Frailty and Outcomes in Multiple Myeloma Patients Eligible for Autologous Hematopoietic Cell Transplantation: A Prospective Study of the Grupo Español De Trasplante Hematopoyético y Terapia Celular (GETH-TC)

Introduction

Frailty in multiple myeloma (MM) patients is linked to higher mortality, disease progression, and treatment toxicity. However, frailty assessments in MM patients have traditionally focused on elderly adults, ineligible for auto-HCT and excluded using arbitrary age cut-offs. In addition, the assessments were often complex and time consuming, limiting their application in clinical practice.

Since April 2022, 15 institutions members of the GETH-TC have participated in a multicenter observational study to assess frailty dynamics in all adult MM patients eligible for auto-HCT and its impact on outcomes. This abstract summarizes prospective findings from this collaborative effort.

Methods

All eligible MM patients for auto-HCT were assessed for frailty after providing informed consent, irrespective of age and comorbidities. Frailty was evaluated at the first consultation and HCT admission using the HCT Frailty Scale, which classifies patients as fit, pre-frail, or frail. Evaluations, conducted by HCT teams without additional external resources, took 8-10 minutes per patient, including Mini-Cog and EQ-5D-3L quality of life (QoL) tests. Prospective data was updated in July 2024.

Results

296 MM patients were included. The median age was 57 (range:31-75)) with 28.3% patents over 64 years old. 53.3% patients were male, 39.1% had a KPS < 90%, and 23.0% had an HCT-CI > 3. At diagnosis, 42.9% were ISS I, 32.7% ISS II, and 24.4% ISS III. Most patients (89.1%) received one line of treatment before auto-HCT, primarily using VTD/VRD (44.2%) and DARA-VRD (23.9%).

The first consultation occurred at a median of 52 days (IQR 25-120) before HCT, mostly before stem cell collection. Initially, 64 (23.2%) patients were classified as fit, 160 (58.0%) as pre-frail, and 52 (18.0%) as frail. Binary logistic multivariate regression analysis (MVA) indicated that frailty phenotype correlated with a KPS < 90% (OR 3.52, P<0.01), an abnormal Mini-Cog (OR 7.28, P<0.01), and ISS II (OR 4.21, P=0.026) or III (OR 7.03, P<0.01). However, frailty did not correlate with age ≥ 65 years (OR 1.06, P=0.89) and comorbidities (HCT-CI > 3: OR 0.66, P=0.50).

At HCT admission, 50 (18.1%) patients were classified as fit, 174 (63.0%) as pre-frail, and 52 (18.8%) as frail. By this time, 53.0% patients were in CR, 45.2% in VGPR or PR, and 1.7% had stable disease, with no differences across frailty states (P=0.21). Interestingly, some patients changed their frailty categories between consultations with 1.6% of fit and 12.5% of pre-frail worsening to frail states, while 59.6% of frail patients remaining frail, with 38.5% of them improving to pre-frail and 1.9% to fit states.

Differences in auto-HCT outcomes of patients across frailty levels at admission were examined. The results showed similar hospitalization durations across frail, pre-frail and fit status (median: 15, 15, and 14 days, P=0.29), but a trend towards higher readmission rates for frail patients than for the rest (13.5% vs. 8.6% and 2.0%, P=0.06). Admission QoL assessments revealed better auto-recorded global health scores in fit patients compared to pre-frail and frail ones (80%, 70%, and 60%, P<0.01).

With a median follow-up of 13 months, 21 (7.6%) patients relapsed, and 14 (5.1%) died: 10 of them due to infections, in a median of 8 months after the HCT and not during the admission. The 1-year relapse incidence was comparable among fit, pre-frail, and frail patients (8.9% vs. 5.6% and 6.5%, P=0.903), and fit patients had higher OS than pre-frail and frail ones (1-year OS: 98.9%, 95.8%, and 84.3%, P<0.01). The observed associations between frailty and lower OS (Frail vs. others: HR 3.27, P=0.047) and higher mortality non-related with relapse (HR 3.24, P=0.075) were confirmed on MVA controlling by age (≥70), KPS (<90%), and comorbidities (HCT-CI>3).

Lastly, the ability of the scale for identifying frail patients at risk for mortality when measured earlier, at first consultation was additionally confirmed.

Conclusions

Frailty of MM patients before auto-HCT had a prevalence of 18%, could be diagnosed irrespective of chronological age, and was associated with worse outcomes and QoL. Results underscore the importance of introducing early frailty assessments in predicting risks, and exploring the effect of pre-transplant interventions for preventing and reversing frailty before auto-HCT in MM patients.