Clinical Performance of the Idylla MSI Test for a Rapid Assessment of the DNA Microsatellite Status in Human Colorectal Cancer

Clinical Performance of the Idylla MSI Test for a Rapid Assessment of the DNA Microsatellite Status in Human Colorectal Cancer

Journal Pre-proof Clinical performance of the Idylla MSI Test for a rapid assessment of the DNA microsatellite status in human colorectal cancer Karen...

558KB Sizes 0 Downloads 17 Views

Journal Pre-proof Clinical performance of the Idylla MSI Test for a rapid assessment of the DNA microsatellite status in human colorectal cancer Karen Zwaenepoel, Julie Holmgaard Duelund, Koen De Winne, Vincent Maes, Christine Weyn, Suzan Lambin, Robina Dendooven, Glenn Broeckx, Torben Steiniche, Patrick Pauwels PII:

S1525-1578(19)30458-1

DOI:

https://doi.org/10.1016/j.jmoldx.2019.12.002

Reference:

JMDI 868

To appear in:

The Journal of Molecular Diagnostics

Received Date: 5 April 2019 Revised Date:

6 November 2019

Accepted Date: 5 December 2019

Please cite this article as: Zwaenepoel K, Duelund JH, De Winne K, Maes V, Weyn C, Lambin S, Dendooven R, Broeckx G, Steiniche T, Pauwels P, Clinical performance of the Idylla MSI Test for a rapid assessment of the DNA microsatellite status in human colorectal cancer, The Journal of Molecular Diagnostics (2020), doi: https://doi.org/10.1016/j.jmoldx.2019.12.002. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Copyright © 2019 Published by Elsevier Inc. on behalf of the American Society for Investigative Pathology and the Association for Molecular Pathology.

Clinical performance of the Idylla MSI Test for a rapid assessment of the DNA microsatellite status in human colorectal cancer

Karen Zwaenepoel,*† Julie Holmgaard Duelund,‡ Koen De Winne,* Vincent Maes,* Christine Weyn,*† Suzan Lambin,* Robina Dendooven,* Glenn Broeckx,* Torben Steiniche‡ and Patrick Pauwels*† From the Laboratory of Pathological Anatomy,* Antwerp University Hospital (UZA), Edegem, Belgium; the Center for Oncological Research Antwerp (CORE),† University of Antwerp, Wilrijk, Belgium; and the Aarhus University Hospital,‡ Aarhus, Denmark

Running head: Clinical performance of the Idylla MSI Test in colorectal cancer. Disclosures: P.P. has received speaker fees from Biocartis and has acted as consultant for Biocartis. The Idylla platforms and Idylla MSI Test Investigational Use Only (IUO) cartridges were provided by Biocartis. Corresponding author: Karen Zwaenepoel, Laboratory of Pathological Anatomy, Antwerp University Hospital, Wilrijkstraat 10, 2650 Edegem, Belgium Email: [email protected]

1 / 32

Abstract In this study we evaluated the clinical performance of the Idylla MSI Test (Investigational Use Only) in 330 colorectal carcinoma samples (all stages). This test is fully automated from formalin-fixed, paraffin-embedded slide to result, and gives a result in less than 2.5 hours. Compared to the Promega MSI Analysis System, Version 1.2, an overall agreement, sensitivity, and specificity of 99.7%, 98.7%, and 100% was reached, respectively. Whereas seven samples were invalid with the Promega MSI Analysis, only two were invalid with the Idylla MSI Test. Compared to the historical immunohistochemistry (IHC) data, an overall agreement, sensitivity, and specificity of 98.7%, 94.4%, and 100% was observed, respectively. Tumor mutation burden analysis of the discordant IHC cases was in favor of the Idylla MSI Test result in three of the four samples. Furthermore, for those cases where the IHC data were invalid or hard to interpret as sole loss of one DNA mismatch repair deficiency marker was observed, Idylla MSI Test results were always valid and accurate. Overall, we have demonstrated that the Idylla MSI Test is an accurate, fast screening assay for the detection of microsatellite status in colorectal cancer patients, with a low amount of invalid results.

2 / 32

Introduction Colorectal cancer (CRC) is the third most common cancer in men and the second in women. Although the majority of CRC cases are sporadic in nature, 5% to 10% of cases are due to inherited autosomal dominant mutations. The most common subtype of hereditary CRC is Lynch syndrome (Hereditary Non-Polyposis Colon Cancer, HNPCC) which accounts for approximately 3% to 5% of all CRC diagnoses.1,2 Consequently, it is imperative that patients with colorectal cancer are screened for Lynch syndrome. Approximately 15% of the CRC samples exhibit loss of DNA mismatch repair (MMR) function and in about one third of this population the tumor development is associated with Lynch syndrome.3 Consequently, loss of MMR function is used as an initial screening marker to identify patients with Lynch syndrome. Investigating the MMR status in colorectal cancer also has prognostic value. Patients with MMR-deficient tumors generally have a better survival. Furthermore, MMR-deficient tumors have been associated with an apparent resistance to treatment with 5-Fluorouracil therapy.4,5 Loss of MMR function leads to microsatellite instability-high (MSI-H) status and can be detected by evaluating the tumor tissue expression of the MMR proteins (MLH1, PMS2, MSH2, and MSH6).6 Alternatively, MSI-H status can be evaluated on tumor DNA.7,8,9 In general, this analysis is performed with at least five microsatellite markers, often mononucleotide or dinucleotide repeat markers. Since dinucleotide repeats have a lower sensitivity and specificity compared to mononucleotide repeats for identifying tumors with an MSI-H phenotype, the revised Bethesda Guidelines of the National Cancer Institute regarding MSI testing and Lynch Syndrome advices the use of more than two mononucleotide markers in the evaluation of MSI.10,11,12,13 Based on available data on the allele frequency of various markers, the MSI Analysis System of Promega investigates five mononucleotide markers that are quasi-monomorphic.

3 / 32

Loss of MMR function leads to DNA replication errors in the tumor tissue. This buildup of somatic mutations results in a high tumor mutation burden (TMB) and an increased expression of neoantigens.

