Evaluation of Microsatellite Instability via High-Resolution Melt Analysis in Colorectal Carcinomas

Abstract

Background/Objectives: Colorectal cancer (CRC) is the third leading cause of cancer death globally, with rising incidence. The immunohistochemistry (IHC) for mismatch repair (MMR) proteins is the first technique used in routine practice to evaluate an MMR status. Microsatellite instability (MSI) may be tested in case of doubt during IHC staining. This study introduces a novel high-resolution melt (HRM) protocol for MSI detection and compares it with traditional fragment length analysis (FLA) via capillary electrophoresis. Methods: A total of 100 formalin-fixed and paraffin-embedded CRC specimens were analyzed using two distinct protocols: one based on FLA (TrueMark MSI Assay kit) and another one based on HRM (AmoyDx^® Microsatellite Instability Detection Kit). Results: Overall, 68 (68.0%) of the cases were MSS, and 32 (32.0%) were MSI-H. HRM analysis was first successfully carried out in all the cases. A perfect concordance in MSI evaluation between HRM and FLA was observed. HRM showed slightly shorter hands-on time and turnaround time. Conclusions: We provided evidence of the validity of this new HRM approach in determining the MSI status of colorectal carcinomas.

1. Introduction

Colorectal cancer (CRC) is the world’s third biggest cause of cancer death, and its incidence is rising in many nations [1,2].

The etiology of colorectal carcinomas is multifactorial and can be attributed to both genetic and environmental factors. Most colorectal carcinomas are sporadic, with both genetic and epigenetic alterations (which would lead to the aberrant methylation of oncosuppressor genes), resulting in an inactivation or alteration of the genes responsible for regulating the pathways of cell proliferation and differentiation, with the resulting promotion of the neoplasm [3]. Sporadic cases share many of their genetic alterations with their hereditary counterparts, and therefore the study of hereditary syndromes, such as familial adenomatous polyposis (FAP) and Lynch syndrome (HNPCC), has helped researchers to understand the molecular pathogenesis underlying sporadic colorectal carcinomas [3].

Three distinct pathways of genomic instability have been identified [4,5]: (i) the Chromosomal Instability (CIN) Pathway (70–85% of all the cases, involving the activation of a proto-oncogene such as KRAS, and the inactivation of oncosuppressors such as APC and TP53 [6]; (ii) Microsatellite Instability (MSI), due to the loss of functionality of the genes/proteins of the mismatch repair system, the main of which are MLH1 (3p21), PMS2 (7p22), MSH2 (2p16), MSH6 (2p16) [7,8]; (iii) the CpG Island Methylator Phenotype [9,10], mainly associated with the hypermethylation of the promoter of numerous genes (including MLH1) and with BRAF pathogenic mutations. 70–85% of all colorectal cancer cases are due to this pathway.

The American College of Gastroenterology and the American Society of Clinical Oncology (2015) advocate using a screening strategy for all the CRC specimens to identify patients for germline testing to diagnose Lynch syndrome [11,12,13]. The first stage is the dMMR (mismatch repair deficiency) analysis, preferably performed using immunohistochemistry (IHC) for the MLH1, MSH2, MSH6, and PMS2 proteins [14,15]. Positive immunostaining for all the proteins is consistent with an intact MMR system; hence, no additional tests are required. Samples exhibiting the immunohistochemistry loss of MMR proteins must subsequently be tested for BRAF mutations. The ACG/ASCO screening algorithm recommends performing BRAF analysis in CRC without MLH1 protein staining, followed by MLH1 promoter methylation analysis if the BRAF V600E mutation is not discovered [12,13]. Moreover, in those cases harboring MLH1 staining loss, the MLH1 promoter methylation analysis can also be performed before the BRAF evaluation [16]. In the case of immunohistochemical doubts, such as the partial loss of staining positivity, a confirmation test via MSI analysis is strongly recommended. The MSI phenotype is usually tested using PCR-based assays [17], and the fragment length analysis (FLA) is to date the widest used, even if other methods, such as automated PCR high-resolution melt curve analysis [18] or the automated microfluidic electrophoretic run chip-based assays [19,20], have also been successfully used.

The primary aim of our study was to validate a high-resolution melt (HRM) analysis method focusing on microsatellite regions distinct from the Bethesda markers for MSI detection in colorectal carcinoma specimens, using fragment length analysis (FLA) as the reference standard. By comparing the performance of HRM to FLA, we also evaluated whether HRM could offer certain advantages, such as reduced hands-on and turnaround times. In this study, we tested a new HRM protocol not including Bethesda satellites and that allows MSI analysis in CRC (AmoyDx^® Microsatellite Instability Detection Kit) and compared it with fragment length assay capillary electrophoresis.

