Lynch-like syndrome is as frequent as Lynch syndrome in early-onset nonfamilial nonpolyposis colorectal cancer

Abstract

Early-onset (<50 years-old) nonpolyposis nonfamilial colorectal cancer (EO NP NF CRC) is a common clinical challenge. Although Lynch syndrome (LS) is associated with EO CRC, the frequency of this syndrome in the EO NF cases remains unknown. Besides, mismatch repair deficient (MMRd) CRCs with negative MMR gene testing have recently been described in up to 60% of cases and termed “Lynch-like syndrome” (LLS). Management and counseling decisions of these patients are complicated because of unconfirmed suspicions of hereditary cancer. To define the prevalence of MMR deficient CRCs, LS and LLS in patients with EO NP NF CRC, we recruited 102 patients with a first diagnosis of NP NF CRC ≤ 50 years old during 2003–2009 who underwent genetic counseling at our institution in Argentina. Tumor immunohistochemical (IHC) MMR for protein expression and microsatellite instability (MSI) status were evaluated, and in those with loss of MLH1 expression by IHC, somatic BRAF V600E mutation and both somatic and germline MLH1 methylation levels were studied. Tumors characterized as MMRd without somatic BRAF mutation nor MLH1 methylation were sent for germline analysis. Twenty one (20.6%) tumors were MMRd. Fourteen of 16 putative LS cases underwent germline testing: 6 pathogenic mutations were identified and 8 cases had no identifiable pathogenic mutations. The prevalence of LS and LLS in this cohort was 5.8% (6/102) and 7.8% (8/102), respectively. As a conclusion we found that 20% of patients with EO NP NF CRC have MMRd tumors, and at least half of these are likely to have LLS.

Introduction

Colorectal cancer (CRC) is the second most commonly diagnosed type of cancer and the third leading cause of cancer-related deaths in most developed countries; age is the main risk factor.¹ However, up to 10–15% of all cases of CRC occur before the age of 50,² and recent epidemiological studies suggest the incidence of early-onset CRC is increasing.^3–7 Several hereditary and nonhereditary diseases (Lynch syndrome, familial adenomatous polyposis, MUTYH associated polyposis, ulcerative colitis, Crohn’s disease) are associated with early-onset CRC, and early detection is essential because of the increased lifetime CRC risk and the potential impact of preventive measures on survival.⁸ Inflammatory bowel disease and colonic polyposis syndromes are easily identified by their phenotypic features, but Lynch syndrome (LS) patients do not have a characteristic clinical phenotype and are often missed, especially in the absence of a family history of cancer. Although early-onset nonfamilial nonpolyposis CRC is a prevalent clinical challenge, the frequency of LS in this setting has not been thoroughly studied.

LS is the most common hereditary cause of CRC, accounting for approximately 1–4% of all cases.⁹ It is an autosomal dominant condition caused by germline mutations in a DNA mismatch repair (MMR) gene. MSH2 and MLH1 account for most of the LS-associated CRCs, but PMS2 and MSH6 mutations are actually more prevalent on a population basis. This syndrome has a marked gene-dependent variable penetrance for CRC and endometrial carcinoma (30–80%), and an increased risk for various other extra-colonic tumors.¹⁰ The diagnosis of LS predicts the natural course of the disease. Annual surveillance colonoscopies and total hysterectomy reduce the mortality related to cancer.^11,12 Additionally, the identification of the causal mutation in one of the MMR genes leads to genetic presymptomatic diagnosis in relatives, and focuses screening measures on mutation carriers.

While the hallmark of this disease is tumor MMR deficiency (MMRd), defined by the presence of microsatellite instability (MSI) and/or absence of MMR protein expression by immunohistochemistry (IHC), a diagnosis of LS requires the presence of a deleterious germline mutation in a DNA MMR gene.¹³ MMRd tumors without a germline mutation in any of the four MMR genes may be as common as 70%.¹⁴ These cases are termed “Lynch-like syndrome” (LLS), and management decisions in these patients and their families are complicated because of unconfirmed suspicions of hereditary cancer. They typically develop MMR deficient CRC not related to methylation of MLH1 or the serrated pathway of carcinogenesis.¹⁵ In addition to the hypothesis regarding possible “cryptic” germline mutations in MMR genes, recent reports suggest that 50–60% of LLS tumors present biallelic somatic inactivation of a MMR gene.^16–19 Studies in this population are needed to assess familial risk and so that appropriate surveillance measures may be developed.

