Article Text
PDF
Abstract
Introduction/Background Ovarian high-grade serous carcinoma (HGSC) is commonly diagnosed at an advanced stage, showing multiple genetically heterogeneous clones existing prior to therapeutic intervention.
Methodology Clonal composition and topology were estimated from whole-genome sequencing data from 510 samples of 148 patients in the prospective, longitudinal, multiregion DECIDER study. Detected subclones were followed in 152 longitudinal circulating tumour DNA (ctDNA) samples, by using targeted sequencing. Multiomics data included H&E-stained images and RNA sequencing data.
Three evolutionary states were discovered in 148 HGSC patients from the DECIDER cohort, that showed distinct treatment responses, mutated pathways and growth patterns.
Results Based on phylogenies, patients were stratified into three evolutionary states with distinct subclonal heterogeneity within and between metastases:
1) Evolving state is mostly monoclonal and shows only some differences between sampled sites.
2) Maintaining state is highly polyclonal within sites, but shows similar, stable subclonal patterns between sampled sites.
3) Adaptive state has high heterogeneity between metastases that show parallel evolution, making them the most divergent.
Tumours belonging to maintaining state had worst prognosis (p=0.008). Prognosis difference was independent of homologous recombination (HRD), which was found equally in each of the three states.
States were characterized by distinct molecular pathways and morphological features. PI3K/AKT pathway was enriched in maintaining state tumours, and it was shown to be targetable with alpelisib in patient-derived organoids.
After treatment, most patients revealed heterogenic subclonal composition with acquired mutations and only few cases underwent selection. Evolutionary states were modified by the changing selection pressure caused by treatments.
Conclusion Our results revealed that HGSC tumours are not all highly heterogenic and genomically unstable at the time of diagnosis. We showed that prognosis is worst for maintaining tumours that grow in mixtures of multiple clones that remain genomically similar between metastases. These tumours had enriched PI3K/AKT alterations, which were successfully targeted by alpelisib monotherapy in organoids. Multiple lines of treatments altered diagnostic states and most tumours revealed acquired mutations at relapses, making the relapsed disease a moving target.
Disclosures The study has been funded by the European Union’s Horizon 2020 Research and Innovation Programme under Grant agreement no. 965193 (DECIDER).