P896: DEVELOPMENT AND VALIDATION OF A FRAILTY PREDICTION MODEL MORE SUITABLE FOR MULTIPLE MYELOMA IN CHINESE PATIENTS: THE TM FRAILTY SCORE.

Abstract

Background: Frailty is associated with adverse clinical outcomes in patients with multiple myeloma (MM). Accurate evaluation of frailty is critical when choosing appropriate chemotherapy regimens for elderly patients. Currently, the IMWG geriatric assessment (GA) is widely used to assess tolerance to chemotherapy for elderly patients. However, the applicability of the IMWG GA in the Chinese MM population is limited. Aims: To developed a new frailty model that is more suitable for Chinese patients with MM. Methods: The applicability of the IMWG GA was evaluated in 167 consecutive patients diagnosed with MM from June 2019 to September 2021 at the First Affiliated Hospital of Sun Yat-sen University. Eight comprehensive geriatric assessment (CGA) domains were analyzed in 135 of these patients to screen items most closely associated with the outcomes, including grade ≥ 3 adverse events (AEs), treatment discontinuation, TTP, and OS. A new frailty prediction model was developed, and its fitness and predictive performance were compared to those of the IMWG GA. Results: The patients in the IMWG GA Frail group had a higher rate of grade ≥3 AEs, a higher rate of treatment discontinuation, and a greater risk of death than those in the Fit group (P<0.05). No significant differences were observed in patients in the Int-fit group who experienced chemotherapy AEs, treatment discontinuation, and TTP compared with the Fit group. By screening items in all 8 CGA domains, multivariate analysis confirmed that the Timed Up and Go test（TUG） and the Mini Nutritional Assessment Short-Form（MNA-SF） were independent prognostic factors for grade ≥3 AEs and OS. We found that the TUG and MNA-SF scores could stratify the risk of grade ≥ 3 AEs in the IMWG GA Int-fit group (P < 0.05). After MNA-SF was added to construct the IMWG GA Plus model, Harrell’s concordance index (C-index) of the IMWG GA Plus model for the prediction of grade ≥3 AEs increased from 0.662 to 0.701. The C-index for treatment discontinuation rose from 0.636 to 0.669. We further combined the TUG test and the MNA-SF to construct the TM frailty scoring system. The KM curve analysis showed that the TM frailty score could well distinguish the risk of AEs between the Fit and Frail groups (P<0.05). The C-index of the TM frailty score was 0.741, 0.690, and 0.702 for grade ≥ 3 AEs, treatment discontinuation, and OS, respectively, suggesting good predictive ability, which was higher than that of the IWMG GA (C-index=0.662, 0.636 and 0.631, respectively) and the IMWG GA Plus model (C-index=0.701, 0.656 and 0.618, respectively). Image:Summary/Conclusion: A novel scoring system—the TM frailty score—was developed by combining the TUG (T) and MNA-SF (M). The TM frailty score is more appropriate than the IMWG GA for evaluating frailty in the Chinese population.