Search

ObjectiveMismatch Repair Deficiency (dMMR) / Microsatellite Instability (MSI) is a key biomarker in colorectal cancer (CRC). Universal screening of CRC patients for dMMR/MSI status is now recommended, but contributes to increased workload for pathologists and delayed therapeutic decisions. Deep learning has the potential to ease dMMR/MSI testing in clinical practice, yet no comprehensive validation of a clinically approved tool has been conducted.DesignWe developed an MSI pre-screening tool, MSIntuit, that uses deep learning to identify MSI status from H&E slides. For training, we used 859 slides from the TCGA database. A blind validation was subsequently performed on an independent dataset of 600 consecutive CRC patients. Each slide was digitised using Phillips-UFS and Ventana-DP200 scanners. Thirty dMMR/MSI slides were used for calibration on each scanner. Prediction was then performed on the remaining 570 patients following an automated quality check step. The inter and intra-scanner reliability was studied to assess MSIntuit’s robustness.ResultsMSIntuit reached a sensitivity and specificity of 97% (95% CI: 93-100%) / 46% (42-50%) on DP200 and of 95% (90-98%) / 47% (43-51%) on UFS scanner. MSIntuit reached excellent agreement on the two scanners (Cohen’s κ: 0.82) and was repeatable across multiple rescanning of the same slide (Fleiss’ κ: 0.82).ConclusionWe performed a successful blind validation of the first clinically approved AI-based tool for MSI detection from H&E slides. MSIntuit reaches sensitivity comparable to gold standard methods (92-95%) while ruling out almost half of the non-MSI population, paving the way for its use in clinical practice.

In some situations, there is a need for rapid mutation tests for guiding clinical decisions and starting targeted therapies with minimal delays. In this study we evaluated the turnaround time before and after the implementation of a fully automated multiplex assay for KRAS and NRAS/BRAF mutation tests (Idylla™ platform, Biocartis) in metastatic colorectal cancer. The objective of this project was to compare the turnaround times in 2017–2018 with the fully automated multiplex assay to the 2016 results with previous methods. Centers with a number of tests for metastatic colorectal cancer > 100 yearly and a usual turnaround time ≥ 3 weeks for mutation detection were selected. Results of 505 KRAS tests and 369 NRAS/BRAF tests were transmitted by 10 centers. The mean turnaround time from test prescription to reception of results was reduced from 25.8 days in 2016 to 4.5 days in 2017–2018. In conclusion, this pilot project shows that the Idylla™ platform for testing KRAS and NRAS/BRAF mutations allows an optimized turnaround time from test prescription to reception of results.

Resume En France, la determination du statut mutationnel des genes RAS a visee predictive de reponse aux traitements cibles anti-EGFR est realisee par les plateformes de genetique moleculaire des cancers creees par l’Institut national du cancer. Cette etude vise a demontrer la faisabilite de ces analyses par une structure liberale de pathologie (MEDIPATH) dans des conditions d’accreditation. Nous avons etudie retrospectivement le statut mutationnel des genes KRAS et NRAS dans 163 cas de cancer colorectal metastatique par technique Cobas ® . Nous avons compare nos resultats a ceux obtenus prospectivement par pyrosequencage et discrimination allelique par le laboratoire de genetique des tumeurs solides du CHU de Nice, (plateforme PACA-EST). Les resultats des deux series etaient concordants : correlation positive de 98,7 % ; correlation negative de 93,1 % ; correlation globale de 95,7 % (Kappa = 0,92). Nous avons analyse le cout de revient d’un test. Cette etude demontre la faisabilite des analyses moleculaires dans une structure liberale. Cette pratique necessitant un niveau eleve de garantie, son accreditation selon la norme NF-EN-ISO15189 est indispensable. La realisation d’analyses moleculaires dans ce contexte permet d’eviter les etapes d’acheminements du prelevement et du resultat, entre le laboratoire de pathologie et la plateforme, ce qui permet de reduire le delai global de rendu du resultat. En conclusion, le transfert de certaines analyses des plateformes vers des structures liberales permettrait aux plateformes d’etre dechargees d’une partie d’analyses courantes, validees et de pouvoir consacrer leurs efforts au developpement de nouvelles analyses continuellement requises pour l’acces a la medecine personnalisee.

In France, determination of the mutation status of RAS genes for predictive response to anti-EGFR targeted treatments is carried out by public platforms of molecular biology of cancer created by the French National Cancer Institute. This study aims to demonstrate the feasibility of these analyses by a private pathology laboratory (MEDIPATH) as per the requirements of accreditation. We retrospectively studied the mutation status of KRAS and NRAS genes in 163 cases of colorectal metastatic cancer using the Cobas