Simplified frailty score for patients with multiple myeloma undergoing autologous stem cell transplant

Abstract Background: Autologous stem cell transplant (autoHCT) is standard of care for eligible patients with newly diagnosed multiple myeloma (MM). However, determining autoHCT eligibility is highly subjective, and requires complex frailty assessment tools. Here, we used a simplified frailty score to evaluate its ability to stratify newly diagnosed MM patients undergoing upfront autoHCT into “frail” and “non-frail” groups. This score was previously used in the setting of B-cell maturation antigen-directed CAR-T therapy (Davis et al. JTCT 2024). Methods: In this single center, retrospective analysis, we included newly diagnosed MM patients who underwent upfront autoHCT between 1988 and 2021, and had available information to calculate the frailty score. A simplified score was calculated for each patient based on the following: age (76-80 years, 1 point; >80, 2 points), Eastern Cooperative Oncology Group [ECOG] Performance Status (≤1, 1 point; ≥2, 2 points) and Hematopoietic Cell Transplant Comorbidity Index [HCT-CI] (≥2, 1 point). A score of ≥2 was defined as frail. High risk cytogenetic abnormalities (HRCA) were defined as del17p, t(4;14), t(14;16), 1q21 gain or amplification (1q+) by fluorescence in-situ hybridization. Primary outcomes were progression-free (PFS) and overall survival (OS). Non-relapse mortality (NRM), a key secondary outcome, was assessed using the competing risks method, with progression as the competing risk. MRD status was determined in bone marrow samples using an 8-color NGF assay with a sensitivity of 1/10⁵ cells (0.001%) on a minimum of 2 million events. Results: We included 2485 patients with a median age of 61 (range 25-83) years. Thirty-two percent of patients (n=803) were ≥65-years-old. Using the simplified frailty score, 628 (25%) patients were identified as frail and 1857 (75%) as non-frail. Fifty-nine percent of the patients in the entire cohort were male, 18% self-identified as black, without a significant difference between the frail and non-frail groups. Twenty-eight percent of patients in the frail group had HRCA, compared to 24% in the non-frail group (p=0.053), and there was no significant difference in the proportion of patients with R-ISS stage III disease between groups (7% vs. 6%; p=0.38). Both groups had a similar proportion of patients with high tumor burden (≥50% plasma cells) in the bone marrow. A lower proportion of patients in the frail group received post-transplant maintenance (63% vs. 70%; p=0.002). Patients in the frail group had a longer median duration of hospitalization for autoHCT compared to the non-frail group (17 vs. 16 days; p<0.001). There was no significant difference between the frail and non-frail groups in pre-transplant hematologic responses (≥very good partial response (VGPR): 54% vs. 52%; p=0.43) or MRD-negativity rate (41% vs. 43%; p=0.44). There was also no difference in either post-transplant responses, including day 100 (≥VGPR: 76% vs. 75%; p=0.42) and best post-transplant (≥VGPR: 84% vs. 84%; p=0.75) hematologic responses, or in MRD-negativity rate at best response (66% vs. 67%; p=0.77). After a median follow-up of 54.6 (range 0.2 - 262.0) months, the incidence of NRM was 24% in the frail group compared with 10% in the non-frail group (p=0.050). Patients in the frail group had significantly shorter median PFS [34.7 months (95% CI 32.1-38.1)] versus the non-frail group [43.0 months (95% CI 39.6-46.0); p=0.002], and shorter OS [85.9 months (95% CI 79.4-97.4) vs. 108.5 months (95% CI 101.9-115.7); p<0.001]. In multivariable analysis, frailty status retained its predictive significance for PFS [hazard ratio 1.20 (95% CI 1.06-1.35; p=0.003)] and OS [hazard ratio 1.26 (95% CI 1.08-1.46; p=0.003)]. Conclusions: A simplified frailty score incorporating age, ECOG and HCT-CI score predicts survival outcomes in patients with newly diagnosed MM undergoing upfront autoHCT.