2. Materials and Methods

The DNA samples were retrieved from anonymized electronic files from the Solid Tumor Molecular Pathology Laboratory, IRCCS Policlinico di S.Orsola Bologna (Italy) to find samples of colorectal carcinomas. A total of 100 consecutive DNA CRC specimens were analyzed at the Solid Tumor Molecular Pathology Laboratory in Bologna using two distinct protocols: a TrueMark MSI Assay kit (Thermo Fisher Scientific, Waltham, MA, USA), based on FLA, and an AmoyDx^® Microsatellite Instability Detection Kit (AmoyDx, Xiamen, China), based on HRM. All the samples were extracted from formalin-fixed and paraffin-embedded (FFPE) specimens, starting from 2 to 4 10 um-thick unstained sections. The area of interest was selected by a pathologist in a final hematoxylin and eosin (H&E) section, for ensuring at least 20% of the neoplastic cells in the area used for the analysis. The extraction was performed using the QuickExtract FFPE DNA Extraction Kit (LGC Biosearch Technology, Steinach, Germany), and then quantified using the Qubit fluorometer (Thermo Fisher Scientific).

2.1. MSI by Fragment Length Analysis

Fragment length analysis (FLA) was performed using the TrueMark MSI Assay kit (Thermo Fisher Scientific); 5 ng of gDNA was used for amplification. The kit allowed the analysis of 15 satellites, 13 mononucleotidic markers (BAT-25, BAT-26, BAT-40, CAT-25, NR-21, NR-22, NR-24, NR-27, ABI-16, ABI-17, ABI-19, ABI-20A, ABI-20B) and 2 pentanucleotide repeats (PentaD, TH-01, sample ID markers). In the samples where a non-neoplastic specimen was available, the FLA was performed both on the neoplastic and control specimens. Briefly, in a total of 10 µL, a mix was made containing approximately 5 ng of gDNA, 4 µL of MSI Assay Master Mix, and 1 µL of MSI Assay Primer Mix. The 10 µL were aliquoted into one well for each sample to be analyzed. The samples were then amplified with the following program: 95 °C for 11 min, [94 °C for 20 secs, 59 °C for 2 min] for 29 cycles, and 60 °C for 25 min. At the end of PCR, 2 µL of PCR product was added to a mixture of Formamide (17 µL) and Size Standard (1 µL—GeneScan™ 600 LIZ™, Thermo Fisher Scientific), and aliquoted to a MicroAmp™ Optical 96-Well Reaction Plate. The samples were then run on a 3500/3500 × L Genetic Analyzer.

Results were analyzed using the GeneMapper ID-X Software tool v1.5 (Thermo Fisher Scientific). The samples were classified as MSS (no unstable markers or if less than 30% of the evaluable markers were unstable), MSI-Low (MSI-L, at least one unstable marker but less than 30% of the analyzed microsatellites), or MSI-High (MSI-H, equal or more than 30% of the evaluable markers were unstable).

2.2. MSI by High-Resolution Melt Analysis

A high-resolution melt analysis was performed using the AmoyDx^® Microsatellite Instability (MSI) Detection Kit. Briefly, a total of 5 ng was used for DNA amplification. Once the reaction mix was set up, each sample was analyzed as follows: each well was already pre-filled with a multiplex of three primers, two satellites, and an internal control (IC). The following 8 mononucleotide repeats were analyzed by the kit: EIF4E3, IFT140, PPP1CC, UBAC2, PRR5-ARHGAP8, ACVR2A, TAOK3, RBM14-RBM4. The plate was then run on a Real-Time PCR System SLAN-96S for a total of 125 min [95 °C for 5 min (1 cycle); 95 °C for 5 s, 54 °C for 15 s, 72 °C for 10 s (60 cycles); 95 °C for 1 min, 45 °C for 3 min, and 35 °C for 1 min (1 cycle); the melting steps were from 35 °C to 58 °C].

The samples were then classified as MSS (0 or <25% of evaluable satellites unstable); and MSI-H (more than 25% of evaluable satellites unstable). In all the cases, only the DNA from the neoplastic specimen was analyzed.

3. Results

Details of each one of the 100 specimens are reported in Supplementary Table S1.

3.1. Fragment Length Analysis vs. HRM

Regarding fragment length analysis, the assay was successfully carried out in all of the cases. Overall, 68 (68.0%) cases were MSS (including 3 cases evaluated as MSI-L), and 32 (32.0%) were MSI-H (

Table 1. MSI results obtained using FLA and HRM techniques.

Sample Size (N = 100)		FLA
		MSI-H	MSI-L	MSS	Total
HRM	MSI-H	32	0		32
	MSI-L/MSS *	0	3	65	68
			68
	Total	32	68		100

MSI-H: high microsatellite instability; MSI-L: low microsatellite instability; MSS: microsatellite stability; HRM: high-resolution melt; FLA: fragment length analysis. * AmoyDx HRM MSI Detection Kit does not distinguish between MSS and MSI-L.

) (Figure 1 and Figure 2).

An HRM analysis was first successfully carried out in 95 of the 100 cases. In the five cases that were not evaluable, 7 ng of DNA, instead of 5 ng, was then used for the re-analysis. Using 7 ng of input DNA, these five cases were successfully evaluated by HRM.

Considering that according to the ESMO recommendations [21], MSI-L and MSS should be viewed as a unique clinical group (i.e., MSS sample), we observed 100% concordance between FLA and HRM analysis in evaluating the MSI status.

