Evaluation of Micro Satellite Instability and Mismatch Repair Status in Different Solid Tumors: A Multicenter Analysis in a Real World Setting

Immune-checkpoint inhibitors (ICIs) play a key role in the treatment of advanced stage colorectal cancer (CRC) patients featuring a deficient DNA mismatch repair (dMMR) system or a high microsatellite instability (MSI-H) profile. However, beyond the established role in CRC patients, ICIs have highly proven efficacy in other solid tumors featuring MSI-H/dMMR status represented by endometrial, gastric, ovarian, prostatic, and pancreatic carcinomas (EC, GC, OC, PrC, and PaC). Our aim was to compare the concordance rates among the Idylla™ MSI test, TapeStation 4200, and immunohistochemical (IHC) analysis in assessing MSI-H/dMMR status in EC, GC, OC, PrC, and PaC patients. The Sanger sequencing-based Titano MSI test was used in discordant cases. One hundred and eighty-five cases (n = 40 PrC, n = 39 GC, n = 38 OC, n = 35 PaC, and n = 33 EC) were retrospectively selected. MMR protein expression was evaluated by IHC. After DNA quality and quantity evaluations, the IdyllaTM and TapeStation 4200 platforms were adopted for the evaluation of MSI status. Remarkably, compared to IHC, the Idylla™ platform achieved a global concordance rate of 94.5% (154/163) for the microsatellite stable (MSS)/proficient MMR (pMMR) cases and 77.3% (17/22) for the MSI-H/dMMR cases. Similarly, a global concordance rate of 91.4% (149/163) and 68.2% (15/22) for MSS/pMMR and MSI-H/dMMR cases was also identified between IHC and the TapeStation 4200 microfluidic system. In addition, a global concordance of 93.1% (148/159) and 69.2% (18/26) for MSS/pMMR and MSI-H/dMMR cases was observed between the Idylla™ and TapeStation 4200 platforms. Discordant cases were analyzed using the Titano MSI kit. Overall, our data pinpointed a central role for molecular techniques in the diagnostic evaluation of dMMR/MSI-H status not only in CRC patients but also in other types of solid tumors.


Introduction
DNA mismatch repair (MMR) is a highly conserved system responsible for restoring mismatching errors, such as single base mismatches, or small insertions and deletions. When these errors are not corrected, owing to a flawed MMR system, genomic steadiness is disrupted during DNA replication and recombination-a phenomenon that eventually gives rise to multiple cancer-associated mutations [1][2][3][4][5]. Generally, a deficient MMR (dMMR) system, is triggered by germline (e.g., Lynch syndrome), somatic, and epigenetic changes, which, in turn, result in the inactivation of MMR genes [1,3,[6][7][8][9]. Thus, identifying MMR defective genes is crucial for cancer patient management.
The status of the MMR system is generally assessed by immunohistochemical analysis of the four proteins involved in MMR: namely, MLH1, MSH2, MSH6 and PMS2 [1,10]. However, evaluation and interpretation of immunohistochemical results may at times be challenging because of intra-and inter-observer variability and pre-analytical and analytical issues [1]. In this scenario, evaluation of microsatellite instability (MSI) through molecular approaches, such as polymerase chain reaction (PCR), genomic sequencing, and capillary electrophoresis, has been proposed as a valuable alternative to overcome some of the issues inherent to immunohistochemistry (IHC) [1,11].
Clinically, MMR/MSI testing has attracted increasing interest following Food and Drug Administration (FDA) approval in 2017 of the anti-programmed cell death protein 1 (PD-1) immune checkpoint monoclonal antibody pembrolizumab for MSI high (MSI-H)/dMMR patients with unresectable or metastatic solid tumors, regardless of age and histotype [29,30].
Beyond the established role of immune checkpoint inhibitors (ICIs) in the management of colorectal cancer (CRC) patients, these anticancer drugs have proven highly effective in other solid tumors featuring MSI-H/dMMR status. Among these, endometrial cancer (EC) displays a higher incidence of MSI-H status in endometrioid histotypes (40%) than in serous ones (2%) [31]. Overall, about 30% of primary ECs, and 13% to 30% of recurrent ECs, are MSI-H/dMMR [32]. Regarding the response rates to ICIs, MSI-H/dMMR advanced EC patients show higher response rates (from 27% to 57%) compared to microsatellite stable (MSS) patients (from 3% to 23%) [32].
A positive response to ICIs has also been reported in gastric carcinoma (GC) patients, who in 20% of cases feature MSI-H [33]. Indeed, MSI-H GC patients show a higher objective response rate (ORR) and disease control rate (DCR) to ICIs than MSS GC patients [34]. On the other hand, a lower frequency (12%) of MSI-H/dMMR status has been seen in unselected ovarian cancers (OCs) [35][36][37]. Regarding treatment response, more studies are needed to validate the efficacy of ICIs in MSI-H/dMMR OCs, suggesting that for this type of cancer, a better treatment option would be a combination therapy of ICIs and other anticancer drugs [38,39]. In intraductal papillary mucinous neoplasm (IPMN) of the pancreas, the incidence of MSI-H/dMMR is lower, dropping to 6.9% [40]. The lowest incidence rates of MSI-H/dMMR status have, instead, been observed in prostatic (PrC, about 3%) and pancreatic (PaC, about 1%) cancers [41,42]. However, despite such low incidence, ICIs have proven highly effective in both PrC and PaC patients featuring an MSI-H/dMMR status [41,42].
Thus, this evidence strongly underlines the importance of expanding the evaluation of MSI-H/dMMR status to different types of solid tumors. Although our research group has demonstrated the feasibility of adopting the fully automated PCR high-resolution melt curve analysis (Idylla TM ) and the automated microfluidic electrophoretic run chipbased assay (TapeStation 4200) to assess MSI-H/dMMR status in CRC patients [11], little is yet known about their performance in other types of tumors. In an attempt to fill this knowledge gap, we compared the concordance rates between these two molecular approaches and IHC analysis in assessing MSI-H/dMMR status in EC, GC, OC, PrC, and PaC patients. The Sanger sequencing-based Titano MSI test (Diatech Pharmacogenetics, Jesi, Italy) was used to confirm discordant cases.

