Integration of Geriatric Measures Alongside Disease-Based Measures to Advance Precision Oncology for Older Veterans with Multiple Myeloma

尽管主要的癌症关注组织，如美国临床学会

Although leading cancer-focused organizations such as the American Society of Clinical Oncology now recommend geriatric assessments for all older adults with cancer, these assessments have not been widely adopted within VA and elsewhere due to limitations in resources, geriatrics expertise, and time in busy oncology clinics. The VA National Precision Oncology Program is profiling tumor genetics in myeloma and other cancers to characterize disease and personalize therapy in unprecedented fashion. However, truly personalized cancer treatment for older Veterans must be based not only on tumor “genotype,” but also on older adult “phenotype.” Dr. Clark DuMontier is a geriatrician and clinical investigator focused on integrating geriatric measures into oncology. The research proposed for this CDA-2 will advance Dr. DuMontier’s long-term goal of developing feasible and valid tools to help VA oncologists integrate frailty and multimorbidity alongside myeloma-specific factors to advance precision oncology for older Veterans with myeloma. The objectives of this application are to electronically measure frailty and multimorbidity in over 5000 Veterans age ³ 65 years with myeloma newly treated in VA from 2004-present using healthcare data that is readily available within VA’s nationally integrated health system. Preliminary findings demonstrate that frailty and multimorbidity can be rapidly assessed using diagnostic and procedural codes and data from the electronic health record (EHR). Aim 1 will determine whether an electronic frailty index, the VA-FI, independently predicts mortality and hospitalizations in older Veterans with myeloma, and whether it modifies the effect of initial therapy on these outcomes. Analyses will include important myeloma-specific factors such as triplet or doublet chemotherapy, stage and cytogenetics, sociodemographic variables, and prognostic labs. Aim 2 will determine whether multimorbidity patterns independently predict mortality and hospitalizations in myeloma, and whether they modify the effect of specific regimens on these outcomes. A machine learning analysis will be applied to 67 chronic conditions measured within the VA database to define these patterns and their impact in older Veterans with myeloma. Aim 3 will develop and validate a predictive risk model for mortality that incorporates frailty and multimorbidity with myeloma-specific factors. Once validated, this model will be translated into a clinical decision support tool embedded in the VA EHR. This tool will help VA oncologists to rapidly estimate frailty and multimorbidity and enhance individualized prognosis and treatment decisions for older Veterans with myeloma. The mentored research and training program described in this CDA-2 will accelerate Dr. DuMontier’s development into an independent VA investigator in geriatric oncology. His mentorship team includes Drs. Jane Driver, Nikhil Munshi, Michael Gaziano, and Mary Brophy—leaders in geriatrics, oncology, and big data science from the New England Geriatric Research Education Clinical Center (GRECC), the Massachusetts Veterans Epidemiology Research and Information Center (MAVERIC), and the VA National Precision Oncology Program. Under the guidance of these mentors and in collaboration with MAVERIC, Dr. DuMontier will pursue training that will provide (1) experience with large database investigation; (2) expertise in advanced statistical analyses and machine learning; (3) project management and multidisciplinary collaboration; and (4) communication, leadership skills, and professional development. This training and mentored research will prepare him to be competitive for a VA Merit Award toward independence as a VA clinician investigator.