Lynch syndrome (LS) is the most common hereditary colorectal cancer (CRC) syndrome and is characterized by an accelerated adenoma-carcinoma sequence, a relatively higher prevalence of flat and subtle CRC precursor lesions, and exceptionally high adenoma miss rates despite intensive colonoscopy surveillance.
Existing LS-specific data suggest that AI can be safely integrated into high-quality surveillance. Meanwhile, use of AI has not yet been demonstrated to aid in improving overall adenoma or advanced neoplasia detection rates when used by expert colonoscopists, and when adequate baseline procedural quality is guaranteed.
Towards the future, efforts are required to develop more LS-specific training datasets and algorithms tuned to adenomas that are flat and have an MSI-driven morphology. Moreover, biologically aligned endpoints and multicenter prospective trials capable of measuring meaningful improvements in cancer prevention in LS carriers are needed. Only then can AI contribute to reducing the persistent burden of interval cancers in this high-risk population.