In this study we analyzed a large and well-annotated cohort of early-onset nonfamilial nonpolyposis CRC patients to determine the frequency of LS and LLS with a comprehensive analysis.

Methods

Patients

We retrospectively recruited 102 patients with a first diagnosis of CRC ≤50 years old during 2003–2009 who underwent genetic counseling at the Oncology Unit of the Argentine Hospital of Gastroenterology “Dr. C. B. Udaondo” in Buenos Aires. Patients had no family history of CRC or other Lynch syndrome associated-neoplasia among first or second-degree relatives, and those with >15 polyps and/or inflammatory bowel disease were excluded.

Demographic and clinical-pathological features were obtained from each patient’s medical record, and family history of cancer in first and second-degree relatives was updated by a personal interview.

This study was approved by the Institutional Review Board, and patients signed a protocol-specific informed consent. CRC tissue samples and germline DNA were obtained from each patient.

DNA isolation

Genomic DNA from each patient was extracted from formalin-fixed paraffin-embedded (FFPE) microdissected tumor tissues using the QiaAmp Tissue Kit (Qiagen, Courtaboeuf, France) according to the manufacturers’ instructions. Peripheral blood DNA was extracted using the QiaAmpDNA blood Mini Kit (Qiagen, Courtaboeuf, France).

Tumor MMR protein expression

One block of FFPE tumor tissue was selected per case and immunostaining was performed using standard protocols. The following mouse monoclonal antibodies were used: anti-MLH1 (BD PharMingen, San Diego, CA), anti-MSH2 (Oncogene ResearchProducts, Cambridge, MA), anti-MSH6 (Serotec, Raleigh, NC), and anti-PMS2 (BD PharMingen, San Diego, CA). A tumor was considered negative for protein expression only if the neoplastic epithelium lacked staining, while healthy colonocytes and/or stromal cells retained the expression of that protein.

Tumor MSI analysis

MSI analysis was carried out using five mononucleotide repeat microsatellite targets (BAT-25, BAT-26, NR-21, NR-24 and NR-27) in a pentaplex PCR system. Primer sequences have been described previously.²⁰ Tumors with instability at ≥2 of these markers were classified as microsatellite unstable (MSI) and those showing instability at ≤1 of these markers were classified as microsatellite stable (MSS). The pathologists scoring immunostaining were blinded to the MSI results, and IHC results were unknown to MSI analysis researchers.

Somatic BRAF V600E mutation analysis

In CRC displaying IHC loss of MLH1 or MLH1/PMS2 protein expression, BRAF V600E mutational analysis was performed by PCR and sequencing as previously reported.²¹

Methylation analyses

In tumors displaying lack of expression of MLH1 or MLH1/PMS2 in the IHC, tumor DNA was modified with sodiumbisulfite using the EZ Methylation Gold Kit (Zymo Research, Orange, CA), and the methylation status of MLH1 was analyzed by quantitative bisulfite pyrosequencing as previously described.²² Based on our previous research, a tumor was considered as methylated at theMLH1promoter when the level of methylation was>12% was found. An analysis of germline MLH1 methylation to rule out a constitutional epimutation²³ was conducted in the subgroup of patients with BRAF wild type tumors displaying MLH1 methylation.

Germline MMR mutational analysis

Tumors with MSI and/or lacking protein MMR expression that also had wild-type BRAF and no MLH1 methylation underwent germline genetic testing of the MMR genes. Peripheral blood samples were sent to Myriad Genetic Laboratories and to the Diagnostic Biomedic Centre at the Hospital Clinic of Barcelona, where mutational analyses of MLH1, MSH2, MSH6, PMS2, and EPCAM were conducted. Assay methods were identical to those used for clinical testing, including full sequencing of all coding regions and exonintron boundaries. Large rearrangement testing for duplications and deletions of MLH1, MSH2, MSH6, PMS2 and the clinically relevant 3′ region of the EPCAM gene were performed. All sequence variations and large rearrangements detected were classified for pathogenicity into the following categories: deleterious mutation, suspected deleterious mutation, variants of uncertain clinical significance (VUS), favor polymorphism, and polymorphism.

