Updated results from the frailty-stratified, randomised controlled bayesian adaptive trial of bortezomib versus lenalidomide in transplant ineligible myeloma (TI NDMM) – the FRAIL-m study (AMaRC 19-01; ALLG MM22)

Abstract Introduction: Real-world survival for TI NDMM patients (pts) is inferior to transplant eligible (TE)-NDMM pts with median OS of 48 vs 68 months respectively (p <0.001), with age >70 years the strongest predictor for early mortality (unpublished, ANZ MRDR). Lenalidomide (R), bortezomib (V) and dexamethasone (d) are available, but no direct comparisons of VRd vs Rd vs Vd exist to rationalise treatment selection or establish the standard of care for TI NDMM pts (60% of all NDMM diagnoses). Frailty adversely impacts the likelihood of completing an intended course of therapy and tension exists between ensuring treatment tolerability with attainment of disease control. FRAIL-M aims to identify effective and deliverable backbone induction regimens to further explore the integration of newer immunotherapeutics into therapeutic paradigm for TI NDMM. Methods: FRAIL-M (ACTRN12619001199101) is a novel, frailty-stratified, adaptive, Phase Ib/II trial that assigns pts to 1 of 3 strata (fit, intermediate-fit and frail) prior to 1:1 randomisation to compare V-based (arm A) and R-based regimens (arm B). Frailty assessment utilises the IMWG Frailty Score, Katz Activity of Daily Living, Lawton Instrument Activity of Daily Living and Charlson Comorbidity Index. The primary objective is to establish, after 4 cycles (EOC4), ORR and safety as indicated by occurrence of deliverability-limiting toxicity (DeLT), defined as any treatment-related toxicity that mandates dose modification (reduction/cessation) of R, V, or d within cycle 1-4. Efficacy and toxicity are jointly monitored using a Bayesian Optimal Phase 2 (BOP2) design. A regimen is deemed unacceptable if the probability of response ≤50% or DeLT rate>35%. Protocol-specified dose escalation is permitted for inefficacy in intermediate-fit/frail strata and dose reduction for unacceptable toxicity in all strata. There are three analysis sets – efficacy evaluable, toxicity evaluable and jointly evaluable. 95% credible intervals (95% CrI) for ORR and the DeLT rate (DeLT-R) are reported. The trial was set-up for 300 pts (50 pts in 6 treatment arms). Accrual opened 02DEC2019, with 201 registered pts by 16JUL2025. Recruitment to Arm 1B (Rd) closed on 3JUN2025 due to operational futility. We report updated results for the frail strata (3A, 3B) and preliminary results for the intermediate-fit strata (2A, 2B). Results: The frail strata comprises 107 pts, 57% males, median age 81 years (63-93). Initially in 3A: V 1.0mg/m2, d 12mg D1, 8, 15 of a 28-day cycle for 12 cycles. The 1st interim analysis (N=10) demonstrated inefficacy (ORR 20%) and 0 DeLTs; thus 3A underwent dose escalation to include R 15mg on D1-21 of each 28-day cycle. 25 pts were evaluable for efficacy ORR 96.0% (95%CrI: 82.8–99.6%), 27 pts evaluable for toxicity, DeLT-R 40.7% (95%CrI: 23.5–58.9%) and 25 pts jointly evaluable (ORR 96.0%, DeLT-R 36.0%). The posterior probability that ORR>50% is >0.99 and DeLT-R≤35% is 0.28.In 3B pts (R 10mg D1-21, d 12mg D1,8,15 of a 28-day cycle) with 44 pts evaluable for efficacy ORR 65.9% (95%CrI: 51.3–78.6%), 41 pts evaluable for toxicity, DeLT-R 29.3% (95%CrI: 16.8–43.9%), and 38 pts jointly evaluable (ORR 78.4%, DeLT-R 21.6%). The posterior probability that ORR>50% is 0.98 and DeLT-R ≤35% is 0.79. The intermediate-fit strata comprises 58 pts, 64% males, median age 77 years (61-80). In 2A (VRd: V 1.0mg/m2, d 20mg on D1, 8, 15, R 15mg on D1-21 of a 28-day cycle), 26 evaluable pts for efficacy ORR 84.6% (95%CrI: 67.5–94.6%), 28 pts evaluable for toxicity DeLT-R 39.3% (95%CrI: 22.5–57.2%) and 24 pts jointly evaluable (ORR 91.7%, DeLT-R 29.2%). The posterior probability that ORR >50% is >0.99 and that DeLT-R ≤35% is 0.33. In 2B, (R 15mg D1-21, d 20mg D1, 8, 15 of a 28-day cycle), 24 pts evaluable for efficacy ORR 70.8% (95%CrI: 51.1–85.9%), toxicity DeLT-R 33.3% (95%CrI: 16.7–52.6%) and jointly (ORR 70.8%, DeLT-R 33.3%). The posterior probability that ORR >50% is >0.98 and that DeLT-R ≤35% is 0.58. Conclusion: All arms demonstrate acceptable efficacy however, in the triplet (VRd) regimen arms toxicity is emerging in 2A and may be unacceptable in 3A. The observed rate in jointly evaluable 3A patients does not meet BOP2 guideline for discontinuation and thus recruitment to all arms continues. These findings highlight the impact of therapeutic deliverability, support frailty-adapted therapy and question the benefit of attempting triplet and even quadruplet induction in the truly frail TI-NDMM.