Background: The PURPLE Translational registry supports analysis of large amounts of structured clinical and molecular data from routine clinical practice, into a single collaborative data repository. We explored whether, by utilising digital health technology, we could potentially enable big data insights into PDAC.
Methods: The PURPLE Translational Pancreatic Cancer registry is a collaborative effort among 27 institutes. Employing an electronic web-based platform for entry of key clinicopathological and outcome data on consecutive patients with PDAC. This federated platform allows de-identified data to be combined and analysed for research purposes, whilst maintaining privacy, confidentiality and data security, with the goal of supporting clinical, genomic, and translational research.
Results: Between January 2016 and June 2019, 1279 PDAC patients with >400 matched biospecimens, 369 resections and 486 biopsy specimens have been entered in PURPLE. Median age at diagnosis was 69 (range 20-94 years), and 681/1279 (53%) were male. Overall 754/1279 (59%) patients presented with localised disease; 350/1279 (27%) were deemed resectable, 148/1279(12%) borderline resectable, and 247/1279 (20%) unresectable. A further 445/1279 (35%) were metastatic and 80 (6%) were not fully staged. Targeted molecular sequencing performed in 143 PDAC patients identified mutations in KRAS in 131 (92%), BRAF in 1 (1%), TP53 in 72 (71%), CDKN2A in 41 (41%), and PIK3CA in 2 (2%). BRCA/ BRCA-like signatures were identified in 5 of 75 cases undergoing Genome Sequencing (WGS). Since September 2019, 20 WGS cases were reviewed by the Molecular Tumour Board and Precision Oncology Program at the Victorian Comprehensive Cancer Centre. Research in circulating tumour DNA, biomarkers, and organoid models is also being supported by the PURPLE translational pancreatic cancer platform.
Conclusion: The comprehensive data collected in the PURPLE registry supports a broad range of research focused on precision oncology, including linking molecular data that can help identify candidates for targeted interventions.