14,15,16

Recently, the efficacy of checkpoint inhibitor drugs, such as anti-PD1/PD-L1

antibodies, has been linked to the presence of tumor neo-antigens. Consequently, FDA has granted an accelerated approval to pembrolizumab, an anti-PD1 antibody, for patients with MMR deficient tumors as determined by MSI or MMR proteins expression testing. 17,18 In this study we assessed the performance of the Idylla MSI Test Investigational Use Only (IUO). This Test is fully automated from formalin-fixed, paraffin-embedded (FFPE) tissue section to result and can generate a result within 2.5 hours. Compared to the Promega MSI Analysis System, Version 1.2, an overall concordance, sensitivity, and specificity of 99.7%, 98.7%, and 100% were found, respectively. In addition, the Idylla MSI Test was compared to immunohistochemical (IHC) testing, demonstrating loss of expression of the MMR proteins.

An overall concordance, sensitivity, and

specificity of 98.7%, 94.4%, and 100% was demonstrated, respectively. Samples with discordance between MMR protein expression and Idylla MSI Test were further analyzed by TMB. TMB results were in agreement with the Idylla MSI Test result in the majority of the cases. Overall the data suggest that the Idylla MSI Test is a good screening Test to accurately and rapidly identify the presence or absence of MSI-H in CRC patients.

Materials and Methods Tissue specimens and study design This study was approved by the Ethical Committee of the University Hospital Antwerp and includes archival clinical FFPE tumor samples from 330 CRC patients referred for MLH1, PMS2, MSH2, and MSH6 immunohistochemistry (IHC) between 2009 and 2018. To include at least 75 MSI-H samples in the study, a partial enrichment of MSI-H samples was made during sample selection.

4 / 32

The study was conducted in two clinical centers: The Aarhus University Hospital (Denmark) and the University Hospital Antwerp (UZA; Belgium). Each center performed analysis of microsatellite status by Idylla MSI Test on 150 to 180 stage I-IV CRC samples. Analysis of microsatellite status with the Promega MSI Analysis System, Version 1.2, was performed at UZA for all samples. From each sample 10 consecutive FFPE tissue sections of 5 µm or 10 µm thickness were prepared. The first and last sections were stained with H&E. Based on histological assessment of H&E staining, consecutive slides of the samples were enriched by manual macrodissection to reach a tumor content of at least 20%. One to five FFPE slides were subsequently tested with the Idylla MSI Test and three FFPE slides were used for the Promega MSI Analysis System, Version 1.2.

Idylla MSI Test The Idylla MSI Test (Biocartis, Mechelen, Belgium) is an in vitro diagnostic test intended for the qualitative detection of a novel panel of seven monomorphic biomarkers (ACVR2A, BTBD7, DIDO1, MRE11, RYR3, SEC31A, SULF2) for identification of microsatellite status in CRC12. The Idylla MSI Test uses FFPE tissue sections from human CRC tissue, from which nucleic acids are liberated by a combination of chemical reagents, enzymes, heat, and high intensity focused ultrasound. Then the seven targets are amplified with biomarker-specific primers. The Specific targets were detected using fluorescent-labeled molecular beacons. These beacons differentially dissociate from the wildtype or mutated amplicons with increasing temperature. The MSI-specific software will first check the validity of the measured fluorescent profiles. Next, a powerful pattern-recognition approach will calculate a similarity score (MSI score) for any given valid biomarker-specific profile. This MSI Score expressed the probability of a pattern being wild-type or mutant. Each biomarker is scored individually and reported as ‘No mutation detected’, ‘Mutation detected’, or ‘Invalid’. The MSI status of a sample can be determined with high confidence if at least five marker-specific fluorescence profiles could be analyzed (otherwise the MSI status of the sample is called ‘Invalid’). 5 / 32

The presence of at least two mutant markers will result in a status being MSI-H, otherwise the status is scored MSS (Figure 1). Microsatellite Instability Low (MSI-L) is not reported by this assay. The Idylla MSI Test automates the entire process from FFPE sample preparation to reporting of microsatellite status. To perform the Idylla MSI Test, a minimal total tissue area of 50mm2 and a maximum total tissue area of 600mm2 was used with a slide thickness of 5 µm and a minimum of 25mm2 and a maximum of 300mm2 was used with a slide thickness of 10 µm. To reach this amount of tissue one to five FFPE slides were used per sample. When an invalid test result was obtained, the test was repeated once with the same tissue input. The Test was performed as per Instructions For Use (IUO) of the manufacturer (Biocartis, Mechelen, Belgium). In this study, the Idylla MSI Investigational Use Only (IUO) Assay was used, as the CE-IVD–labeled Idylla MSI Test was not yet on the market.

Promega MSI analysis The Promega MSI Analysis System, Version 1.2 (Promega Corporating, Madison, US) (Promega MSI Analysis) contains five mononucleotide markers (BAT-25, BAT-26, NR-21, NR-24, MONO-27) and two pentanucleotide markers (PentaC, Penta D). The assay includes fluorescently-labeled primers for PCR amplification of FFPE-extracted DNA followed by capillary electrophoreses. In this study per case three FFPE tissue sections of 5 to 10 µm were used. If needed to reach a tumor content of at least 20%, manual macrodissection was performed prior to gDNA isolation. gDNA isolation was performed on a QIAcube instrument using the QIAamp DNA Mini QIAcube Kit according to the instructions of the manufacturer (Qiagen, Venlo, Netherlands). gDNA concentration was measured with a NanoDrop 2000 instrument. If the concentration exceeded 120 ng/µL the sample was diluted 1 over 6 in AE buffer (Qiagen). Promega MSI Analysis was performed according to the instructions of the manufacturer with 2 µL gDNA as input material. This method has been validated at UZA for routine use according to ISO 15189. To cope with the complexity of the method positive and negative 6 / 32

control samples instead of healthy tissue gDNA were added to evaluate the technical performance of the assay. For the PCR amplification a Veriti 96-Well Thermal Cycler (Thermo Fisher Scientific, Waltham, MA) was used and for the capillary electrophoresis an ABI 3130 Genetic Analyzer (Thermo Fisher Scientific, Waltham, MAS) was used. Data were analyzed with the GeneMapper Software Analyzer (Thermo Fisher Scientific, Waltham, MA) (Figure 1). Samples were classified MSI-H when ≥2 mononucleotide marker out of five (BAT-25, BAT-26, NR-21, NR-24, MONO-27) were altered. In case not all five markers could be evaluated and less than four markers were evaluated as stable, the test was considered invalid and one repeat test was performed. To successfully evaluate a marker, a minimal peak amplitude threshold of 100RFU was applied.

