Improved microsatellite instability detection in colorectal cancer patients by a combination of fourteen markers especially DNMT3a,DCD,and MT1X

Abstract

BACKGROUND:

Microsatellite instability (MSI) results from genetic and epigenetic changes. Studying Microsatellite instability can help in treatment and categorization of colorectal cancer (CRC) patients.

OBJECTIVES:

We aimed to investigate whether 14 genomic markers consisting of BAT-62, BAT-60, BAT-59a, BAT-56a, BAT-56b, DCD, RIOX, RNF, FOXP, ACVR, CASP2, HSP110, MT1X, and DNMT3a can increase the detection rate of MSI in CRC.

METHODS:

Samples were stratified by pentaplex panel (Promega) and 14 markers using multiplex PCR and fragment analysis. In MSI+ samples, to identify the pattern of BRAF V600E mutation and MLH1 promoter methylation, ARMS-scorpion, and Methylation-Specific High-Resolution Melting Curve analysis, were applied respectively.

RESULTS:

Totally, 35 MSI+ cases identified by 14 marker panel. Only 18 cases of them were detected by both panels which are pentaplex and 14 marker. On the other hand, 17 new MSI+ cases just were identified by 14 markers panel. The highest diagnostic value among 14 markers is related to three makers, namely DCD, MT1X, and DNMT3a. In MSI+ cases, the rate of MLH1 promoter methylation was insignificant, ( $P$ value $=$ 0.3979) while the rate of observed BRAFV600E mutation was significantly higher ( $P$ value $=$ 0.0002).

CONCLUSION:

Fourteen marker panel showed higher sensitivity in comparison with the pentaplex panel increasing the detection rate of MSI+ cases up to 1.94 fold. Three markers namely DNMT3a, DCD, and MT1X of 14 marker panel were the best among them showing excellent diagnostic value. A combination of these markers showed 100% sensitivity and specificity in the studied group. In contrary to the markers in the pentaplex panel, these markers had the ability to detect MSI without any bias for the clinicopathological features. These markers will help to identify more end-stage MSI+ tumors which are located distal colon.

Keywords

Colorectal Neoplasm Microsatellite Instability Biomarker tumor DNA methyltransferase 3A

1. Introduction

Genomic instability is the hallmark of all types of cancers. On this basis, colorectal cancer is categorized into two main groups. Chromosomal instability (CIN) and microsatellite instability (MSI) constitute the predominant tumorigenic pathways in CRC. About 80% to 85% of all CRC cases show CIN. These tumors have a poor prognosis. On the other hand, another group represents genome sequence instability which can be evaluated by analyzing the deletion mutations in the microsatellite that is microsatellite instability (MSI) [1]. MSI analysis test is recommended for all colorectal cancer (CRC) patients at any stage [2]. Approximately 15% of sporadic colorectal cancers (CRCs) suffer from instability in their sequence of the genome. This phenotype is also seen in more than 90% of patients affected with Lynch syndrome. Initially, this test was recommended for screening Lynch syndrome [3], but its role as a predictive marker has increased its application [4]. Approved drugs for the treatment of MSI-positive cases of colorectal cancer are Nivolumab and PD-1 inhibitor Pembrolizumab. While Nivolumab is used in MSI+ cases, Pembrolizumab is used in advanced stages or refractory cases of MSI+ solid tumors. Therefore, MSI is the first guidance for immunotherapy regardless of tumor type [5, 6].

The maintenance of the genome sequence is accomplished by genetic and epigenetic mechanisms, DNA repair, replication genes, DNA methylation, histone modification, and non-coding RNAs. Defects in these mechanisms could result in genome sequence instability. Deletion and insertion mutations in microsatellite repeats which is called microsatellite instability are signs of genome sequence instability [7]. The dysfunction in the Mismatch repair (MMR) system has the main role in causing this instability [8].

A clinically in-use MSI analyzing system has been introduced in 2004 composing of five mononucleotide repeats, BAT-25, BAT-26, NR-21, NR-24, and MONO-27 [9]. These mononucleotide markers displayed better sensitivity to Bethesda panel (NCI) comprising of BAT-25, BAT-26, DSS346, D2S123, and DI7S250. These reference panels were recommended for surveying the usefulness of any new panel. These reference markers identify MSI traits in tumors which is characterized by defective mismatch repair. Tumor samples that are scored MSI high (MSI-H) by the above-mentioned panels are usually observed in the proximal colon in stage II [3].

Increased detection rate of MSI phenotype in precancerous lesions which are suspected of having Lynch syndrome by long mononucleotide repeats (LMRs) in comparison with NCI and Promega panels, indicate that NCI and Promega markers do not usually mutate in the primary stages of tumor development [10].

Good markers are markers that show instability due to any defect in the function of sequence maintenance systems which is disrupted in the first stages of tumor development and showing sensitivity and specificity more than 80%. In 1998, NCI proposed that if more than 30%–40% of markers in a patient sample showed instability, it can be considered as MSI-H. This criterion is helpful in the absence of comprehensive verifying tests [3]. Ideal markers showed a mutation rate of 100% in MSI-H cases and no mutation in MSS tissue specimens [11].

This study aimed to evaluate the detection capability of a new set of short tandem repeats (STRs) mostly located in cancer-related genes for distinguishing MSI phenotypes. For this purpose, we evaluated 14 STRs with the highest reported sensitivity. Long mononucleotide repeats (LMR), BAT-62, BAT-60, BAT-59a, BAT-56a, and BAT-56b were selected from experimental biochemistry study [12]. DCD, RIOX, RNF, FOXP, and ACVR genes with a mutation frequency of more than 75% in MSI-H samples were chosen from whole-exome sequencing studies [13]. CASP2, HSP110, and MT1X as markers having a sensitivity of more than 90% were compared with the conventional Pentaplex MSI analysis system [11, 14, 15]. Finally, DNMT3a was chosen due to its biological role in tumorgenesis and long mononucleotide repeat in its 3’UTR [16].

2. Materials and methods

2.1 Sample preparation and DNA extraction

In this study, 200 tumor samples from patients affected with primary colorectal cancer in addition to their adjacent normal tissues were collected during surgery in the colorectal surgery ward of Imam Khomeini Hospital Hospital (Tehran, Iran) and Cancer Institute (Tehran, Iran) from 2013 to 2018. All tumor specimens were analyzed pathologically. The study protocol and amendments were approved by the Ethics Committee of Pasteur Institute of Iran (Ethics code: IR.PII.REC.1398.041). All patients signed informed consent in compliance with the Pasteur Institute declaration. Normal DNA samples were also taken from peripheral blood samples of age and sex-matched normal volunteers. All volunteers signed the informed consent before donating blood samples too.

Figure 1.

Microsatellite instability analysis by 14 mononucleotide repeats.