Study Design
For our comparative analysis, we retrospectively selected a total of n = 199 formalin fixed paraffin embedded (FFPE) (n = 40 PrC, n = 40 GC, n = 39 OC, n = 40 PaC, and n = 40 EC) tumor samples, analyzed by IHC for MMR protein expression at the Fondazione IRCCS Casa Sollievo della Sofferenza (San Giovanni Rotondo, Italy) from 2015 to 2020. The percentage of neoplastic cells and the pathological status of each patient were evaluated by two expert pathologists. Briefly, for the molecular evaluation of MSI, tumor and a corresponding normal tissue specimen were collected after revision of original hematoxylinand-eosin (H&E)-stained sections. All information regarding human material was managed using anonymous numerical codes, and all samples were handled in compliance with the Declaration of Helsinki (http://www.wma.net/en/30publications/10policies/b3/ last accessed 25 May 2021).
MMR protein expression was categorized as (i) retained (i.e., proficient MMR; pMMR), when a moderate to strong nuclear protein expression was detected in tumor cells as well as in internal control; and (ii) lost (i.e., dMMR), when a complete loss of nuclear expression in tumor cells was observed, but retained in normal cells [43]. Tumor samples showing absence of immunoreactions in internal controls were classified as "inadequate" for IHC evaluation and analysis and excluded from the study. A total of 185 cases were submitted to molecular approach (Figure 1).

Figure 1.
We retrospectively selected a total of n = 185 formalin fixed paraffin embedded (n = 40 prostate, n = 39 gastric, n = 38 ovarian, n = 35 pancreatic, and n = 33 endometrial) tumor samples, previously analyzed by immunohistochemistry for mismatch repair protein expression. In all samples, microsatellite instability (MSI) status was evaluated by Idylla™ MSI test and TapeStation 4200. Discordant cases were further analyzed by Titano MSI test analysis.

DNA Extraction and Qualification
DNA extraction and qualification were performed as previously reported [11]. Overall, four 5-µm thick sections obtained from tumor tissues and corresponding normal mucosae were adopted. The neoplastic cell area was manually microdissected. Then, DNA was extracted with the Mini Amp kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. Finally, it was eluted in 30 µL of DNAse-and RNAsefree water (Thermo Fisher Scientifics, Waltham, MA, USA) and stored at −20 °C until DNA was qualified by the TapeStation 4200 microfluidic platform. In particular, the genomic ladder and sample buffer (Agilent Genomic ScreenTape, Agilent Technologies) were performed on the Genomic Screen Tape device (Agilent Technologies) according to the manufacturer's instructions ( Figure 1).

Microfluidic Analysis for MSI Status Evaluation
TapeStation 4200 analysis was performed as previously reported [11]. Five amplification reaction mixtures were prepared to analyze patients' MSI status by using the Bethesda panel, starting from 20 ng of extracted DNA on the TapeStation 4200 system, as previously validated [11]. Briefly, for each patient, 1 µL of each amplified product from both tumor and normal tissues of each reaction mixture and 3 µL of D1000 Buffer (Agilent Technologies) were automatically charged on a solid device comprising 16 nanocapillaries (D1000 ScreenTape) and analyzed on the TapeStation 4200 platform. Results were inspected with proprietary software (TapeStation Analysis Software, Agilent Technologies) (Figure 1).

