Abstract

Background

Mismatch repair (MMR)-deficiency analysis is increasingly recommended for all endometrial cancers, as it identifies Lynch syndrome patients, and is emerging as a prognostic classifier to guide adjuvant treatment. The aim of this study was to define the optimal approach for MMR-deficiency testing and to clarify discrepancies between microsatellite instability (MSI) analysis and immunohistochemical (IHC) analysis of MMR protein expression.

Patients and methods

Six hundred ninety- six endometrial cancers were analyzed for MSI (pentaplex panel) and MMR protein expression (IHC). Agreement between methodologies was calculated using Cohen’s Kappa. MLH1 promoter hypermethylation, dinucleotide microsatellite markers and somatic MMR and POLE exonuclease domain (EDM) gene variants (using next-generation/Sanger sequencing) were analyzed in discordant cases.

Results

MSI was found in 180 patients. Complete loss of expression of one or more MMR proteins was observed in 196 cases. A PMS2- and MSH6-antibody panel detected all cases with loss of MMR protein expression. The results of MSI and MMR protein expression were concordant in 655/696 cases (kappa = 0.854, P < 0.001). Ambiguous cases (n = 41, 6%) included: subclonal loss of MMR protein expression (n = 18), microsatellite stable or MSI-low cases with loss of MMR protein expression (n = 20), and MSI-low or MSI-high cases with retained MMR protein expression (n = 3). Most of these cases could be explained by MLH1 promoter hypermethylation. Five of seven cases with solitary loss of PMS2 or MSH6 protein expression carried somatic gene variants. Two MSI-high cases with retained MMR protein expression carried a POLE-EDM variant.

Conclusion

MSI and IHC analysis are highly concordant in endometrial cancer. This holds true for cases with subclonal loss of MMR protein expression. Discordant MMR-proficient/MSI-high cases (<1%), may be explained by POLE-EDM variants.

Introduction

A defect in DNA mismatch repair (MMR) leads to the accumulation of mismatches, insertions and deletions in repeated sequences—a phenomenon named microsatellite instability (MSI). Approximately 20–30% of sporadic endometrial cancers (ECs) display MSI as a consequence of somatic promoter hypermethylation and silencing of MLH1 [1]. Defective MMR due to pathogenic germline variants in MMR genes causes Lynch syndrome (LS), a tumor predisposition syndrome that accounts for 2% of ECs [2].

Determination of MMR-deficiency in EC may be important for several reasons. First, recent studies have suggested that tumor molecular features, including MMR-deficiency, may improve prognostication and help guide adjuvant therapy for EC patients [3, 4]. Second, accurate assessment of MMR-deficiency is essential to identify patients with EC caused by LS. However in contrast to colorectal cancer, where consensus guidelines for MMR-deficiency testing have been published [5], there is no general agreement on screening EC patients for LS [6, 7]. Finally, recent studies have shown that MMR-deficiency in colorectal and urothelial cancer is predictive of response to immunotherapy [8, 9], suggesting that MMR-deficient ECs may also benefit from these therapeutics.

MMR-deficiency can be detected by either MSI analysis and/or immunohistochemical (IHC) staining, typically for four MMR proteins. The National Cancer Institute microsatellite panel was optimized and correlated with IHC analysis (∼95%) to detect MMR-deficiency in colorectal cancer [10, 11]. IHC alone has become standard practice in multiple institutions. Experience in this setting is that while some tumors show uniform and widespread loss of MMR protein expression, cases with subclonal loss of MMR protein expression are also observed [11, 12]. Such cases present with two populations of tumor cells; one with retained expression, and another with abrupt and complete regional loss of MMR protein expression [12]. Small studies have shown high agreement between MSI and loss of MMR protein expression in EC [13–15], while others have described subclonal loss of MMR protein expression [16–20]. However, studies identifying the frequency of such staining patterns in large patient series are sparse.

In this study, we sought to establish the optimum method for MMR-deficiency testing by comparison of MSI with IHC analysis in a large series of ECs. We also investigated the frequency of subclonal loss of MMR protein expression and the number of potential LS cases. Cases showing disagreement between methodologies and those with subclonal loss of MMR protein expression were further characterized.

Methods

Study population

The population comprised, 854 ECs from the PORTEC-1 and -2 clinical trials based on availability of formalin-fixed paraffin-embedded slides and sufficient tumor material for DNA isolation [21, 22]. Further details are summarized in

and , available at Annals of Oncology online.