Tissues obtained after surgery were immediately frozen and stored at $-$ 70 ${}^{\circ}$ C. DNA was purified by DNAeasy kit (Qiagen, Netherlands) according to the manufacturer’s instructions. The concentration and purity of extracted DNA were analyzed by NanoDrop 2000C Spectrophotometer (Thermo Scientific, USA). DNA of peripheral blood of normal controls was extracted by the salting-out method [17].

2.2 Microsatellite instability analysis by MSI analysis system

MSI phenotype in 200 cancerous and matched normal tissues was analyzed by the commercially available MSI detection kit (Promega, USA) based on the manufacturer’s protocol. MSI Analysis System (Promega) is based on the analysis of mononucleotide repeats namely BAT-25, BAT-26, NR-21, NR-24, MONO-27, Penta C, and Penta D. It is based on co-amplification of markers while 5’end of primers was labeled with different fluorophore dyes and fragment analysis in automated micro-capillary electrophoresis [18]. Scoring MSI status is based on accepted consensus in the National Cancer Institute workshop [3]. High MSI tumor samples showed alleles with different sizes in more than 30% of studied locations in comparison with their adjacent normal tissue [3]. According to the protocol of the Promega kit if two or more markers were unstable they were considered as MSI-H. If one unstable marker was observed, the cases were categorized as MSI-L while in MSI-S cases no unstable marker should be observed. Pentanucleotide repeats such as Penta C and Penta D were used for sample identification. PCR program for amplification of samples was initial incubation of samples in 95 ${}^{\circ}$ C for 15 minutes leading to 30 cycles of 94 ${}^{\circ}$ C for 30 s, 58 ${}^{\circ}$ C for 40 s, 72 ${}^{\circ}$ C for 40 s leading to final elongation at 60 ${}^{\circ}$ C for 30 minutes, and final cooling of samples in 4 ${}^{\circ}$ . DNA fragments were separated by ABI Sequencer (ABI PRISM ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ 3100 Genetic Analyzer applied Biosystems, Foster City, California, USA). Size and fragment files were assessed by GeneMarker Software version 1.6 (SoftGenetics LLC, State College, PA, USA) using internal controls 600.

2.3 Microsatellite instability analysis by 14 mononucleotide repeats

Fourteen studied mononucleotide repeats were BAT62, BAT60, BAT59a, BAT56a, BAT56b, FOXP, DNMT, ACVR, HSP, CAT25, RIOX, RNF, MT1X, and DCD. Colorectal cancer tissues and normal adjacent tissue in addition to healthy blood DNA samples were analyzed by this panel, Fig. 1. PCR was done in three multiplex groups. Group one was composed of BAT62, BAT59a, BAT60, and BAT56a while group two was composed of BAT56, MT1X, DCD, CASP, and DNMT. Group three comprised of HSP, FOXP, RIOX, RNF, and ACVR. For amplification of these locations, primers were designed by primer3 online software and for targeted amplification, primer sequences were blasted on the NCBI database (www.ncbi.nlm.nih.gov/blast) and the second structures were analyzed on online IDT software (Coralville, Iowa, USA). Primers were synthesized by Metabion Company (Germany). The 5’ end of each forward primer was labeled with Fam/TAMRA/ROX fluorophores. The sequence of primers is given in Supplementary Table 1.

Multiplex TEMPase hot start 2x Master Mix (Ampliqon, Denmark) in addition to 0.4 pmol/L of each primer were used for amplification of each fragment. PCR was carried out according to the following protocol: initial denaturation at 95 ${}^{\circ}$ C for 15 minutes leading to 35 cycles at 94 ${}^{\circ}$ C for 30 s, 58 ${}^{\circ}$ C for 40 s and 72 ${}^{\circ}$ C for 40 s leading to final elongation at 72 ${}^{\circ}$ C for 10 minutes. DNA fragments were separated by ABI Sequencer (ABI PRISM ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ 3100 Genetic Analyzer Applied Biosystems, Foster City, California, USA). Size and fragment files were assessed by GeneMarker Software version 1.6 (SoftGenetics LLC, State College, PA, USA) using internal control 560. Analysis of blood samples from healthy controls was used to determine qusimonomorphic and polymorphic markers and allele variation range. Markers that had $\pm$ 3 bp variation around the most frequent allele were considered as qusimonomorphic markers [24].

2.4 MLH1 promoter methylation analysis

MLH1 promoter was studied with the methylation-specific MS-HRM method. To design primers for studying the methylation of hMLH1gene, the upstream sequences were retrieved from NCBI database. Primers were designed according to the manually converted DNA sequence. Primer-BLAST was carried out for evaluating the specificity of each primer (http:// blast.ncbi.nlm.nih.gov/blast). About 1 $\mu$ g of DNA was converted by GOLD Zymoresearch kit (Zymoresearch, USA). Fully methylated DNA and Non-methylated DNA (HUM Diagnostics Company, Iran) were used as controls and 25%, 50%, and 75% controls were made by diluting different amounts of these controls. These controls were used in each run of the experiment. The reaction mixture consisted of 10 $\mu$ l of Master Mix (AmpliconRealQ Plus, Denmark), 1 $\mu$ l of bisulfite modified DNA, and 0.5 pmol of each primer in the final volume of 20 $\mu$ l. The reaction was done in StepOnePlus™instrument (Applied Biosystems™, USA). For each run of MS-HRM, five wells were allocated to control samples (0%, 25%, 50%, 75%, and 100%). Methylation status in each sample could be estimated by comparing them with control peaks. All PCR reactions were carried out using the following conditions; Initial hold at 95 ${}^{\circ}$ C for 45 s leading to 45 cycles of 95 ${}^{\circ}$ C for 20 s, 60 ${}^{\circ}$ C for 30 s, and final denaturation step from 60 ${}^{\circ}$ C to 95 ${}^{\circ}$ C with a ramp temperature of 0.1 ${}^{\circ}$ C/s. The HRM curve of each bisulfite DNA was analyzed using HRM ABI software version 3.0.1. The methylation profile for each sample was analyzed by comparing the normalized curve of each sample with the normalized curve of controls [19].

2.5 BRAF V600E detection

To study BRAFV600E mutation, V600E-BRAF Q-PCR kit (HumDiagnostics, Iran) was used. Mutation detection in this kit is based on ARMS scorpion methods. About 50 ng of tumor DNA was amplified in a final volume of 10 $\mu$ l using the following components: 1 $\mu$ l (50 ng) of Primer-Probe Detection system, 5 $\mu$ l of Master Mix (RealQPlus, Ampliqon, Denmark) in the final volume of 10 $\mu$ l. PCR reactions were performed in triplicate besides positive and negative controls provided by the manufacturer. Real-time PCR cycling was performed in the Applied Biosystems™ StepOnePlus™ system. The mixture was initially preheated to 95 ${}^{\circ}$ C for 15 minutes followed by 40 cycles of 95 ${}^{\circ}$ C for 5 seconds, 59 ${}^{\circ}$ C for 10 seconds, 64 ${}^{\circ}$ C for 15 seconds, 72 ${}^{\circ}$ C for 20 seconds leading to a melt curve from 30 ${}^{\circ}$ C to 95 ${}^{\circ}$ C rising by 0.5 ${}^{\circ}$ C per second.

