Found 1 Presentation For Request "920P"
920P - Blind validation of MSIntuit, an AI-based pre-screening tool for MSI detection from colorectal cancer H&E slides
- Magali Svrcek (Paris, France)
Abstract
Background
MisMatch Repair deficiency (dMMR) / Microsatellite instability (MSI) is a crucial biomarker in colorectal cancer (CRC) predictive of response to immunotherapy and Lynch syndrome. Universal screening of newly diagnosed CRC for dMMR/MSI status is now recommended. With the growing numbers of biomarkers screened in clinical practice, dMMR/MSI testing, currently diagnosed by immunohistochemistry (IHC) and/or polymerase chain reaction (PCR), contributes to increasing pathologists workload and delaying therapeutic decisions. Artificial intelligence (AI) models detecting MSI tumors directly from H&E slides have shown promise in improving MSI patients' diagnosis. MSIntuit, an AI-based pre-screening tool developed by Owkin for MSI detection from H&E whole slide images (WSI) of surgical resection specimens from patients with CRC, outputs if the patient is likely to be MSI and should get further testing or not. In this study, we performed a one-shot blind validation of MSIntuit on a large external cohort of WSI from consecutive CRC.
Methods
H&E WSI of 600 consecutive resected CRC (including n=123 dMMR/MSI cases) diagnosed at Medipath pathology laboratories in 2017/2018 were studied. dMMR status was assessed using IHC for the 4 MMR proteins, and confirmed by PCR for doubtful cases. Slides were digitized with 2 scanners (Phillips UFS, Roche DP200) and 30 dMMR/MSI WSI were used to calibrate the tool on each scanner. To assess performance in the most robust way, inference was done on the remaining 570 patients blinded to their status. Automated quality check (QC) discarded WSI that did not meet the tool requirements (large blurry/artifact regions, too few tumor tissue).
Results
QC led to the exclusion of 5% / 2% of WSI digitized with DP200 / UFS scanners, respectively. MSIntuit reached sensitivity/specificity of 0.97 [0.93-1.0] / 0.46 [0.42-0.5] on DP200 scanner and 0.95 [0.9-0.98] / 0.47 [0.43-0.51] on UFS scanner.
Conclusions
We performed the first successful blind validation of an AI-based tool for MSI detection from H&E WSI. MSIntuit reaches sensitivity comparable to gold standard methods (92-95%) while reducing almost by 40% the number of patients to screen with standard techniques, paving the way for MSIntuit use in clinical practice.
Legal entity responsible for the study
Owkin.
Funding
Owkin.
Disclosure
M. Svrcek: Financial Interests, Personal, Advisory Role, Consulting or Advisory Role: BMS, Astellas, MSD Oncology, Sanofi; Financial Interests, Personal, Other, Travel, accommodations, expenses: BMS, Ventana/Roche; Financial Interests, Personal, Advisory Role, Consulting: Owkin. C. Saillard, R. Dubois, N. Loiseau, F. Brulport, J. Guillon, M. Auffret, M. Sefta, A. Fouillet: Financial Interests, Institutional, Full or part-time Employment: Owkin; Financial Interests, Institutional, Stocks/Shares: Owkin. A. Kamoun: Financial Interests, Institutional, Stocks/Shares: Owkin. P. Courtiol: Financial Interests, Institutional, Stocks/Shares: Owkin. S. Rossat: Financial Interests, Institutional, Full or part-time Employment: Medipath. F. Renaud: Financial Interests, Personal, Advisory Role: BMS, MSD. G. Wainrib: Financial Interests, Institutional, Leadership Role: Owkin; Financial Interests, Institutional, Full or part-time Employment: Owkin; Financial Interests, Institutional, Ownership Interest: Owkin. All other authors have declared no conflicts of interest.