Stratifying ‘frail’ Patients Defined By the Simplified Frailty Scale in Multiple Myeloma: Correlation with Clinical Outcomes

• Published in: Blood 2024-11, Vol.144, p.3342-3342
• Main Authors: Kato, Yukiyasu, Suzuki, Tomotaka, Kondo, Aki, Matsunaga, Naohiro, Kikuchi, Takaki, Kanamori, Takashi, Yoshida, Takashi, Kayukawa, Satoshi, Asao, Yu, Yano, Hiroki, Sasaki, Hirokazu, Asano, Arisa, Kinoshita, Shiori, Narita, Tomoko, Sanda, Takaomi, Ri, Masaki, Komatsu, Hirokazu, Iida, Shinsuke
• Format: Article
• Language: English
• ISSN: 0006-4971
• DOI: 10.1182/blood-2024-201310

Introduction: Frailty assessment is crucial for predicting the prognosis and ensuring appropriate care for older patients with multiple myeloma (MM). The Simplified Frailty Scale (SFS) comprises performance status (PS) (0-2 points), age (0-2 points), and Charlson Comorbidity Index (CCI) (0 or 1 point). Scores of 0-1 define non-frail patients, whereas scores of 2-5 define frail patients; frail patients have a significantly worse prognosis (Facon T et al. Leukemia 2020). However, because the frail group includes those with SFS scores of 2-5 (a 4-point range), it is considered heterogeneous. As the treatment strategy for older patients with MM may be stratified according to frailty status, reliable frailty evaluation is necessary. We aimed to evaluate the potential for further stratifying frail patients using the SFS and its clinical applicability. Methods: This study included patients with newly diagnosed MM aged ≥ 70 years who were ineligible for transplantation and were diagnosed between 2012 and 2022. The clinical data of 110 patients treated at Nagoya City University Hospital were used as the training cohort. The validation cohort included 203 patients treated at three affiliated hospitals. The clinical information was retrospectively reviewed. Missing PS data were estimated based on medical records. Overall survival (OS) was estimated using the Kaplan-Meier method and compared using the log-rank test. Univariate and multivariate analyses of OS were performed using Cox proportional hazard models. Results: In the training cohort with a median follow-up of 3.5 years (IQR 1.8-6.0), the median age was 76.5 years (range 70-92), with PS ≥2 (32%) and CCI ≥2 (24%). Seventy-two patients were classified as frail (SFS ≥ 2), and their OS was significantly shorter than that of non-frail patients (p = 0.031). Several cutoff points were tested to further stratify frail patients into moderate and severe categories (SFS 2 vs. 3-5, 2-3 vs. 4-5, 2-4 vs. 5). The best stratification was achieved when moderate frailty was defined as SFS 2 or 3. Severe frailty was defined as SFS 4 or 5. The median OS for non-, moderate-, and severe-frailty patients was 7.0, 5.3, and 2.8 years, respectively. Univariate analysis showed hazard ratios (HRs) of 1.74 (95% confidence interval [CI] 0.92-3.3, non- vs. moderate-frail) and 2.38 (95% CI 1.20-4.73, non- vs. severe-frail). The validity of these cutoff points was evaluated in the validation cohort (median follow-up 2.9 years [IQR 1.8-4.7]),