient-level clinical data via the cross-industry request site https://vivli.org . Full details of Roche’s approach to sharing individual patient-level clinical data via the Vivli site can be found at https://vivli.org/ourmember/roche . In summary, access requests are assessed by an independent review panel managed by Vivli. The panel considers the scientific merit of each application and compliance with patient consent and data-sharing policies, with the timeline for request approval (usually 2–5 months) influenced by the number of data contributors, the number of studies and the requestor’s availability to respond to comments. Once approved, the data are anonymized and made available for a minimum of 12 months. For up-to-date details on Roche’s Global Policy on the Sharing of Clinical Information and how to request access to related clinical study documents, see https://go.roche.com/data_sharing . Anonymized records for individual patients across more than one data source external to Roche cannot, and should not, be linked due to a potential increase in risk of patient reidentification.

Competing interests

A.D.C. discloses consultancy (AbbVie, Arcellx, AstraZeneca, BMS, Caribou, Genentech/Roche, GSK, iTeos, Janssen, Kite, Legend, Moderna, Pfizer, Prothena, Regeneron, Sanofi), research funding (Genentech/Roche, GSK, Janssen, Novartis), patents and royalties (Novartis) and advisory committees (Ichnos, Novartis). J.R. discloses consultancy (AbbVie, BMS, Forus, Genentech, Janssen, Karyopharm, Kite/Arcellx, Menarini, Pfizer, Regeneron, Sanofi, Takeda, UB Therapeutics) and speaker’s bureau (Adaptive Biotechnologies, BMS, Janssen, Pfizer, Sanofi). S.T. discloses consultancy (GSK), research funding (BMS, Genentech, GSK, Janssen, K36 Therapeutics, Pfizer, Roche), honoraria (AstraZeneca, GSK, Janssen) and advisory committees (AstraZeneca, BMS, GSK, Janssen, Kite/Arcellx, Pfizer, Roche, Sanofi). A.L. discloses employment (Memorial Sloan Kettering Cancer Center), consultancy (Arcellx, Janssen, Pfizer), honoraria (Pfizer) and patents and royalties (Serametrix). J.P.L. has no conflicts of interest or relevant competing interests. S.K.T. discloses research funding (AbbVie, Ascentage Pharma, BMS, Cellectar Biosciences, Genentech, Janssen, Sanofi). S.J.H. discloses research funding (Johnson & Johnson), honoraria (Celgene, Janssen, Johnson & Johnson, Kite/Gilead, Novartis), advisory committees (Celgene, Haemalogix, Janssen, Kite/Gilead, Novartis) and other interests (MSAG chair Myeloma Australia). L.J.C. discloses consultancy (AbbVie, Amgen, AstraZeneca, BMS, Caribou, Genentech, Johnson & Johnson, Pfizer, Sanofi), research funding (AbbVie, Amgen, AstraZeneca, BMS, Caribou, Genentech, Johnson & Johnson) and honoraria (AbbVie, Adaptive Biotechnologies, Amgen, AstraZeneca, BMS, Caribou, Genentech, Johnson & Johnson, Pfizer, Sanofi). A.S. discloses consultancy (BMS, Celltrion, Haemalogix, Janssen, Rheumagen), research funding (AbbVie, Antengene, Haemalogix, Janssen), honoraria (AbbVie, Antengene, BMS, Haemalogix, Janssen, Pfizer, Roche, Sanofi, Thermo Fisher), advisory committees (Antengene, BMS, Haemalogix, Janssen, Pfizer, Roche, Sanofi) and speaker’s bureau (BMS, Janssen, Pfizer, Sanofi, Thermo Fisher). R.F. discloses consultancy (AbbVie, Adaptive, Amgen, Apple, BMS/Celgene, GSK, Janssen, Karyopharm, Pfizer, RA Capital, Regeneron, Sanofi), equity (Antengene), patents and royalties (FISH in MM), and advisory committees (Caris Life Sciences). P.A.F. discloses consultancy (AbbVie, BMS, GSK, Legend, Pfizer, Sanofi), research funding (Johnson & Johnson), advisory committees (Johnson & Johnson, Pathos) and speaker’s bureau (Sanofi). J.G.B. discloses consultancy (AstraZeneca, BMS, Galapagos, Johnson & Johnson, Karyopharm, Kite, Kyowa Kirin, Pfizer, Regeneron, Roche, Sebia), research funding (AstraZeneca, BMS, C4 Therapeutics, Caribou Biosciences, CARsgen, Cartesian Therapeutics, Genentech, Gracell, GSK, Ichnos Sciences, Johnson & Johnson, Juno Therapeutics, K36 Therapeutics, Karyopharm, Kite, Moderna, Pfizer, Regeneron, Roche, Sanofi) and honoraria (Janssen). R.K. discloses research funding (Pfizer), honoraria (Johnson & Johnson, Pfizer, Sanofi) and advisory committees (Forus, GSK, Johnson & Johnson, Pfizer, Sanofi). N.J.B. discloses consultancy (AbbVie, BMS, Genentech, Johnson & Johnson, Pfizer, Roche, Sanofi), honoraria (AbbVie, BMS, Genentech, Johnson & Johnson, Pfizer, Roche, Sanofi) and advisory committees (AbbVie, BMS, Genentech, Johnson & Johnson, Pfizer, Roche, Sanofi). P.R.-O. discloses consultancy (AbbVie, AstraZeneca, BMS, H3 Biomedicine, GSK, Johnson & Johnson, Menarini Stemline, Oncopeptides, Pfizer, Regeneron, Roche, Sanofi), research funding (Pfizer), honoraria (BMS, GSK, Johnson & Johnson, Pfizer), advisory committees (BMS, Johnson & Johnson, Regeneron) and other interests (Pfizer). M.-V.M. discloses consultancy (Amgen, Johnson & Johnson, Kite, Menarini Stemline, Roche) and honoraria (Amgen, BMS, GSK, Johnson & Johnson, Kite, Menarini Stemline, Oncopeptides, Pfizer, Regeneron, Roche, Sanofi). T.T. discloses employment (Roche/Genentech) and equity (Gilead, Roche/Genentech). D.S. discloses employment (Roche/Genentech) and equity (Roche/Genentech). W.L. discloses employment (Roche/Genentech). R.N. discloses employment (Roche/Genentech) and equity (Roche/Genentech). V.C. discloses employment (Roche/Genentech). C.W. discloses past employment (Roche/Genentech). J.C. discloses past employment (Roche/Genentech) and equity (Roche/Genentech). A.K. discloses consultancy (AbbVie, Adaptive, Arcellx, BMS, GSK, Janssen, Roche, Sanofi), equity (Arcellx, BMX) and advisory committees (Johnson & Johnson, Pfizer).