MSI assay

DNA was isolated as previously described [23]. In cases with subclonal loss of MMR protein expression, tissue sections were used to microdissect the differentially expressed tumor areas. Tumor MSI status was determined as previously reported (

, available at Annals of Oncology online) [4]. Tumors initially classified as MSS or MSI-L with concomitant loss of MMR protein expression underwent evaluation of three dinucleotide repeat markers [24], and reclassified as MSI-H if instability was detected at two dinucleotide markers.

IHC analysis

IHC staining for MLH1, PMS2, MSH2, and MSH6 was performed on all tumors in which MSI status was successfully determined (

, available at Annals of Oncology online). The slides were evaluated in three categories as retained, loss and subclonal loss of protein expression with stromal- and/or lymphocytic cells as internal controls [16]. The cases with subclonal loss of protein expression were re-evaluated to determine the percentage of tumor cells with loss of MMR expression.

Methylation-specific PCR for MLH1

Tumors with loss of MLH1 protein expression underwent testing for hypermethylation status of the MLH1 5′ regulatory region by methylation-specific PCR, as previously described [25].

Somatic variant screening

Subject to DNA availability and quality, tumors in which the results of MSI analysis and MMR protein expression were discordant underwent targeted next generation sequencing (NGS) of MMR and POLE genes using the Ion Proton™ System (ThermoFisher, MA) as previously described (

, available at Annals of Oncology online) [26]. Three additional cases were similarly analyzed using the Ion AmpliSeq Comprehensive Cancer Panel (ThermoFisher) at The Welcome Trust Center for Human Genetics. Frameshift variants in the polycytosine tract in exon 5 of MSH6 were analyzed using Sanger sequencing [27].

Results

Combined analysis of MMR protein expression and MSI was possible in 696 (81%) ECs (

, available at Annals of Oncology online). The frequencies of MSS, MSI-H and MSI-L were 74%, 24% and 2%, respectively. Among the 516 tumors assessed as MSS, 496 (96%) showed retained expression of all four MMR proteins (Table 1). The remaining 20 MSS cases showed loss of MMR protein expression as follows: combined MLH1 and PMS2 loss (n = 14), combined MSH2 and MSH6 loss (n = 3), solitary MSH6 loss (n = 3, Figure 1A–D). Of the 11 cases assessed as MSI-L, six displayed combined loss of MLH1 and PMS2 expression (complete in four cases, subclonal in two cases), two cases showed solitary loss of PMS2, a further two cases had solitary MSH6 loss and one case retained expression of all four MMR proteins.
Figure 1

Representative images of MMR protein expression in EC. MMR protein expression of a MSS case with subclonal loss of MLH1 and PMS2 protein expression (A–D, case 13), a MSI-H case with complete loss of MLH1 and PMS2 and subclonal loss of MSH6 protein expression (E–H), a MSS case with loss of MSH6 protein expression (I–L, case 38). (A–E–I) MLH1 protein expression, (B–F–J) PMS2 protein expression, (C–G–K) MSH2 protein expression and (D–H–L) MSH6 protein expression. Scale bar represents 50 µM.

Figure 1

Representative images of MMR protein expression in EC. MMR protein expression of a MSS case with subclonal loss of MLH1 and PMS2 protein expression (A–D, case 13), a MSI-H case with complete loss of MLH1 and PMS2 and subclonal loss of MSH6 protein expression (E–H), a MSS case with loss of MSH6 protein expression (I–L, case 38). (A–E–I) MLH1 protein expression, (B–F–J) PMS2 protein expression, (C–G–K) MSH2 protein expression and (D–H–L) MSH6 protein expression. Scale bar represents 50 µM.

Table 1

Details on the MSI status and MMR protein expression in early-stage EC (n = 696)

MSI status MMR protein expression
 
Count 
MLH1 PMS2 MSH6 MSH2 Protein expression 
MSS Retained 496 
MSS Subclonal loss 
MSS Subclonal loss 
MSS Loss 
MSS Loss 
MSS Loss 
MSI-L Retained 
MSI-L Subclonal loss 
MSI-L Loss 
MSI-L Loss 
MSI-L Loss 
MSI-H Retained 
MSI-H Subclonal loss 
MSI-H Loss/subclonal loss 
MSI-H Loss 130 
MSI-H Loss 10 
MSI-H Loss 
MSI-H Loss 
MSI status MMR protein expression
 
