Tumor analysis of MMR genes in Lynch-like syndrome: challenges associated to results interpretation

doi:10.21203/rs.3.rs-3223657/v1

Download PDF

Brief Communication

Tumor analysis of MMR genes in Lynch-like syndrome: challenges associated to results interpretation

https://doi.org/10.21203/rs.3.rs-3223657/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Up to 70% of suspected Lynch syndrome patients harboring MMR deficient tumors lack identifiable germline pathogenic variants in MMR genes, being referred to as Lynch-like syndrome (LLS). Previous studies have reported biallelic somatic MMR inactivation in 15–95% LLS-associated tumors. However, translating tumor testing results into patient management remains controversial. Our aim is to assess the challenges associated to the implementation of tumoral analyses in routine genetic testing workflows. Here we present the clinical characterization of 229 LLS patients. MMR testing was performed in 39 available tumors, and results were analyzed using two variant allele frequency (VAF) thresholds (≥ 5% and ≥ 10%). More biallelic somatic MMR inactivating events were identified at VAF ≥ 5% than ≥ 10% (35.9% vs. 25.6%), although the rate of non-concordant results regarding immunohistochemical pattern increased (30.8% vs. 20.5%). Standardized protocols for the analysis and interpretation of tumoral MMR testing are needed to improve management of LLS individuals.

Health sciences/Health care/Diagnosis/Genetic testing

Health sciences/Medical research/Genetics research

Lynch syndrome

Lynch-like syndrome

mismatch repair genes

mismatch repair-deficiency

tumor testing

clinical management

Lynch syndrome (LS) accounts for ~ 3% of colorectal cancer (CRC) cases, and also predisposes to endometrium, ovary and other neoplasms¹. LS is caused by germline inactivation of one of the mismatch repair (MMR) genes (MLH1, MSH2, MSH6, PMS2) or EPCAM gene deletions, and its molecular diagnosis mainly relies on MMR-deficiency (MMRd) screening in tumors (microsatellite instability (MSI) assessment and/or MMR immunohistochemical (IHC) staining) along with germline MMR testing². Up to 60–70% of patients harboring MMRd tumors not linked to somatic methylation of MMR genes lack identifiable germline pathogenic variants (PV)^3,4. These cases are referred to as Lynch-like syndrome (LLS). Recent studies have reported somatic biallelic inactivation of MMR genes at frequencies ranging from 15–95% of LLS tumors (reviewed in Table S1), likely representing sporadic events. Additionally, MMRd can result from germline biallelic inactivation of MUTYH and MSH3 or germline PVs in the exonuclease domains of POLD1 and POLE genes^5–8.

Clinical management of LLS is challenging, since they represent a heterogeneous group of hereditary and sporadic conditions. LLS relatives’ present lower risk of CRC compared to those from LS individuals, but higher than sporadic cases³. In this sense, we conducted a clinical and molecular characterization of our LLS cohort and performed tumor testing in patients with available tumors. Our aim was to assess the viability of implementing tumor analyses in routine genetic testing workflows to improve the diagnosis and management of LLS patients.

Patients

The complete clinical cohort comprises 229 LLS cases (52.4% males and 47.6% females): 105 previously reported⁹ and 124 newly identified patients with MMRd and/or MSI, but no germline MMR PVs nor somatic hypermethylation of MLH1 promoter (Figure S1). Clinical and tumor data are displayed in Table S2.