Germline MUTYH gene mutation analysis

Patients with MMR deficient tumors and negative test for MMR genes were screened for MUTYH mutations by complete sequencing of all MUTYH exons and intron-exon boundaries.²⁴

Statistical analysis

Data were analyzed using SPSS v17 software. Quantitative variables were analyzed using the Student’s t test. Qualitative variables were analyzed using either the Chi Square test or the Fisher’s test when appropriate. Overall survival associated with MMR deficiency was calculated by using the Kaplan–Meier method (log rank test). A two sided p-value of <0.05 was regarded as significant.

Results

Clinical characteristics and demographics

One hundred two patients under 50 years of age with a primary nonfamilial nonpolyposis CRC were assessed. Clinical, histopathological and MMR characterization of this cohort are summarized in Table 1. The mean age of the patients at diagnosis was 37.29 years (SD 8.36); 51 were women (50%). Twenty-six (25.5%) tumors were located proximal to the splenic flexure, 33 (32.4%) were located in the colon distal to the splenic flexure, and 43 were in the rectum (42.2%). The majority of cases (65; 63.7%) were diagnosed at advanced stages (III-IV). Poorly differentiated tumors were seen in 13 (12.7%) patients, 33 (33.3%) had mucinous features and 54 (52.9%) had pathological features of MSI.

Table 1.

Clinical, pathological and molecular features of patients with mismatch repair deficiency

Clinical, pathological or molecular feature	Cohort (n = 102)	MMR deficient1 (n = 21) (20.6%)	MMR proficient2 (n = 81) (79.4%)	p Value
Age at diagnosis, y
mean (standard deviation)	37.29 (8.36)	34.52 (10.47)	38.01 (7.63)	0.088
Age range	(12–50)	(12–49)	(17–50)
Sex, n(%)
Female	51 (50.0)	11 (52.4)	40 (49.4)	0.807
Male	51 (50.0)	10 (47.6)	41 (50.6)
Tumor location, n (%)
Rectum	43 (42.2)	4 (19.0)	39 (48.1)	0.001
Distal to splenic flexure	33 (32.4)	5 (23.8)	28 (34.6)
Proximal to splenic flexure	26 (25.5)	12 (57.1)	14 (17.3)
Synchronous or metachronous CRC, n(%)
Yes	4 (3.9)	2 (9.5)	2 (2.5)	0.187
No	98 (96.1)	19 (90.5)	79 (97.5)
Synchronous adenomas, n(%)
0	69 (67.6)	13 (61.9)	56 (69.1)	0.553
1–5	17 (16.7)	5 (23.8)	12 (14.8)
No or Incomplete colonoscopy	14 (13.7)	2 (9.5)	12 (14.8)
TNM tumor stage, n(%)
I-II	37 (36.3)	11 (52.4)	26 (32.1)	0.085
III-IV	65 (63.7)	10 (47.6)	55 (67.9)
Tumor differentiation, n(%)
Well or moderate	89 (87.3)	18 (85.7)	71 (87.7)	0.728
Poor	13 (12.7)	3 (14.3)	10 (12.3)
Mucinous component, n (%)
>50%	34 (33.3)	14 (66.7)	20 (24.7)	<0.001
<50%	68 (66.7)	7 (33.3)	61 (75.3)
Tumorinfiltrating lymphocytes, n (%)
Yes	26 (25.5)	16 (76.2)	10 (12.3)	<0.001
No	76 (74.5)	5 (23.8)	71 (87.7)
Medullary growth pattern, n (%)
Yes	9 (8.8)	3 (14.3)	6 (7.4)	0.386
No	93 (91.2)	18 (85.7)	75 (92.6)
TumorswithCrohn’sreaction, n (%)
Yes	10 (9.8)	5 (23.8)	5 (6.2)	0.029
No	92 (90.2)	16 (76.2)	76 (93.8)
Pathology suggestive of MSI3, n (%)
Yes	54 (52.9)	19 (90.5)	35 (43.2)	<0.001
No	48 (47.1)	2 (9.5)	56.8)

¹MSI and/or loss of expression of MMR proteins by immunohistochemistry.

