Performance of time to discontinuation and time to next treatment as proxy measures of progression-free survival, overall and by treatment group.

Abstract

e19135 Background: Real-world data sources have historically provided limited information about the occurrence of disease progression. As a result, progression-free survival (PFS), a standard effectiveness outcome measure in oncology, has been estimated using proxies such as time-to-discontinuation (TTD) or time-to-next-treatment (TTNT). Curation of unstructured electronic medical record (EMR) data make it possible to calculate PFS from directly observed progression. The aim of this research was to compare TTD and TTNT as proxy indicators with PFS determined from direct statements regarding progression in the medical record and to compare effect sizes on treatment-based groups across these endpoints. This report cites findings from the metastatic breast cancer (mBC) setting. Methods: Previously curated EMR data were collected from Concerto’s Definitive Oncology Dataset for adult, female, hormone receptor-positive/human epidermal growth factor receptor-negative mBC patients diagnosed 2008 or later who received at least one line of systemic therapy. Patients were grouped based on any receipt vs. no receipt of an aromatase inhibitor (AI) in first line (1L). Kaplan-Meier and Cox regression methods were used to calculate PFS, TTD, and TTNT in 1L and OS from start of 1L. Results: The study included 378 patients, of which 138 (36.5%) were de novo stage IV or metastatic. The mean patient age was 60.3 years (SD 13.3), 57.1% were Caucasian, and 68.3% were postmenopausal. Conclusions: Estimates of TTD and TTNT aligned closely, suggesting that nearly all patients who discontinued initial therapy received subsequent treatment. Notably, TTNT and TTD were both meaningfully shorter than PFS, suggesting that their use as proxy indicators may systematically underestimate PFS. Further, analysis based on receipt of AI indicates that comparative effects on TTD or TTNT may produce quite different results than those for PFS and OS. [Table: see text]