Tumor testing

In 41/229 LLS patients, tumor samples were available for molecular testing and fulfilled the following inclusion criteria: (1) MSI and/or MMRd tumors; (2) no germline (likely) PVs in MLH1, MSH2, MSH6, PMS2 or EPCAM identified, which were sequenced based on their MMR expression pattern by immunohistochemistry; and (3) no somatic MLH1 promoter hypermethylation (Table S3). DNA extracted from colorectal or endometrial carcinomas was analyzed using our I2HCP next-generation sequencing (NGS) panel adapted for somatic testing¹⁰. All variants identified in coding and adjacent intronic regions (i.e. ±20 bp) at variant allele frequencies (VAFs) ≥ 5% and covered at ≥ 30X were retrieved from the following genes: MLH1 (NM_000249.3), MLH3 (NM_001040108.1), MSH2 (NM_000251.1), MSH3 (NM_002439.3), MSH6 (NM_000179.2), MUTYH (NM_001128425.1), PMS2 (NM_000535.5), POLD1 (NM_001256849.1) and POLE (NM_006231.2). Variant classification was performed using vaRHC, an in-house developed tool for the semi-automation of variant interpretation according to ACMG/AMP and ClinGen’s gene-specific guidelines¹¹. Whenever possible, loss of heterozygosity (LOH) was assessed in tumor DNA by analyzing the alternative allele ratio of germline heterozygous SNPs previously genotyped by Sanger sequencing or NGS in the routine clinical setting¹². Variants identified in tumors at VAFs ≥ 20% were re-tested in blood-derived DNA by Sanger sequencing to discard constitutional variants previously missed, as recommended by ESMO PMWG¹³.

Our clinical series comprises 229 LLS individuals. 34.9% had a family history of LS-related tumors. The mean age at first cancer diagnosis was 53.9 years (range 16–85). The majority were diagnosed with CRC (90.8%), predominantly in the proximal colon (58.9%). Table 1 summarizes demographics, tumor and molecular data, and Table S2 displays the complete information from the series.

Table 1

	All	Male	Female
	Total n (%)	Total n (%)	Total n (%)
DEMOGRAPHICS
Complete cohort	229 (100%)	120 (52.4%)	109 (47.6%)
Mean current age (range)	64 (21-95)	64 (21-95)	64 (26-94)
Deceased	49 (21.4%)	27 (22.5%)	22 (9.6%)
Family history
Yes	79 (34.9%)	42 (35.0%)	37 (33.9%)
Amsterdam II	13 (6.1%)	5 (4.2%)	8 (7.3%)
No	150 (65.5%)	78 (65.0%)	72 (66.1%)
TUMORAL DATA
Mean age at first cancer	53.9 (16-85)	54.8 (16-85)	53.0 (24-81)
Mean age at MMRd/MSI tumor	54.7 (16-85)	56.7 (16-85)	53.6 (24-81)
Individuals with family history	54.8 (21-78)	56.0 (21-78)	53.4 (24-78)
Individuals without family history	54.7 (16-85)	55.5 (16-85)	53.7 (30-81)
Individuals with CRC	207 (90.8%)	116 (96.7%)	91 (83.5%)
Location of CRC
Proximal colon	122 (58.9%)	57 (49.1%)	65 (71.4%)
Distal colon	60 (29.0%)	42 (36.2%)	18 (19.8%)
Multiple locations	22 (10.6%)	15 (12.9%)	7 (7.7%)
Unknown	3 (1.4%)	2 (1.7%)	1 (1.1%)
Individuals with EC	21 (9.2%)	0 (0%)	21 (19.3%)
Individuals with other LS-related tumors	13 (5.7%)	7 (5.8%)	6 (5.5%)
Individuals with other LS-non related tumors	31 (13.5%)	19 (15.7%)	12 (11.0%)
Individuals with multiple tumors	58 (25.3%)	33 (27.5%)	25 (22.9%)
Stage at diagnosis of MMRd/MSI tumors
I	36 (15.7%)	13 (10.8%)	23 (21.1%)
II	110 (48.0%)	68 (56.7%)	42 (38.5%)
III	53 (23.1%)	25 (20.8%)	28 (25.7%)
IV	10 (4.4%)	6 (5.0%)	4 (3.7%)
unknown	20 (8.7%)	8 (6.7%)	12 (11.0%)
MOLECULAR DATA
MMRd	205 (89.5%)	104 (86.7%)	101 (92.7%)
Loss MLH1-PMS2	122 (59.5%)	60 (57.7%)	62 (61.4%)
Loss PMS2	7 (3.4%)	5 (4.7%)	2 (2.0%)
Loss MSH2-MSH6	52 (25.4%)	27 (26.0%)	25 (24.8%)
Loss MSH6	24 (11.7%)	12 (11.5%)	12 (11.9%)
MSI	128 (55.9%)	66 (28.8%)	62 (56.9%)

