Assistant Professor McMaster University, Hamilton, Ontario, Canada Hamilton, Ontario, Canada
Introduction: Existing prognostic tools in multiple myeloma (MM) are designed specifically for clinicians and incorporate variables often unknown to MM patients (i.e. cytogenetics). Developing prognostic tools for patient use represents an important step in empowering patients to understand their prognosis and engage in shared decision making. Incorporating patient reported outcomes (PROs) may represent an opportunity for developing prognostic tools. The implementation of standardized cancer symptom assessment in clinics in Ontario, Canada since 2007 provides a unique opportunity to evaluate the prognostic ability of PROs.
The objective of our study was to develop and validate a prognostic score for overall survival among transplant ineligible newly-diagnosed multiple myeloma (TI NDMM) patients incorporating PROs.
Methods: This was a retrospective population-based, prognostic study using administrative data of all patients with TI NDMM in Ontario diagnosed between Jan 2007-Dec 2019. Index date was defined as one year following diagnosis to ensure only patients who did not receive a transplant in first line were included. Patients were randomly selected for model derivation (75%) and validation (25%). The derivation cohort was used to develop a multivariable Cox proportional hazards regression model with backward stepwise variable selection to predict the risk of death one year following index date. Baseline covariates at index date included demographic characteristics, clinical information, health care utilization, performance status, and PROs using the Edmonton Symptom Assessment Score, which measures 9 common cancer symptoms (e.g. pain, dyspnea) on a 10-point scale, where 7-10 is high burden.
Results: There were 2356 TI NDMM patients identified. The median age was 75 with 42.2% female sex. The majority of patients (90.3%) had received a novel drug (IMID and/or PI) following diagnosis. The following factors were associated with an increased risk of death: age >80 (HR 1.11), previous history of congestive heart failure (HR 1.52), CRAB symptoms at diagnosis (HR 1.61), distance of >50 km to cancer center (HR 1.25), radiation received in the year prior (HR 1.48), no novel drugs received in the year prior (HR 1.36), emergency department (HR 1.55) or hospitalization (HR 2.13) in the prior 6-months, poor performance status (ECOG 3-4 HR 1.76) and increasing number of symptoms with high burden (HR 1.56) in the prior 6-months. Model discrimination in the validation cohort was high with C-statistic of 0.74, and calibration plots indicated strong agreement between observed and predicted risks.
Conclusions: To our knowledge, this represents one of the first prognostic models developed in NDMM designed specifically for patients, using easily available variables that can be easily completed by patients themselves including PROs. This survival prognostic tool may improve communication regarding prognosis and shared decision making among older adults with MM and their health care providers.
