Does daratumumab plus lenalidomide/dexamethasone benefit frail myeloma patients?

A frailty re-analysis of an existing trial

MAIA was a randomized, open-label phase 3 trial in transplant-ineligible newly diagnosed multiple myeloma (NDMM). Patients received either daratumumab plus lenalidomide/dexamethasone (D-Rd) or lenalidomide/dexamethasone alone (Rd), and were treated until disease progression or unacceptable toxicity. Daratumumab is an antibody targeting CD38 on myeloma cells; lenalidomide and dexamethasone are the standard oral backbone.

This paper does not report a new trial. It re-examines the same 737 randomized patients (368 D-Rd, 369 Rd) after a median follow-up of 64.5 months, asking whether the benefit of adding daratumumab holds within frailty groups. Frailty was scored retrospectively from three baseline variables: age, Charlson Comorbidity Index, and Eastern Cooperative Oncology Group performance status. Patients were labeled fit, intermediate, or frail, with fit plus intermediate combined into a "non-frail" group.

That retrospective construction is the main interpretive limit. The authors note MAIA was designed before the International Myeloma Working Group frailty scoring system was released, and that a post hoc analysis using that system was not possible here. So the subgroups describe baseline characteristics, not a prospectively assigned frailty definition.

Nearly half the trial population was frail

Of the 737 randomized patients, 146 (19.8%) were classified fit, 250 (33.9%) intermediate, and 341 (46.3%) frail; the combined non-frail group contained 396 patients (53.7%). Treatment arms were reasonably balanced within each subgroup, for example 172 frail patients on D-Rd and 169 on Rd.

Frail patients were older than the non-frail group (median 77.0 vs 71.0 years). This matters because age also drives dosing: dexamethasone was reduced to 20 mg for patients older than 75 years or with body mass index below 18.5 kg/m², and lenalidomide was reduced to 10 mg when creatinine clearance was 30–50 mL/min. Frailty status and dose intensity therefore travel together in this dataset, and the subgroup results should be read with that overlap in mind.

Progression-free survival improves in every frailty group

The clearest result is progression-free survival (PFS), the time until disease progression or death. D-Rd beat Rd in all subgroups: fit (median not reached vs 43.5 months; HR 0.57; 95% CI 0.35–0.91; P = 0.0183), intermediate (73.7 vs 37.4 months; HR 0.43; P < 0.0001), non-frail overall (73.7 vs 37.8 months; HR 0.48; P < 0.0001), and frail (52.2 vs 30.4 months; HR 0.64; 95% CI 0.48–0.85; P = 0.0018).

The direction is consistent, but the size is not uniform. The hazard ratio is smallest in the intermediate group and largest in the frail group, meaning the relative risk reduction is somewhat smaller for frail patients even though the absolute gap in median PFS (about 22 months) is substantial. Estimated 60-month PFS rates were 45.0% with D-Rd versus 26.5% with Rd in frail patients, compared with 59.6% versus 38.5% in fit patients.

The paper also reports that PFS benefit appeared across subgroups split by ISS disease stage and by lenalidomide starting dose (25 vs <25 mg), including in the frail subgroup. Those are supporting analyses, not the primary basis for the frailty conclusion.

Overall survival separates the frail group from the rest

Overall survival (OS) is where the subgroups diverge. In fit patients, median OS was not reached in either arm (HR 0.48; 95% CI 0.23–0.97; P = 0.0375). In intermediate patients it was not reached versus 66.4 months (HR 0.52; P = 0.0012), and in the combined non-frail group not reached versus 69.8 months (HR 0.53; 95% CI 0.37–0.76; P = 0.0004).

In frail patients, median OS was not reached with D-Rd versus 50.4 months with Rd, and the hazard ratio point estimate favored D-Rd at 0.79. But the confidence interval crossed 1 (0.58–1.06) and the P value was 0.1128, so this analysis does not establish a statistically significant OS benefit in the frail subgroup. The authors describe the frail OS benefit as "less pronounced" rather than absent.