MMRd: Mismatch repair deficient; MSI: Microsatellite instability; CRC: Colorectal cancer; EC: Endometrial cancer; LS: Lynch syndrome

Out of 41 LLS tumors, two samples (4.9%) were discarded due to failed DNA amplification. By using a VAF threshold ≥ 5%, in 14/39 cases (35.9%) double somatic hits consistent with the IHC patterns were identified: in 10/14 (71.4%) two somatic MMR PVs were detected, while 4/14 (28.6%) showed one somatic MMR PV along with LOH of the same MMR gene. Interestingly, some patients harbored more than two somatic hits affecting the same MMR gene. The observed VAFs suggest the presence of cell clones sharing a first hit and evolving with different somatic second hits (cases 5, 104, 208). Furthermore, in 11/39 (28.2%) patients, one somatic hit consistent with the tumor MMR IHC pattern was detected. LOH could not be evaluated in nine of them due to the lack of informative SNPs, since pre-NGS genetic testing algorithms were directed to only one or two MMR genes. In 12/39 (30.8%) cases, the somatic hits identified disagreed with the IHC pattern. Additionally, four patients with one compatible somatic hit harbored additional somatic PVs in other MMR genes, non-concordant with the tumor MMR IHC pattern, making interpretation challenging (cases 45, 61, 78, 194) (Fig. 1A; Table S3).

To address interpretation ambiguities and due to the lack of orthogonal validation, VAF detection threshold was raised to ≥ 10% to minimize potential false positives due to formalin fixation artifacts. With this adjustment, 10/39 patients (25.6%) harbored compatible double somatic hits according to their IHC MMR status (compared to 35.9% in the previous scenario). The rate of non-concordant results also decreased significantly (30.8–20.5%). However, the number of patients with no somatic hits identified increased (5.1–28.2%) (Fig. 1A-1C).

Our study did not find any candidate PVs in POLE, POLD1, MSH3 and MUTYH genes.

As summarized in Table S1, there is a wide heterogeneity in the percentage of biallelic MMR inactivation reported, which can be partly attributable to differences in the experimental designs and data analysis, making it difficult to draw unbiased comparisons. Besides, the limitations derived from the interpretation of somatic analyses have not yet been clearly depicted. Considering hypermutability in MMRd tumors, ambiguous interpretations may arise, which evidences the necessity of standardized protocols. Our study shows that a single tumor may present several hits involving different MMR genes. Furthermore, in 30.8% of cases (VAF ≥ 5%) or 20.5% of cases (VAF ≥ 10%), the predominant MMR hit disagreed with the IHC pattern. These cases were considered inconclusive.

A remarkable finding of this study is illustrated in case 198 (Fig. 2A). Routine tumor screening identified PMS2 loss of expression (Fig. 2B), prompting only PMS2 germline mutational analysis. However, MLH1 c.113A > G; p.(Asn38Ser) PV was detected at VAF = 59.1% in tumor DNA, and subsequently confirmed in blood DNA (Fig. 2C), resulting in the reclassification of patient 198 as LS. This finding highlights one of the drawbacks of former genetic testing strategies, which targeted the candidate MMR gene according to the IHC pattern. More recently, the advent of NGS technologies has resulted in the implementation of multi-gene panels in most diagnostic workflows, allowing simultaneous analysis of all MMR genes in patients with suspected LS. Nevertheless, the analysis of PMS2 using NGS-based approaches still poses challenges in the diagnostic setting due to the presence of highly homologous pseudogenes¹⁴.

It is worth mentioning that our tumoral mutational screening has not addressed copy-number variants (CNVs) due to technical limitations. The use of gene panels specifically designed for FFPE samples could increase the rate of double somatic hits by assessing CNVs, LOH and other mechanisms of somatic inactivation unexplored in our series.

Paired somatic-germline testing has been repeatedly suggested as a useful approach to increase the diagnostic yield by identifying germline PV causing LS as well as biallelic MMR somatic mutations in LLS tumors. This strategy also holds potential for detecting somatic mosaicism, although mosaic mutations in MMR genes appear to be rare^15,16.