Count 
MLH1 PMS2 MSH6 MSH2 Protein expression 
MSS Retained 496 
MSS Subclonal loss 
MSS Subclonal loss 
MSS Loss 
MSS Loss 
MSS Loss 
MSI-L Retained 
MSI-L Subclonal loss 
MSI-L Loss 
MSI-L Loss 
MSI-L Loss 
MSI-H Retained 
MSI-H Subclonal loss 
MSI-H Loss/subclonal loss 
MSI-H Loss 130 
MSI-H Loss 10 
MSI-H Loss 
MSI-H Loss 

Mismatch repair protein expression was scored as following: 0—Complete loss; 1—Retained; 2—Subclonal loss. MMR—mismatch repair, MSS—microsatellite stable, MSI-L/H—microsatellite unstable with low or high frequency.

The majority of MSI-H cases (130 of 169, 77%) showed complete loss of MLH1 and PMS2 expression (Table 1). Sporadic MSI due to MLH1 hypermethylation was observed in 97% of these 130 MSI-H cases. Eight MSI-H cases showed areas of subclonal loss of MLH1 and PMS2 (Figure 1E–H), and six cases displayed subclonal loss of MSH6 in addition to complete loss of MLH1 and PMS2 protein expression (Figure 1I–L). In 10 cases (6%), combined loss of MSH2 and MSH6 protein expression was observed. The remaining MSI-H tumors showed solitary loss of PMS2 (n = 8) or MSH6 (n = 5) protein expression, or retained expression of all MMR proteins (n = 2).

Overall, concordance between MSI and IHC analysis was observed in 655 of 696 cases (94%, kappa = 0.854; 95% CI 0.811–0.897, P < 0.001). A PMS2- and MSH6-antibody panel was as effective as the four-antibody panel in detecting MMR protein abnormalities. Twenty-seven concordant cases without MLH1 promoter hypermethylation were identified as potential LS, but the underlying defect was not further tested. Discordant cases (n = 41, 6%) included: subclonal loss of MMR expression (n = 18), MSS or MSI-low cases with loss of MMR expression (n = 20), and MSI-low or MSI-high cases with retained MMR protein expression (n = 3). Details on the sample analysis of discordant cases are shown in

, available at Annals of Oncology online.

All cases with subclonal loss of MMR protein expression (n = 18) were evaluated in more detail by analyzing MSI in mono- and dinucleotide markers, MLH1 promoter hypermethylation and somatic MMR- and POLE-exonuclease domain (EDM) variants in microdissected tumor areas (Table 2). Among 16 tumors with subclonal MLH1 and PMS2 loss, 14 had areas of differential expression that were sufficiently large to permit microdissection. Among these, MSI testing of microdissected areas was concordant with IHC analysis in 11 cases; tumor areas with retained MMR expression were MSS, whereas areas with loss of MMR expression showed MSI-H. A further three tumors showed microsatellite stability of markers in microdissected areas regardless of MMR protein expression (cases 13–15, Table 2). All 14 cases were found to have somatic promoter hypermethylation of MLH1. One case with subclonal loss of MSH2 and MSH6 protein expression showed microsatellite stability in the differently expressed areas (case 17, Table 2). Unfortunately, both cases with subclonal loss of MSH2 and MSH6 protein expression had limited DNA available, and could not be analyzed in more detail.

Table 2

Details on ECs with subclonal loss of MMR protein expression

Case Subclonal loss of protein expression % tumor with loss of expression MSI statusa MSI statusb
 
Area with retained expression Area with focal loss of expression No. of affected dinucleotidesc MLH1 hyper- methylationd Gene variant 
MLH1, PMS2 90 MSI-H – MSI-H – Yes – 
MLH1, PMS2 90 MSI-H – MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 25–50 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-L MSS MSI-H – Yes – 
10 MLH1, PMS2 25–50 MSI-L MSS MSI-H – Yes – 
11 MLH1, PMS2 >75 MSS MSS MSI-H – Yes – 
12 MLH1, PMS2 25–50 MSS MSS MSI-H – Yes – 
13 MLH1, PMS2 25–50 MSS MSS MSS Yes – 
14 MLH1, PMS2 25–50 MSS MSS MSS Failed Yes – 
15 MLH1, PMS2 10 MSS MSS MSS Failed None 
16 MLH1, PMS2 10 MSS MSS MSI-H – Failed None 
17 MSH2, MSH6 25–50 MSS MSS MSS – – – 
18 MSH2, MSH6 10 MSS – – – – – 
Case Subclonal loss of protein expression % tumor with loss of expression MSI statusa MSI statusb
 
