Introduction Previous studies have shown that available tools such as ‘Adjuvant Online!” are not able to accurately predict the prognosis of patients aged 65 years or older with breast cancer. Furthermore, all available tools predict prognosis in terms of recurrence-free survival or overall survival, whilst the risk of other-cause mortality is often high in the older patient with breast cancer. This is highly relevant as it may influence treatment decisions. Patient characteristics such as comorbidity and various geriatric variables have shown to be predictive for these outcomes and could enhance the precision of prognostic tools for this target population. The objective of this study was to develop a prediction tool for recurrence, survival and other-cause mortality for older patients with breast cancer who received locoregional treatment, with incorporation of patient-, tumor- and geriatric variables. The tool additionally predicts expected benefits of systemic treatment. Methods Data from the large population-based FOCUS cohort was used, consisting of consecutive breast cancer patients in the South-Western part of the Netherlands, diagnosed between 1997 and 2004, aged 65 years and older. It contains detailed information on tumor characteristics, treatment, comorbidity and geriatric parameters. We developed a risk prediction model using a Cox proportional hazards regression model for overall survival and cause-specific Cox proportional hazards models for recurrence and other-cause mortality (defined as mortality without recurrence). The included predictors were derived from the PREDICT tool (consisting of age and various tumor variables), since this tool was previously shown to have the best performance in older adults so far. Predictors were complemented with comorbidity and geriatric variables. Discrimination accuracy was evaluated using time-dependent area under the curve (AUC). The potential annual benefit of chemotherapy was calculated assuming a relative risk of chemotherapy on recurrence of 0.7, derived from data from the most recent updates of the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Additional benefit of endocrine treatment will be included in further development of the tool. Results A total of 2,744 patients were included for the initial development. For all patients, 5-year follow-up was complete with a high event-rate including 343 recurrences and 831 total deaths of which 586 without recurrence. The strongest predictors for overall survival and non-recurrence mortality were age (HR = 2.14, 95% CI: 1.89 - 2.43 and HR = 2.87, 95% CI: 2.46 - 3.35, respectively) and dementia (HR = 1.52, 95% CI: 1.16 - 1.99 and HR = 1.9, 95% CI: 1.49 - 2.65, respectively), and for recurrence, nodal status (HR = 1.80, 95% CI: 1.45 - 2.24) and tumor grade (HR = 2.96, 95% CI: 1.88 - 4.66). The time-dependent AUC at 5 years for recurrence-specific and other-cause mortality were 0.78 (95% CI: 0.76 - 0.81), and 0.75 (95% CI: 0.72 - 0.78), respectively. The AUC for overall survival was 0.75 (95% CI: 0.72 - 0.78). External validation is currently being performed in a large dataset retrieved from the national cancer registry (N= 13,631). These results will be presented during the symposium. Conclusion We have developed a model for predicting 5-year recurrence, other-cause mortality and overall survival, including expected benefits of adjuvant treatment, for older patients with breast cancer, with a good discrimination performance within a large-population based cohort. To our knowledge, this is the first model specifically designed for the older population, including competing risk as a predicted outcome and with incorporation of geriatric variables.

Citation Format: Willeke van der Plas-Krijgsman, Daniele Giardiello, Hein Putter, Ewout W Steyerberg, Esther Bastiaannet, Anne M Stiggelbout, Simon P Mooijaart, Johanneke EA Portielje, Gerrit J Liefers, Nienke A de Glas. The PORTRET-tool: A prediction tool for older patients with breast cancer that predicts recurrence, survival and other-cause mortality [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-08.