Comprehensive geriatric assessment prior to autologous stem cell transplant as a predictor of patient outcomes: A retrospective review.

Abstract

e23023 Background: While autologous stem cell transplant (ASCT) has been shown to be safe and effective in the geriatric population, age alone is not a sufficient predictor of outcomes. Minimal data exists regarding the use of a comprehensive geriatric assessment (CGA) tool prior to ASCT as a predictor of patient outcomes. This study aims to assess our institution’s pre-ASCT CGA on its ability to predict transplant-related outcomes in the elderly. Methods: We retrospectively analyzed all patients age 65 years of age and older (n = 68) who received ASCT at Thomas Jefferson University Hospital between 2012 and 2018 for multiple myeloma (MM) (n = 52), AL amyloidosis (n = 1), combined MM/amyloidosis (n = 2), B-NHL (n = 11), T cell lymphoma (n = 2), and Hodgkin lymphoma (n = 1). The CGA consists of 13 components, culminating in an overall frailty status. Kaplan-Meier survival estimates were used to assess overall survival (OS) and progression free survival (PFS), and multivariable logistic regression and Cox proportional hazard regression were used to determine associations between CGA variables and relevant post-ASCT outcomes. Results: Among the 68 patients (43 males, 25 females) who received ASCT, the median age was 68 years (range 65-78). Fifty-one patients (75%) underwent pre-transplant CGA. Among the entire cohort, 2-year OS and PFS were 95.4% and 79.6%, respectively. Median OS was not reached, and median PFS was 48.5 months. There was no difference in OS between patients who underwent CGA and those who did not (P = 0.937). Multivariate analysis revealed that none of the variables in the CGA were predictive of patient outcomes, such as OS, length of stay on ASCT admission, 30-day readmissions, and falls. Conclusions: This single-institution data suggests that ASCT is safe and highly effective in geriatric patients. Undergoing a CGA was not associated with improved outcomes, and none of the CGA components including frailty status were predictive of outcomes in this preliminary data set. Our data suggest that further refinement of the CGA may improve its predictive ability. Future studies examining additional outcomes, including patient-reported quality-of-life outcomes, are warranted.