Immunohistochemistry for Mismatch Repair protein expression: MLH1, PMS2, MSH2, and MSH6 At UZA, to evaluate the expression of four MMR proteins, IHC analysis was performed on the Omnis instrument (Agilent, Santa Clara, CA) for samples obtained between 2016 to 2018 and on a PT Link and Autostainer Link 48 instrument (Agilent, Santa Clara, CA) for samples obtained prior to 2016. Following primary antibodies were used for this analysis: MLH1 clone ES05 (IR0079, Agilent), PMS2 clone EP51 (IR087, Agilent), MSH2 clone FE11 (IR085, Agilent), and MSH6 clone EP49 (IR086, Agilent). All IHC assays were performed using the target retrieval solution high pH for 30 min at 95 °C and EnVision FLEX detection DAB system of Agilent. The MSH2 antibody (ready to use solution) was incubated for 30 min on the Autostainer Link 48 and for 20 min in combination with a mouse linker on the Omnis, the MSH6 antibody (ready to use solution) was incubated for 20 min on both the Autostainer Link 48 and Omnis, the MLH1 antibody (ready to use solution) was incubated for 30 min in combination with a mouse linker on both the Autostainer Link 48 and the Omnis and the PMS2 antibody (ready to use solution) was incubated for 30 min on the Autostainer Link 48 and for 30 min in combination with a rabbit linker on the Omnis.

7 / 32

At University Hospital Aarhus immunohistochemistry for MSI evaluation was performed on the Dako Omnis instrument (Agilent, Santa Clara, CA) for samples obtained between 2015 to 2018 and prior that on the Ventana XT (Ventana Medical Systems, Tucson, AZ). The following antibodies and assays were used on the Omnis instrument: MLH1 clone ES05 (20 min, IR079, Agilent), MSH2 clone FE11 (20 min, IR085, Agilent), MSH6 clone EP49 (20 min, IR086, Agilent), and PMS2 clone EP51 (15 min, IR087, Agilent) in combination with the target retrieval solution high pH (30 min at 95 °C) and EnVision FLEX detection DAB system of Agilent. On the Ventana XT instrument MLH1 clone M1 (12 min, 790-4535, Roche), MSH2 clone G219-1129 (8 min, 760-4265, Roche), MSH6 clone 44 (12 min, 760-4455, Roche), and PMS2 clone EPR3947 (32 min, 760-4531, Roche) antibodies were used in combination with the CC1 pretreatment solution (40 min for MLH1, MSH2, and MSH6 and 64 min for PMS2). At both sites, the IHC protocols for the four MMR markers were validated prior to use according to ISO 15189. Samples were considered to demonstrate loss of MMR function when paired loss of expression of MLH1 and PMS2 or paired loss of MSH2 and MSLH6 expression was observed. Samples with sole loss of PMS2 and MSH6 were considered ‘suggestive for MMR deficient’. Samples with sole loss of MLH1 or MSH2 were considered to be with unclear MMR status (since this is more likely due to poor immunostaining). Samples without loss of expression of any MMR proteins were considered to have an intact MMR function. When the internal control (stromal cells, lymphocyte aggregates) did not show expression, the IHC analysis was considered invalid.

Tumor mutation burden analysis For TMB analysis the Oncomine Tumor Mutation Load Assay (Thermo Fisher Scientific, Waltham, MA) was used with an Ion Chef System and an IonS5 System following the instructions of the manufacturer. In short 10 ng gDNA was used per amplicon library preparation. Quantification of the library preparation was executed using the Ion Library TaqMan Quantification kit. A pool with 50 pM of each library was created and loaded on an Ion 540 Chip using the Ion Chef System and sequenced

8 / 32

using the IonS5 system. Analysis was performed in the Ion Reporter software using the Tumor Mutation Load – w2.0 – DNA – single sample workflow. For samples with poor quality (proportion of SNP with deamination >10), the filter settings for the TMB calculations were adjusted to 10%, 12.5%, or 15% (instead of 5%) allele frequency. The TMB analysis of samples were considered valid if the proportion of SNP with deamination was <10. Samples with 0 to 12.5 mutations per Mb were considered to have a low TMB, samples with 12.6 to 25 mutations/Mb were considered to have a moderate TMB, and samples with >25 mutations/Mb were considered to have a high TMB.

Statistics The diagnostic accuracy (concordance) is reported as percent agreement (OPA, PPA, NPA) with MSIH or MSS status as determined by the reference test (Promega MSI Analysis or IHC testing). Following calculations were used:

OPA = 100% x (MSI-H with Idylla MSI Test and reference assay + MSS with Idylla MSI Test and reference assay) /( MSI-H + MSS with reference assay),

PPA = 100% x (MSI-H with Idylla MSI Test and reference assay / MSI-H with reference assay) and

NPA = 100% x (MSS with Idylla MSI Test and reference assay / MSS with reference assay).

