Colorectal Cancer (CRC) incidence and the impact of detecting it early, make CRC diagnosis a primary focus in the oncology community. Indeed, up to 95% of patients can be cured by tumor resection if the CRC is detected at an early stage. The most widespread techniques for CRC diagnosis are colonoscopy and stool tests. However, these existing approaches fail at generating a high level of compliance among the patients, because of the costs, invasive nature and/or of the inconvenience of the tests. Thus, there is urgent need for a sensitive and specific, non-invasive solution for early CRC diagnosis. Over the past year, thanks to the ERC PoC grant Monomark, we further developed a simple blood-based test (Hamm A., et al. Gut, 2016, international patent application WO2013/110817) to address this unmet medical need. The test is based on the measurement of a 23-gene expression signature in a specific subpopulation of cells, in casu human monocytes, extracted from standard blood samples. In partnership with DNAlytics, Biogazelle and UZ Leuven, we collected about 150 samples in order to develop the assay in a technological setting fitting routine requirements. The entire workflow has been optimized and turned into a standardized non-invasive screening solution for CRC under the name "ColonoKit". Isolation of the monocytes can now be easily performed in the clinical labs of most hospitals according to a quick and user-friendly standard procedure. Moreover, we validated a transportation kit that allows the storage and shipping of the monocytes at ambient temperature to remote labs for gene expression analysis. Several technical variations were also introduced in the gene expression analysis procedure to meet the highest international standards. The data produced with the new standardized industrial laboratory procedure enabled the building of a new classification model with good predictive performance that are slightly under what was previously observed (sensitivity about 74% and specificity about 89%). However, compared to the original study, disease specificity was lost since we could also detect non-cancerous inflammatory conditions and other cancer types. In parallel, to disclose the diagnostic power of our standardized diagnostic platform head to head with the well-established iFOB diagnostic test, we prospectively recruited 253 patients with a positive stool test (iFOB+). In this previously unexplored setting, we regret to say that neither the new clinical grade diagnostic platform nor the original model have been able to efficiently identify the CRC patients within the iFOB+ population (i.e. quasi random classification). This unexpected drop in performance is now under investigation with ad hoc analyses on the entire workflow and with additional studies aiming at estimating the influence of bowel preparation and fasting regimen on genes and model predictions. If the new analyses and studies will succeed in rescuing the performance within an acceptable range, the main goal will be to position our test as a second line screening after a patient has resulted iFOB positive. In such a perspective, introducing ColonoKit in the screening campaign as a confirmatory step before colonoscopy could result in a decrease of up to 50% in the number of useless or non-urgent colonoscopies thus reducing the social and economic burden of false positive patients that undergo a colonoscopy.