As our work demonstrates, interpretation of tumor analyses in individuals with LLS is controversial, and no consensus has been reached regarding the clinical management of these individuals and their first-degree relatives. Double somatic MMR inactivation could explain up to one third of MMRd tumors of our series. It seems reasonable to recommend follow-up upon family history of CRC in these cases, according to the current guidelines^17–19. In unexplained LLS cases, some studies advocate for intermediate risk colorectal surveillance protocols^2,3,18, while others argue that they should be approached as potential LS patients^19,20.

In the era of multi-gene testing, a germline cause is ruled out with high certainty when non-informative results are obtained. This fact, added to the difficulties in the interpretation of somatic analyses, the wide variety in the reported results, and the lack of standardized protocols, poses an interrogation in the current utility of MMR tumor testing in the routine diagnostic algorithm of suspected LS cases.

Acknowledgements

We thank CERCA Program / Generalitat de Catalunya for their institutional support. We also wish to thank all patients, families and members of the ICO Hereditary Cancer Program.

Grant support

Supported by Carlos III National Health Institute and Ministerio de Ciencia e Innovación, funded by FEDER funds – a way to build Europe [PI19/00553; PID2019-111254RB-I00]; CIBERONC [CB16/12/00234]; and the Government of Catalonia (Secretariat for Universities and Research of the Department of Business and Knowledge; grant 2021SGR01112).

Conflict of interest

Authors declare no conflict of interest.

Contributorship statement

P.R., N.D., J.V., C.L. and M.P. conceived, designed, and planned the study; P.R., J.V., C.L. and M.P. contributed to the acquisition, analysis, and/or interpretation of the molecular data; N.D., M.N., Ju.Ba., T.R.C., N.T., C.C., Jo.Br. and G.C. provided samples and clinical data; P.R., N.D. and M.P. drafted the manuscript. All authors critically reviewed the manuscript for intellectual content and approved the final version.

Data Availability Statement

All the data generated in this study can be found within the published article and its supplementary files.

Ethical Approval Statement

Informed written consent for both diagnostic and research purposes was obtained from all patients. The study protocol was approved by the ethics committee of Bellvitge Biomedical Research Institute (IDIBELL; PR278/19).