Associations of the discordant cases with the sample characteristics (tissue age, amount of necrosis, amount of tissue, fraction of tumor cells) was investigated by the Mann-Whitney test using the SPSS version 21 software. A P-value > 0.05 (two-tailed) was considered as significant. 9 / 32

Results Overall performance of the Idylla MSI Test To evaluate the performance of the Idylla MSI Test, 330 CRC samples over two testing sites were selected. Technical inclusion criteria include the presence of at least 20% tumor content and the availability of at least 50mm2 tissue from 1 to 5 5 µm FFPE tissue slides or at least 25 mm² tissue from 1 to 5 10µm FFPE tissue sections. In 157 cases macrodissection of the tumor tissue was needed to meet the 20% tumor content criteria. In the majority of cases (n=307) tissue from the primary lesion was used. Patient and sample characteristics are described in Table 1. Of the 330 enrolled patients, 54.6% were males and 45.5% were females with a mean age of the patient between 60 and 70 years. Of the 330 samples, staging information was available for 314 samples. 6.7%, 19.4%, 43.3%, and 25.8% were stage I, stage II, stage III, and stage IV, respectively. Information on the presence of necrosis was based on the pathological assessment of the H&E–stained slide. In the majority of cases, the degree of necrosis was below 5% of the tumor tissue area. All cases with a higher amount of necrosis (5% to 50%) gave a valid MSI result with the Idylla MSI Test. Archival FFPE tissue blocks were used in this study. On average the age of the FFPE blocks was three years. All tissue blocks older than three years gave valid result with the Idylla MSI Test illustrating that the test can easily process older FFPE samples.

Concordance of Idylla MSI Test with Promega MSI Analysis Three-hundred and thirty CRC tissue samples were analyzed with both Promega MSI Analysis and Idylla MSI Test on consecutive samples (Figure 1). When a test failed, it was repeated once to exclude technical errors that might have occurred during the manual steps. Eighteen samples 10 / 32

needed to be repeated for Promega MSI Analysis (5.4%) whereas only three samples needed retesting for the Idylla MSI Test (0.9%). In total seven samples still gave invalid results with the Promega MSI Analysis (2.12%) after retesting. Only two of them displayed an invalid result with the Idylla MSI Test (0.61%) after retesting. The seven invalid samples were excluded for the agreement analysis (Table 2). For 323 samples the results of the Idylla MSI Test were compared to the Promega MSI Analysis. Table 3 presents the absolute values for sample status obtained per Idylla MSI Test and Promega MSI Analysis. The overall percentage agreement (OPA) is 99.7% (with 98.3% and 100% lower and upper limit of the 95% confidence intervals (CI), respectively). The positive percentage agreement (PPA) is 98.7% (with 92.9% and 99.8% lower and upper limit of the 95% CI, respectively) and the negative percentage agreement (NPA) is 100% (with 98.5% and 100% lower and upper limit of the 95% CI, respectively). One discordant case was identified. This sample was initially scored invalid by Promega MSI Analysis, and MSI-H after retesting (five unstable mononucleotide markers) whereas the Idylla MSI Test, analyzing seven monomorphic biomarkers, labeled the sample microsatellite stable (MSS) (with seven negative biomarkers) . Retesting of this sample starting from the original FFPE block gave an MSI-H result with the Idylla MSI Test (four of the seven biomarkers positive). For the 75 samples with concordant MSI-H results, a detailed analysis of the amount of positive markers for each assay is described in Table 4.

Concordance of Idylla MSI Test with immunohistochemistry analysis Of all the 330 samples selected for Idylla MSI Test analysis, historical IHC data for MLH1, PMS2, MSH2, and MSH6 expression were available. For two cases the IHC results were invalid as staining of at least one MMR expression marker failed (Table 5). These failures were detected by evaluating stromal, inflammatory, or non-neoplastic epithelial cells surrounding the malignant cells. Both cases produced a valid result with the Idylla MSI Test. 11 / 32

IHC evaluation of MLH1, PMS2, MSLH2, and MSH6 resulted in an unclear MMR status for seven cases since sole loss of MLH1 or MSH2 expression was observed. Nine cases were suggestive for MMR deficiency as sole loss of PMS2 and MSH6 expression was observed (Table 5 and Table 6). With the Idylla MSI Test a valid result was obtained in all 16 cases. Four of the nine samples suggestive for MMR deficiency contained MSI-H and three of the seven samples with unclear MMR status contained MSI-H. Promega MSI Analysis confirmed these findings. To assess the concordance between the Idylla MSI Test and the historical IHC data, 20 samples were excluded: two samples with an invalid IHC result, 16 samples with either an unclear or suggestive IHC conclusion and two cases with an invalid Idylla MSI Test result. The OPA is 98.7% (with 96.7% and 99.5% lower and upper limit of the 95% CI, respectively). The PPA is 94.4% (with 86.7% and 97.8% lower and upper limit of the 95% CI, respectively) and the NPA is 100% (with 98.4% and 100% lower and upper limit of the 95% CI, respectively). In total, four discordant cases were identified (Table 7). In these cases, the conclusion of the IHC staining pointed towards loss of MMR function whereas the Idylla MSI Test did not detect MSI-H. In one of these cases, paired loss of MSH2 and MSH6 expression was found whereas in the other three cases, paired loss of MLH1 and PMS2 was observed (Table 8). Reexamination of the IHC slides of the discordant cases showed that the loss of MLH1 and PMS2 expression in samples B034 and B078 was not homogenous (±66% and 80% to 90% of tumor cells with loss).

Analysis of the discordant cases by tumor mutation burden determination Comparing IHC and Idylla MSI Test results, four discordant cases were identified (Table 7). These cases all contain loss of MMR expression but were MSS according to the Idylla MSI Test. Three of the four cases were also MSS according to the Promega MSI Analysis (Table 8). No association of the discordant samples with any of the sample characteristics (tissue age, amount of necrosis, amount of tissue, percentage tumor content) could be detected (Table 1). 12 / 32

To further investigate the presence of a defect in the DNA mismatch repair machinery in these samples the TMB load was determined. Table 8 shows the amount of mutations per megabase detected by the Oncomine Tumor mutation load assay. For one case the TMB load was high, pointing towards a defect in the DNA repair machinery. In this case the TMB result disagrees with the initial Idylla MSI Test result and is in agreement with the IHC (and Promega MSI Analysis) result. In the three other cases the TMB was low indicative for an intact DNA mismatch repair function. Hence, for the majority of cases, the TMB results were in agreement with the Idylla MSI Test results (Table 8).

