How Effective is Fixed-Duration Cevostamab in Multiple Myeloma?

In a conversation with CancerNetwork, Adam D. Cohen, MD, director of Myeloma Immunotherapy and professor of medicine at the Hospital of the University of Pennsylvania, spoke about data from the phase 1 GO39775 trial (NCT03275103) assessing cevostamab, a first-in-class FcRH5/CD3 bispecific antibody, in relapsed or refractory multiple myeloma.¹ At the recommended phase 2 dose of 160 mg every 3 weeks, the objective response rate (ORR) was 44.3% (n = 74/167), with 25.7% (n = 43/167) of patients achieving a very good partial response (VGPR) or better.¹ The trial tested cevostamab as fixed-duration therapy given for 17 cycles over about 12 months.

Cohen discussed the rationale behind the study design, the trial’s efficacy and safety findings, and where the agent may fit as it moves into earlier lines of therapy. Altogether, the data validate FcRH5 as a “great target” for multiple myeloma therapy, according to Cohen.

CancerNetwork: What was the background of this phase 1 trial assessing cevostamab in relapsed or refractory multiple myeloma?

Cohen: The background stems from the fact that despite all these amazing new agents we have for treating multiple myeloma—and we seem to keep getting new ones every year—the majority of patients are still relapsing, unfortunately. Especially in the late-relapse setting, patients have already become refractory to standard proteasome inhibitors, immunomodulatory agents, and CD38 antibodies. If they have had a BCMA-targeted treatment, there are not a lot of great options at that point. There remains this unmet need for these later-line patients to have novel targets or approaches, and that was the background for this study.

This trial assessed an FcRH5/CD3 bispecific antibody. What makes FcRH5 an attractive target to pursue, and how does its biology position cevostamab differently from the existing bispecifics?

FcRH5 appears to be a great target for multiple myeloma. It is expressed on normal plasma cells, which are, of course, the precursors that turn into multiple myeloma, and it is expressed ubiquitously on multiple myeloma cells, although the level of expression can vary from patient to patient. It has limited expression on other normal tissues; it is expressed on a subset of mature B cells, but not elsewhere. That speaks to its good safety profile. In addition, FcRH5 expression appears to be independent of expression of other markers like BCMA or GPRC5D, the targets of some of our other currently available immune therapies, so this drug could work even if patients have lost or mutated those other receptors. That is what makes it an attractive target. Cevostamab is a typical T-cell–engaging bispecific antibody: by targeting CD3, it attaches to T cells by attaching to FcRH5 on the myeloma cells, and it then brings the T cells to the myeloma cells and activates them to kill.

In this trial, patients received cevostamab as fixed-duration therapy. What was the clinical rationale for capping treatment with this agent, and what did the data show about patients maintaining responses off therapy?

At the time this study was designed, in 2017, this was a novel or radical idea. Almost all the trials, especially in late-line relapsed/refractory myeloma, gave continuous therapy. Once patients started treatment, they continued indefinitely unless they progressed or had unacceptable toxicity, and that includes the currently available bispecific antibodies that target BCMA or GPRC5D. What we have learned, subsequently, is that continuous therapy might not be the best approach, particularly for T-cell–engaging therapy. If you continuously activate T cells over and over again indefinitely, that may lead to T-cell dysfunction over time. It certainly seemed to increase the risk of infections, particularly with the BCMA-targeted bispecifics, where you see profound hypogammaglobulinemia and perhaps ongoing T-cell dysfunction.

The rationale was: what if we treat for a fixed duration? In this trial, that was 17 cycles, with the drug given every 3 weeks, so about 51 weeks, or roughly 1 year of treatment. The thought was that if patients were in a good response, you could stop treatment, which might allow recovery of their endogenous immunity and a lower long-term infection risk. It might also lower the risk of antigen escape because you are not keeping constant pressure on the FcRH5 target, potentially meaning fewer patients relapsing with FcRH5-low or -negative disease. That was the rationale at the time. It turned out to be prescient because fixed-duration treatment is now becoming something of a [trend] in [multiple] myeloma therapy. There is this push to try to do more fixed-duration treatments based on some of these same findings with continuous dosing.

