Detection of DNA mismatch repair (MMR) deficiencies by immunohistochemistry can effectively diagnose the microsatellite instability (MSI) phenotype in endometrial carcinomas

Highlights

• MMR IHC and MSI assay are equivalent methodologies for the assessment of MSI phenotype in endometrial cancers (EC).

• MMR IHC advantages over MSI assay include low cost, fast turnaround, and ability to be performed on FFPE biopsies.

• Determination of MSI phenotype may be helpful in the classification and stratification of endometrial tumors.

Abstract

Background

A proportion of endometrial carcinomas (ECs) are associated with deficient DNA mismatch repair (MMR). These tumors are characterized by high levels of microsatellite instability (MSI). Identification of MSI is important in identifying women who should be tested for Lynch syndrome and identifying a phenotype that may have specific prognostic and predictive implications. Genomic characterization of ECs has shown that MSI tumors form a distinct subgroup. The two most common methodologies for MSI assessment have not been compared in EC.

Methods

Pentaplex mono and di-nucleotide PCR for MSI testing was compared to MMR IHC (presence/absence of MLH1, MSH2, MSH6, PMS2) in a cohort of patients with EC. Concordance, Kappa statistic, sensitivity, specificity, positive and negative predictive values were obtained on the cross-tabulation of results.

Results

Comparison of both MSI and MMR status was complete for 89 cases. Overall agreement between methods (concordance) was 93.3% (95% CI[85.9%–97.5%]). A one-sided test to determine whether the accuracy is better than the “no information rate,” which is taken to be the largest class percentage in the data, is significant (p < 0.00001). Unweighted Kappa was 0.84, along with the sensitivity (88.5%), specificity (95.2%), PPV (88.5%), and NPV (95.2%). The balanced accuracy (i.e. the average between sensitivity and specificity) was 92%.

Discussion

We show the equivalence of MSI testing and MMR IHC. We advocate the implementation of MMR IHC in future EC classification schemes, enabling stratification of cases for future clinical trials as well as assisting identification of Lynch syndrome, so that screening and risk reducing interventions can be undertaken.

Introduction

Microsatellite instability (MSI) describes a molecular phenotype that arises secondary to defects in the post-replicative DNA mismatch repair (MMR) system [1], [2]. Microsatellite regions are repetitive sequences throughout the genome consisting of mono-, di- or higher order nucleotide repeats. These repeats are more frequently transcribed incorrectly as DNA polymerases cannot bind efficiently. The MMR system is responsible for the identification and correction of these errors. Defects in the system can be secondary to genetic or epigenetic mechanisms, through germline or somatic mutations in one of the MMR genes (four highest frequency = MLH1, MSH2, MSH6, PMS2), or methylation of a promoter region, most commonly MLH1, respectively. MSI has been well characterized in colorectal cancer (CRC) but is also commonly recognized within specific histologic subtypes of ovarian cancer and endometrial cancers [3], [4], [5], [6], [7].

Identification of the MSI-high phenotype may trigger hereditary testing for Lynch Syndrome (LS) for the index case, and if confirmed, will initiate testing for first degree relatives. Detection of LS enables an individual to consider changes in lifestyle, screening, and/or risk reducing surgery [8], [9], [10], [11], [12], [13]. The MSI-high or hypermutated phenotype was one of the four categories of EC distinguished by genomic characterization in The Cancer Genome Atlas (TCGA)[14], and categorization of MSI status proposed as an early step in molecular classification of tumors. Identification of MSI has both prognostic and predictive implications in colorectal carcinomas and similar associations in gynecologic malignancies are being elucidated [7], [14], [15], [16], [17], [18], [19], [20]. Categorization of tumors can be achieved by either immunohistochemistry for MMR-associated proteins or MSI assay.

In colorectal cancers and more recently in endometrial cancers and endometriosis-associated ovarian cancer MSI phenotype is commonly tested by mismatch repair protein immunohistochemistry (MMR IHC) in the post-operative pathology specimen. Standardization of methods and published guidelines in interpretation has greatly improved consistency [21], [22], [23], [24], [25], [26]. MMR IHC cost is ~$20–130/MMR protein tested (usually two-six proteins tested).

In contrast, in the research setting MSI is assessed utilizing PCR amplification of a set of nucleotide repeat markers. These microsatellite markers are then compared between tumor and normal DNA to detect somatic changes. Historically, choice of microsatellite markers varied widely; thus in 1998 a National Cancer Institute consensus panel recommended the assessment of two mononucleotide markers (BAT25 and BAT26) and three dinucleotide markers (D5S346, D17S250 and D2S123). MSI-H (MSI-high) is assigned if size alterations or shifts are observed in two or more markers, and MSI-L (MSI-low) or low microsatellite instability phenotype if just one marker shows instability. If none of the markers show instability the phenotype is considered MSS (MicroSatellite Stable). There are no FDA-approved MSI tests, and private, academic or commercial labs must validate their own assays independently. This has led to variation in techniques and challenges in interpreting data from across different centers. The PCR MSI assay requires normal DNA for comparison, which limits the number of cases easily tested. Minimal cost estimates from an in-house lab include DNA extraction and MSI assay sum to ~$40/case but commercial cost is over $400 with additional costs accrued for interpretation (~$25–50) and microdissection (~$200–250).

Although a high level of concordance of MSI assay testing and MMR IHC has been demonstrated in CRCs, data in other Lynch-associated cancers is absent or sparse. We wished to determine the level of agreement between these two methodologies in endometrial cancers, with the ultimate goal of supporting MMR IHC in research as the primary means of determining MSI phenotype.