OBJECTIVES: High quality, recent, oncology real world data (RWD) sources with clinically meaningful depth and completeness in Europe and Japan have previously been limited. Our objective was to develop oncology datasets for retrospective and prospective research.

METHODS: We use electronic health records (EHR) sources including both structured and unstructured information (eg, clinic notes, pathology reports). Sites include a range of regionally relevant cancer care providers (ie, NHS Trusts in the UK, hospitals and community clinics in Germany, cancer centers in Japan). Pre-specification of variables (eg, ECOG performance status, biomarker test result definitions) and outcomes (eg, mortality, progression, response) was established. Data curation with ongoing longitudinal follow-up is refreshed every 90 days. De-identified patient-level data is made secure for analysis in a trusted research environment (TRE) in accordance with all local laws and regulations.

RESULTS: Retrospective data from patients diagnosed with cancer between Jan 2015 and Dec 2023, and across disease stages, was successfully curated into research-ready datasets and pre-specified prospective follow-up of cohorts began in Jan 2024. Early insights into follow-up through 31 March 2024 are available. The prevalence of biomarker positivity in Europe breast cancer (n=402) was: HER2+ 14%, ER+ 79%, and PR+ 65%. The prevalence rate in Europe of biomarker positive results (as a % of those tested) in NSCLC (n=730) was: EGFR+ 10%, PDL1+ 54%, KRAS+ 37%. Among patients in Japan with colorectal cancer (n=390) and gastric cancer (n=421), we found 3.7% had an ECOG score of 2 or above at the time of their first treatment.

CONCLUSIONS: We demonstrated the feasibility of a secure and compliant environment to curate and combine real world patient-level oncology data across country borders for analysis that include Europe and Japan in addition to existing US data. Flatiron Health international EHR-derived datasets were developed with intention being fit-for-purpose for treatment comparative-effectiveness research.