For decades, the combination of induction chemoimmunotherapy followed by autologous stem cell transplantation (ASCT) has been the accepted standard of care for transplant-eligible patients with newly diagnosed multiple myeloma (NDMM).1 Many international centers also use tandem ASCT (two consecutive ASCTs within six months of each other) in patients with high-risk disease.2 Triplet regimens such as lenalidomide, bortezomib, and dexamethasone (RVd) followed by ASCT had been the standard based on a number of studies demonstrating improvement in progression-free survival (PFS) but not overall survival (OS).3 The lack of OS benefit has largely been attributed to the successful salvage in relapse. Nationally, the use of ASCT is diminishing, with only about 30% of transplant-eligible patients actually receiving transplant in the frontline setting.4
In recent years, quadruplet therapy incorporating an anti-CD38 monoclonal antibody (daratumumab or isatuximab) into induction therapy has become standard, based on the results from the GRIFFIN and PERSEUS trials.5,6 As the growing armamentarium of treatment modalities for MM patients continues to extend survival, the follow-up required to assess PFS and OS endpoints has become prohibitively long. As such, achieving minimal residual disease (MRD) negativity has become an important prognostic and predictive indicator of disease behavior and therapeutic efficacy.7,8 In April 2024, the U.S. Food and Drug Administration’s Oncologic Drugs Advisory Committee voted 12 to 0 in favor of using MRD as an endpoint to support accelerated approvals, with the understanding that PFS and OS must still be considered for maintenance of approval.9 With the emergence of more potent and efficacious induction regimens and the growing clinical utility of MRD assessments, it is time to rethink the widespread applicability of historical standards.
The phase III MIDAS trial was a response-adapted, randomized trial investigating the role of transplant in patients who are able to achieve MRD negativity with quadruplet induction therapy and the role of tandem versus single ASCT in patients who remained MRD-positive after induction.10 MRD was assessed by next-generation sequencing (NGS), when able, and otherwise by flow cytometry. In this study, 757 transplant-eligible NDMM patients underwent six 28-day cycles of induction with a quadruplet regimen of isatuximab, carfilzomib, lenalidomide, and dexamethasone (Isa-KRd). Sixty-six percent of patients (n=499) were MRD-negative to 10-5 after induction and were randomized to either ASCT followed by two cycles of Isa-KRd (cohort A) or six additional cycles of Isa-KRd without transplant (cohort B). The remaining patients who were MRD-positive after induction were randomized to either single ASCT followed by two cycles of Isa-KRd (cohort C) or tandem ASCT (cohort D). The primary endpoint for the study was MRD-negative status at 10-6 sensitivity before maintenance therapy. Median follow-up was 16.8 months in cohorts A and B, and 16.3 months in cohorts C and D.
Among patients randomized to cohorts A and B, 86% and 84% (adjusted relative risk, 1.02; 95% CI 0.95-1.10; p=0.64), respectively, were MRD-negative to 10-6 prior to maintenance initiation. The surprisingly similar rates of deep MRD negativity in the two arms challenge the long-held belief that transplant is necessary to achieve deep disease clearance early in treatment. This suggests that for patients who can achieve MRD negativity with modern induction regimens, a prolonged induction approach may be adequate to yield equivalent response and avoid the toxicities of high-dose melphalan. Among patients who were MRD-positive after induction (n=252), MRD negativity to 10-6 was achieved in 40% and 32% (adjusted relative risk, 0.82; 95% CI 0.58-1.15; p=0.31), respectively, after single or tandem ASCT, indicating that tandem transplant did not provide a clear benefit over single ASCT despite suboptimal response to induction. Patients carrying a t(11;14) translocation were more heavily represented in cohorts C and D, suggesting a delay in achieving MRD negativity for this group. Post-hoc analysis showed that prior to starting maintenance, 63% of patients with t(11;14) were MRD-negative to 10-6, compared to 78% of those without the translocation. This is an important observation, because while this abnormality is thought to predict standard-risk disease behavior, it may take such patients longer to achieve a deep response.