²MSS and normal expression of MMR proteins by immunohistochemistry.

³Signet ring cells and/or Crohn’s-like lymphocytic reaction and/or tumor infiltrating lymphocytes and/or medullary growth pattern and/or anaplastic tumor.

MMR deficiency analysis

Tumor MMRd was evaluated by MSI and IHC, and defined by the presence of MSI and/or loss of IHC expression in any of the four MMR proteins. Twenty-one (20.6%) CRCs showed MMRd. The multivariate analysis showed that the independent variables associated with MMRd were: CRC < 31 years old (OR = 5.16; 95%CI: 1.29–20.66; p = 0.020), tumor proximal to splenic flexure (OR = 4.09; 95%CI: 1.26–13.27]; p = 0.019), and MSI suggestive pathology (OR = 13.38; 95% IC: 2.45–73.08]; p = 0.003). Age at diagnosis of CRC was dichotomized selecting the cut-off point of 31 years old based on ROC (receiving operating characteristics) curve analysis (Table 2).

Table 2.

Clinico-pathological features associated with mismatch-repair deficiency

Variable, n (%)	MMR deficient1	MMR proficient2	Adjusted OR (95% CI)	p Value
Age, y
≤31	8 (38.1)	17 (21.0)	5.16 (1.29 to 20.66)	0.020
>31	13 (61.9)	64 (79.0)
Tumor Location
Proximal to splenic flexure	12(57.1)	14(17.3)	4.09 (1.26 to 13.27)	0.019
Distal to splenic flexure	9(42.9)	67(82.7)
MSI pathology 3
Yes	19(90.5)	35(43.2)	13.38 (2.45 to 73.08)	0.003
No	2(9.5)	46(56.8)

¹MSI and/or loss of expression of MMR proteins by immunohistochemistry.

²MSS and normal expression of MMR proteins by immunohistochemistry.

³Signet ring cells and/or Crohn’s-like lymphocytic reaction and/or tumor infiltrating lymphocytes and/or medullary growth pattern and/or anaplastic tumor.

MMR analysis results are summarized in Table 3 and in Supporting Information Figure S1. Regarding the IHC results, 20 of the 21 MMR deficient tumors had loss of expression of MMR proteins: 8 (40%) failed to express MLH1 and PMS2, 1 (5%) showed isolated loss of MLH1, 8 (40%) losses of MSH2 and MSH6, and 3 (15%) isolated loss of PMS2. One of the three cases of isolated PMS2 lacked PMS2 expression both in tumor cells and in normal colonic surrounding tissue.

Table 3.

