BACKGROUND: The clinical utility of comprehensive genomic profiling (CGP) for guiding treatment has gradually become the standard-of-care procedure for colorectal carcinoma (CRC). Here, we comprehensively assess emerging targeted therapy biomarkers using CGP in primary CRC.METHODS: A total of 575 primary CRCs were sequenced by ACTOnco® assay for genomic alterations, tumour mutational burden (TMB), and microsatellite instability (MSI).RESULTS: Eighteen percent of patients were detected as MSI-High (MSI-H), and the remaining cases were classified as microsatellite stable (MSS). Driver mutation prevalence in MSS CRCs were APC (74%), TP53 (67%), KRAS (47%), PIK3CA (21%) and BRAF (13%). The median TMBs for MSI-H and MSS patients were 37.8 mutations per mega base (mut/Mb) and 3.9 mut/Mb, respectively. Forty-seven percent of MSI-H CRC harboured at least one loss-of-function mutations in genes that may hamper immune checkpoint blockade. Among MSS RAS/RAF wild-type CRCs, 59% had at least one actionable mutation that may compromise the efficacy of anti-EGFR therapy. For late-stage CRC, 51% of patients are eligible for standard care actionability and the remaining 49% could be enrolled in clinical trials with investigational drugs.CONCLUSIONS: This study highlights the essential role of CGP for identifying rational targeted therapy options in CRC. <<<

Clinical workflows in oncology rely on predictive and prognostic molecular biomarkers. However, the growing number of these complex biomarkers tends to increase the cost and time for decision-making in routine daily oncology practice; furthermore, biomarkers often require tumour tissue on top of routine diagnostic material. Nevertheless, routinely available tumour tissue contains an abundance of clinically relevant information that is currently not fully exploited. Advances in deep learning (DL), an artificial intelligence (AI) technology, have enabled the extraction of previously hidden information directly from routine histology images of cancer, providing potentially clinically useful information. Here, we outline emerging concepts of how DL can extract biomarkers directly from histology images and summarise studies of basic and advanced image analysis for cancer histology. Basic image analysis tasks include detection, grading and subtyping of tumour tissue in histology images; they are aimed at automating pathology workflows and consequently do not immediately translate into clinical decisions. Exceeding such basic approaches, DL has also been used for advanced image analysis tasks, which have the potential of directly affecting clinical decision-making processes. These advanced approaches include inference of molecular features, prediction of survival and end-to-end prediction of therapy response. Predictions made by such DL systems could simplify and enrich clinical decision-making, but require rigorous external validation in clinical settings. <<<