Validity of a two-antibody testing algorithm for mismatch repair deficiency testing in cancer; a systematic literature review and meta-analysis

Abstract

Reflex mismatch repair immunohistochemistry (MMR IHC) testing for MLH1, PMS2, MSH2 and MSH6 is used to screen for Lynch syndrome. Recently MMR-deficiency (MMRd) has been approved as a pan-cancer predictive biomarker for checkpoint inhibitor therapy, leading to a vast increase in the use of MMR IHC in clinical practice. We explored whether immunohistochemical staining with PMS2 and MSH6 can be used as a reliable substitute. This two-antibody testing algorithm has the benefit of saving tissue, cutting costs and saving time. PubMed, Embase and Cochrane library were systematically searched for articles reporting on MMR IHC. The weighed percentage of cases with isolated MLH1 or MSH2 loss or combined MLH1/MSH2 loss alone was analyzed using a random effects model meta-analysis in R. The search yielded 1704 unique citations, of which 131 studies were included, describing 9014 patients. A weighed percentage of 1.1% (95% CI 0.53–18.87, I = 87%) of cases with isolated MLH1 or MSH2 loss or combined MLH1/MSH2 loss alone was observed. In the six articles with the main aim of investigating the two-antibody testing algorithm all MMRd cases were detected with the two-antibody testing algorithm, there were no cases with isolated MLH1 or MSH2 loss or combined MLH1/MSH2 loss alone. This high detection rate of MMRd of the two-antibody testing algorithm supports its use in clinical practice by specialized pathologists. Staining of all four antibodies should remain the standard in cases with equivocal results of the two-antibody testing algorithm. Finally, educational sessions in which staining pattern pitfalls are discussed will continue to be important.