Discussion In this study we have evaluated the performance of the Idylla MSI Test, a fast and easy to use Test to evaluate the microsatellite status on FFPE material of colorectal cancer patients. Compared to the Promega MSI Analysis, an OPA of 99.7%, a PPA of 98.7%, and an NPA of 100% was achieved with the Idylla MSI Test. These results are in agreement with the findings of De Craene et al using the prototype Idylla MSI Test cartridges .19 One discordant sample was found (Table 3). Retesting starting from the FFPE block revealed an MSI-H result with the Idylla and gave an invalid result with the Promega assay. A sample mix up could be at the origin of this initial discrepancy. Alternatively, a difference in tumor content could be the cause of these differences as during the original testing macrodissection was not required whereas in the duplicate testing macrodissection needed to be performed to increase the tumor content. The Idylla MSI Test analyses seven markers and the Promega Test analyses five markers. Table 4 illustrates that three or more markers were positive with both assays in the majority of cases with MSI-H (100% with Idylla  98.7% with Promega), suggesting comparable sensitivities of both assays. Of the 330 samples analyzed, seven cases gave an invalid result with the Promega MSI Analysis and two cases gave an invalid result with the Idylla MSI Test after retesting. The Idylla MSI Test is highly compatible with fragmented DNA 13 / 32

obtained from FFPE material as it has been specifically designed to amplify very small amplicons. The low degree of invalid results illustrates that the Idylla MSI Test can cope with less optimal sample conditions such as a high proportion of necrotic tissue, limited amounts of tissue, and older tissue blocks. The Idylla MSI Test offers several practical benefits compared to the Promega MSI Analysis: sample batching is not required to improve economical and organizational aspects and a turnaround time of less than one day is achievable since the assay takes less than 2.5h. Further, the Idylla MSI Test contains a limited amount of manual steps and is less prone to human errors. This was clearly reflected during the repeat testing performed in this study: only one of the three initial invalid Idylla MSI Test results (0.3% of total samples) became valid whereas 11 of the 18 initial invalid Promega MSI Analysis results (or 3.33% of total samples) became valid upon retesting. According to the instructions for use, the Idylla MSI Test requires a minimum total tissue area of 50mm2 whereas the Promega kit requires at least 1ng of genomic DNA. This might suggest that the Promega kit is more suited for samples with limited material such as small biopsy samples and post treatment rectal resections. In this study, 19 biopsy samples were however included. Two of them gave an invalid result with the Idylla MSI Test. The Promega test also gave an invalid result for both cases. A more extensive study with limited material is therefore needed to test which assay is better suited for restricted material. In contrast to the Promega MSI Analysis, the Idylla MSI Test does not require a separate analysis of non-malignant patient tissue. A limitation of this study is that patient-specific controls were not analyzed with the Promega MSI Analysis. This practice can lead to an erroneous missclassification of some markers as unstable. In specimens with limited material such as colorectal biopsy samples non-malignant material is however not always available. Compared to the historically obtained IHC data, an OPA of 98.7%, a PPA of 94.4%, and an NPA of 100% was achieved with the Idylla MSI Test. For two cases no valid conclusion could be drawn from the IHC data. In both cases, a valid Idylla MSI Test result could be obtained. Similarly, the two cases

14 / 32

with an invalid Idylla MSI Test results gave a valid result with IHC. Consequently IHC is a good alternative in the limited cases where the Idylla MSI Test fails, and vice versa. When MSI-H status is detected, generally loss of expression of the set MLH1 and PMS2 or of the set MSH2 and MSH6 is observed. However, sole loss of one marker can take place in some cases of Lynch Syndrome that do not display MSI-H. In this study, 14 cases showed loss of one mismatch protein marker and two cases showed the loss of an unconventional set of markers, namely MLH1 and MSH6 or MLH1 and MSH2. For these 16 cases (4.85%) the Idylla MSI Test produced a valid result. Seven cases were scored as MSI-H and nine cases were scored as MSS. Promega MSI Analysis confirmed these findings. Consequently samples with loss of one MMR expression marker not always display MSI-H. The Idylla MSI Test can serve as an alternative method when the IHC data are difficult to interpret. It must be noted that some rare cases of Lynch Syndrome with MSH6 mutations do not display MSI-H. Fortunately, this observation is less pronounced when enough mononucleotide markers are used for MSI testing.20 In agreement with these observations loss of MSH6 IHC staining has been described in the setting of prior chemotherapy or radiotherapy.10 In four cases a discrepancy between IHC data and Idylla MSI Test data was observed. In one case Promega MSI Analysis matches the IHC result being indicative for MSI-H. This sample also contains a high tumor mutation load, further suggesting that a defect of the DNA repair machinery is likely present. In the other three samples the Promega MSI Analysis matches the MSS result of the Idylla MSI Test. The tumor mutation load of these three samples was low. Reanalysis of the IHC slides revealed that the two MSS cases with MLH1 and PMS2 loss (B034 and B078) display incomplete loss of both proteins (±66% and 80% to 90% loss). These data are in line with international literature suggesting that a loss in >90% of the tumor cells correlates well with MMR function deficiency.21,22 The case with loss of MSH2 and MSH6 expression (D126) might contain a mutation in the MSH2 gene explaining the discordant IHC/MSI result.

15 / 32

At present NCCN recommends screening for MMR status either by IHC or by PCR.23 IHC is often preferred as it can be performed quickly and the cost of four IHC tests is relatively low. Although this phenomenon was not observed in this study, literature shows that IHC produces false negative results in 5% to 11% of cases.10,12,24 In this study, however, some false positive IHC results were observed. When IHC is used as a prescreen for Lynch this is not an issue as follow up testing by a genetic center will be performed. When the microsatellite status is used as a predictive test for immunotherapy

17,18

, accuracy and a short turn-around time are key. Consequently both false

negative and false positive results should be avoided. With the Idylla MSI Test one false negative and two invalid results were observed. With the IHC test three false positives, two invalid, and 16 unclear or only suggestive results were observed. This study has been restricted to CRC. However, microsatellite instability is also observed in about 15% of gastric and endometrial cancers and at lower frequencies in several other cancers.25 Approval for anti-PD1 immune-checkpoint inhibition has been granted for all patients with solid tumors containing MSI-H or being mismatch repair deficient.8,26 Several reports suggest that in endometrial cancer the allelic shifts at several loci can be shorter and microsatellite instability is hence more difficult to evaluate with a PCR assay.27,28 Taking these data in account, the clinical performance of the Idylla MSI Test in other solid tumors should be evaluated as well. In conclusion, we have demonstrated that the Idylla MSI Test is extremely powerful to identify the microsatellite status in human colorectal cancer samples. The test is fully automated and gives accurate results within 2.5h.

