Mel200 or BuMel conditioning in myeloma: is there a winner? Free

In this issue of Blood, Lahuerta et al¹ present the mature results of the GEM12 trial of 200 mg/m² melphalan (Mel200) vs busulfan-plus-melphalan (BuMel) conditioning before autologous stem cell transplantation (ASCT) in multiple myeloma (MM). The progression-free survival (PFS) was significantly longer with BuMel in patients with International Staging System (ISS) II/III myeloma and in those with del1p or t(14;16), as determined by fluorescence in situ hybridization (FISH) cytogenetics, whereas Mel200 was better in ISS I disease.

The results from large phase 3 trials, like GEM12, represent a gold standard in clinical research and would ideally settle the question of optimal conditioning in patients with newly diagnosed MM (NDMM) who are undergoing autologous transplantation. Yet, several issues merit further discussion. These relate to the relative contribution of each component of the ASCT sequence to PFS, challenges with the precise delineation of myeloma risk groups, and rapid improvements in first-line MM therapy.

ASCT in NDMM encompasses a 3-part package, namely induction, high-dose conditioning plus ASCT, and post-ASCT therapy (maintenance ± consolidation); each of these steps contributes to the overall response rate, PFS, and overall survival. Before the advent of novel agents, many efforts focused on improving the high-dose component particularly since the standard-of-care (SOC) in myeloma involved the use of single agent melphalan conditioning, in contrast to the combination high-dose regimens given in other malignancies. Numerous combinations of melphalan and other agents were studied, including BuMel. Several retrospective studies, matched comparisons, and meta-analyses suggested that BuMel conditioning was superior to Mel200, including in higher-risk populations.^2-4 However, these reports often used older induction regimens, and maintenance therapy was not uniform. This heterogeneity was also noted in a previous smaller randomized trial that compared Mel200 with BuMel in which the final analysis failed to demonstrate that BuMel had a significant advantage.⁵ Importantly, the GEM12 trial incorporated both modern induction/consolidation with the combination of bortezomib, lenalidomide, and dexamethasone (VRD) and lenalidomide-based maintenance.

The results in both arms of the GEM12 trial were excellent and did not differ significantly despite a numerically longer PFS for BuMel (89 months) when compared with 73.1 months for Mel200 (P = .3); the overall survival was similar in both cohorts. With full acknowledgment of the caveats of cross-trial comparisons, these PFS outcomes compared favorably with those reported for the widely accepted SOC in patients with NDMM established in the DETERMINATION study.⁶ The transplant arm in DETERMINATION received 3 cycles of a VRD combination, ASCT with Mel200, 2 cycles of VRD consolidation, and then lenalidomide maintenance until progression; this approach yielded a median PFS of 67.5 months. However, the GEM12 trial administered 6 cycles of VRD before randomization to Mel200 or BuMel, followed by 2 more VRD cycles as consolidation. In addition to more total cycles, that is, 8 compared with 5 to 6 cycles in other VRD regimens, GEM12 used higher doses of the agents in VRD. Lahuerta et al acknowledge that their “reinforced VRD” induction/consolidation may have contributed to the longer PFS in both conditioning arms. It is possible that “reinforced VRD” also muted the benefit of BuMel.

The identification of specific higher-risk subgroups in which BuMel outperformed Mel200 also has been sought by different investigators. In GEM12, patients with ISS II/III experienced a significantly longer PFS with BuMel of 76 months vs 57 months with Mel200. However, patients with ISS II and III disease have different outcomes, as exemplified by the DETERMINATION trial in which the median PFS was 62.5 months in patients with ISS II and 35.9 months for those with ISS III. The grouping of these 2 distinct prognostic groups together in GEM12, albeit as part of a predefined adjusted analysis, makes it difficult to ascertain at which point the advantage of BuMel lies. In contrast, patients with ISS I disease had a significantly longer median PFS with Mel200 (not yet reached) than with BuMel (100 months).

A significant PFS advantage for BuMel was also observed in 2 small cytogenetics subsets, namely those with FISH positivity for del1p or t(14;16), which accounted for 6.1% and 3.7%, respectively, of all trial patients, but not in the commonly reported high-risk grouping of t(14;16), t(4;14), and del17p.¹ The complexity of myeloma biology has recently led to significant revision of the myeloma risk classification system,⁷ and there are no data available yet on the impact of any conditioning regimen in more precisely defined high-risk groups.

As patients live longer, late toxicities, particularly secondary primary malignancies (SPMs), have assumed increasing importance. The raw incidence rates were similar in both the Mel200 and BuMel arms (6.5 % and 9.2%, respectively).¹ These rates seem to be somewhat lower than those noted in other BuMel series.^5,8 However, the SPM rates can be calculated in different ways, and the use of cumulative incidence rates or standardized incidence ratios may enable better cross-trial comparisons. Although more evaluation is needed, the potentially shorter duration of lenalidomide maintenance, as directed by bone marrow minimal residual disease testing in GEM12, might offer a mitigation strategy for this complication.

Finally, the emerging SOC in patients who undergo ASCT, based on the PERSEUS trial, adds the anti-CD38 monoclonal antibody daratumumab both to the more intensive VRD dosing regimen (4 induction and 2 consolidation cycles) and to lenalidomide maintenance; the median PFS with this regimen is anticipated to exceed previous values, including in high-risk patients.⁹ As the field continues to evolve rapidly, whether ASCT will even remain a cornerstone of initial therapy in the future has been called into question. The unprecedented results of chimeric antigen receptor (CAR) T-cell therapy in advanced relapse have led to an ongoing international trial in NDMM that directly compares Mel200 + ASCT vs CAR T-cell therapy that targets B-cell maturation antigen after induction with daratumumab + VRD and later followed by lenalidomide maintenance.¹⁰ 

In the end, the question that remains is should the GEM12 trial findings change the current ASCTpractice? The study confirms that both BuMel and Mel200, in conjunction with “reinforced VRD” and lenalidomide maintenance, yield a comparatively long PFS. In this trial, however, Mel200 was the winner in patients with ISS I disease. As the authors suggest, BuMel could be considered in selected high-risk patients if given with “reinforced VRD” in jurisdictions where anti-CD38 antibodies are not available.

However, a clear-cut, broader indication for BuMel has not emerged and will likely remain undefined in the era of increasingly sophisticated immunotherapy.

Conflict-of-interest disclosure: D.E.R. reports receiving research funding and honoraria from Bristol Myers Squibb and Janssen.