Selected Domains within a Comprehensive Geriatric Assessment in Older Patients with Non-Hodgkin Lymphoma are Highly Associated with Frailty.

Background: The incidence of Non-Hodgkin Lymphoma (NHL) is increasing, particularly among older patients who tend to have worse outcomes and can be predisposed to increased toxicities and less treatment tolerance. Therefore, a thorough pre-treatment assessment is essential. A comprehensive geriatric assessment (CGA) can be used to evaluate the older patient considering chemotherapy and is the preferred evaluation tool. However, a formal CGA is laborious, complex and time-consuming.

Objectives: To characterize older adults with NHL and determine the CGA variables with the greatest association to frailty in order to propose a more simplified assessment.

Methods: We performed a cross-sectional study using data collected from CGAs in NHL patients > 65 years admitted to our oncology service, from September 2015 to August 2017. Our evaluation parameters included: polypharmacy, a screening tool of older people's prescriptions (STOPP), the Lawton scale, Barthel index, Katz index, gait speed, a Timed Up and Go (TUG) test, a Mini-Mental state examination (MMSE), the Yesavage and Gijon scales, a Mini-nutritional assessment (MNA), a Geriatric Syndromes assessment, and a Cumulative Illness Rating Scale-Geriatric (CIRS-G). The formal CGA was comprised of nine domains; frailty was defined as an impairment in > 2 domains. Each parameter was individually compared with frailty, and the results were used to build different multivariate models using logistic regression analyses to obtain the variables with the highest frailty association.

Results: A total of 253 patients were included. Their median age was 75.4 years (range 65-92), and 62.1% had > 1 impaired domain, with 39.9% considered frail. Bivariate analysis showed strong associations with age > 85 and all the geriatric parameters except for STOPP. Our final multivariate analysis resulted in 5 domains (the use of > 5 medications, a Lawton < 7, TUG > 20, Yesavage > 5, and the presence of at least one geriatric syndrome) being significantly associated with frailty and performing similarly to a CGA.

Conclusion: In our population of older NHL patients, an abbreviated evaluation based of only five domains, polypharmacy, TUG, Lawton scale, Yesavage scale and the presence of at least one geriatric syndrome, had similar performance to a formal CGA in determining frailty.