2.6 Statistical analysis of data

Data were evaluated using descriptive statistics. The performance of each marker was analyzed by calculating the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NNP), the area under the ROC curve (AUC), likelihood ratio (LR), and diagnostic odds ratio (DOR) by statistical software GraphPad InStat, version 5.01 (La Jolla, CA, USA) according to the standard definitions. Two-tailed p-value and confidence interval for diagnostic accuracy indexes were calculated by Fisher’s exact test. The correlation between clinicopathological features and markers was explored with the unpaired t-test. P-values less than 0.05 were considered statistically significant. Gold Standard to score MSI-H, was considered instability in $\geqslant$ 30% of markers in the pentaplex panel and 14 markers. The MSI-H samples were considered as MSI+, and the MSI-L and MSI-S ones were considered as MSI- samples.

Figure 2.

Polymorphic markers had larger allelic size changes in MSI-H tumors; A: The average change in tumor allele size of MSI-High samples was determined for the polymorphic markers and pentaplex panel. The average size shift was significantly greater in polymorphic markers, which increased confidence in MSI classification B: The average change in tumor allele size of MSI-High samples was determined for the qusimonomorphic markers and Pentaplex panel. The average size shift was not significantly different from Pentaplex panel.

3. Result

3.1 About 9% of cases were identified to be MSI-H by By Promega Pentaplex method

Since the pentaplex test is accepted for the analysis of MSI in clinics, at the first step tumor samples from 200 patients were analyzed with this panel by comparing the genotypes of tumor samples with their own healthy tissue DNA. It was discovered that 18 cases (9%) were MSI-H while 12 cases (6%) were MSI-L. In this regard, 10 cases (5%) showed polymorphism which exempts a case from the MSI-H category and 9 cases from the MSI-L category when it was matched to their own healthy tissue DNA profiles. Four markers namely NR21 (7 cases), BAT25 (2 cases), BAT26 (one case), and NR24 (one case) showed polymorphism while MONO-27 did not show any polymorphism. All MSI+ specimens were identifiable without any need for comparison with their normal tissues since unstable alleles were outside the QMVR range which is determined by the kit. The adjustment process with their healthy tissue DNA samples was necessary for differentiating MSI-L cases from polymorphisms.

3.2 The MSI detection rate by 14 marker panel was 17.5%

To determine polymorphic and quasimonomorphic markers, DNA of 200 healthy blood samples were analyzed. For each marker, the analysis of 400 alleles were created. Based on the distribution of alleles around the most frequent alleles, 8 polymorphic markers namely BAT-62, BAT-60, BAT-59a, BAT-56a, BAT-56b, FOXP, ACVR, and DNMT3a and 6 quasi monomorphic markers namely DCD, MT1X, RIOX, RNF1, CASP2, and HSP110 were scored. The distribution ranges for each marker are presented in Supplementary Table 2. For quasi monomorphic markers, no allele was placed outside the QMVR range. The lowest (2 bp) and highest (53 bp) allele distributions were associated with RNF and BAT59 markers, respectively.

The profile of each tumor and healthy tissue was assessed and it was shown that 35 cases (17.5%) were MSI-H. If at least 4 markers in patients were unstable ( $\geqslant$ 30%), they were considered as MSI+ though at least 6 markers from 14 markers (42%) showed instability in MSI+ samples. If the number of unstable markers was less than 4 ( $<$ 30%), samples were categorized as MSI-L. in this regard, 25 cases (12.5%) were MSI-L. On the other hand, 140 cases (70%) were identified as MSI-S which showed no unstable marker. The number of allelic shifts for all markers compared with pentaplex markers is shown in Fig. 2. In comparison with pentaplex markers, quasimonomorphic markers had a lower average size of allelic shift while polymorphic markers showed a higher average size significantly. BAT-59 showed the highest allelic shifts (55 bp) whilst RNF showed the lowest allelic shift (6 bp). In quasimonomorphic markers, no polymorphism was found beyond the range of QMVR.

Table 1
Value of diagnostic accuracy of each marker based on 35 MSI+ and 165 MSI- samples (labeled by 14 marker panel)

Marker	Sensitivity (%)	CI (%)	Specificity (%)	CI (%)	AUC (%)	CI (%)	DOR	CI	LR	PV
BAT25	45.71	28.83–63.35	98.79	95.69–99.85	72.25	61.1–83.36	68.63	14.64–321.8	37.71	$<$	0.0001
BAT26	45.71	28.83–63.35	99.39	96.67–99.98	72.55	61.43–83.68	138.1	17.32–1101	75.43
NR21	62.86	44.92–78.53	95.76	91.45–98.28	79.31	69.26–89.36	38.20	13.75–106.1	14.82
NR24	34.29	19.13–52.21	99.39	96.67–99.98	66.84	55.42–78.26	85.57	10.62–689.5	56.57	$<$	0.001777
NR27	48.57	31.38–66.01	100.0	97.79–100	74.29	63.38–85.29	313.1	18.06–5427	$\geqslant$ 80.14	$<$	0.0001
BAT62	100	87.66–100	90.70	85.33–94.59	95.35	92.60–98.10	540.6	31.51–9275	10.75
BAT60	96.00	81.65–99.91	86.00	79.95–90.85	91.24	86.10–96.37	166.5	21.60–1284	6.91
BAT59	60.71	40.58–78.50	96.51	92.56–98.71	78.61	67.25–89.98	42.76	14.05–130.2	17.40
BAT56a	66.67	49.03–81.44	96.51	92.56–98.71	81.59	72.02–91.15	162.0	34.13–769.0	19.11
BAT56b	42.86	24.46–62.82	96.51	92.56–98.71	69.68	57.30–82.07	75.11	19.17–294.2	12.29	$<$	0.0008507
DNMT	97.14	85.08–99.93	97.58	93.91–99.34	97.36	96.89–100.8	1369	1482–12640	40.07	$<$	0.0001
FOXP	60.00	42.11–76.13	93.83	88.94–97.00	76.91	66.64–87.19	18.78	7.597–46.41	9.72
ACVR	60.00	42.11–76.13	100.0	97.78–100.0	80.00	69.71–90.29	456.4	26.33–7909	$\geqslant$ 99
DCD	94.29	80.84–99.30	99.39	96.67–99.98	96.84	92.3–110.4	2706	238.2–30740	155.57
MT1X	97.14	85.08–99.93	98.79	95.69–99.85	97.97	94.62–101.3	2771	244.1–31450	80.14
RNF	51.43	33.91–68.62	100.0	97.79–100.0	75.71	64.85–86.58	349.9	20.19–6065	$\geqslant$ 84.86
RIOX	51.43	33.99–68.62	100.0	97.79–100.0	75.71	64.85–86.58	349.9	20.19–6065	$\geqslant$ 84.86
HSP	54.29	36.65–71.17	100.0	97.79–100.0	77.14	0.65–0.88	391.2	22–6783	$\geqslant$ 66	$<$	0.0002065
CASP	60.00	42.11–76.13	100.0	97.79–100	80.00	69.71–90.29	490.8	28.23–8531	$\geqslant$ 99	$<$	0.0001

AUC: Area under Curve; CI: Confidence Interval; PV: Predictive Value; LR: Likelihood Ratio; SEN: Sensitivity; CI: Confidence Interval; SPE: Specificity; DOR: Diagnostic Odds Ratio.