16 / 32

Acknowledgments We thank Christophe Hermans for his help with preparing tissue sections for analysis and all members of UZA pathology laboratory involved in generating the historical IHC data (staining and evaluating the slides). Author Contributions K.Z. is the guarantor of this work and, as such, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

17 / 32

References 1. Lynch HT, Snyder CL, Shaw TG, Heinen CD, Hitchins MP: Milestones of Lynch syndrome: 18952015. Nat Rev Cancer 2015, 15:181-194 2. Yamamoto H, Imai K: Microsatellite instability: an update. Arch Toxicol 2015, 89:899-921 3. de la Chapelle A, Hampel H: Clinical relevance of microsatellite instability in colorectal cancer. J Clin Oncol. 2010, 28:3380-3387 4. Roth AD, Tejpar S, Delorenzi M, Yan P¨, Fiocca R, Klingbiel D, Dietrich D, Biesmans B, Bodoky G, Barone C, Aranda E, Nordlinger B, Cisar L, Labianca R, Cunningham D, Van Cutsem E, Bosman F. Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60-00 trial. J Clin Oncol. 2010, 28:466474. 5. Sinicrope FA, Sargent DJ: Molecular pathways: microsatellite instability in colorectal cancer: prognostic, predictive, and therapeutic implications. Clin Cancer Res. 2012 , 18:1506-1512 6. 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 7. Zhang L: Immunohistochemistry versus microsatellite instability testing for screening colorectal cancer patients at risk for hereditary nonpolyposis colorectal cancer syndrome. Part II. The utility of microsatellite instability testing. J Mol Diagn 2008, 10:301-307 8. Ryan E, Sheahan K, Creavin B, Mohan HM, Winter DC: The current value of determining the mismatch repair status of colorectal cancer: A rationale for routine testing. Crit Rev Oncol Hematol 2017, 116:38-57 9. Sepulveda AR, Hamilton SR, Allegra CJ, Grody W, Cushman-Vokoun AM, Funkhouser WK, Kopetz SE, Lieu C, Lindor NM, Minsky BD, Monzon FA, Sargent DJ, Singh VM, Willis J, Clark J, Colasacco C, Rumble RB, Temple-Smolkin R, Ventura CB, Nowak JA: Molecular Biomarkers for the Evaluation of Colorectal Cancer: Guideline Summary From the American Society for Clinical Pathology, 18 / 32

College of American Pathologists, Association for Molecular Pathology, and American Society of Clinical Oncology. J Oncol Pract 2017, 13:333-337 10. Bartley AN, Hamilton SR, Alsabeh R, Ambinder EP, Berman M, Collins E, Fitzgibbons PL, Gress DM, Nowak JA, Samowitz WS, Zafar SY; Members of the Cancer Biomarker Reporting Workgroup, College of American Pathologists: Template for reporting results of biomarker testing of specimens from patients with carcinoma of the colon and rectum. Arch Pathol Lab Med 2014, 138:166-170 11. Murphy KM, Zhang S, Geiger T, Hafez MJ, Bacher J, Berg KD, Eshleman JR: Comparison of the microsatellite instability analysis system and the Bethesda panel for the determination of microsatellite instability in colorectal cancers. J Mol Diagn 2006, 8:305-311 12. Zhao H, Thienpont B, Yesilyurt BT, Moisse M, Reumers J, Coenegrachts L, Sagaert X, Schrauwen S, Smeets D, Matthijs G, Aerts S, Cools J, Metcalf A, Spurdle A; ANECS, Amant F, Lambrechts D: Mismatch repair deficiency endows tumors with a unique mutation signature and sensitivity to DNA double-strand breaks. Elife. 2014, 3:e02725 13. Asad Umar, C. Richard Boland, Jonathan P. Terdiman, Sapna Syngal, Albert de la Chapelle, Josef Rüschoff, Richard Fishel, Noralane M. Lindor, Lawrence J. Burgart, Richard Hamelin, Stanley R. Hamilton, Robert A. Hiatt, Jeremy Jass, Annika Lindblom, Henry T. Lynch, Païvi Peltomaki, Scott D. Ramsey, Miguel A. Rodriguez-Bigas, Hans F. A. Vasen, Ernest T. Hawk, J. Carl Barrett, Andrew N. Freedman, Sudhir Srivastava: Revised Bethesda Guidelines for Hereditary Nonpolyposis Colorectal Cancer (Lynch Syndrome) and Microsatellite Instability. J Natl Cancer Inst. 2004, 96: 261–268. 14. Stadler ZK, Battaglin F, Middha S, Hechtman JF, Tran C, Cercek A, Yaeger R, Segal NH, Varghese AM, Reidy-Lagunes DL, Kemeny NE, Salo-Mullen EE, Ashraf A, Weiser MR, Garcia-Aguilar J, Robson ME, Offit K, Arcila ME, Berger MF, Shia J, Solit DB, Saltz LB: Reliable Detection of Mismatch Repair Deficiency in Colorectal Cancers Using Mutational Load in Next-Generation Sequencing Panels. J Clin Oncol 2016, 34:2141-2147 19 / 32