DNA Extraction and Qualification
DNA extraction and qualification were performed as previously reported [11]. Overall, four 5-µm thick sections obtained from tumor tissues and corresponding normal mucosae were adopted. The neoplastic cell area was manually microdissected. Then, DNA was extracted with the Mini Amp kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. Finally, it was eluted in 30 µL of DNAse-and RNAse-free water (Thermo Fisher Scientifics, Waltham, MA, USA) and stored at −20 • C until DNA was qualified by the TapeStation 4200 microfluidic platform. In particular, the genomic ladder and sample buffer (Agilent Genomic ScreenTape, Agilent Technologies) were performed on the Genomic Screen Tape device (Agilent Technologies) according to the manufacturer's instructions ( Figure 1).

Microfluidic Analysis for MSI Status Evaluation
TapeStation 4200 analysis was performed as previously reported [11]. Five amplification reaction mixtures were prepared to analyze patients' MSI status by using the Bethesda panel, starting from 20 ng of extracted DNA on the TapeStation 4200 system, as previously validated [11]. Briefly, for each patient, 1 µL of each amplified product from both tumor and normal tissues of each reaction mixture and 3 µL of D1000 Buffer (Agilent Technologies) were automatically charged on a solid device comprising 16 nanocapillaries (D1000 ScreenTape) and analyzed on the TapeStation 4200 platform. Results were inspected with proprietary software (TapeStation Analysis Software, Agilent Technologies) ( Figure 1).

Idylla™ MSI Assay
The Idylla™ MSI test was performed as previously reported [11]. This test consists of a fully automated RT-qPCR system able to detect microsatellite instability directly from FFPE human cancer tissue sections. Overall, a total of four 5-micron thick slides obtained from tumor specimens were used for MSI analysis, according to the manufacturer's instructions. Briefly, tumor areas were scraped from each sample with a sterile blade and inserted into an MSI cartridge, as previously described [11]. Overall, DNA was automatically extracted by a combination of HIFU, enzymatic/chemical digestion and heat, and then amplified into 5 PCR chambers. The Idylla MSI assay then analyzed homopolymers in 7 biomarkers (ACVR2A, BTBD7, DIDO1, MRE11, RYR3, SEC31A, and SULF2). Fluorescent signals, generated by fluorescently labeled molecular beacon probes, were automatically inspected by proprietary software able to analyze a minimum allele frequency of 10% and to calculate a probability score (MSI score) for all the tested biomarkers. Overall, samples with ≥2 of the 7 mutated biomarkers and <2 of the 7 mutated biomarkers were classified as MSI-H and MSS, respectively ( Figure 1).

Discordant Cases Evaluted with the Titano MSI Test
The Titano MSI test analysis was performed as previously reported [11]. Briefly, the Titano MSI test (Diatech Pharmacogenetics) was used to investigate discordant cases between IHC and the 2 molecular approaches adopted in the study. DNA obtained from tumor and corresponding normal mucosa samples for each case were analyzed with MSI Titano kit. In brief, this test detects MSI status in cancer tissues through multiplex amplifications with fluorescent primers and subsequent DNA fragment analysis on an automated sequencer. An input of optimal 20 ng of extracted DNA is generally required to analyze variations in the number of microsatellite loci of 10 different molecular targets (BAT25, BAT26, D2S123, D17S250, D5S346, BAT40, D18S58, NR21, NR24, TGFβRII). In essence, variations are identified by comparing the peak profiles generated by capillary electrophoresis runs of tumor and corresponding normal tissue samples from each patient (Figure 1).

Figure 2.
Venn diagram representing the concordance rate among the technologies adopted in the present study. Briefly, dark gray circle represents fully concordant cases among the three different assays; dark gray plus dark orange circles represent the concordant cases between TapeStation 4200 and IHC; dark gray and dark blue circles represent the concordant cases between Idylla TM and IHC.

