Redefining the EPCAM Phenotype: Deletion Architecture Determines Lynch Syndrome Cancer Risk

EPCAM

Germline EPCAM deletions whose risk depends on involvement of the 3' region and extension into MSH2

This retrospective study compares the tumour phenotype of 662 unrelated probands tested by multigene panel, grouped by deletion architecture: isolated EPCAM full-gene deletions (n = 22), isolated EPCAM 3' deletions (n = 96), contiguous deletions extending into MSH2 (n = 64 and 116), MSH2 gross deletions (n = 155) and MSH2 nonsense or frameshift variants (n = 209). Isolated EPCAM full-gene deletions were not significantly associated with Lynch syndrome-associated cancers (prevalence 13.6%), versus 54.6% for isolated 3' deletions and 64.1-72.2% for MSH2-involving groups. Relative to panel-negative individuals, isolated 3' deletions were strongly associated with colorectal cancer (OR 17.59) but not endometrial cancer (OR 0.76). MSH2-involving groups were associated with both (OR 8.50-16.03 for colorectal, 17.31-30.47 for endometrial). The authors derive three risk groups based on deletion boundaries and MSH2 involvement.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

A very practical result for disclosure: an isolated EPCAM full-gene deletion sparing MSH2 does not appear to confer a Lynch phenotype, and an isolated 3' deletion points to a colorectal-predominant risk. This requires a laboratory able to define deletion boundaries and MSH2 involvement, which not every report provides. The isolated full-gene deletion group includes only 22 probands, and ascertainment through testing indications does not measure penetrance.

Analysis by Dr Thibaut Benquey