Figure 3.

The sensitivity and specificity of markers of Pentaplex panel.

3.3 Three groups of markers with different diagnostic values were identified in 14 marker panel

MSI-H cases were considered MSI+ while MSI-L and MSI-S cases were defined as MSI-. Thirty-five cases were scored MSI+ while only 18 cases of them were previously categorized MSI+ by the pentaplex panel. Since all MSI+ samples detected by the pentaplex panel were identified by 14 markers panel, we categorized data into two groups. In this regard, 18 cases were positive by both panels (pentaplex+ 14M+) while 17 cases were just positive by 14 marker panel (pentaplex- 14M+). Six cases that were marked as MSI-L by the pentaplex panel were labeled as MSI-H by 14 marker panel. While by the pentaplex panel just one marker showed instability in these 6 cases ( $<$ 30%), by 14-marker panel at least 7 markers ( $\geqslant$ 50%) showed instability in all MSI+ cases.

Figure 4.

The sensitivity and specificity of 14 markers panel. A. The first group of markers, B. The second group of markers, C. The third group of markers.

Based on 35 MSI+ and 165 MSI- cases (21 MSI-L, 144 MSI-S by 14 marker panel), diagnostic accuracy criteria for all short tandem repeats (STR) were calculated (Table 1). The Sensitivity and specificity of Pentaplex panel were defined as 51.4% (CI95%, 33.9–68.6%) and 99.3% (CI95%, 96.6–99.9%) respectively. BAT25, BAT26, NR21, NR24, and NR-27 showed the sensitivity of 45.71%, 45.71%, 62.86%, 34.29%, and 48.57% respectively while their specificity were 98.79%, 99.39%, 95.76%, 99.39% and 100% respectively. The sensitivity and specificity of the Pentaplex panel are shown in Fig. 3.

All fourteen studied locations can be classified into three groups according to their sensitivity and specificity (Fig. 4A). The first group of markers namely FOXP, HSP110, CASP2, ACVR, RNF, BAT59, BAT56a, BAT56b, and RIOX similar to the pentaplex panel showed medium sensitivity and high specificity. The sensitivity of FOXP, HSP110, CASP2, ACVR, RNF, BAT59, BAT56a, BAT56b, and RIOX were 60%, 40%, 60%, 60%, 51.43%, 60.71%, 66.67%, 42.86%, and 51.43% respectively whereas their specificity were 93.83%, 100%, 100%, 100%, 100%, 96.51%, 96.51%, 96.51%, and 100% respectively (Fig. 4A).

The second group of markers namely BAT62 and BAT60 showed high sensitivity and low specificity (Fig. 4B). The sensitivity of BAT62 and BAT60 was 100% and 96% while their specificity was 90.7% and 86% respectively.

Table 2

Comparison of sensitivity value of each marker between 18 [Pentaplex+] and 17 [Pentaplex-] MSI+ samples

Marker	Pentaplex+	CI (%)	Pentaplex-	CI (%)	$p$ -value
BAT-25	77.78	52.36–93.59	11.76	1.458% to 36.44%	$<$ 0.0001
BAT-26	88.89	65.29–98.62	0	0.0000 to 0.1951
NR-21			19.05	0.05446 to 0.4191
NR-24	72.22	46.52–90.31	0	0.0000 to 0.1951
NR-27	94.44	72.71–99.86
BAT56A	88.80	65.29–98.62	45.70	28.83–63.35
BAT56b		65.0–98.0	47	22.90–72.00	0.0066
BAT59	94.40	72.7–99.8	58.80	32.90–81.50	0.0111
BAT60			76.40	50.10–93.19	0.1366
BAT62		72.71–99.86	64.70	38.33–85.79	0.0280
FOXP			17.60	37.99–43.43	$<$ 0.0001
ACVR	88.80	65.29–98.62	11	14.58–36.44
DNMT	100.0	81.47–100.0	94.0	71.31–99.85	0.9377
CASP	94.40	72.71–99.86	17.60	37.99–43.43	$<$ 0.0001
HSP			11.70	14.5–36.4
RNF	88.80	65.29–98.62	5.8	0.1488–28.69
RIOX
MT1X	100.0	81.47–100.0	94.40	72.71–99.86	0.99377
DCD	94.40	72.71–99.86			0.9680

CI: Confidence Interval.

Table 3

Comparison of features between MSI+ cases [Pentaplex+] and [Pentaplex-] samples

Chemo-radiotherapy before surgery
Variable	Total ( $n=$ 35)	18 Pentaplex+	17 Pentaplex-	$p$ -value
Age	54.68 $\pm$ 13.8	51.39 $\pm$ 14.4	58.18 $\pm$ 12.7	0.1495
Gender
Male (Percentage)	19 (54.3%)	9 (47.4)	10 (52.6%)	0.6131
Female (Percentage)	16 (45.7%)	9 (56.2)	7 (43.8)	0.8762
Tumor stage
Stage II (Percentage)	17 (48.6%)	12 (70.6%)	5 (29.4%)	0.0155*
Stage III (Percentage)	9 (25.7%)	3 (33.3%)	6 (66.7%)	0.1765
Stage IV (Percentage)	9 (25.7%)	3 (33.3%)	6 (66.7%)	0.1765
Tumor location
Ascending colon (Percentage)	15 (42.8%)	9 (60%)	6 (40%)	0.2890
Transverse colon (Percentage)	4 (11.4%)	2 (50%)	2 (50%)	1.0
Descending colon (Percentage)	4 (11.4%)	4 (100%)	0 (0%)	0.06
Sigmoid colon (Percentage)	4 (11.4%)	1 (25%)	3 (75%)	0.2
Rectum (Percentage)	7 (20%)	0 (0%)	7 (100%)	0.0008*
Chemo-radiotherapy No. (Percentage)	8 (22.8%)	2 (25%)	6 (75%)	0.0486*

* Indicates $p$ value equal or less than 0.05.

The third group of markers namely DCD, DNMT3a, and MT1X showed high sensitivity and specificity (Fig. 4C). The sensitivity of DCD, DNMT3a, and MT1X was 94.29%, 97.14%, and 97.14% respectively while their specificity was 99.39%, 97.58% and 98.79%, respectively. Except for BAT56b, all studied regions within these 14 markers showed a diagnostic sensitivity equal to or greater than the Pentaplex panel.