3.2. Hands-On Time and Turnaround Time

We also compared both techniques regarding hands-on time (HOT) and turnaround time (TAT) (

Table 2. Technical protocols comparison between FLA and HRM.

	FLA MSI	HRM * MSI
DNA	5 ng/µL	5 ng/µL (7 ng/µL in 5 samples)
Sample preparation	15 min	25 min
Time for PCR (thermal cycler or Real-Time machine)	120 min	125 min
Sample preparation for the capillary run	15 min	/
Fragment analysis	55 min	/
Data analysis (per sample)	2–5 min	1–3 min
TOTAL
TAT	~210 min	~150 min
HOT	~30 min	~25 min

FLA: fragment length analysis; HRM: high-resolution melt analysis; TAT: turnaround time; HOT: hands-on time. * Performed by the AmoyDx ^® Microsatellite Instability kit.

). The FLA protocol was faster in sample preparation (15 min) but needs a preparation step after the PCR reaction (PCR product preparation for the capillary run). On the contrary, the HRM protocol had a slightly longer pre-PCR protocol (about 25 min), but after loading, the Real-Time machine had no other “wet” steps. Overall, both of the techniques are rapid and not time-consuming with a similar hands-on-time—about 30 min for FLA and about 25 min for HRM—and a total turnaround time of 210 min for FLA and 150 min for HRM.

4. Discussion

The microsatellite instability state (MSI) is a phenotype resulting from alterations in the mismatch repair (MMR) system. An MSI condition reflects a malfunction of MMR proteins (MMR deficiency—dMMR) such as, for example, MLH1, PMS2, MSH2, and MSH6. The evaluation of MMR/MSI status is considered a standard clinical practice in the work-up of several tumors, such as colorectal carcinomas, gastric cancers, and endometrial carcinomas. Alterations in the MMR system can therefore be assessed either by assessing the expression of the MLH1, PMS2, MSH2, and MSH6 proteins via immunohistochemistry, or by assessing the MSI status. In general, the most frequently recognized approach for MSI evaluation depends on the fragment length analysis of the Bethesda panel, which covers two mononucleotide (BAT-25 and BAT-26) and three dinucleotide (D5S346, D2S123, and D17S250) repetitions [22] or, alternatively, a panel encompassing five poly-A mononucleotide repeats (BAT-25, BAT-26, NR-21, NR-24, and NR-27) [23]. Several studies have demonstrated that alternative methods can be successfully used for MSI evaluation [24,25,26]. One option is the fully automated PCR Idylla MSI assay (Biocartis) which evaluates MSI using seven repeat markers within 150 min per sample, including DNA extraction [18,27,28,29,30,31]. Another possibility is the use of automated microfluidic electrophoretic run chip-based assays that ensure a simple, fast, and high throughput (up to 15 samples in 10 min) approach for MSI evaluation [20]. Also, droplet-digital PCR (ddPRC) was successfully used for MSI detection, consisting of a rapid and cost-effective multiplex assay suitable for large-scale patient testing; albeit, the data must be interpreted by a specialist laboratory [29]. More recently, MSI analysis using NGS has been introduced [32], although its use in clinical practice is not yet widespread, mainly due to the fact that there is no universal consensus on the precise definition of MSI-high (MSI-H), as different gene panels and different analysis algorithms may lead to slightly different results. For this reason, MSI via an NGS test has to be verified against MMR-IHC or MSI by PCR [32].

In this study, we compared an HRM-based strategy based on microsatellite regions distinct from the Bethesda markers with a fragment length analysis (FLA), the gold standard method for MSI analysis.

There were no inconsistent findings across the board for the approaches. The unique cases with a minimal discrepancy between the two methods were categorized as MSS by HRM but MSI-L by FLA. Nevertheless, MSI-L patients ought to be categorized alongside MSS instances in accordance with standards [21], so there would be a 100% agreement considered between the two approaches.

It was shown that the FLA method required less DNA, which could be helpful for biopsies and other samples with low cellularity. Even though HRM required more starting DNA, all the instances were successfully evaluated utilizing it.

FLA took a little more hands-on time because it comprises both a post-PCR phase (plate preparation for capillary sequencer loading) and an amplification stage (PCR setup). The HRM approach proved to be incredibly fast, requiring the operator’s assistance only during the PCR setup phase, which took about 5 min. Overall, the HRM strategy required less time (150 vs. 210 min) to produce results. Furthermore, considering that in molecular pathology labs, NGS is increasingly being used over the conventional Sanger method because of its multi-gene and high-sensitivity approaches, Sanger sequencers are becoming less common than Real-Time PCR machines.

Finally, we provided evidence of the validity of an HRM approach focusing on microsatellite regions distinct from the Bethesda markers in determining the MSI status of colorectal carcinomas. Subsequent investigations may examine the suitability of this technique for assessing MSI in other neoplasms, like endometrial carcinoma, where MSI evaluation is essential for the molecular categorization of these malignancies [33,34].