References

1. Gandhi UH Outcomes of patients with multiple myeloma refractory to CD38-targeted monoclonal antibody therapy Leukemia 2019 33 2266 2275 10.1038/s41375-019-0435-7 30858549 PMC6820050 Gandhi, U. H. et al. Outcomes of patients with multiple myeloma refractory to CD38-targeted monoclonal antibody therapy. Leukemia 33 , 2266–2275 (2019). 30858549 10.1038/s41375-019-0435-7 PMC6820050

2. Mateos M-V LocoMMotion: a study of real-life current standards of care in triple-class exposed patients with relapsed/refractory multiple myeloma—2-year follow-up (final analysis) Leukemia 2024 38 2554 2560 10.1038/s41375-024-02404-6 39322709 PMC11588650 Mateos, M.-V. et al. LocoMMotion: a study of real-life current standards of care in triple-class exposed patients with relapsed/refractory multiple myeloma—2-year follow-up (final analysis). Leukemia 38 , 2554–2560 (2024). 39322709 10.1038/s41375-024-02404-6 PMC11588650

3. Berdeja JG Ciltacabtagene autoleucel, a B-cell maturation antigen-directed chimeric antigen receptor T-cell therapy in patients with relapsed or refractory multiple myeloma (CARTITUDE-1): a phase 1b/2 open-label study Lancet 2021 398 314 324 10.1016/S0140-6736(21)00933-8 34175021 Berdeja, J. G. et al. Ciltacabtagene autoleucel, a B-cell maturation antigen-directed chimeric antigen receptor T-cell therapy in patients with relapsed or refractory multiple myeloma (CARTITUDE-1): a phase 1b/2 open-label study. Lancet 398 , 314–324 (2021). 34175021 10.1016/S0140-6736(21)00933-8