15. Kim JE, Chun SM, Hong YS, Kim KP, Kim SY, Kim J, Sung CO, Cho EJ, Kim TW, Jang SJ: Mutation Burden and I Index for Detection of Microsatellite Instability in Colorectal Cancer by Targeted Next-Generation Sequencing. J Mol Diagn 2019, 21:241-250 16. Patel SA, Longacre TA, Ladabaum U, Lebensohn A, Lin AY, Haraldsdottir S: Tumor Molecular Testing Guides Anti-PD-1 Therapy and Provides Evidence for Pathogenicity of Mismatch Repair Variants. Oncologist. 2018, 23: 1395-1400. 17. Le DT, Durham JN, Smith KN, Wang H, Bartlett BR, Aulakh LK, Lu S, Kemberling H, Wilt C, Luber BS, Wong F, Azad NS, Rucki AA, Laheru D, Donehower R, Zaheer A, Fisher GA, Crocenzi TS, Lee JJ, Greten TF, Duffy AG, Ciombor KK, Eyring AD, Lam BH, Joe A, Kang SP, Holdhoff M, Danilova L, Cope L, Meyer C, Zhou S, Goldberg RM, Armstrong DK, Bever KM, Fader AN, Taube J, Housseau F, Spetzler D, Xiao N, Pardoll DM, Papadopoulos N, Kinzler KW, Eshleman JR, Vogelstein B, Anders RA, Diaz LA Jr: Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science 2017, 357:409-413 18. Le DT, Uram JN, Wang H, Bartlett BR, Kemberling H, Eyring AD, Skora AD, Luber BS, Azad NS, Laheru D, Biedrzycki B, Donehower RC, Zaheer A, Fisher GA, Crocenzi TS, Lee JJ, Duffy SM, Goldberg RM, de la Chapelle A, Koshiji M, Bhaijee F, Huebner T, Hruban RH, Wood LD, Cuka N, Pardoll DM, Papadopoulos N, Kinzler KW, Zhou S, Cornish TC, Taube JM, Anders RA, Eshleman JR, Vogelstein B, Diaz LA Jr: PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N Engl J Med 2015, 372:2509-2520 19. De Craene B, Van de Velde J, Rondelez E, Vandenbroeck L, Peeters K, Vanhoey T, Elsen N, Vandercruyssen G, Melchior L, Willemoe GL, Jones A, Hongxiang L, Henrique R, Teixeira M, Zhao H, Lambrechts D, Decanniere K, Sablon E, Maertens GG: Detection of microsatellite instability (MSI) in colorectal cancer samples with a novel set of highly sensitive markers by means of the Idylla MSI Test prototype. Journal of Clinical Oncology 2018, 36(15_suppl):e15639-e15639 20. You JF, Buhard O, Ligtenberg MJ, Kets CM, Niessen RC, Hofstra RM, Wagner A, Dinjens WN, Colas C, Lascols O, Collura A, Flejou JF, Duval A, Hamelin R: Tumours with loss of MSH6 expression are 20 / 32

MSI-H when screened with a pentaplex of five mononucleotide repeats. Br J Cancer. 2010, 103:1840-1845 21. Shia J, Stadler Z, Weiser MR, Rentz M, Gonen M, Tang LH, Vakiani E, Katabi N, Xiong X, Markowitz AJ, Shike M, Guillem J, Klimstra DS: Immunohistochemical staining for DNA mismatch repair proteins in intestinal tract carcinoma: how reliable are biopsy samples? Am J Surg Pathol 2011, 35:447-454 22. Sarode VR, Robinson L: Screening for Lynch Syndrome by Immunohistochemistry of Mismatch Repair Proteins: Significance of Indeterminate Result and Correlation With Mutational Studies. Arch Pathol Lab Med. 2019,143:1225-1233 23. Provenzale D, Gupta S, Ahnen DJ, Bray T, Cannon JA, Cooper G, David DS, Early DS, Erwin D, Ford JM, Giardiello FM, Grady W, Halverson AL, Hamilton SR, Hampel H, Ismail MK, Klapman JB, Larson DW, Lazenby AJ, Lynch PM, Mayer RJ, Ness RM, Regenbogen SE, Samadder NJ, Shike M, Steinbach G, Weinberg D, Dwyer M, Darlow S: Genetic/familial high-risk assessment: Colorectal Version 1.2016, NCCN Clinical Practice Guidelines in Oncology: J Natl Compr Canc Netw 2016, 14:1010–1030 24. Bartley AN1, Luthra R, Saraiya DS, Urbauer DL, Broaddus RR: 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 25. Bonneville R, Krook MA, Kautto EA, Miya J, Wing MR, Chen HZ, Reeser JW, Yu L, Roychowdhury S1: Landscape of Microsatellite Instability Across 39 Cancer Types. JCO Precision Oncol 2017, (ePub) doi: 10.1200/PO.17.00073 26. Abida W, Cheng ML, Armenia J, Middha S, Autio KA, Vargas HA, Rathkopf D, Morris MJ, Danila DC, Slovin SF, Carbone E, Barnett ES, Hullings M, Hechtman JF, Zehir A, Shia J, Jonsson P, Stadler ZK, Srinivasan P, Laudone VP, Reuter V, Wolchok JD, Socci ND, Taylor BS, Berger MF, Kantoff PW, Sawyers CL, Schultz N, Solit DB, Gopalan A, Scher HI: Analysis of the Prevalence of Microsatellite

21 / 32

Instability in Prostate Cancer and Response to Immune Checkpoint Blockade. JAMA Oncol. 2019, 5:471-478. 27. Wang Y, Shi C, Eisenberg R, Vnencak-Jones CL: Differences in Microsatellite Instability Profiles between Endometrioid and Colorectal Cancers: A Potential Cause for False-Negative Results? J Mol Diagn 2017, 19:57-64 28. Djordjevic B, Broaddus RR: Laboratory Assays in Evaluation of Lynch Syndrome in Patients with Endometrial Carcinoma. Surg Pathol Clin 2016, 9:289-299

22 / 32

Figure Legend Figure 1: Read-out from the Idylla MSI Test and Promega MSI Analysis for a MSS (A, B) and MSI-H result (C, D).

23 / 32

Table 1: Patient and sample characteristics.

