Should you repeat mismatch repair testing in cases of tumour recurrence? An evaluation of repeat mismatch repair testing by the use of immunohistochemistry in recurrent tumours of the gastrointestinal and gynaecological tracts

Abstract

Aims The role of mismatch repair (MMR) testing has evolved from identifying Lynch syndrome patients to predicting response to immune checkpoint inhibitors. This has led to requests from clinicians to retest recurrences of MMR‐proficient primary tumours in the hope that the recurrence may show a different MMR status and qualify the patient for treatment. We aimed to determine whether repeat testing is warranted. Methods and results We evaluated recurrent tumours (local recurrences or metastases) from 137 patients with MMR‐proficient primary tumours of the gastrointestinal and gynaecological tracts. The local recurrences and metastases all occurred at least 30 days after resection of the primary tumour. We used a combination of a tissue microarray and whole slide staining to perform immunohistochemistry (IHC) for PMS2, MLH1, MSH2, and MSH6, and compared the results with the MMR status of the primary tumour. Three of 137 (2%) initially showed a discordant staining pattern. However, further investigation showed that these discordances were attributable to some of the known pitfalls associated with MMR IHC interpretation – post‐radiotherapy loss of MSH6 expression and subclonal loss of MLH1 staining. We did not identify any cases with a genuine discordance in MMR status. Conclusion We conclude that repeat MMR IHC testing of recurrences is not warranted, as MMR status does not change relative to that of the primary tumour.

Patterns of DNA mismatch repair protein expression for primary and recurrent colorectal cancer at an advanced surgical unit: A retrospective audit

“…Moreover, our findings show that MMR profiles are not fixed from primary to locally recurrent or metastatic CRC, with 10.6% of cases showing discordant MMR protein expression. While discordance of MMR profiles between primary CRC and corresponding metastatic or recurrent disease has been previously described, the frequency of discordance is incompletely understood, with estimates from 0% to 31% being reported [7,[24][25][26][27][28]. Importantly, both a loss and a gain of MMR expression from primary to recurrent or metastatic disease have been described, with one study reporting preserved MMR profiles in 18.8% of paired metastatic lesions with MMR protein-deficient primary tumours [25].…”

“…The effects of neoadjuvant or adjuvant therapies on MLH1, PMS2 or MSH2 expression, however, is not well established and may be insignificant [14,30]. A paper by Aird et al reported that although 2% of primary and corresponding metastatic CRC cases displayed discordant MMR profiles, all could be explained by either MLH1 promoter hypermethylation or adjuvant radiotherapy [26].…”

“…Aberrant MMR protein expression following adjuvant radiotherapy or subclonal loss of MLH1 due to promoter hypermethylation are important considerations when interpreting MMR IHC [26]. Specifically, MSH6 expression may be altered by neoadjuvant chemotherapy or radiotherapy, raising the possibility of spurious MSH6 expression in patients treated with these therapies [14,[30][31][32].…”

“…Specifically, our results show one case of isolated MSH2 loss, two cases of isolated MSH6 loss, one case of MLH1, PMS2 and MSH2 loss, one case of MLH2 and PMS2 loss and one case of MSH2 and MSH6 loss (Table 6). Of these, the two cases with loss of MSH6 expression from primary to recurrent disease may be explained by adjuvant radiotherapy and the single case with a loss of MLH1 and PMS2 by MLH1 promoter hypermethylation; however, neither could be confirmed [26]. The two cases with loss of MSH2 expression without the expected loss of MSH6 staining on recurrent disease are difficult to interpret [20].…”

“…This may be significant, as MMR profiles may change over time and are now understood to have diagnostic, therapeutic and prognostic implications [22,23]. A small number of studies have reported MMR discordance between primary and metastatic or recurrent disease to be less than 5%; however, rates of up to 31% have been published [7,[24][25][26][27][28]. Adjuvant or neoadjuvant therapy and MLH1 hypermethylation can explain some instances of discordant profiles; however, it is possible that other mechanisms make a contribution.…”