4. Chari A Talquetamab, a T-cell-redirecting GPRC5D bispecific antibody for multiple myeloma N. Engl. J. Med. 2022 387 2232 2244 10.1056/NEJMoa2204591 36507686 Chari, A. et al. Talquetamab, a T-cell-redirecting GPRC5D bispecific antibody for multiple myeloma. N. Engl. J. Med. 387 , 2232–2244 (2022). 36507686 10.1056/NEJMoa2204591

5. Lesokhin AM Elranatamab in relapsed or refractory multiple myeloma: phase 2 MagnetisMM-3 trial results Nat. Med. 2023 29 2259 2267 10.1038/s41591-023-02528-9 37582952 PMC10504075 Lesokhin, A. M. et al. Elranatamab in relapsed or refractory multiple myeloma: phase 2 MagnetisMM-3 trial results. Nat. Med. 29 , 2259–2267 (2023). 37582952 10.1038/s41591-023-02528-9 PMC10504075

6. Moreau P Teclistamab in relapsed or refractory multiple myeloma N. Engl. J. Med. 2022 387 495 505 10.1056/NEJMoa2203478 35661166 PMC10587778 Moreau, P. et al. Teclistamab in relapsed or refractory multiple myeloma. N. Engl. J. Med. 387 , 495–505 (2022). 35661166 10.1056/NEJMoa2203478 PMC10587778

7. Munshi NC Idecabtagene vicleucel in relapsed and refractory multiple myeloma N. Engl. J. Med. 2021 384 705 716 10.1056/NEJMoa2024850 33626253 Munshi, N. C. et al. Idecabtagene vicleucel in relapsed and refractory multiple myeloma. N. Engl. J. Med. 384 , 705–716 (2021). 33626253 10.1056/NEJMoa2024850

8. Lee H Mechanisms of antigen escape from BCMA- or GPRC5D-targeted immunotherapies in multiple myeloma Nat. Med. 2023 29 2295 2306 10.1038/s41591-023-02491-5 37653344 PMC10504087 Lee, H. et al. Mechanisms of antigen escape from BCMA- or GPRC5D-targeted immunotherapies in multiple myeloma. Nat. Med. 29 , 2295–2306 (2023). 37653344 10.1038/s41591-023-02491-5 PMC10504087

9. Philipp N T-cell exhaustion induced by continuous bispecific molecule exposure is ameliorated by treatment-free intervals Blood 2022 140 1104 1118 10.1182/blood.2022015956 35878001 PMC10652962 Philipp, N. et al. T-cell exhaustion induced by continuous bispecific molecule exposure is ameliorated by treatment-free intervals. Blood 140 , 1104–1118 (2022). 35878001 10.1182/blood.2022015956 PMC10652962

10. Verkleij CPM T-cell characteristics impact response and resistance to T-cell-redirecting bispecific antibodies in multiple myeloma Clin. Cancer Res. 2024 30 3006 3022 10.1158/1078-0432.CCR-23-3333 38687588 Verkleij, C. P. M. et al. T-cell characteristics impact response and resistance to T-cell-redirecting bispecific antibodies in multiple myeloma. Clin. Cancer Res. 30 , 3006–3022 (2024). 38687588 10.1158/1078-0432.CCR-23-3333

11. Elkins K FcRL5 as a target of antibody–drug conjugates for the treatment of multiple myeloma Mol. Cancer Ther. 2012 11 2222 2232 10.1158/1535-7163.MCT-12-0087 22807577 Elkins, K. et al. FcRL5 as a target of antibody–drug conjugates for the treatment of multiple myeloma. Mol. Cancer Ther. 11 , 2222–2232 (2012). 22807577 10.1158/1535-7163.MCT-12-0087

12. Ise T Elevation of soluble CD307 (IRTA2/FcRH5) protein in the blood and expression on malignant cells of patients with multiple myeloma, chronic lymphocytic leukemia, and mantle cell lymphoma Leukemia 2007 21 169 174 10.1038/sj.leu.2404445 17051241 Ise, T. et al. Elevation of soluble CD307 (IRTA2/FcRH5) protein in the blood and expression on malignant cells of patients with multiple myeloma, chronic lymphocytic leukemia, and mantle cell lymphoma. Leukemia 21 , 169–174 (2007). 17051241 10.1038/sj.leu.2404445