Area with retained expression Area with focal loss of expression No. of affected dinucleotidesc MLH1 hyper- methylationd Gene variant 
MLH1, PMS2 90 MSI-H – MSI-H – Yes – 
MLH1, PMS2 90 MSI-H – MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 25–50 MSI-H MSS MSI-H – Yes – 
MLH1, PMS2 75 MSI-L MSS MSI-H – Yes – 
10 MLH1, PMS2 25–50 MSI-L MSS MSI-H – Yes – 
11 MLH1, PMS2 >75 MSS MSS MSI-H – Yes – 
12 MLH1, PMS2 25–50 MSS MSS MSI-H – Yes – 
13 MLH1, PMS2 25–50 MSS MSS MSS Yes – 
14 MLH1, PMS2 25–50 MSS MSS MSS Failed Yes – 
15 MLH1, PMS2 10 MSS MSS MSS Failed None 
16 MLH1, PMS2 10 MSS MSS MSI-H – Failed None 
17 MSH2, MSH6 25–50 MSS MSS MSS – – – 
18 MSH2, MSH6 10 MSS – – – – – 

Grey colored lines are unexplained/failed cases.

a

First round of MSI analysis.

b

Second round of MSI analysis in differently expressed areas (three cases had too small areas).

c

Instability in dinucleotide markers was only analyzed in discordant cases after the second round of MSI analysis.

d

MLH1 promoter hypermethylation was only assessed in cases with loss of/subclonal loss of MLH1 protein expression.

MSS, microsatellite stable; MSI-L/H, microsatellite unstable with low or high frequency.

Analysis of microdissected material from the six MSI-H cases with subclonal loss of MSH6 in addition tocomplete MLH1 and PMS2 protein loss demonstrated frameshift variants in the polycytosine tract of MSH6 in areas with MSH6 loss and stable polycytosine tracts in areas with retained MSH6. Five of these cases displayed MLH1 promoter hypermethylation.

We proceeded to perform detailed analysis of the 23 cases with discordant MSI status and MMR protein expression by examination of dinucleotide markers, MLH1 promoter hypermethylation and/or NGS of the MMR- and POLE genes (Table 3). The two MSI-H cases (cases 19–20) with retained MMR protein expression had a POLE-EDM variant, p.(V411L), and p.(A428T). The POLE-EDM p.(V411L) mutant case also harbored a truncating [p.(R563*)] and missense p.(R107W) variant in PMS2. The solitary MSI-L case with retained MMR protein expression (case 21) showed stability in dinucleotide markers and no somatic MMR or POLE-EDM gene variants.

