Overview
Multiple myeloma is the third most common blood cancer in Japan, requiring robust evidence to guide care across diverse patient groups. However, existing Japanese data sources—such as health insurance claims and traditional cancer registries—often lack granular clinical details, including high-risk cytogenetic test results, detailed treatment sequencing, and minimal residual disease status. To address these limitations, a team of oncology clinical experts, statisticians, and data engineers curated an electronic health record (EHR)-derived dataset that combines structured records (such as lab values and prescriptions) with unstructured clinical documentation (such as physician notes and cytogenetic reports).
In an initial evaluation of 36 patients with multiple myeloma, the dataset demonstrated exceptional clinical depth and completeness. Overall, 94% of patients had documented FISH testing for key genetic markers, with 59% harboring high-risk alterations. The dataset also successfully tracked complex treatment pathways, capturing autologous stem cell transplantation in 47% of patients and triple-class drug exposure in 42%.
Why this matters
This EHR-derived dataset fills critical gaps in Japan’s healthcare research infrastructure by providing the deep clinical and genetic detail needed for advanced multiple myeloma research. This approach is scalable across cancer hospitals in Japan; as it expands, this dataset can potentially enable researchers to evaluate real-world treatment sequences, construct external control arms for novel therapies (such as CAR T-cell therapy and bispecific antibodies), and inform regulatory strategies across hematologic malignancies in Japan.