Total

IDYLLA INVALID (n=2)

Discordant with PROMEGA (n=1)

Discordant with IHC (n=4)

Sex Male 180 (54.6%) 2 Female 150 (45.5%) Patient Age 18 <30y 1 (0.3%) 30 <40y 6 (1.8%) 1 40 <50y 17 (5.2%) 50 <60y 60 (18.2%) 60 <70y 72 (21.8%) 1 70 <80y 106 (32.1%) 80 <90y 51 (15.5%) 90 ≤100y 17 (5.2%) Stage of disease Stage I 22 (6.7%) Stage II 64 (19.4%) Stage III 143 (43.3%) Stage IV 85 (25.8%) unknown 16 (4.8%) 2 Tumor location Primary tissue 307 (93%) 2 Metastatic lesion 4 (1.2%) Unspecified 19 (5.8%) Tumor necrosis 0-5% 269 (80.9%) 2 5-10% 29 (8.79%) 10-20% 19 (5.8%) 20-50% 13 (3.9%) Macrodissection required to reach minimum of 20% tumor content YES 157 (47.6%) NO 173 (52.4%) 2 Tumor content in tissue used for analysis 20 <30% 168 (50.9%) 1 30 <40% 77 (23.3%) 1 40 <50% 34 (10.3%) 50 <60% 25 (7.6%) 60 <70% 15 (4.5%) 70 <80% 11 (3.3%) Age of the tissue block 0 <1y 39 (11.8%) 1 1 <2y 70 (20.9%) 1 2 <3y 52 (15.8%) 3 <4y 60 (18.2%) 4 <5y 56 (17%) ≥5y 54 (16.4%) Total Tissue Area (assuming 5µm slide thickness) 50 <100 mm2 62 (18,2%) 2 24 / 32

1

1

2 2

2 2

2

1

1 1

1

3 1

1

4

1

1 3

1

2 2

1

3

1

1

1

100 <200 mm2 200 <400 mm2 400 ≤600 mm2

83 109 76

(25.2%) (33.0%) (23,0%)

1 2

25 / 32

Table 2: Amount of valid test results with Idylla MSI Test and Promega MSI Analysis.

IDYLLA Valid Invalid Total

PROMEGA Valid 323 0 323

Invalid 5 2 7

26 / 32

Total 328 2 330

Table 3: Concordance between Idylla MSI Test and Promega MSI Analysis. PROMEGA MSS MSI-H MSS 247 1* MSI-H 0 75 Total 247 76 * Upon retesting, this sample was also MSI-H with the Idylla MSI Test.

IDYLLA

27 / 32

Total 248 75 323

Table 4: Amount of positive biomarkers with Idylla MSI Test and Promega MSI Analysis for the 75 concordant MSH-I samples.

IDYLLA 2/7 3/7 4/7 5/7 6/7 7/7 Total

2/5 1 1 (1.3%)

3/5 2 1 1 4 (5.3%)

Promega 4/5 8 4 1 13 (17.3%)

28 / 32

5/5 1 2 10 24 8 45 (60%)

4/4 1 1 8 2 12 (16%)

Total 0 (0%) 1 (1.3%) 5 (6.7%) 21 (28%) 36 (48%) 12 (16%) 75

Table 5: Amount of valid test results with Idylla MSI Test and IHC. IHC Valid Valid – suggestive Valid IDYLLA and clear for MMR unclear MMR Invalid† Total result deficiency * status * Valid 310 9 7 2 328 Invalid 2 0 0 0 2 Total 312 9 7 2 330 *Suggestive for MMR deficiency as sole loss of PMS2 or MSH6 was observed, unclear MMR status as sole loss of MLH1 or MSH2 was observed. †Invalid as at least one of the four markers did not show expression in the internal control tissue (lymphocytes, stromal cells).

29 / 32

Table 6: Detailed analysis of the four MMR proteins on the 16 samples deemed unclear or only suggestive by IHC. Sample IHC PMS2 IHC MLH1 IHC MSH6 IHC MSH2 IHC name Conclusion* B027 Loss No loss No loss No loss suggestive B039 No loss No loss Loss No loss suggestive B046 No loss No loss Loss No loss suggestive B052 No loss No loss No loss Loss unclear B066 Loss No loss No loss No loss suggestive B119 No loss Loss Loss No loss suggestive B121 No loss Loss No loss No loss unclear B125 No loss Loss No loss No loss unclear D073 No loss Loss No loss Loss unclear D110 Loss No loss No loss No loss suggestive D127 No loss Loss No loss No loss unclear D135 No loss Loss No loss No loss unclear D137 No loss No loss Loss No loss suggestive D138 No loss No loss Loss No loss suggestive D149 No loss Loss No loss No loss unclear D150 No loss No loss Loss No loss suggestive * suggestive: suggestive for MMR deficiency, unclear: unclear MMR status.

30 / 32

IDYLLA MSI-H MSI-H MSI-H MSS MSS MSS MSS MSS MSI-H MSS MSI-H MSS MSS MSS MSI-H MSI-H

Table 7: Concordance between Idylla MSI Test and IHC results.

IDYLLA MSS MSI-H Total

MMR function intact 238 0 238

IHC MMR deficient 4 68 73

31 / 32

Total 242 68 310

Table 8: Detailed analysis of the four discordant samples – TMB analysis in relation to the IHC, Idylla, and Promega MSI Analysis results. TMB Sample IHC Mutations Tumor name /Mb Load B034 4.28 LOW MMR deficient (PMS2/MLH1 loss)† B040* 572.25 VERY HIGH MMR deficient (PMS2/MLH1 loss) B078 5.88 LOW MMR deficient (PMS2/MLH1 loss)† D126 9.07 LOW MMR deficient (MSH2/MSH6 loss) * Upon retesting, this sample was also MSI-H with the Idylla MSI Test. †Reanalysis of the IHC slide shows heterogeneous loss of PMS2/MLH1.

32 / 32

IDYLLA

PROMEGA

MSS MSS MSS MSS

MSS MSI-H MSS MSS

Promega MSI Analysis

Idylla MSI Test A

C

B

D