Clinical, histological, molecular and genetic features of patients with MMRd

Case	Age/sex	CRC relation to splenic flexure	TNM	MSI status	Protein loss by IHC	BRAF V600E Mutation status	MLH1 Methylation	Gene tested	Result	Interpretation1	Classification
D9	33 M	Proximal	II	MSI	MLH1/PMS2	wt	1%	MMRg + MUTYH	c.1852_1854del AAG of MLH1	Deleterious	LS
18	43 M	Proximal	II	MSI	PMS2			PMS2	c.861_864del ATTA of PMS2	Deleterious	LS
A17	15 F	Proximal	III	MSS	PMS2			PMS2	c.24-?_2589 +?del of PMS2	Deleterious	CMMR-D
B6	46 F	Proximal	II	MSI	MSH2/MSH6			MSH2	c.1840G>T of MSH2	Deleterious	LS
23	30 F	Distal	III	MSI	MSH2/MSH6			MMRg + MUTYH	c.1760–30G of MSH2	Suspected deleterious	LS
A25	37 F	Proximal	II	MSI	MLH1/PMS2	wt	2%	MMRg + MUTYH	c.l942C>T of MLH1	Deleterious	LS
A19	12 M	Distal	IV	MSI	MSH2/MSH6			MMRg + MUTYH	Negative		LLS
C13	25 F	Proximal	III	MSI	MSH2/MSH6			MMRg + MUTYH	Negative		LLS
D14	30 F	Proximal	II	MSI	MSH2/MSH6			MMRg + MUTYH	Negative		LLS
A20	40 M	Distal	III	MSI	MLH1/PMS2	wt	2%	MMRg + MUTYH	Negative		LLS
C4	31 F	Distal	III	MSI	MSH2/MSH6			MSH2, MSH6, EPCAM + PM MUTYH	Negative		LLS
24	34 M	Proximal	1	MSI	MSH2/MSH6			MSH2, EPCAM + PM MUTYH	Negative		LLS
D30	36 M	Distal	III	MSI	PMS2	wt		MLH1, PMS2 + PM MUTYH	Negative		LLS
D25	41 M	Proximal	II	MSI	MLH1	wt	1%	MLH1 + PM MUTYH	c.1852_1853delAAinsGC of MLH1	Polymorphism 29,30	LLS
C16	42 M	Proximal	II	MSI	MLH1/PMS2	wt	25%				Sporadic CRC
C12	43 M	Proximal	II	MSI	MLH1/PMS2	wt	26%				Sporadic CRC
D18	43 F	Distal	II	MSI	MLH1/PMS2	wt	27%				Sporadic CRC
D32	48 F	Distal	II	MSI	MLH1/PMS2	wt	51%				Sporadic CRC
A6	49 F	Proximal	III	MSI	MLH1/PMS2	mutated	88%				Sporadic CRC
D6	19 M	Proximal	IV	MSI	MSH2/MSH6			Not tested			MMR deficient
D29	28 F	Distal	III	MSI	Normal expression of 4 proteins			Not tested			MMR deficient

M: male, F: female, MSI: microsatellite instability, MSS: microsatellite stability, IHC: immunohistochemistry of mismatch repair proteins, −: loss of protein expression in IHC, +: normal expression of proteins in IHC, wt: wild type, MLH1 Met: methylation levels (%) of MLH1’s promoter region, MMRg: all 5 mismatch repair genes (MLH1, MSH2, MSH6, PMS2 and EPCAM), MYH: full sequencing of MUTYH, PM MUTYH: prevalent MUTYH mutations (G393D and Y176C) LS: Lynch syndrome, LLS: Lynch-like syndrome.

¹Information about variant classification is detailed in Supporting Information Table S2.

MSI was found in 20 of the 21 MMR deficient tumors; there was only 1 CRC with MSI and normal expression of MMR proteins, and 1 CRC with MSS and loss of PMS2 expression, showing a high correlation between both methods assessment of DNA MMR status (p < 0.001).

Out of 9 MLH1 deficient tumors, 1 evidenced a V600E BRAF mutation and MLH1 promoter methylation (88%). Other 4 were BRAF wild type and had somatic hypermethylation (methylation levels: 25–51%) with no germline MLH1 methylation; these were considered sporadic CRC associated with a CpG island methylator phenotype (CIMP).^22,25 Consequently, 16 of the 21 MMRd CRCs were considered putative LS cases.

DNA MMR gene mutations

Germline genetic testing was available on 14 of the 16 individuals with putative LS. Six pathogenic/suspected pathogenic mutations in a MMR gene were identified: 2 in MLH1 (c.1852_1854delAAG and c.1942C>T), 2 in MSH2 (c.1840G>T and c.1760–3C>G) and 2 in PMS2 (c.861_864delATTA and c.24-?_2589 +?del). Of these, 3 are classified as Class V Pathogenic in ClinVar and InSIGHT Database (c.1852_1854delAAG,c.1942C>T, c.861_864delATTA). Although the variant c.1840G>T is still not reported in InSIGHT nor ClinVar, it meets the PVS1, PM1, PM2, PP4, and PP5 ACMG Criteria,²⁶ therefore it can be classified as Pathogenic. Regarding the variant c.1760–3C>G in MSH2, it is also not reported in InSIGHT nor ClinVar, and it meets the PM2 and PP4 ACMG Criteria,²⁶ therefore it would be classified as a variant of uncertain significance. Nevertheless, this mutation consists of a nucleotide substitution in a noncoding intervening sequence (IVS) occurring 3 bases pairs before the beginning of the exon 12; nucelotides at this position are highly conserved and are usually necessary for proper mRNA processing,²⁷ and therefore we consider this specific mutation pathogenic. At last, the variant c.24-?_2589 +?del in PMS2 represents a large homozygous deletion of PMS2 exons 2–15, defining a Constitutional Mismatch Repair Deficiency (CMMR-D) syndrome; phenotypic and molecular characterization of this case was already reported.²⁸