Figure 2.
Venn diagram representing the concordance rate among the technologies adopted in the present study. Briefly, dark gray circle represents fully concordant cases among the three different assays; dark gray plus dark orange circles represent the concordant cases between TapeStation 4200 and IHC; dark gray and dark blue circles represent the concordant cases between Idylla TM and IHC.  Discordant cases among the TapeStation 4200 system, Idylla TM platform, and IHC approach were successfully analyzed by applying the Titano MSI kit. The results showed that in 14 out of 27 (51.8%) cases, the Titano kit and IHC showed concordant results (n = 10 MSS/pMMR and n = 4 MSI-H/dMMR); whereas in 13 out 27 cases (48.2%), a discordant result was reported (n = 7 MSS/dMMR and n = 6 MSI-H/pMMR). As far as the comparison between the Titano MSI kit and Idylla TM is concerned, overall, 22 concordant cases (81.5%; n = 13 MSS and n = 9 MSI-H) were detected. Only five discordant cases (n = 4 MSS for Titano MSI kit and MSI-H for Idylla TM and n = 1 MSI-H for Titano MSI kit and MSS for Idylla TM ) were reported. The results showed that in 11 out of 27 (40.7%) cases, the Titano MSI kit and TapeStation 4200 showed concordant results (n = 6 MSS and n = 5 MSI-H); whereas in 16 out 27 cases (59.3%) a discordant result was reported (n = 11 MSS for Titano MSI kit and MSI-H for TapeStation 4200 and n = 5 MSI-H for Titano MSI kit and MSS for TapeStation 4200). The results are summarized in Table 6.
Although IHC remains the gold standard approach for MSI biomarker analysis in solid tumors, our data indicate that both Idylla and TapeStation molecular platforms are able to analyze MSI status just as efficiently as IHC. Indeed, a high concordance rate was observed between the Idylla™ platform and IHC (94.5% and 77.3% for the for MSS/pMMR and MSI-H/dMMR cases, respectively), and between the TapeStation 4200 microfluidic system and IHC (91.4% and 68.2% for the MSS/pMMR and MSI-H/dMMR cases, respectively). Of note, overall high concordance rates were also seen between the TapeStation 4200 and Idylla™ platforms (93.1% and 69.2% for MSS and MSI-H cases, respectively). In addition, in line with previous data, GCs (28.9%) and ECs (27.0%) showed the highest rates of dMMR cases, followed by PrCs (7.5%) and OCs (5.3%). In our experience, all PaC cases showed a pMMR profile. Similarly, MSI-H status was more frequent in GCs (28.2% and 28.2%, by Idylla TM and TapeStation 4200, respectively) and ECs (33.3% and 24.3%, by Idylla TM and TapeStation 4200, respectively) than in OCs (7.3% and 13.2%, by Idylla TM and TapeStation 4200, respectively), PrCs (2.5% and 7.5%, by Idylla TM and TapeStation 4200, respectively), and PaCs (0.0% and 5.7%, by Idylla TM and TapeStation 4200, respectively).
Hence, our data clearly highlight that molecular analysis represents a valid upfront approach to evaluate MSI status in various types of cancer patients who would definitely benefit from ICI treatments. Indeed, we have demonstrated that MMR/MSI testing has acquired a central role not only in CRC management but also in the clinical stratification of different solid tumors [29,30]. In fact, patients displaying MSI-H/dMMR status respond well to immunotherapy regimens [29,30]. It is against this background that molecular analysis could represent a valuable strategy to overcome the well-known limitations of IHC, including intra-and inter-observer variability and pre-analytical and analytical issues [1]. In particular, a recent study has highlighted that a pMMR IHC profile may be retained in cases featuring antigenically intact non-functioning proteins despite the presence of an MSI-H status [44]. In addition, the high global concordance (91.4% and 68.2% for MSS/pMMR and MSI-H/dMMR cases, respectively) observed between the fully automated PCR highresolution melt curve analysis (Idylla TM ) and the automated microfluidic electrophoretic run chip-based assay highlights that these user-friendly, rapid, and sensitive molecular techniques are interchangeable and, therefore, all equally efficient. Moreover, in all cases they seemed to outperform the gold standard IHC approach in selecting MSI-H/dMMR advanced patients for ICI treatments. Remarkably, for each solid tumor analyzed, they were both able to identify a higher percentage of patients eligible for ICI administration than did IHC. Although these results clearly indicate the practical advantages of integrating these approaches into clinical practice, further studies are definitely warranted to confirm whether their implementation in clinical practice might actually help oncologists to choose the best treatment option for their patients.
Among the 27 discordant cases, 13 tumors showed different IHC and Titano profiling. Twelve of these tumors were endometrial and prostate adenocarcinomas. Of note, different molecular mechanisms characterize MSI in the different organs and may impact microsatellite instability testing and MMR IHC [45]. Focusing on EC, high agreement between IHC and MSI analysis has been described. However, several factors can affect this view: (i) MSH6 negative tumors may show MSS or MSI-L at MSI analysis [46]; (ii) POLE mutated tumors may present an MSI profile due to the accumulation of mutations in microsatellite loci despite preserved MMR protein expression; (iii) subclonal loss of MMR protein expression is present in around 3% of EC cases; a minimal microsatellite shift characterizes MSI EC and may interfere with the diagnostic interpretation of the results [47]; (iv) the use of obsolete IHC evaluation criteria may impact MMR evaluation [48]; and (v) the different PCR-based approaches are affected by intrinsic limitations that adequate training of the user and a high quality sample can overcome [15]. Overall, our data pinpointed a central role for molecular techniques in the diagnostic evaluation of MSI-H/dMMR status not only in CRC patients but also in other types of solid tumors. Future efforts should be focused on designing and developing clinical trials to validate the actual ability of molecular techniques to stratify patients with different types of solid tumors, including breast cancer [49,50], who may benefit from ICIs. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.