Development and validation of a Medicare health outcomes survey (MHOS) frailty index in multiple myeloma (MM).

Abstract

e21519 Background: Increasing age is associated with poorer outcomes in MM. However, biological aging (BA) is heterogeneous and does not correlate perfectly with chronological age (CA). A recently developed frailty of CA, functional status, and comorbidities is more predictive of toxicity, progression-free and overall survival (OS) than CA alone (Palumbo et al Blood 2015). We sought to apply the Rockwood Accumulation of Deficits approach to develop and validate a frailty index (FI) for older adults with MM. Methods: MHOS, a random survey of HRQOL (health related quality of life) in Medicare Advantage HMO enrollees, data were obtained from 1,747,042 non-cancer patients (pts) and 340 with newly diagnosed MM 1997-2011 from the SEER-MHOS linked dataset. In creating the FI, we considered only variables available throughout all 4 survey versions used since its initiation in 1998. We identified 34 variables including items related to function/mobility, comorbidities, and mental health. The FI was calculated as the number of deficits present, divided by the number total number of possible deficits, with possible scores ranging from 0 to 1 and higher values indicating more frail. Missing data points were excluded from the FI calculation. Cox regression analysis was performed to determine the association of the FI with OS, controlling for CA, gender, and race. The FI was first derived in the non-cancer cohort then applied in the MM cohort. Results: The median age was 72 for non-cancer pts and 75 for pts with MM; the median FI was 0.308 and 0.368, respectively. Among non-cancer pts, each 10% increase in FI (~3-4 more deficits) was associated with a 35% increased risk for death (aHR 1.352, 95% CI 1.350-1.354, p < 0.001); among pts with MM it was associated with a 15% increased risk (aHR 1.147, 1.065-1.236, p < 0 .001). Median OS for pts with MM in the lowest tertile of FI was 48 months (95% CI 42-56) compared to 23 months (17-33) for the highest tertile. Conclusions: Here we present a MHOS-based FI. The FI was more prognostic among non-cancer pts than among pts with MM, presumably because stage and tumor characteristics also contribute to OS in MM pts. Further exploration of the MHOS-FI in MM and other cancers is warranted.