For identifying markers in which sensitivities were not affected by sample clinicopathological features, the sensitivity of all markers were compared between two MSI+ groups. All MSI+ cases identified by pentaplex and 14 marker panels are represented in Table 1. Six markers between the two groups had no significant difference, BAT-60, BAT-59, BAT-62, DNMT3a, DCD, and MT1X. The specificity of these markers was 92.73%, 96.36%, 95.76%, 97.58%, 99.39%, and 98.79% respectively. Area under the curves (AUC) of these marker were 89.22%, 86.75%, 87.8%, 97.36%, 96.84% and 97.97% respectively. The AUC is a pair of diagnostic sensitivity and specificity values helping to estimate how high the discriminative power of a marker is.

Combination of three LMRs like BAT-59, BAT-60, and BAT-62 modified the sensitivity and specificity to 97.14% and 96.36% respectively. Combination of two quasimonomorphic markers namely MT1X and DCD increased the sensitivity and specificity to 100% and 99.39% respectively while the combination of DCD, DNMT3a, and MT1X increased both sensitivity and specificity to 100%. The sensitivity of FOXP, ACVR, CASP2, HSP110, RNF, and RIOX was as low as the pentaplex panel making them unsuitable for detecting MSI. Decreased sensitivity of LMRs in pentaplex-, and MSI+ group mostly was not statistically significant but it was considerable. DCD, DNMT3a, and MT1X had the same sensitivity in both groups indicating that these markers were superior to all studied markers.

As it is shown in Table 2, the sensitivities of the Pentaplex panel markers and BAT-56a, BAT56b, FOXP, ACVR, CASP2, HSP110, RNF1, and MT1X were significantly different between two groups ( $P$ value $<$ 0.0001).

3.4 Fourteen marker panel identified more MSI+ cases in distal colon and/or end stages of tumors

To identify the correlation between studied panels and clinical features, data were arranged into two groups of [pentaplex+ 14markers+] and [pentaplex- 14markers+] which is shown in Table 3 too. The average age of MSI+ cases (identified by the pentaplex panel) was 51.39 $\pm$ 14.4 and this for MSI+ cases who were discovered by 14 marker panel was 54.68 $\pm$ 13.8 while the average age of 17 patients who were just identified by 14 marker panel was 58.18 $\pm$ 12.7. The average ages of patients between the two groups were not different significantly.

While male to female ratio in patients identified by both panels was the same, in the second group (pentaplex-, 14marker panel+) 58.8%, of cases were male and 41.2% were female. Gender distributions were not significantly different between these groups. Percentages of cases in stages II of cancer in the first and the second panel was significantly different ( $P$ value $=$ 0.0155). A considerable number of patients in stage III and IV of colorectal cancer could not be identified by Pentaplex panel while 14 marker panel identified them.

Most tumors that were not identified by the Pentaplex panel are located in the rectum ( $P$ value $=$ 0.0001). Detection of MSI phenotype in subjects who underwent neoadjuvant chemoradiotherapy before surgery was significantly different between two panels ( $P$ value $=$ 0.0486) while they were mostly identified by 14 marker panel.

3.5 Methylation in MLH1 promoter was not a significant finding in MSI positive cases

We examined MLH1 promoter methylation in all 35 MSI+ cases. The age of all samples ranged from 28 to 76 years. The cut-off value of methylation was defined according to several references as 12.5% which means samples showing methylation level less than 12.5% in High-Resolution Melting curve analysis were considered as non-methylated.

After studying the MLH1 promoter region, it was shown that 19 out of 35 MSI+ samples (60%) ( $P$ -value $=$ 0.3979) were hypermethylated in the MLH1 promoter region. Methylated samples were more frequently observed in tumors located in the proximal colon ( $P$ -value $=$ 0.001) while in tumors located in rectum ( $P$ value $=$ 0.126) and distal colon ( $P$ -value $=$ 0.091) methylated samples was not higher significantly. Different clinical features of patients in methylated and non-methylated groups are represented in Supplementary Table 3.

3.6 About two third of MSI positive cases showed BRAFV600E mutation

Using BRAF mutation detection kit it was shown that 25 out of 35 (68.57%) cases were positive for BRAFV600E ( $P$ -value $=$ 0.0002) significantly. In tumors located in the proximal and distal colon, the rate of mutated BRAFV600Ev was significantly higher than non-mutated ones ( $P$ -value $=$ 0.007) and ( $P$ -value $=$ 0.048) respectively. Additionally, the rate of this mutation in stage III and stage IV of cancer was higher than tumors in stage II. Pretreated samples showed a higher rate of this mutation in comparison with non-treated ones ( $P$ -value $=$ 0.007). The distribution of BRAFV600E mutation in MLH1 methylated samples was not significantly different ( $P$ -value $=$ 0.36). Demographic data of MSI+ cases according to the status of BRAFV600E mutation is given in Supplementary Table 4.

4. Discussion

One of the limitations of PCR based MSI analyzing is low analytical sensitivity due to contamination of tumor DNA with DNA from non-tumor cells [20] while intra-tumoral heterogeneity increased in cancer progression leading to false-negative results in the end stages cancer [20].

While the MSI test is a part of the Lynch syndrome diagnosis algorithm and prognosis predictor, it is a predictive biomarker for the efficiency of chemotherapy in stage II and efficacy of immune checkpoint inhibitors in mCRC [6]. MSI status is among the predictors of efficacy of immune blockers indicating the response to these drugs in colorectal cancer and other types of solid tumors. MSI test could generally be the indicator of genome instability, while IHC tests can only show the deficiency in the expression of MMR proteins [21]. Therefore using MSI biomarkers like what we have suggested in this study could be unique since the same biomarker has been applied to guide the immune therapy regardless of the location and the stage of the tumor.

The current clinical in-use panel, Pentaplex panel, in concordance with the IHC test detects instability in samples that have defective MMR system. MSI+ tumors detected by this panel are mostly located in the proximal colon and are typically in stage II [3]. According to the spectrum of different factors influencing sequence maintenance of the genome [7], MSI is an early event in tumor development and it is rational that the MSI+ trait is not confined to distinct stages and location of the tumor. Hence the higher rate of MSI+ cases in the proximal colon and stage II of cancer means that some of the MSI+ cases are overlooked by this panel [22].

This study aimed to investigate whether it may be possible to improve the detection rate of MSI in colorectal cancer patients. In this regard, we tested 14 markers obtained from different previous studies which had been reported separately. These markers showed higher sensitivity in comparison with the Pentaplex panel while detecting 35 tumor samples as MSI-H (17.5% of patients) compared to 18 tumor samples which have been detected by the Pentaplex panel (9% of patients). Therefore, a 1.94 fold increase in the detection rate of the MSI-H trait was achieved by using this newly introduced panel.