Table 3

Details on ECs with discordant MSI status and MMR protein expression

Case Protein expression MSI status No. of affected dinucleotidesa MLH1 hyper- methylationb Somatic variantsc Variant allele frequency Classification of pathogenicityd 
19 Retained MSI-H – – POLE p.(V411L), PMS2 p.(R107W), p.(R563*) 0.36, 0.37, 0.42 –, class 3, class 5 
20 Retained MSI-H – No POLE p.(A428T) 0.46 – 
21 Retained MSI-L – None – – 
22 MLH1-PMS2 loss MSI-L Yes – – – 
23 MLH1-PMS2 loss MSI-L Yes – – – 
24 MLH1-PMS2 loss MSI-L Yes – – – 
25 MLH1-PMS2 loss MSI-L Failed Yes – – – 
26 MLH1-PMS2 loss MSS Failed No POLE p.(P286R) 0.69 – 
27 MLH1-PMS2 loss MSS Failed None – – 
28 MLH1-PMS2 loss MSS Yes – – – 
29 MLH1-PMS2 loss MSS Failed Yes – – – 
30 MLH1-PMS2 loss MSS Yes – – – 
31 MLH1-PMS2 loss MSS Yes – – – 
32 MLH1-PMS2 loss MSS Yes – – – 
33 MLH1-PMS2 loss MSS Yes – – – 
34 PMS2 loss MSI-L – PMS2 p.(M1?), p.(V302Tfs*4) 0.40, 0.27 class 3, class5 
35 PMS2 loss MSI-L Failed – – – – 
36 MSH6 loss MSI-L – MSH6 p.(R495*), p.(N975Kfs*10) 0.35, 0.41 class 5, class 5 
37 MSH6 loss MSI-L – MSH6 p.(R922*) 0.47 class 5 
38 MSH6 loss MSS – MSH6 p.(S445Nfs*7) 0.44 class 5 
39 MSH6 loss MSS – MSH6 p.(Q415H) 0.43 class 3 
40 MSH6 loss MSS Failed – – – – 
41 MSH2-MSH6 loss MSS Failed – – – – 
Case Protein expression MSI status No. of affected dinucleotidesa MLH1 hyper- methylationb Somatic variantsc Variant allele frequency Classification of pathogenicityd 
19 Retained MSI-H – – POLE p.(V411L), PMS2 p.(R107W), p.(R563*) 0.36, 0.37, 0.42 –, class 3, class 5 
20 Retained MSI-H – No POLE p.(A428T) 0.46 – 
21 Retained MSI-L – None – – 
22 MLH1-PMS2 loss MSI-L Yes – – – 
23 MLH1-PMS2 loss MSI-L Yes – – – 
24 MLH1-PMS2 loss MSI-L Yes – – – 
25 MLH1-PMS2 loss MSI-L Failed Yes – – – 
26 MLH1-PMS2 loss MSS Failed No POLE p.(P286R) 0.69 – 
27 MLH1-PMS2 loss MSS Failed None – – 
28 MLH1-PMS2 loss MSS Yes – – – 
29 MLH1-PMS2 loss MSS Failed Yes – – – 
30 MLH1-PMS2 loss MSS Yes – – – 
31 MLH1-PMS2 loss MSS Yes – – – 
32 MLH1-PMS2 loss MSS Yes – – – 
33 MLH1-PMS2 loss MSS Yes – – – 
34 PMS2 loss MSI-L – PMS2 p.(M1?), p.(V302Tfs*4) 0.40, 0.27 class 3, class5 
35 PMS2 loss MSI-L Failed – – – – 
36 MSH6 loss MSI-L – MSH6 p.(R495*), p.(N975Kfs*10) 0.35, 0.41 class 5, class 5 
37 MSH6 loss MSI-L – MSH6 p.(R922*) 0.47 class 5 
38 MSH6 loss MSS – MSH6 p.(S445Nfs*7) 0.44 class 5 
39 MSH6 loss MSS – MSH6 p.(Q415H) 0.43 class 3 
40 MSH6 loss MSS Failed – – – – 
41 MSH2-MSH6 loss MSS Failed – – – – 

Grey colored lines are unexplained/failed cases.

a

Instability in dinucleotide markers was only analyzed in discordant MSI-L and MSS cases.

b

MLH1 promoter hypermethylation was only assessed in cases with loss/focal loss of MLH1 protein expression.

c

Mismatch repair gene variants were analyzed in those cases without MLH1 promoter hypermethylation.

d

Classification of pathogenicity was based upon InSiGHT guidelines. Class 3 variants are predicted to be pathogenic by at least two out of three protein prediction programs used: (Align GVGD, SIFT and MutationTaster).

MSS, microsatellite stable; MSI-L/H, microsatellite unstable with low or high frequency.

One of 12 cases classified as MSS or MSI-L despite combined loss of MLH1 and PMS2 protein expression, showed mobility shifts in the dinucleotide markers (case 24, Table 3). Analysis of the MLH1 promoter was successful in 11 of these cases, and revealed promoter hypermethylation in 10 cases, while the single case lacking MLH1 promoter hypermethylation was found to harbor a pathogenic POLE-EDM variant, p.(P286R).

Of the two MSI-L tumors with solitary PMS2 loss (cases 34–35, Table 3), only one had sufficient DNA quality for further analysis. This confirmed MSI in the dinucleotide markers and revealed two likely pathogenic somatic PMS2 variants, a start loss [p.(Met1?)] and a frameshift variant [p.(Val302Thrfs*4)].

Four of five cases classified as MSS/MSI-L with solitary MSH6 loss were informative for further analysis. All four showed microsatellite stable dinucleotide markers. Three tumors carried two (n = 1) or one (n = 2) pathogenic MSH6 variants, while one tumor carried one somatic VUS predicted to affect function by two out of three protein prediction software used (Table 3). Case 41 with loss of MSH2 and MSH6 protein expression and a MSS phenotype had limited DNA, and was therefore excluded for further analysis.

Discussion

Accurate identification of MMR-deficiency in EC may be important to identify patients with a higher risk of recurrence [3, 4, 28], and those whose tumors may be a consequence of LS. Similarly to two small studies, we demonstrated high agreement (94%) between MSI and IHC analysis in 696 ECs [13, 14]. Most discordant cases involved loss of MMR protein expression and a MSS/MSI-L phenotype and could be explained by MLH1 promoter hypermethylation or MMR variants. In addition, subclonal loss of MMR protein expression generally corresponded to MLH1 promoter hypermethylation and subclonal MSI within microdissected area of the tumor.