Estimated 60-month OS rates still differed: 53.6% with D-Rd versus 42.7% with Rd among frail patients, against 83.5% versus 72.5% among fit patients. A rate difference at one timepoint and a non-significant hazard ratio are not the same quantity, and the paper does not claim they are.

Deeper responses appear even in frail patients

Response depth moved in the same direction as PFS. Rates of complete response or better were higher with D-Rd in every subgroup: fit 60.3% vs 26.9%, intermediate 54.7% vs 27.9%, non-frail 56.6% vs 27.5%, and frail 44.8% vs 33.1% (P = 0.0279 for the frail comparison).

Minimal residual disease (MRD) negativity at the 10⁻⁵ threshold, meaning no detectable disease at that sensitivity, was also more frequent with D-Rd: 36.8% vs 9.0% in fit, 38.3% vs 9.8% in intermediate, 37.8% vs 9.5% in non-frail, and 25.6% vs 13.0% in frail patients (P = 0.0039). The frail gap is narrower than in the other groups, but it remains in favor of D-Rd.

Longer treatment duration with daratumumab

Median duration of study treatment was longer with D-Rd than Rd in every subgroup: fit 58.0 vs 31.0 months, intermediate 51.8 vs 20.1 months, non-frail 54.6 vs 25.8 months, and frail 39.1 vs 20.7 months. The authors interpret this as evidence of improved disease control and long-term tolerability of daratumumab in frail patients.

Discontinuation patterns support a tolerability reading but with nuance. Overall discontinuation was lower with D-Rd across subgroups, and discontinuation due to adverse events was also lower with D-Rd (for example 17.3% vs 24.1% in frail patients). However, permanent discontinuation of lenalidomide or dexamethasone occurred more often with D-Rd, and lenalidomide dose modifications were more frequent with D-Rd. Median daratumumab relative dose intensity stayed at or above 97.8% across frailty subgroups, so the added antibody was delivered close to plan.

Deaths were most common in the frail subgroup: 51.8% of treated frail patients died (47.6% D-Rd, 56.0% Rd), versus 24.1% of fit patients. Treatment-emergent adverse events leading to death occurred in 14.9% of frail patients on D-Rd and 14.5% on Rd.

The frail plus stage III subgroup needs caution

When the authors subdivided by ISS stage, OS benefit with D-Rd was primarily seen in the non-frail subgroups. In the frail plus ISS stage III subgroup, OS appeared to favor Rd at timepoints under 2 years from randomization. The paper reports that the proportion of deaths was similar between arms by the 2-year mark and slightly higher with Rd overall, and that deaths due to adverse events (most often infections or cardiac disorders) were numerically higher with D-Rd while deaths from disease progression or other causes were numerically higher with Rd.

This is a small, exploratory slice of an already retrospective subgroup, and the paper does not present it as a reason to withhold D-Rd. It does show that the frail group is not uniform: early mortality risk and disease control pull in different directions, and the balance may depend on stage and comorbidity.

Disease control is the robust claim for frail patients

Read together, the results support a specific priority rather than a blanket statement. For transplant-ineligible NDMM patients classified as frail by these baseline variables, the well-supported benefit of D-Rd is longer disease control: a PFS hazard ratio of 0.64, a median PFS gain of roughly 22 months, and higher rates of complete response and MRD negativity. The OS hazard ratio of 0.79 points the same way but its confidence interval includes no difference, so survival should be described as a favorable but unproven signal in this subgroup.

The contrast that matters most is between the non-frail and frail groups: OS benefit is statistically secure in fit, intermediate, and combined non-frail patients (HR 0.48, 0.52, and 0.53), while the frail group carries the largest PFS hazard ratio and the only non-significant OS result. Because frailty was assigned retrospectively from age, comorbidity, and performance status, and because frail patients were older and more often dose-reduced, these subgroup estimates describe associations within the trial rather than a tested frailty-specific treatment strategy.

Sources used in this answer