This new panel identified all 18 MSI+ cases detected by the Pentaplex panel. Comparing clinical features of 17 MSI+ cases which could not be detected by the Pentaplex panel and 18 MSI+ cases which were identified by this panel, showed that the Pentaplex panel had not enough sensitivity for identifying cases that were located in the distal colon and rectum and cases which were in higher stages of cancer (stage III and IV). Also, tumors undergoing the previous chemo-radiotherapy will be overlooked by the Pentaplex panel. Similar to the previous studies, MSI+ tumors which were scored by Pentaplex panel were mostly in stage II and located in the proximal colon [23]. In contrary to the previous findings, this study showed that the cause of decreased detection rate of MSI phenotype in stage III, and IV and descending colon and rectum is the limitation of Pentaplex markers not a lower rate of MSI in the distal colon and higher stages of cancer.

Probably pentaplex markers are not primary targets of the MSI system since markers mutated in primary clones of tumor cells are detectable in subclones of cancer in all stages. Through cancer evolution and increasing tumor heterogeneity, the diagnostic sensitivity of some marker which are not primary targets of mutation decreases [24].

Methylation of MLH1 promoter is a marker for differentiating Lynch syndrome from sporadic MSI-H samples. In the previous studies, it was shown that methylated MLH1 promoter is significantly observed in MSI-H tumors especially the ones which are located in the proximal colon [25]. Since 15 samples of MSI+ cases (42.8%) in our study were located in the distal colon and rectum and 25 MSI+ cases (57.2%) located in the proximal colon, difference between methylated and non-methylated MSI+ cases was not significant in different parts of colons except for the proximal colon. This shows that disabling the MMR system by MLH1 promoter methylation is the main mechanism of the MSI+ phenotype in the proximal colon tumors.

In contrary to methylation, positive and negative cases of BRAFV600E mutation were differently distributed in MSI+ cases which are following the previous studies [25]. Most MSI+ cases which were not detected by the Pentaplex marker panel were suffering from stage III, and IV and BRAFV600E mutation was detected in these cases significantly which is concordant with an increased rate of BRAF mutations in end stages of cancer [26]. Previous studies showed that BRAF mutation was mostly seen in proximal colon tumors [27] while in our study MSI+ cases in the distal colon showed this mutation too.

Treatment modalities in MSI positive and negative cases are different. Through the emergence of the monoclonal antibodies for blocking immune checkpoints in MSI patients, the importance of analyzing MSI status for CRC and other types of cancer has been highlighted too. Misdiagnosis of MSI status can have adverse effects on the patients. The false-positive results, mostly because of polymorphisms in pentaplex markers’ regions, would be toxic and costly for the patients without any clinical effect. False-negative results, mostly because of limitations of the pentaplex panel for diagnosing the MSI status in some stages and locations of tumor, and the previous treatment of the patients would deprive the patients of the benefits of new immune blocker drugs [28]. While one-third of colorectal tumors are located in the distal colon and rectum and about half of them are in stage III and IV [29], finding a biomarker panel that diagnoses MSI-H in these types of CRC would be beneficial for the prescription of immune blockers for these patients.

Among investigated markers, five markers namely BAT-62, BAT-60, DNMT3a, MT1X, and DCD had ideal proposed conditions of a marker suggested by NCI (1998). These criteria consist of sensitivity and specificity of more than 80%. On the other hand, a combination of three markers namely DNMT3a, DCD, and MT1X had similar sensitivity to all 14 markers.

From 200 samples which were analyzed by Pentaplex panel, non-informative markers were observed in 10 samples (5%) which lead to misclassification of one MSI-H and nine MSI-L. The rate of polymorphism in these markers makes comparing tumor and healthy profile necessary for classification of samples especially in the cases which showed two unstable markers increasing the chance of misclassification.

Qusimonomorphic markers namely DCD, MT1X, HSP110, CASP2, RIOX, and RNF had no variant out of QMVR rang in our studied population and could be analyzed without matched normal samples.

All 35 MSI-H cases scored by 14 markers could be categorized into two groups, the ones who were detected by Pentaplex either and the cases not detectable by Pentaplex panel (17 cases). Comparing the sensitivities of markers in two groups of MSI-H cases indicated three categories of markers; the first category consisted of FOXP, ACVR, CASP2, HSP110, RNF, and RIOX which showed a high sensitivity for identifying MSI-H samples which were also pinpointed by Pentaplex panel and low sensitivity for the second group who are 17 cases not detected by Pentaplex and just detected by 14 markers. Concordance between these markers and Pentaplex panel shows that this set of markers is suitable for analyzing MSI in samples which are in stage II and located in the proximal colon. These types of colon cancer are mostly due to defective MMR mechanism induced by MLH1 promoter methylation or mutation in genes of the MMR system.

CASP2 and HSP110 were described as ideal markers in the previous studies [11, 20] because of showing 100% instability in MSI+ samples, no mutation in MSI- samples, and quasimonomorphic allele patterns in all populations. These markers were previously compared with Promega/NCI panel but in this study, they were compared with 14 marker panel and showed non-ideal characteristics. The sensitivity of CASP2 and HSP110 were 60% and 54.29% respectively while their AUCs were 0.8, and 0.77, and diagnostic odds ratio (DOR) were 490, and 391 respectively. Excellent markers should have AUC more than 0.9 whereas in good markers, this parameter should be higher than 0.8 which cannot be fulfilled by the above-mentioned markers [30].

The second category of markers consisted of LMRs, BAT-62, BAT-60, BAT-59a, BAT-56a, and BAT-56b. The sensitivity of these markers reduced to about half for detecting the MSI phenotype in the second group as 17 MSI-H cases who were not diagnosed by the Pentaplex panel. These markers showed a high sensitivity for detecting 18 MSI-H diagnosed by Pentaplex panel [10]. Comparing sensitivity between these two groups indicated LMRs are not specifically targeted locations for analyzing MSI. Long repeats of mononucleotides increase the chance of trapping mutations in MSI-H cases but are passenger mutations, not driver ones.

Three markers consisted of DCD, DNMT3a, and MTIX showed the same sensitivity in both groups. Additionally, their sensitivity is not affected by sample features (stage, location, grade, treatment, etc.). The detection ability of these markers is not limited to specific locations in the colon therefore they are probably common targets of instability in different locations of the colon and rectum. Their ability to be a marker of instability is not confined to distinct stages of cancer. These markers may be the primary targets of instability in any stage of tumors and are detectable in all stages of cancer.

Dermcidin or DCD is a natural antibiotic in the body which is effective in the growth and survival of the cells, especially during oxidative stress conditions [31]. In one study on breast cancer samples, due to association between increased level of this gene and invasiveness, this gene was regarded as poor survival marker [32].

Another gene that is sensitive to oxidative stress is MTIX. This gene encoding metallothionein protein is responsible for protecting the cells from DNA damage, oxidation stress, and apoptosis. Expression of this protein leads to resistance to platinum-based drugs and alkalizing agents [15]. Multistep progression of cancer suggests that oxidative stress is responsible for cancer initiation, promotion, and progression [33]. Disruption of genes which are against oxidative stress is frequently seen in cancer since their expression is induced for encountering reactive oxygen species (ROS). STR repeats in open reading regions of genes are prone to insertion and deletion mutations by the MSI system [7]. Reduced expression of these genes resulted in increased apoptosis, decreased proliferation, and a favorable prognosis, which are concordant with the characteristics of MSI+ tumors.