From a safety standpoint, there were no new safety signals for Isa-KRd, ASCT, or tandem ASCT. The advantage of following a risk-adapted approach for patients who achieve a deep response with initial induction lies mainly in the ability to avoid melphalan-associated toxicities including but not limited to mucositis, gastrointestinal toxicities, febrile neutropenia, secondary malignancies, and profound immunosuppression. The use of tandem ASCT further compounds these toxicities and prolongs the period of severe immunosuppression. Initial analyses from the STaMINA trial had shown no additional PFS benefit from tandem ASCT following triplet induction with RVd, but long-term follow-up and as-treated analyses showed a PFS benefit (49.4% vs. 39.7%, p=0.01) from tandem ASCT, primarily driven by patients with high-risk disease. The MIDAS trial challenges whether these results would be recapitulated in the era of quadruplet induction and the ability to stratify treatment based on response. MRD analysis of patients in the STaMINA trial showed that MRD negativity at one year correlated with PFS (hazard ratio [HR] = 3.61, p<0.0001) and OS (HR=3.36, p<0.001), suggesting that attaining a deeper response may be a better predictor of survival outcomes than the treatment modalities used.11 Long-term follow-up data of the MIDAS trial will be very important to determine whether these impressive MRD findings will translate to similar survival outcomes.
As the treatment landscape for MM continues to evolve at a rapid pace and MRD testing continues to improve, we must adapt our practices to reflect these advancements yet be careful about applying conclusions too broadly. We need to become comfortable using surrogate endpoints to predict survival but maintain a conscientious effort to revisit these practices based on long-term follow-up data and relevant subgroup analyses. In particular, we must exercise caution when applying conclusions to patients with high-risk cytogenetic abnormalities (HRCAs) or ultra high-risk myeloma (two or more HRCAs). For example, in the FORTE trial, patients with high-risk cytogenetics and those who took longer to achieve MRD negativity had a higher risk of non-sustained MRD negativity.12 Similarly, in the MASTER trial — an MRD-directed treatment cessation trial using a carfilzomib-based quadruplet — patients with two or more HRCAs had shorter PFS, with initial MRD negativity not as predictive of long-term outcomes.13,14 The MIDAS trial is an important step in the direction of more personalized care of MM patients, but early adaptation should be limited to the standard-risk population until longer follow-up has been reported. While it may be premature to widely adopt this strategy, it is a promising start that can at least open the door to a more flexible approach for patients who have tenuous performance status or lack social support to proceed with transplant. Hesitation to implement this framework may stem from the concept that immune reconstitution and alterations in the bone marrow microenvironment likely affect the durability of deep remissions post-transplant.15 Long-term PFS data will clarify whether those effects are necessary to sustain remissions or whether MRD negativity and — more importantly — sustained MRD negativity can reliably start to guide management decisions for most patients.
In Brief
The role of ASCT for NDMM in the era of quadruplet induction and novel immunotherapies such as chimeric antigen receptor T-cell therapy and bispecific antibodies is not well-defined. It is crucial to obtain more data through randomized and response-adapted clinical trials so that we can determine which patients are more likely to benefit from high-dose chemotherapy in addition to the novel treatments that are now available to them. Real-world application of these trials will also require strong consideration of quality-of-life measures, cost of treatments, and accessibility for broader patient populations. MRD is a valuable and important endpoint to determine strong early efficacy signals, but sustained MRD negativity is more reliably predictive of outcomes, and follow-up PFS and OS data is paramount to validating this as a surrogate endpoint for decision-making. The MIDAS trial is an elegant example of response-directed therapy that can be used to deepen response after induction or minimize toxicity in patients who have already achieved deep responses. The results need to be interpreted with caution in patients with high-risk disease and until longer follow-up confirms these impressive initial trends.
Disclosure Statement
Dr. Lipof has received honoraria from Sanofi and has served on the advisory board of AstraZeneca. Dr. Chung has received research funding from AbbVie, Bristol Myers Squibb, Caelum Biosciences, CARsgen Therapeutics, Cellectis, Johnson & Johnson, K36 Therapeutics, and Merck and has served on the advisory boards of Bristol Myers Squibb and Johnson & Johnson. Dr. Kumar indicated no relevant conflicts of interest. Dr. Pan has received honoraria from Sanofi. Dr. Chari has received grant/research support from Janssen and has served as a consultant and/or advisor for Abbvie, Adaptive, Amgen, Antengene, Bristol Myers Squibb, Forus, Genetech/Roche, GlaxoSmithKline, Janssen, Karyopharm, Millenium/Takeda, and Sanofi/Genzyme.
