Development of a multi-institutional integrated resource platform linking primary and secondary biospecimens in gynecologic cancers

Background: Gynecologic cancers exhibit substantial molecular and histopathologic heterogeneity, highlighting the need for infrastructures that integrate high-quality biospecimens with multi-omics and functional resources. To address this need, we established a seven-center framework that collects standardized primary biospecimens and generates secondary resources for gynecologic cancers, supported by integrated clinical and pathological datasets. Methods: Primary biospecimens-including serum, plasma, buffy coat, urine, ascites, frozen tumor tissues, and formalin-fixed paraffin-embedded (FFPE) blocks-were collected under Standard Operating Procedures (SOPs), with pre-analytical variables recorded using the Standard PRE analytical Code (SPREC). Frozen tissues supported high-depth whole-genome sequencing (WGS) and the development of patient-derived xenograft (PDX) and organoid models. FFPE blocks were used to construct tissue microarrays (TMAs). Human ovarian surface epithelial (HOSE) cells were subjected to standard immortalization procedures to establish stable cell lines. Results: As of November 5, 2025, 372 patients (161 endometrial, 101 ovarian, 104 cervical cancers, and 6 rare tumors) had been enrolled, yielding more than 6,000 primary biospecimens. WGS was performed using tumor-normal paired samples when available, generating 403 datasets: 215 tumor genomes (60×/90×) and 188 normal genomes (30×), representing 191 endometrial, 35 ovarian, and 177 cervical cancer cases. TMAs were constructed with 373 endometrial, 260 cervical, 490 ovarian cores. In addition, 17 PDX models were established (7 endometrial, 8 cervical, 2 ovarian), along with two organoid models and five IHOSE cell lines. Importantly, the primary biospecimens, WGS datasets, and PDX models were all generated from the same patients, creating a patient-matched resource that enables direct integration of genomic, pathologic, and functional data at the individual-patient level. Conclusions: This multi-institutional platform integrates standardized primary biospecimens with a broad set of secondary resources, providing a rare patient-linked structure in gynecologic cancers. The combined availability of WGS datasets, TMAs, PDX and organoid models, and IHOSE cell lines supports genomic-pathologic correlation, biomarker development, and functional studies of therapeutic response. Its scalable design offers a durable infrastructure for advancing gynecologic cancer research.