Lack of correlation between MSH3 immunohistochemistry and microsatellite analysis for the detection of elevated microsatellite alterations at selected tetranucleotide repeats (EMAST) in colorectal cancers.

Immunohistochemical evaluation of mismatch repair protein (MMR) expression is an important screening tool in diagnostic pathology, where it is routinely used to identify subsets of colorectal cancers (CRCs) with either inherited or sporadic forms of microsatellite instability (MSI). MSH3 is not included in current MMR panels, although aberrant MSH3 expression is reported to occur in 40-60% of CRCs and is associated with elevated microsatellite alterations at selected tetranucleotide repeats (EMAST) and a worse prognosis. In this study, we applied MSH3 immunohistochemistry and tetranucleotide MSI analysis to a cohort of 250 unselected CRCs to evaluate the potential use of the methods in routine practice. Partial, complete, and focal loss of nuclear MSH3 and its cytoplasmic mislocalization were evident in 67% of tumors, whereas MSI was evident in two to six of a panel of six tetranucleotide repeats in 46% of cases. However, concordance between MSH3 immunohistochemistry and tetranucleotide MSI results was only 61%, indicating the unsuitability of this combination of tests in routine pathology practice. MSH3 immunostaining was compromised in areas of tissue crush and autolysis, which are common in biopsy and surgical samples, potentially mitigating against its routine use. Although tetranucleotide MSI is clearly evident in a subset of CRCs, further development of validated sets of tetranucleotide repeats and either MSH3 or other immunohistochemical markers will be required to include EMAST testing in the routine evaluation of CRCs in clinical practice.