In terms of what the data showed for patients who completed all 17 cycles and ended in a response, there were about 26 such patients, and 17 of them were still in an ongoing response at the time of data cutoff. That includes several patients with more than 30 months of ongoing remission; I have one patient who is 4 years out now without any other therapy and still in remission. For a subset of patients, particularly those with the deepest responses, complete responders especially, you may be able to stop therapy. That is what I think this trial helped teach us.

Some of the response rates appear to be lower in patients previously treated with a BCMA-targeted bispecific compared with prior CAR T-cell therapy or antibody-drug conjugates. What does that gradient show about resistance mechanisms and how prior therapy type should factor into selecting a next-line agent like cevostamab?

That is one of the key questions in [multiple] myeloma therapy today: how do we sequence all these agents, particularly all these T-cell–directed agents? Just to summarize, the ORR at the recommended phase 2 dose of 160 mg every 3 weeks was 44.3%, with 25.7% having VGPRs or better. The responses were certainly higher in patients who had not had prior BCMA-targeted therapy—[it was] 60.6% in that setting—compared with lower responses if patients had received a prior BCMA-targeted therapy.

It seemed to be lowest in patients who had a prior bispecific antibody, with a 17.2% ORR in that group vs 36.2% [in those with a prior BCMA-targeted] CAR T-cell product and 43.8% with a prior [BCMA-targeted] antibody-drug conjugate. Take that with a grain of salt because these numbers are relatively small, but perhaps going directly from one bispecific right to the next, especially while progressing on the last one, may not be the optimal way to sequence these therapies. In this trial, because it was designed in 2017, only a 4-week washout was required between bispecific antibodies, so some patients progressing on teclistamab-cqyv [Tecvayli] or talquetamab-tgvs [Talvey] went right onto cevostamab. Those T cells may have been somewhat dysfunctional or exhausted, contributing to that lower response rate.

Nowadays, we might design the trial with a longer washout period or think about intervening therapy that is not T-cell dependent, such as an immunomodulatory drug or a CELMoD. That might be restorative for immune function before coming back with a second bispecific. I think what this teaches us is that some gap between our T-cell–directed therapies, if possible, would likely allow the second T-cell–directed therapy to work a little better.

Cevostamab did not appear to produce the same dysgeusia, weight loss, or nail and skin toxicities that are associated with GPRC5D-directed agents like talquetamab. How significant is that quality-of-life distinction when you are counseling patients on treatment choice?

That is a great point, and because FcRH5 has a different normal tissue distribution compared with GPRC5D, the target of talquetamab, you do see very different toxicity profiles overall. Cevostamab’s toxicity profile is very favorable. It does have the same cytokine release syndrome and risk of early neurotoxicity that you can see with other bispecifics and CAR T-cell products, but once you get past that initial step-up period, it is very well tolerated, with some cough or allergic-type symptoms, some risk of infections—although that seems lower than what is seen with BCMA-directed agents—and the absence of these other GPRC5D-related toxicities.

Talquetamab and other GPRC5D-directed drugs are great drugs, and they can work when you do not have anything else to reach for, but those toxicities do affect patients. They do not like being on the drug; they can develop weight loss, loss of taste, and skin peeling, and there is a lot of supportive care required with GPRC5D agents. Not having to deal with that is an advantage, and that is where cevostamab may play a nice role, especially if BCMA-directed agents move up into second-line and perhaps even frontline therapy in the future. To have something that could work without those GPRC5D-related toxicities down the road, that could be a nice option for patients.

Given the heavily pretreated patient population of this trial, how do you see this agent’s role evolving as it might move into earlier lines of therapy? How might efficacy and durability change in a less refractory population?