DNMT3a is responsible for the De novo methylation pattern in different cell lines. In this process, it has interaction with DNMT1 and DNMT3b. It is reasonable that a cell line that is moving toward cancer will dedifferentiate by reduced activity of DNMT3a. Dysfunctional DNMT3a leads to abnormal methylation patterns like global hypomethylation besides promoter hypermethylation of tumor suppressor genes [16] which is usually seen in MSI+ colon cancers [19, 34].

Scoring samples in this study was based on the proposed criteria of NCI. To prevent any bias in our judgment, we excluded IHC method as the gold standard since IHC findings are limited to the stage and location of tumor. For a conclusive evaluation, the only method which can be used as the gold standard is Next Generation Sequencing (NGS).

The disadvantage of polymorphic markers for analyzing the instability is matching the fragmentation profiles of both tumoral and normal tissues. DNMT3a in contrary to DCD and MT1X is a polymorphic marker. DNA quality is an important pre-analytical factor in MSI tests. Fragmented DNA which is observed in tumor tissues is also a serious challenge to obtain PCR products without any smear and sharp electrograph.

5. Conclusion

In any stage of cancer we can pinpoint the clonal cancer cells though mutations which are related to certain features of tumors happen mostly in subclones of cancer cells. Analyzing MSI with markers which are related to certain features resulted in missing some MSI+ cases. In this regard, DNMT3a, DCD, and MT1X were shown to have no correlation with any clinical or pathological features and probably are mutated in first clones of MSI+ cells [35]. Our findings suggest that these markers can be used for MSI analysis aiming to define MSI+ samples for precision medicine.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-203226.

sj-docx-1-cbm-10.3233_CBM-203226.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-203226.docx

Funding

This study was funded as Ph.D. student project by Pasteur Institute of Iran (96/0110/2265) and grant number 822 by Pasteur Institute of Iran.

Ethical issues

This study had been approved by Ethics Committee of Pasteur Institute of Iran (ethical code: IR.PII.REC.1398.041). All patients had informed about the research and signed the written consent before operation.

Competing interest

The authors declare that they have no competing interests.

Data Availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author contribution

Conception: Ali Khaligh, Ladan Teimoori-Toolabi, Mohammad Sadegh Fazeli, Habibiollah Mahmoodza-deh

Interpretation or analysis of data: Ali Khaligh, Ladan Teimoori-Toolabi, Amirhosein Mehrta-sh, Setareh Kompanian, Sirous Zeinali

Preparation of the manuscript: Ali Khaligh

Revision for important intellectual content: Ladan Teimoori-Toolabi

Supervision: Ladan Teimoori-Toolabi, Sirous Zeinali

Footnotes

Acknowledgments

We hereby thank the staff of operation room in Imam Khomeini Hospital and Cancer Institute Hospital. Also we hereby wanted to thank Prenatal Diagnosis (PND) laboratory in Molecular Medicine Department of Pasteur Institute of Iran.

References

Müller

M.F.

Ibrahim

A.E.

and Arends

M.J.

, Molecular pathological classification of colorectal cancer, Virchows Archiv 469 (2016), 125–134.

Benson

A.B.

Venook

A.P.

Al-Hawary

M.M.

Cederquist

Chen

Y.-J.

Ciombor

K.K.

Cohen

Cooper

H.S.

Deming

and Engstrom

P.F.

, NCCN guidelines insights: colon cancer, version 2.2018, Journal of the National Comprehensive Cancer Network 16 (2018), 359–369.

Boland

C.R.

Thibodeau

S.N.

Hamilton

S.R.

Sidransky

Eshleman

J.R.

Burt

R.W.

Meltzer

S.J.

Rodriguez-Bigas

M.A.

Fodde

and Ranzani

G.N.

, A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer, Cancer Research 58 (1998), 5248–5257.

Gelsomino

Barbolini

Spallanzani

Pugliese

and Cascinu

, The evolving role of microsatellite instability in colorectal cancer: A review, Cancer Treatment Reviews 51 (2016), 19–26.

Ryan

Sheahan

Creavin

Mohan

and Winter

, The current value of determining the mismatch repair status of colorectal cancer: a rationale for routine testing, Critical Reviews in Oncology/Hematology 116 (2017), 38–57.

Zhao

Jiang

and Li

, Mismatch repair deficiency/microsatellite instability-high as a predictor for anti-PD-1/PD-L1 immunotherapy efficacy, Journal of Hematology and Oncology 12 (2019), 54.

Kim

T.-M.

Laird

P.W.

and Park

P.J.

, The landscape of microsatellite instability in colorectal and endometrial cancer genomes, Cell 155 (2013), 858–868.

Romaniuk

Lyndin

Smiyanov

Sikora

Rieznik

Kuzenko

Budko

Moskalenko

Karpenko

and Gladchenko

, Primary multiple tumor with affection of the thyroid gland, uterus, urinary bladder, mammary gland and other organs, Pathology-Research and Practice 213 (2017), 574–579.

Bacher

J.W.

Flanagan

L.A.

Smalley

R.L.

Nassif

N.A.

Burgart

L.J.

Halberg

R.B.

Megid

W.M.A.

and Thibodeau

S.N.

, Development of a fluorescent multiplex assay for detection of MSI-High tumors, Disease Markers 20 (2004), 237–250.

10.

Bacher

J.W.

Sievers

C.K.

Albrecht

D.M.

Grimes

I.C.

Weiss

J.M.

Matkowskyj

K.A.

Agni

R.M.

Vyazunova

Clipson

and Storts

D.R.

, Improved detection of microsatellite instability in early colorectal lesions, PLoS One 10 (2015), e0132727.

11.

Findeisen

Kloor

Merx

Sutter

Woerner

S.M.

Dostmann

Benner

Dondog

Pawlita

and Dippold

, T25 repeat in the 3’ untranslated region of the CASP2 gene: a sensitive and specific marker for microsatellite instability in colorectal cancer, Cancer Research 65 (2005), 8072–8078.

12.

Megid

W.A.

Ensenberger

M.G.

Halberg

R.B.

Stanhope

S.A.

Kent-First

M.G.

Prolla

T.A.

and Bacher

J.W.

, A novel method for biodosimetry, Radiation and Environmental Biophysics 46 (2007), 147–154.

13.

Cortes-Ciriano

Lee

Park

W.-Y.

Kim

T.-M.

and Park

P.J.

, A molecular portrait of microsatellite instability across multiple cancers, Nature Communications 8 (2017), 15180.

14.