Importantly, the present study demonstrated that <3% of cases displayed subclonal loss of MMR protein expression. The fact that MSI and MLH1 promoter hypermethylation were commonly found in areas with subclonal loss of MLH1 and PMS2 protein expression indicates sporadic intratumor heterogeneity [16, 18, 19]. However, MLH1 germline epimutations cannot be totally excluded [2]. Subclonal loss of MSH6 expression, either in conjunction with or without MSH2 protein expression was also previously observed in EC but the underlying molecular mechanisms remain unclear [12, 20]. In accordance with our findings, subclonal loss of MSH6 in cases with complete loss of MLH1 and PMS2 protein expression has been related to secondary MSI events in MSH6 [20, 29]. Although numbers are limited, subclonal loss of MMR protein expression is not associated with LS.

Our data suggest that cases with subclonal loss of MMR protein expression are best classified as MMR-deficient, even though the areas with retained expression are MSS. With regard to MMR-deficiency as a prognostic or predictive marker, it remains to be determined whether subclonal loss of MMR protein expression has the same biological behavior as tumors with MMR-proficiency. In view of the limited numbers of cases with subclonal loss of MMR protein expression (∼3%), (inter)national collaborations are essential to obtain sufficient cases for such an analysis. Pending these future studies, we suggest for uniformity to classify tumors with 10% subclonal loss of MMR protein expression, as being MMR-deficient.

The interpretation of MSI-L cases remains controversial in EC and it is uncertain whether such cases are best considered as MSS or MSI. Similar numbers of DNA slippage events were observed in MSS and MSI-L ECs [30]. To date, no extensive research on the clinical implications of MSI-L in ECs has been performed, and the number of MSI-L cases in our study (n = 11) was too low to permit such an analysis as well. However, most of these showed loss of MMR protein expression and would generally be regarded as abnormal by strategies that rely on IHC alone. Noteworthy, several studies have also shown MSI-L and MSS in association with loss of MMR expression and/or pathogenic germline MMR variants [13, 14, 31].

Our study shows high agreement between IHC and MSI analysis, but not 100%. Of note, other assessments of DNA defects by IHC analysis, e.g. HER2 gene amplification only reaches 69–98% agreement [32]. Assessment of MMR protein expression is preferred over MSI analysis for the following reasons: lower costs, widely available, and determination of affected MMR gene. Our findings confirm the utility of testing MMR-deficiency using a PMS2- and MSH6-IHC approach [33], which can be followed by MLH1- and MSH2-IHC in case loss of PMS2 or MSH6 was observed. To overcome suboptimal fixation, drawback of IHC analysis, pathologists can rely on IHC analysis in pre-operative EC specimen [23]. IHC with standard well accepted techniques would appear adequate to identify EC patients with LS and to serve as a biomarker for trials of EC patients harboring MMR-deficiency.

It is debatable whether not screening for germline MMR variants is a limitation of this study. Of note, 5% of all cases in this study can be classified as potential LS (no MLH1 promoter hypermethylation). Somatic screening of the discordant cases did show somatic variants but not in all cases. However, we cannot exclude the possibility of missed large genomic rearrangements within the tested genes, which is a limitation of NGS. Further analysis would improve understanding the molecular basis of the discordant cases, however, this study did not aim to determine the sensitivity and specificity of the two methodologies to identify LS. MSI and IHC analysis are highly concordant therefore germline testing is not needed to conclude which approach is best suitable for identifying patients with LS.

In conclusion, MSI and IHC analysis are highly concordant, also in cases with subclonal loss of MMR expression, therefore, an IHC approach is sufficient for determining MMR-deficiency in EC. Pathologists should be aware of the MMR protein expression patterns, including subclonal loss, to ensure correct classification in daily diagnostic pathology.

Acknowledgements

We acknowledge all members of the PORTEC study group and the many patients who participated in the trials, and M. Ventayol Garcia from the Leiden University Medical Center for excellent technical support.

Funding

Dutch Cancer Society (KWF-UL2012-5719 and KWF-UL2012-5542).

Disclosure

The authors have declared no conflicts of interest.