Idos G, Valle L. Lynch Syndrome. GeneReviews(®). Published online February 4, 2021. Accessed November 14, 2021. https://www.ncbi.nlm.nih.gov/books/NBK1211/
Yurgelun MB, Hampel H. Recent Advances in Lynch Syndrome: Diagnosis, Treatment, and Cancer Prevention. Am Soc Clin Oncol Educ Book. 2018;38(38):101-109. doi:10.1200/EDBK_208341
Rodríguez-Soler M, Pérez-Carbonell L, Guarinos C, et al. Risk of cancer in cases of suspected lynch syndrome without germline mutation. Gastroenterology. 2013;144(5). doi:10.1053/J.GASTRO.2013.01.044
Buchanan DD, Rosty C, Clendenning M, Spurdle AB, Win AK. Clinical problems of colorectal cancer and endometrial cancer cases with unknown cause of tumor mismatch repair deficiency (suspected Lynch syndrome). Appl Clin Genet. 2014;7:183-193. doi:10.2147/TACG.S48625
Plaschke J, Krüger S, Jeske B, et al. Loss of MSH3 protein expression is frequent in MLH1-deficient colorectal cancer and is associated with disease progression. Cancer Res. 2004;64(3):864-870. doi:10.1158/0008-5472.CAN-03-2807
Castillejo A, Vargas G, Castillejo MI, et al. Prevalence of germline MUTYH mutations among Lynch-like syndrome patients. Eur J Cancer. 2014;50(13):2241-2250. doi:10.1016/J.EJCA.2014.05.022
Elsayed FA, Kets CM, Ruano D, et al. Germline variants in POLE are associated with early onset mismatch repair deficient colorectal cancer. Eur J Hum Genet. 2015;23(8):1080-1084. doi:10.1038/EJHG.2014.242
Sutcliffe EG, Bartenbaker Thompson A, Stettner AR, et al. Multi-gene panel testing confirms phenotypic variability in MUTYH-Associated Polyposis. Fam Cancer. 2019;18(2):203-209. doi:10.1007/S10689-018-00116-2
Dámaso E, González-Acosta M, Vargas-Parra G, et al. Comprehensive Constitutional Genetic and Epigenetic Characterization of Lynch-Like Individuals. Cancers (Basel). 2020;12(7):1-32. doi:10.3390/CANCERS12071799
Rofes P, González S, Navarro M, et al. Paired Somatic-Germline Testing of 15 Polyposis and Colorectal Cancer–Predisposing Genes Highlights the Role of APC Mosaicism in de Novo Familial Adenomatous Polyposis. The Journal of Molecular Diagnostics. 2021;23(11):1452-1459. doi:10.1016/J.JMOLDX.2021.07.024
Munté E, Feliubadaló L, Pineda M, et al. vaRHC: an R package for semi-automation of variant classification in hereditary cancer genes according to ACMG/AMP and gene-specific ClinGen guidelines. Alkan C, ed. Bioinformatics. Published online March 14, 2023. doi:10.1093/BIOINFORMATICS/BTAD128
Vargas-Parra GM, González-Acosta M, Thompson BA, et al. Elucidating the molecular basis of MSH2-deficient tumors by combined germline and somatic analysis. Int J Cancer. 2017;141(7):1365-1380. doi:10.1002/IJC.30820
Kuzbari Z, Bandlamudi C, Loveday C, et al. Germline-focused analysis of tumour-detected variants in 49,264 cancer patients: ESMO Precision Medicine Working Group recommendations. Ann Oncol. 2023;34(3):215-227. doi:10.1016/J.ANNONC.2022.12.003
De Vos M, Hayward BE, Picton S, Sheridan E, Bonthron DT. Novel PMS2 pseudogenes can conceal recessive mutations causing a distinctive childhood cancer syndrome. Am J Hum Genet. 2004;74(5):954-964. doi:10.1086/420796
Pastrello C, Fornasarig M, Pin E, Berto E, Pivetta B, Viel A. Somatic mosaicism in a patient with Lynch syndrome. Am J Med Genet A. 2009;149A(2):212-215. doi:10.1002/AJMG.A.32620
Geurts-Giele WR, Rosenberg EH, Rens A van, Leerdam ME van, Dinjens WN, Bleeker FE. Somatic mosaicism by a de novo MLH1 mutation as a cause of Lynch syndrome. Mol Genet Genomic Med. 2019;7(7). doi:10.1002/MGG3.699
van Leerdam ME, Roos VH, van Hooft JE, et al. Endoscopic management of Lynch syndrome and of familial risk of colorectal cancer: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy. 2019;51(11):1082-1093. doi:10.1055/A-1016-4977
Stjepanovic N, Moreira L, Carneiro F, et al. Hereditary gastrointestinal cancers: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up†. Ann Oncol. 2019;30(10):1558-1571. doi:10.1093/ANNONC/MDZ233
Monahan KJ, Bradshaw N, Dolwani S, et al. Guidelines for the management of hereditary colorectal cancer from the British Society of Gastroenterology (BSG)/Association of Coloproctology of Great Britain and Ireland (ACPGBI)/United Kingdom Cancer Genetics Group (UKCGG). Gut. 2020;69(3):411-444. doi:10.1136/GUTJNL-2019-319915
Nugroho PP, Ghozali SAS, Buchanan DD, Pisano MI, Reece JC. Risk of cancer in individuals with Lynch-like syndrome and their families: a systematic review. J Cancer Res Clin Oncol. 2023;149(1):25-46. doi:10.1007/S00432-022-04397-0

There is no duality of interest

SupplementalFigure128072023.tif
Supplemental Figure 1: Diagram displaying the individuals included in the reported LLS cohort. One hundred and five cases were previously described (see reference ), and 124 new cases were included in the present study after a thorough revision of newly identified cases since the previous publication.
SupplementalTable101082023scaled.xlsx
SupplementalTable201082023scaled.xlsx
SupplementalTable301082023scaled.xlsx

Download PDF

Version 1

posted

You are reading this latest preprint version

Tumor analysis of MMR genes in Lynch-like syndrome: challenges associated to results interpretation

Status:

Version 1

Abstract

Figures

Introduction

Patients and Methods

Patients

Tumor testing

Results and Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1