13. Polson AG Expression pattern of the human FcRH/IRTA receptors in normal tissue and in B-chronic lymphocytic leukemia Int. Immunol. 2006 18 1363 1373 10.1093/intimm/dxl069 16849395 Polson, A. G. et al. Expression pattern of the human FcRH/IRTA receptors in normal tissue and in B-chronic lymphocytic leukemia. Int. Immunol. 18 , 1363–1373 (2006). 16849395 10.1093/intimm/dxl069

14. Jiang D Chimeric antigen receptor T cells targeting FcRH5 provide robust tumour-specific responses in murine xenograft models of multiple myeloma Nat. Commun. 2023 14 3642 10.1038/s41467-023-39395-4 37339964 PMC10282100 Jiang, D. et al. Chimeric antigen receptor T cells targeting FcRH5 provide robust tumour-specific responses in murine xenograft models of multiple myeloma. Nat. Commun. 14 , 3642 (2023). 37339964 10.1038/s41467-023-39395-4 PMC10282100

15. Li J Membrane-proximal epitope facilitates efficient T cell synapse formation by anti-FcRH5/CD3 and is a requirement for myeloma cell killing Cancer Cell 2017 31 383 395 10.1016/j.ccell.2017.02.001 28262555 PMC5357723 Li, J. et al. Membrane-proximal epitope facilitates efficient T cell synapse formation by anti-FcRH5/CD3 and is a requirement for myeloma cell killing. Cancer Cell 31 , 383–395 (2017). 28262555 10.1016/j.ccell.2017.02.001 PMC5357723

16. Stewart AK Phase I study of the anti-FcRH5 antibody–drug conjugate DFRF4539A in relapsed or refractory multiple myeloma Blood Cancer J. 2019 9 17 10.1038/s41408-019-0178-8 30718503 PMC6362066 Stewart, A. K. et al. Phase I study of the anti-FcRH5 antibody–drug conjugate DFRF4539A in relapsed or refractory multiple myeloma. Blood Cancer J. 9 , 17 (2019). 30718503 10.1038/s41408-019-0178-8 PMC6362066

17. Kumar S International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma Lancet Oncol. 2016 17 e328 e346 10.1016/S1470-2045(16)30206-6 27511158 Kumar, S. et al. International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma. Lancet Oncol. 17 , e328–e346 (2016). 27511158 10.1016/S1470-2045(16)30206-6

18. Lee DW Current concepts in the diagnosis and management of cytokine release syndrome Blood 2014 124 188 195 10.1182/blood-2014-05-552729 24876563 PMC4093680 Lee, D. W. et al. Current concepts in the diagnosis and management of cytokine release syndrome. Blood 124 , 188–195 (2014). 24876563 10.1182/blood-2014-05-552729 PMC4093680

19. Mailankody S GPRC5D-targeted CAR T cells for myeloma N. Engl. J. Med. 2022 387 1196 1206 10.1056/NEJMoa2209900 36170501 PMC10309537 Mailankody, S. et al. GPRC5D-targeted CAR T cells for myeloma. N. Engl. J. Med. 387 , 1196–1206 (2022). 36170501 10.1056/NEJMoa2209900 PMC10309537

20. Ferreri CJ Real-world experience of patients with multiple myeloma receiving ide-cel after a prior BCMA-targeted therapy Blood Cancer J. 2023 13 117 10.1038/s41408-023-00886-8 37558706 PMC10412575 Ferreri, C. J. et al. Real-world experience of patients with multiple myeloma receiving ide-cel after a prior BCMA-targeted therapy. Blood Cancer J. 13 , 117 (2023). 37558706 10.1038/s41408-023-00886-8 PMC10412575

21. Tan CR Real-world evaluation of teclistamab for the treatment of relapsed/refractory multiple myeloma (RRMM): an International Myeloma Working Group Study Blood Cancer J. 2025 15 53 10.1038/s41408-025-01259-z 40175336 PMC11965530 Tan, C. R. et al. Real-world evaluation of teclistamab for the treatment of relapsed/refractory multiple myeloma (RRMM): an International Myeloma Working Group Study. Blood Cancer J. 15 , 53 (2025). 40175336 10.1038/s41408-025-01259-z PMC11965530