References

1
Cancer Genome Atlas Research N
,
Kandoth
C
Schultz
N
et al.  .
Integrated genomic characterization of endometrial carcinoma
.
Nature
 
2013
;
497
:
67
73
.
2
Lynch
HT
Snyder
CL
Shaw
TG
et al.  .
Milestones of Lynch syndrome: 1895–2015
.
Nat Rev Cancer
 
2015
;
15
:
181
194
.
3
Talhouk
A
McConechy
MK
Leung
S
et al.  .
A clinically applicable molecular-based classification for endometrial cancers
.
Br J Cancer
 
2015
;
113
:
299
310
.
4
Stelloo
E
Nout
RA
Osse
EM
et al.  .
Improved risk assessment by integrating molecular and clinicopathological factors in early-stage endometrial cancer – combined analysis of PORTEC cohorts
.
Clin Cancer Res
 
2016
;
22
:
4215
4224
.
5
Umar
A
Boland
CR
Terdiman
JP
et al.  .
Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability
.
J Natl Cancer Inst
 
2004
;
96
:
261
268
.
6
Oncoline. Guidelines for hereditary nonpolyposis colorectal cancer. www.oncoline.nl (6 October 2016, date last accessed).
7
American College of Obstetricians and Gynecologists
. Lynch Syndrome Practice Bulletin. No 47. Obstet Gynecol.
2014
; 124: 1042–1054.
8
Le
DT
Uram
JN
Wang
H
et al.  .
PD-1 blockade in tumors with mismatch-repair deficiency
.
N Engl J Med
 
2015
;
372
:
2509
2520
.
9
Castro
MP
Goldstein
N.
Mismatch repair deficiency associated with complete remission to combination programmed cell death ligand immune therapy in a patient with sporadic urothelial carcinoma: immunotheranostic considerations
.
J Immunother Cancer
 
2015
;
3
:
58
.
10
Shia
J.
Immunohistochemistry versus microsatellite instability testing for screening colorectal cancer patients at risk for hereditary nonpolyposis colorectal cancer syndrome. Part I. The utility of immunohistochemistry
.
J Mol Diagn
 
2008
;
10
:
293
300
.
11
Chapusot
C
Martin
L
Puig
PL
et al.  .
What is the best way to assess microsatellite instability status in colorectal cancer? Study on a population base of 462 colorectal cancers
.
Am J Surg Pathol
 
2004
;
28
:
1553
1559
.
12
Joost
P
Veurink
N
Holck
S
et al.  .
Heterogenous mismatch-repair status in colorectal cancer
.
Diagn Pathol
 
2014
;
9
:
126
.
13
Bartley
AN
Luthra
R
Saraiya
DS
et al.  .
Identification of cancer patients with Lynch syndrome: clinically significant discordances and problems in tissue-based mismatch repair testing
.
Cancer Prev Res (Phila)
 
2012
;
5
:
320
327
.
14
McConechy
MK
Talhouk
A
Li-Chang
HH
et al.  .
Detection of DNA mismatch repair (MMR) deficiencies by immunohistochemistry can effectively diagnose the microsatellite instability (MSI) phenotype in endometrial carcinomas
.
Gynecol Oncol
 
2015
;
137
:
306
310
.
15
de Leeuw
WJ
Dierssen
J
Vasen
HF
et al.  .
Prediction of a mismatch repair gene defect by microsatellite instability and immunohistochemical analysis in endometrial tumours from HNPCC patients
.
J Pathol
 
2000
;
192
:
328
335
.
16
Watkins
JC
Nucci
MR
Ritterhouse
LL
et al.  .
Unusual mismatch repair immunohistochemical patterns in endometrial carcinoma
.
Am J Surg Pathol
 
2016
;
40
:
909
916
.
17
Lassen
Ring K
Bruegl
AS
Batte
BAL
et al.  .
A prospective evaluation of universal tumor testing strategies for Lynch syndrome in endometrial cancer
.
J Clin Oncol
 
2014
;
32
: Abstract 5512.
18
Kato
A
Sato
N
Sugawara
T
et al.  .
Isolated loss of PMS2 immunohistochemical expression is frequently caused by heterogenous MLH1 promoter hypermethylation in Lynch syndrome screening for endometrial cancer patients
.
Am J Surg Pathol
 
2016
;
40
:
770
776
.
19
Pai
RK
Plesec
TP
Abdul-Karim
FW
et al.  .
Abrupt loss of MLH1 and PMS2 expression in endometrial carcinoma: molecular and morphologic analysis of 6 cases
.
Am J Surg Pathol
 