Just to highlight how sick and advanced this population was, there was a median of 6 prior lines, and over half the patients had a prior BCMA-directed therapy. [This was] a tough group to treat, and yet we still saw the efficacy that was reported. This drug is moving forward primarily in combination in less heavily pretreated patients. There have been data in combination with pomalidomide [Pomalyst] and dexamethasone; cevostamab plus pomalidomide/dexamethasone has been reported in a less heavily treated group of patients—maybe 1 to 3 prior lines—with response rates over 80% in preliminary data.² That has led to a phase 3 registration trial called Cevolution [NCT07555938], which has just recently opened worldwide, comparing cevostamab plus pomalidomide and dexamethasone vs investigator’s choice of various agents currently approved in that 1-to-3-prior-line space, to get a sense of the activity in a less heavily treated group of patients in a larger study with this pomalidomide/dexamethasone combination.³

There are also a lot of other interesting combinations being explored with this agent. At Penn, we have an investigator-initiated trial [NCT05801939] giving cevostamab as a consolidation approach after BCMA CAR T-cell therapy: patients get a BCMA CAR T-cell product, and then 10 to 12 weeks later, they start cevostamab.⁴ The idea is that by switching antigen targeting to FcRH5, you might eliminate residual-resistant myeloma cells that have downregulated or lost BCMA, leading to more durable responses. We showed some early preliminary data at the 2025 American Society of Hematology (ASH) Annual Meeting & Exposition that looked promising, and we hope to present updated data this year.

There are also combinations with other bispecifics, a dual-bispecific approach, and combinations with immunomodulatory drugs and CELMoDs. There is also a subcutaneous formulation coming that has shown equivalent preliminary data to the intravenous formulation, which is always a bit easier for patients. I think the next couple of years will tell us where this is going to end up, but it is probably not just going to be for very late-line patients. Hopefully, we will have access to it in less heavily treated patients as well.

What do you hope others take away from the results of this trial?

I think the key takeaways are, one, it validates FcRH5 as a good target for [multiple] myeloma therapy. This is the first trial to do that, and it is clearly an active agent against a good target. Two, fixed-duration therapy, as we have talked about, is a good concept that this trial is helping to establish, and hopefully, that will become the future. We will be able to tailor the duration of therapy to the response, or to the setting, and not have to keep people on therapy forever.

Quality of life is important, especially in a late-line, maybe incurable setting; drugs that do not cause a lot of nuisance or quality-of-life toxicities can be very important. Lastly, this gets back to the sequencing question: how do we optimally use all of these different T-cell–directed agents, and what is the optimal timing? I think we need more studies and more data, and some of these questions may have to be answered in the real-world setting. We encourage centers to pool their data. The International Myeloma Working Group has an immunotherapy database where we are trying to capture how all of these agents are being used in the real world, and those kinds of studies can be important for optimizing sequencing going forward.

References

1. Cohen AD, Richter J, Trudel S, et al. FcRH5×CD3 bispecific antibody cevostamab in relapsed or refractory multiple myeloma: a phase 1 trial. Nat Med. Published online July 22, 2026. doi:10.1038/s41591-026-04522-3
2. Mian HS, Riley CH, Popat R, et al. Cevostamab plus pomalidomide (pom) and dexamethasone (dex) in relapsed/refractory multiple myeloma (RRMM): phase I dose-expansion results from the CAMMA 1 study. Presented at the 22nd Annual International Myeloma Society Meeting and Exposition; September 17-20, 2025; Toronto, Canada. Abstract OA-14.
3. Popat R, Cohen AD, Krishnan A, et al. Cevolution: a randomized phase III trial comparing the efficacy and safety of cevostamab plus pomalidomide and dexamethasone versus standard of care in relapsed/refractory multiple myeloma. Presented at the European Hematology Association 2026 Congress; June 11-14, 2026; Stockholm, Sweden. Abstract 1544.
4. Cohen A, Susanibar-Adaniya S, Garfall 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. 2025;146(suppl 1):699. doi:10.1182/blood-2025-699