A view from the periphery: trial design, commercial competition, and the access gap in multiple myeloma

Introduction

Multiple myeloma research has moved faster in the past five years than in the previous two decades combined. DREAMM-7, DREAMM-8, MajesTEC-3, MajesTEC-9, and the MonumenTAL trials are genuine milestones.^1–5 As a hematologist practicing in Türkiye, I read these results with real admiration. However, I also read them with growing unease.

The problem of trial proliferation without meaningful comparison

One of the most important unanswered questions in relapsed/refractory multiple myeloma is surprisingly simple. How do the new therapies compare with one another? We now have bispecific antibodies against B-cell maturation antigen (BCMA) and G protein-coupled receptor family C group 5 member D (GPRC5D).¹ We have BCMA-directed chimeric antigen receptor (CAR) T-cell therapies.^6,7 We have an antibody-drug conjugate.^2,3 None of these has been tested head-to-head in a randomized phase 3 trial. Not one. This is not solely a question of sponsor interest. CAR T-cell therapies, bispecific antibodies, and antibody-drug conjugates differ in logistics, time to treatment initiation, infrastructure requirements, toxicity profile, route of administration, and the patient populations for whom each is feasible, so a head-to-head comparison is a genuine methodological and practical challenge, not only a missed opportunity.

Instead, sponsors run sequential trials within their own programmes, each compared against a conventional standard of care. MajesTEC-3 compared teclistamab plus daratumumab against the physician’s choice.¹ This was reasonable, given when the trial was designed. DREAMM-7 compared belantamab mafodotin plus bortezomib-dexamethasone against a daratumumab triplet.² This, too, was clinically defensible at the time it was conceived. These trials were not reckless. They answered real questions, with the comparators available to them.

But consider DURGA-4, now recruiting. This phase 3 trial compares a novel dual-targeted CAR T-cell therapy against standard regimens, including daratumumab-carfilzomib-dexamethasone and daratumumab-pomalidomide-dexamethasone. By the time this trial reports, those control arms will likely be outdated. Bispecific antibodies and existing CAR T-cell therapies are already recommended in treatment guidelines in many high-income countries, though regulatory approval, reimbursement, and real-world accessibility do not always move together. A triplet control arm in 2026 may simply reflect the years that separate trial design from trial reporting, and the standard of care, regulatory landscape, and drug availability at the time DURGA-4 was designed. But when this pattern recurs across an entire disease area, each new trial measuring itself against a comparator that is no longer current by the time results appear, the choice stops looking neutral, whatever the intention behind it.

The DREAMM trials raise a related concern. DREAMM-7 and DREAMM-8 were both designed, funded, and run by the same sponsor.^2,3 Follow-up studies and successor trials, DREAMM-9 and DREAMM-10, continue within the same programme rather than testing belantamab against a competing BCMA-directed agent. This also makes strategic sense. It does not advance the field as far as it could.

The MajesTEC trials follow the same pattern. MajesTEC-3 showed an 83% reduction in risk of progression or death at three-year follow-up, an excellent result.¹ MajesTEC-9 now compares teclistamab monotherapy against the physician’s choice, again within the same programme, again without a direct comparator from a competing class.⁴ Each new trial measures itself against yesterday’s standard, never against today’s best alternative.

The post-protocol therapy problem

This issue becomes worse when we consider post-protocol treatment. KarMMa-3 is the clearest case.⁶ Patients were randomized to idecabtagene vicleucel versus standard regimens. Progression-free survival (PFS) strongly favored ide-cel. But more than half of the patients in the control arm crossed over to receive ide-cel after progression. No significant overall survival (OS) benefit had emerged at the reported follow-up, despite an 18-month PFS advantage. Crossover of this magnitude complicates the intention-to-treat analysis of OS and dilutes the observed treatment effect; statistical methods exist to adjust for treatment switching, but they were not the primary analysis here. When more than half of a control arm eventually receives the experimental therapy, OS, the endpoint that matters most to patients, becomes difficult to interpret using standard methods alone.

The smoldering myeloma

There is a further development that concerns me. Several signals point toward bispecific antibodies and CAR T-cell therapies moving into smoldering myeloma within the next five years.⁸ CAR-PRISM is a phase 2 trial, and its encouraging early results are not yet confirmatory evidence of a new standard of care; several other early-phase trials are underway in high-risk smoldering disease, but whether this becomes standard practice within five years depends on the phase 3 data still to come. The clinical rationale is not unreasonable. Treating earlier may delay or prevent progression.

But the framing matters. Smoldering myeloma patients are largely asymptomatic. No organ damage. No renal failure. No hypercalcemia. Treating them with agents in the 300 000 to 500 000 US dollar range, whether the one-time acquisition cost of CAR T-cell therapy or the cost of an ongoing course of bispecific antibody treatment, cost structures that are not interchangeable but that both remain unreimbursed or severely restricted across most of the world, is not a purely clinical decision.^9,10 It carries an economic and ethical weight that the field has not seen.

For physicians in Türkiye, or in Egypt, India, or Brazil, this is a clinical reality we face.^11–13 Our patients with active, symptomatic relapsed disease already struggle to access these drugs.¹⁴ Moving the evidence base earlier, without building the access infrastructure, widens the gap rather than closing it.

What should change

I am not arguing that these trials are without value. I am arguing that the sum of these decisions, control arm selection, crossover policy, successor trial design, and therapeutic expansion strategy, produces a pattern the myeloma community should examine, regardless of whether commercial or scientific considerations drove any single choice. That is a problem the myeloma community is well placed to address.

First, precompetitive collaboration between sponsors should be encouraged to enable cross-agent randomized comparisons. Academic cooperative groups have historically been pioneers for such designs. They should be empowered and funded to lead head-to-head comparisons of bispecific and CAR T therapies, rather than leaving the phase 3 space entirely to industry-sponsored trials.

Second, trial registries should require justification whenever a proposed control arm is not the current evidence-based standard. Regulatory agencies and ethics committees are well placed to enforce this. They should do so consistently, not selectively. By current evidence-based standard, I mean the regimen best supported by recent randomized evidence for a given line of therapy and disease context, not simply the most recently approved agent. I recognize this will vary with prior treatment exposure, refractoriness, cytogenetic risk, and local access to CAR T-cell or bispecific therapy, and that no single comparator fits every patient; a required justification, rather than a fixed rule, is what I am proposing.

Third, the crossover design needs reconsideration. Where crossover is permitted, statistical adjustment for treatment switching should be pre-specified, not added afterward. Overall survival should not be quietly dropped as an endpoint simply because crossover makes it harder to interpret. Difficult is not the same as uninformative.

Finally, access planning should become a required part of phase 3 trial design in hematologic malignancies. A trial that produces a practice-changing result carries a responsibility. If that result will only be actionable in fifteen or twenty high-income countries for the next decade, the responsibility has not been met.

The myeloma field has produced a huge amount of data. It can also produce equitable science; these two goals are not in competition with each other.

Author contributions

Mehmet Baysal (Conceptualization, Writing—original draft, Writing—review & editing) and Seval Akpınar (Conceptualization, Writing—review & editing)

Funding

No funding, grants, equipment, or drug support were received for this work.

Conflicts of interest

None declared.