2015
;
39
:
993
999
.
20
Graham
RP
Kerr
SE
Butz
ML
et al.  .
Heterogenous MSH6 loss is a result of microsatellite instability within MSH6 and occurs in sporadic and hereditary colorectal and endometrial carcinomas
.
Am J Surg Pathol
 
2015
;
39
:
1370
1376
.
21
Creutzberg
CL
van Putten
WL
Koper
PC
et al.  .
Surgery and postoperative radiotherapy versus surgery alone for patients with stage-1 endometrial carcinoma: multicentre randomised trial. PORTEC Study Group. Post operative radiation therapy in endometrial carcinoma
.
Lancet
 
2000
;
355
:
1404
1411
.
22
Nout
RA
Smit
VT
Putter
H
et al.  .
Vaginal brachytherapy versus pelvic external beam radiotherapy for patients with endometrial cancer of high-intermediate risk (PORTEC-2): an open-label, non-inferiority, randomised trial
.
Lancet
 
2010
;
375
:
816
823
.
23
Stelloo
E
Nout
RA
Naves
LC
et al.  .
High concordance of molecular tumor alterations between pre-operative curettage and hysterectomy specimens in patients with endometrial carcinoma
.
Gynecol Oncol
 
2014
;
133
:
197
204
.
24
Loukola
A
Eklin
K
Laiho
P
et al.  .
Microsatellite marker analysis in screening for hereditary nonpolyposis colorectal cancer (HNPCC)
.
Cancer Res
 
2001
;
61
:
4545
4549
.
25
Bosse
T
ter Haar
NT
Seeber
LM
et al.  .
Loss of ARID1A expression and its relationship with PI3K-Akt pathway alterations, TP53 and microsatellite instability in endometrial cancer
.
Mod Pathol
 
2013
;
26
:
1525
1535
.
26
Jansen
AML
van Wezel
T
van den Akker
BEWM
et al.  .
Combined mismatch repair and POLE/POLD1 defects explain unresolved suspected Lynch syndrome cancers
.
Eur J Hum Genet
 
2015
;
24
:
1089
1092
.
27
Planck
M
Halvarsson
B
Palsson
E
et al.  .
Cytogenetic aberrations and heterogeneity of mutations in repeat-containing genes in a colon carcinoma from a patient with hereditary nonpolyposis colorectal cancer
.
Cancer Genet Cytogenet
 
2002
;
134
:
46
54
.
28
Diaz-Padilla
I
Romero
N
Amir
E
et al.  .
Mismatch repair status and clinical outcome in endometrial cancer: a systematic review and meta-analysis
.
Crit Rev Oncol Hematol
 
2013
;
88
:
154
167
.
29
Goodfellow
PJ
Buttin
BM
Herzog
TJ
et al.  .
Prevalence of defective DNA mismatch repair and MSH6 mutation in an unselected series of endometrial cancers
.
Proc Natl Acad Sci U S A
 
2003
;
100
:
5908
5913
.
30
Kim
TM
Laird
PW
Park
PJ.
The landscape of microsatellite instability in colorectal and endometrial cancer genomes
.
Cell
 
2013
;
155
:
858
868
.
31
Goodfellow
PJ
Billingsley
CC
Lankes
HA
et al.  .
Combined microsatellite instability, MLH1 methylation analysis, and immunohistochemistry for lynch syndrome screening in endometrial cancers from GOG210: an NRG Oncology and Gynecologic Oncology Group Study
.
J Clin Oncol
 
2015
;
33
:
4301
4308
.
32
Yan
M
Schwaederle
M
Arguello
D
et al.  .
HER2 expression status in diverse cancers: review of results from 37,992 patients
.
Cancer Metastasis Rev
 
2015
;
34
:
157
164
.
33
Mojtahed
A
Schrijver
I
Ford
JM
et al.  .
A two-antibody mismatch repair protein immunohistochemistry screening approach for colorectal carcinomas, skin sebaceous tumors, and gynecologic tract carcinomas
.
Mod Pathol
 
2011
;
24
:
1004
1014
.

Author notes

Clinical Trial tissue samples
Translational research was performed on tissue samples from two randomized clinical trials (PORTEC-1 and PORTEC-2). PORTEC-2 trial register number is ISRCTN16228756 and PORTEC-1 was conducted before time of trial registries. Both trials were supported by grants from the Dutch Cancer Society (CKTO 90-01 and CKTO 2001-04).

Supplementary data