An additional case with a neutral sequence variation (polymorphism) in MLH1 was detected.^29,30 Germline genetic variants are listed in Table 3, and information about ClinVar, InSIGHT, and ACMG classification and their frequency in GnomAD Database are detailed in Supporting Information Table S2.

The remaining 8 patients with MMRd tumors without germline mutation were classified as LLS. Full mutational analysis of MLH1, MSH2, MSH6, PMS2, EPCAM and MUTYH was performed in 4 of them. In the remaining 4 cases genetic testing was guided by the IHC result, and additional G393D and Y176C prevalent MUTYH mutations were analyzed (Table 3).

The overall results are summarized in Figure 1.

Figure 1.

Overall results in patients with nonfamilial, nonpolyposis CRC < 50, diagnosed between 2003 and 2009 (N = 102).

Discussion

In this study, we determined the frequency of LS and LLS in an early-onset nonfamilial nonpolyposis CRC cohort.

Twenty one patients had MMRd CRCs; of the 14 putative LS cases genetically tested, almost half had demonstrable LS (6/14, 43%). The remaining 8 were considered LLS (8/14, 57%). These LLS patients presented MSI, abnormal DNA MMR protein IHC, BRAF wild type, no MLH1 methylation, absence of pathogenic germline mutations in any of the MMR genes, and are reported to have an intermediate risk for CRC and extra-colonic cancers between the ratios for LS and sporadic CRC.¹⁵ There are at least 4 possible explanations for LLS cases: a) unknown genetic mutations other than the MMR genes in the germline that may drive MSI, b) cryptic germline mutations in the DNA MMR genes not identified by the current detection methods, c) biallelic somatic mutations in a MMR gene causing the MMRd, or d) mosaicism not detectable in lymphocyte DNA. All of these explanations might play a role.³¹ A limitation of this study is that deep sequencing of the tumor DNA to detect all somatic mutations was not undertaken.

The first explanation is based on recent publications demonstrating that MUTYH and POLE germline mutations may be causes of LLS.^32–34 We screened for MUTYH mutations in the 8 LLS patients. Full MUTYH sequencing and MLPA were done in 5 patients, and the common mutations G393D and Y176C were sought in the remaining 3. No MUTYH mutations were identified. Neither POLE nor POLD1 were analyzed, which is a limitation or our study.

The second explanation is based on our limited understanding of gene promoter function, intronic sequence variants, and some variants of unknown significance (VUS) in MMR genes. Although we performed full germline mutation analysis at two state-of the art centers, we could have missed germline mutations within the MMR genes, which would turn LLS patients into LS patients.

The third explanation related to biallelic somatic mutations in a MMR gene is based on reports suggesting this may be true for almost 70% of LLS cases.^16–18 Although we did not address this possibility, biallelic somatic mutations in this setting have been related to CRC in elderly patients. Interestingly our LLS cases were all diagnosed at young ages (12–41 years old, with a mean age of 31 at diagnosis).

Finally, mosaicism occurs when a somatic mutation occurs during embryogenesis, is passed on to some clones and not others, so mutations be absent or present in low copy number in the lymphocyte DNA. It is possible that deep sequencing of germline DNA could uncover some such mutations.

