Daratumumab provides equivalent survival benefits in first-line and second-line treatment of transplant-ineligible multiple myeloma: A computational analysis of Phase 3 trials in first-line and second-line settings. Free

Introduction

Daratumumab is a human IgGκ monoclonal antibody against CD38, with approvals for transplant-ineligible multiple myeloma (TIE-MM) in both newly diagnosed and relapsed/refractory settings. The MAIA trial demonstrated that adding daratumumab to first-line (1L) lenalidomide-dexamethasone (Rd) significantly improves progression-free (PFS) and overall survival (OS), despite 46% of control arm patients receiving daratumumab after progression. However, it remains uncertain whether 1L use of daratumumab offers intrinsically greater benefit than second-line (2L) use or simply ensures access to an effective drug. A similar question was tested in the phase 3 SONIA trial for advanced breast cancer, which randomized patients to receive CDK4/6 inhibitors in 1L or 2L. While 1L intensification prolonged time to first progression (PFS1), there was no difference in PFS2 or OS, and treatment duration and adverse event rates were higher (Sonke et al., Nature 2024). This raises a relevant question in TIE-MM: does daratumumab improve outcomes more when used in 1L versus 2L? Here we investigate this question by analyzing published Kaplan-Meier distributions from phase 3 trials of daratumumab in the 1L and 2L settings.

Methods

We analyzed PFS and OS data from the MAIA (1L), CASTOR (2L), and POLLUX (2L) phase 3 trials of daratumumab-containing regimens in TIE-MM. Using PFS distributions from the 2L trials, we computed expected 1L PFS under the null hypothesis that daratumumab's duration of PFS benefit is equal at 1L and 2L. Differences between predicted (null) and observed 1L PFS were tested by Cox proportional hazards. For OS, we investigated the potential confounding from incomplete crossover in MAIA, where 54% of control-arm patients did not receive daratumumab (nor another CD38 mAb) after progression. We used 2L survival distributions to simulate a 100% crossover cohort where all control patients receive 2L daratumumab, testing the null hypothesis that daratumumab extends OS equally whether given at 1L or 2L. Differences between observed 1L daratumumab OS and this simulated 1L control with complete crossover were tested by Cox proportional hazards. This evaluates whether 1L use of daratumumab improves OS by more than is expected by providing daratumumab to all patients at 2L.

Results

The PFS benefit exhibited by daratumumab in the 2L CASTOR and POLLUX trials fully accounts for the magnitude of PFS improvement observed in the 1L MAIA trial. Predicted 1L daratumumab PFS curves computed from CASTOR and POLLUX data did not significantly differ from observed PFS of 1L daratumumab in MAIA (log-rank p=0.458 [CASTOR], p=0.694 [POLLUX]), with predicted D-Rd vs Rd hazard ratios within 0.03 of the observed value in MAIA. Models also predicted a 24-month reduction in median time on daratumumab with 2L versus 1L use, without reducing its PFS benefit. Simulated OS for a 1L control cohort (Rd) where all patients receive 2L daratumumab showed no significant OS difference compared to the observed 1L D-Rd cohort (log-rank p=0.548 [CASTOR], p=0.811 [POLLUX]). These results suggest that observed OS differences in MAIA primarily reflect limited daratumumab access in the control arm, rather than intrinsic superiority of 1L over 2L use of daratumumab.

Conclusions

This analysis of phase 3 trial results suggests that daratumumab confers equivalent PFS and OS benefits when used in 1L or 2L settings for patients with TIE-MM. The OS advantage observed in MAIA can be explained by limited access to 2L daratumumab in the control arm, rather than intrinsic superiority of 1L use. Although CASTOR and POLLUX cohorts were younger, and often post-transplant, their PFS was significantly shorter than in MAIA, indicating more advanced disease. Thus, these 2L results are unlikely to overestimate efficacy of 1L daratumumab and falsely suggest equivalence, despite demographic differences. While real-world attrition may favor 1L use for individuals who are at risk of being unable to access a 2L therapy, these results suggest that for patients likely to receive subsequent therapy, reserving daratumumab for 2L is a clinically valid strategy that may reduce drug exposure, toxicity, and financial burden without compromising outcomes. Equivalent OS outcomes of 1L or 2L use of daratumumab supports individualized shared decision-making with patients with TIE-MM on daratumumab timing, rather than a uniform strategy of 1L daratumumab for all patients.