22. Touzeau C Efficacy and safety of teclistamab in patients with relapsed/refractory multiple myeloma after BCMA-targeting therapies Blood 2024 144 2375 2388 10.1182/blood.2023023616 39172760 Touzeau, C. et al. Efficacy and safety of teclistamab in patients with relapsed/refractory multiple myeloma after BCMA-targeting therapies. Blood 144 , 2375–2388 (2024). 39172760 10.1182/blood.2023023616

23. Cohen AD Efficacy and safety of cilta-cel in patients with progressive multiple myeloma after exposure to other BCMA-targeting agents Blood 2023 141 219 230 10.1182/blood.2022015526 36095849 PMC10562529 Cohen, A. D. et al. Efficacy and safety of cilta-cel in patients with progressive multiple myeloma after exposure to other BCMA-targeting agents. Blood 141 , 219–230 (2023). 36095849 10.1182/blood.2022015526 PMC10562529

24. Dima D Outcomes of teclistamab (Tec) in patients with relapsed/refractory multiple myeloma (RRMM) with prior exposure to BCMA-directed therapy (BCMA-DT): a multicenter study from the U.S. Multiple Myeloma Immunotherapy Consortium Blood 2024 144 897 10.1182/blood-2024-203409 PMC12198414 40562770 Dima, D. et al. Outcomes of teclistamab (Tec) in patients with relapsed/refractory multiple myeloma (RRMM) with prior exposure to BCMA-directed therapy (BCMA-DT): a multicenter study from the U.S. Multiple Myeloma Immunotherapy Consortium. Blood 144 , 897 (2024). 10.1038/s41408-025-01314-9 PMC12198414 40562770

25. Raje N Monitoring, prophylaxis, and treatment of infections in patients with MM receiving bispecific antibody therapy: consensus recommendations from an expert panel Blood Cancer J. 2023 13 116 10.1038/s41408-023-00879-7 37528088 PMC10394080 Raje, N. et al. Monitoring, prophylaxis, and treatment of infections in patients with MM receiving bispecific antibody therapy: consensus recommendations from an expert panel. Blood Cancer J. 13 , 116 (2023). 37528088 10.1038/s41408-023-00879-7 PMC10394080

26. Cohen A Phase 2 study of cevostamab consolidation following BCMA CAR T cell therapy: preliminary safety, efficacy, and correlative data from the “STEM” (Sequential T Cell-Engagement for Myeloma) trial Blood 2025 146 699 10.1182/blood-2025-699 Cohen, A. et al. Phase 2 study of cevostamab consolidation following BCMA CAR T cell therapy: preliminary safety, efficacy, and correlative data from the “STEM” (Sequential T Cell-Engagement for Myeloma) trial. Blood 146 , 699 (2025).

27. Harrison S Tumor clearance, T-cell fitness, and minimal residual disease (MRD) outcomes in patients with relapsed/refractory multiple myeloma (RRMM) treated with cevostamab plus pomalidomide and dexamethasone: biomarker analyses from CAMMA 1 Arm B Blood 2025 146 252 10.1182/blood-2025-252 Harrison, S. et al. Tumor clearance, T-cell fitness, and minimal residual disease (MRD) outcomes in patients with relapsed/refractory multiple myeloma (RRMM) treated with cevostamab plus pomalidomide and dexamethasone: biomarker analyses from CAMMA 1 Arm B. Blood 146 , 252 (2025).

28. Okano M Proposed guidelines for diagnosing chronic active Epstein–Barr virus infection Am. J. Hematol. 2005 80 64 69 10.1002/ajh.20398 16138335 Okano, M. et al. Proposed guidelines for diagnosing chronic active Epstein–Barr virus infection. Am. J. Hematol. 80 , 64–69 (2005). 16138335 10.1002/ajh.20398

29. National Cancer Institute Common Terminology Criteria for Adverse Events v4.0 (National Cancer Institute, 2009); https://dctd.cancer.gov/research/ctep-trials/for-sites/adverse-events/ctcae-v4.pdf

30. Lee DW ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells Biol. Blood Marrow Transplant 2019 25 625 638 10.1016/j.bbmt.2018.12.758 30592986 PMC12180426 Lee, D. W. et al. ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells. Biol. Blood Marrow Transplant 25 , 625–638 (2019). 30592986 10.1016/j.bbmt.2018.12.758 PMC12180426