Transplantation for Older Adults-More Questions than Answers

Preconditioning Frailty Phenotype Influences Survival and Relapse for Older Allogeneic Transplantation Recipients

Hematologic malignancies disproportionately affect older adults. Hematopoietic cell transplantation (HCT) is potentially curative, but poor overall survival (OS) has limited its use in older adults. Fried's frailty phenotype (FFP) is a geriatric assessment tool that combines objective and subjective performance measures: gait speed, grip strength, activity level, exhaustion, and weight loss. People meeting ≥3 criteria are classified as frail; 1 or 2 criteria, as pre-frail; and 0 criteria, as fit. To evaluate the association of pre-HCT FFP with post-HCT outcomes, we assessed FFP prior to conditioning for 280 HCT recipients age ≥60 years with acute leukemia or a myeloid neoplasm at 3 institutions. When analyzing survival by age group, patients age ≥70 years had inferior OS compared to patients age 60 to 69 years (P = .002), with corresponding OS estimates of 38.9% (95% confidence interval [CI], 27.8% to 49.9%) and 59.3% (95% CI, 51.9% to 65.9%). Nonrelapse mortality (NRM) also was significantly higher in the older patients (P = .0005); the 2-year cumulative incidences of NRM were 38.5% (95% CI, 27.5% to 49.2%) and 17.2% (95% CI, 12.3% to 22.8%), for older and younger recipients, respectively. The cumulative incidences of relapse did not differ by age group (P = .3435). Roughly one-third (35.5%) of the patients were fit, 57.5% were pre-frail, and 7.5% were frail, with corresponding 2-year OS estimates of 68.4% (95% CI, 57.9% to 76.8%), 45.5% (95% CI, 37.4% to 53.2%), and 45.8% (95% CI, 23.4% to 65.8%) (P = .013). FFP was not significantly associated with NRM, but being frail or pre-frail was associated with a higher rate of disease-related deaths (33.3% and 27.3%, respectively, compared with 17.4% for fit patients; P = .043). In univariate modeling of restricted mean survival time with a 3-year horizon (RMST_3y), the factors that were significantly associated were FFP, age, Karnofsky Performance Status (KPS), Disease Risk Index (DRI), and HCT-specific Comorbidity Index (HCT-CI). Of those factors, only FFP (P = .006), age (P = .006), KPS (P = .004), and DRI (P = .005) were significantly associated in multivariate modeling of RMST_3y. Estimates of RMST_3y were computed and 5 risk-groups were created with survival ranging from 31.4 months for those who were age 60 to 69 years, fit, had KPS 90 to 100, and low/intermediate-risk DRI compared to 10.5 months for those who had high-risk features for all the evaluated factors. In univariate and multivariate analyses for restricted mean time to relapse with a 3-year horizon (RMRT_3y), FFP (pre-frail versus fit, P = .007; frail versus fit, P = .061) and DRI (P = .001) were the only significant factors. Predicted RMRT_3y was longest (30.6 months) for those who were fit and had low/intermediate-risk DRI scores and shortest (19.1 months) for those who were frail and had high-risk or very high-risk DRI scores. Both age and FFP impact survival after HCT. Incorporation of FFP into pre-HCT evaluations may improve decision-making and counseling regarding HCT risk for older adults. Our findings support future trials designed to reverse frailty, such as pre-HCT supervised exercise programs, and correlative analyses to unravel the connection of frailty and relapse to generate future targets for intervention. Finally, exploration of novel HCT platforms to reduce relapse in pre-frail and frail patients, as well as reduce NRM in adults age >70 years, are warranted.