Markovic

Antic

Dimitrijevic

Zogovic

Bojic

Svorcan

Markovic

and Krivokapic

, Microsatellite instability affecting the T17 repeats in intron 8 of HSP110, as well as five mononucleotide repeats in patients with colorectal carcinoma, Biomarkers in Medicine 7 (2013), 613–621.

15.

Morandi

de Biase

Visani

Monzoni

Tosi

Brulatti

Turchetti

Baccarini

Tallini

and Pession

, T [20] repeat in the 3’-untranslated region of the MT1X gene: a marker with high sensitivity and specificity to detect microsatellite instability in colorectal cancer, International Journal of Colorectal Disease 27 (2012), 647–656.

16.

Ren

Gao

and Song

, Structural basis of DNMT1 and DNMT3A-mediated DNA methylation, Genes 9 (2018), 620.

17.

Teimoori-Toolabi

Vahedi

Mollahajian

Kamali

Hajizadeh-Sikaroodi

Zeinali

Tabrizian

Olfati

Rashtak

and Malekzadeh

, Three common CARD15 mutations are not responsible for the pathogenesis of Crohn’s disease in Iranians, Hepato-Gastroenterology 57 (2010), 275–282.

18.

Sutter

Gebert

Bischoff

Herfarth

and von Knebel Doeberitz, Molecular screening of potential HNPCC patients using a multiplex microsatellite PCR system, Molecular and Cellular Probes 13 (1999), 157–165.

19.

Afgar

Fard-Esfahani

Mehrtash

Azadmanesh

Khodarahmi

Ghadir

and Teimoori-Toolabi

, MiR-339 and especially miR-766 reactivate the expression of tumor suppressor genes in colorectal cancer cell lines through DNA methyltransferase 3B gene inhibition, Cancer Biology and Therapy 17 (2016), 1126–1138.

20.

Buhard

Lagrange

Guilloux

Colas

Chouchène

Wanherdrick

Coulet

Guillerm

Dorard

and Marisa

, HSP110 T17 simplifies and improves the microsatellite instability testing in patients with colorectal cancer, Journal of Medical Genetics 53 (2016), 377–384.

21.

Duffy

M.J.

and Crown

, Biomarkers for predicting response to immunotherapy with immune checkpoint inhibitors in cancer patients, Clinical Chemistry 65 (2019), 1228–1238.

22.

Kim

T.-M.

and Park

P.J.

, A genome-wide view of microsatellite instability: old stories of cancer mutations revisited with new sequencing technologies, Cancer Research 74 (2014), 6377–6382.

23.

Boland

C.R.

and Goel

, Microsatellite instability in colorectal cancer, Gastroenterology 138 (2010), 2073–2087. e3.

24.

Krøigård

A.B.

Larsen

M.J.

Lænkholm

A.-V.

Knoop

A.S.

Jensen

J.D.

Bak

Mollenhauer

Kruse

T.A.

and Thomassen

, Clonal expansion and linear genome evolution through breast cancer progression from pre-invasive stages to asynchronous metastasis, Oncotarget 6 (2015), 5634.

25.

Yao

Wang

Jiang

Liu

Wang

Sun

and Zhao

, MLH1 promoter methylation frequency in colorectal cancer patients and related clinicopathological and molecular features, PloS One 8 (2013), e59064.

26.

Bläker

Alwers

Arnold

Herpel

Tagscherer

K.E.

Roth

Jansen

Walter

Kloor

and Chang-Claude

, The association between mutations in BRAF and colorectal cancer–specific survival depends on microsatellite status and tumor stage, Clinical Gastroenterology and Hepatology 17 (2019), 455–462. e6.

27.

and Li

, BRAF mutation is associated with poor clinicopathological outcomes in colorectal cancer: a meta-analysis, Saudi Journal of Gastroenterology: Official Journal of the Saudi Gastroenterology Association 23 (2017), 144.

28.

Cohen

Hain

Buhard

Guilloux

Bardier

Kaci

Bertheau

Renaud

Bibeau

and Fléjou

J.-F.

, Association of primary resistance to immune checkpoint inhibitors in metastatic colorectal cancer with misdiagnosis of microsatellite instability or mismatch repair deficiency status, JAMA Oncology 5 (2019), 551–555.

29.

Gomez

Dalal

Raw

Roberts

and Lyndon

, Anatomical distribution of colorectal cancer over a 10 year period in a district general hospital: is there a true “rightward shift”, Postgraduate Medical Journal 80 (2004), 667–669.

30.

Šimundić

A.-M.

, Measures of diagnostic accuracy: basic definitions, Ejifcc 19 (2009), 203.

31.

Esposito

Schiattarella

G.G.

Perrino

Cattaneo

Pironti

Franzone

Gargiulo

Magliulo

Serino

and Carotenuto

, Dermcidin: a skeletal muscle myokine modulating cardiomyocyte survival and infarct size after coronary artery ligation, Cardiovascular Research 107 (2015), 431–441.

32.

Bancovik

Moreira

D.F.

Carrasco

Yao

Porter

Moura

Camargo

Fontes-Oliveira

C.C.

Malpartida

M.G.

and Carambula

, Dermcidin exerts its oncogenic effects in breast cancer via modulation of ERBB signaling, BMC Cancer 15 (2015), 70.

33.

Calaf

G.M.

Urzua

Termini

and Aguayo

, Oxidative stress in female cancers, Oncotarget 9 (2018), 23824.

34.

and Bekaii-Saab

, CpG island methylation, microsatellite instability, and BRAF mutations and their clinical application in the treatment of colon cancer, Chemotherapy Research and Practice 2012 (2012).

35.

Goranova

T.E.

Ohue

Shimoharu

and Kato

, Dynamics of cancer cell subpopulations in primary and metastatic colorectal tumors, Clinical and Experimental Metastasis 28 (2011), 427–435.

Improved microsatellite instability detection in colorectal cancer patients by a combination of fourteen markers especially DNMT3a,DCD,and MT1X

Abstract

BACKGROUND:

OBJECTIVES:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Sample preparation and DNA extraction

2.3 Microsatellite instability analysis by 14 mononucleotide repeats

2.4 MLH1 promoter methylation analysis

2.5 BRAF V600E detection

2.6 Statistical analysis of data

3.1 About 9% of cases were identified to be MSI-H by By Promega Pentaplex method

3.2 The MSI detection rate by 14 marker panel was 17.5%

Table 1 Value of diagnostic accuracy of each marker based on 35 MSI+ and 165 MSI- samples (labeled by 14 marker panel)

3.5 Methylation in MLH1 promoter was not a significant finding in MSI positive cases

3.6 About two third of MSI positive cases showed BRAFV600E mutation

4. Discussion

5. Conclusion

Supplementary data

sj-docx-1-cbm-10.3233_CBM-203226.docx - Supplemental material

Funding

Ethical issues

Competing interest

Data Availability statement

Author contribution

Footnotes

Acknowledgments

References

Table 1
Value of diagnostic accuracy of each marker based on 35 MSI+ and 165 MSI- samples (labeled by 14 marker panel)