In conclusion, MMRd rates and prevalences of LS and LLS in our study are similar to the ranges reported in 21 other studies on the incidence of these syndromes in early-onset nonpolyposis CRC (Table 4).^35–50 Of note, the incidence of LS was higher when the age at CRC diagnosis used as a cutoff to define “early-onset” was lower, when there was a family history of CRC, and when the molecular algorithm for LS was complete. Comparable results of the reviewed literature and our study are summarized in Supporting Information Table S1.

Table 4.

The incidence of MMRd and MMR germline mutation carriers in people with early onset CRC reported in different cohorts

Author (year)	Age of cutoff of CRC diagnosis	MMRd/total cohort (%)	Germline mutations/total cohort (%)
Dunlop (1997) 35	<35	13/23 (56.5%)*	6/23 (26%)
Ho (2000)36	<50	33/124 (26.6%)*	8/124 (6.4%)
Terdiman (2002)37	<35	28/40 (70%)*	13/40 (32.5%)
Durno (2005)38	<25	8/11 (73%)*	6/11 (54.5%)
Niessen (2006)39	<50	50/209 (24%)***	25/281 (9%)
Goel (2010)40	<50	16/75 (21%)***	-
Giráldez (2010)41	<50	20/140 (14%)***	11/140 (8%)
Wright (2011)42	<50	33/214 (15%)**	10/214 (4.6%)
Limburg (2011)43	<50	-	11/195 (5.6%)
Steinhagen (2012)44	<50	38/198 (19%)**	10/198 (5.1%)
Tanskanen (2013)45	<40	14/38 (37%)*	12/38 (32%)
Mork (2015)46	<35	45/205 (22%)	23/205 (11%)
De Voer (2016)47	<45	-	4/45 (9%)
DeRycke (2017)48	<50	-	24/333 (7%)
Pearlman (2017)49	<50	48/450 (10%)	40/430 (9%)
Stoffel (2018)50	<50	41/204 (20%)	23/204 (11%)

MMRd: tumor MMR deficiency tested by MSI (*), IHC of MMR proteins (**), or both (***).

The methodological strengths of our study focused on early-onset CRC was the enriched homogeneous analyzed cohort. Patients were personally interviewed and had no family history of Lynch-associated tumors, no polyposis syndromes, nor inflammatory bowel disease. Although neither POLE/POLD1germline mutations nor somatic mutation or LOH status of the MMR genes in CRC tissue were analyzed in LLS patients, this study is still one of the most comprehensive of its kind.

Conclusions

Our intriguing results will be prospectively validated, but our exploratory hypothesis is that 1/5, 1/13 and 1/17 patients with early-onset, nonpolyposis, nonfamilial CRC have MMR-deficient CRC, LLS and LS, respectively. Therefore, if we consider the increasing incidence of early onset CRC, the current availability of molecular diagnostic techniques, and the paradigm-shifting approaches to treatment of LS and related hypermutated tumors with immune checkpoint therapies, every young individual with CRC should be investigated genetically, even without a family history of cancer. The detection rate of mutations in a MMR gene in this scenario may be as low as 42%, underscoring the high percentage of LLS patients. Although bi-allelic somatic inactivation may account for 50–70% of these cases that are apparently not LS, a mutation in another CRC-predisposing gene may be also the cause of at least some of these MMRd tumors, especially in the early-onset setting. The application of whole exome and whole genome sequencing studies of germline and tumoral DNA in LLS patients may contribute to our understanding of this entity and differentiate the sporadic from the hereditary CRC cases, so that we may categorize familial risk and surveillance approaches appropriately.

What’s new?

Lynch syndrome, diagnosed through a germline mutation in DNA mismatch repair (MMR) genes, is the most common hereditary cause of colorectal cancer (CRC). Lynch-like syndrome (LLS) cases with MMR deficiency but no MMR germline mutations may however be common. In this retrospective study of 102 patients with early-onset non-familial non-polyposis CRC, Lynch was as frequent as Lynch-like syndrome (5.8% and 7.8%, respectively). The data suggest that every young CRC patient should be investigated genetically, even without a family history of cancer. Although biallelic somatic inactivation may be the cause of LLS, a mutation in other CRC-predisposing genes should be considered.