If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, IsraelRecanati Genetics Institute, Beilinson Hospital, Rabin Medical Center, Petach Tikva, Israel
Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, IsraelSusanne Levy Gertner Oncogenetics Unit, The Danek Gertner Institute of Human Genetics, Chaim Sheba Medical Center, Tel-Hashomer, Israel
Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, IsraelGastroenterology Institute, Beilinson Hospital, Rabin Medical Center, Petach Tikva, Israel
Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, IsraelGastroenterology Institute, Beilinson Hospital, Rabin Medical Center, Petach Tikva, Israel
Gastroenterology Institute, Rambam Health Care Campus, Haifa, IsraelThe Ruth and Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel
Routine screening for establishing Lynch syndrome (LS) in young individuals diagnosed with adenomas is not recommended due to its low yield, and limited sensitivity of the employment of immunohistochemistry for DNA mismatch-repair proteins on polyps. Hence we aimed to evaluate the yield of germline mutational analysis in diagnosis of LS in a young Israeli cohort with colorectal adenomatous polyps.
Methods
Data were retrospectively collected on consecutive patients, age ≤ 45 years, who underwent colonoscopy with removal of at least one adenoma during 2015–2020, and subsequently genetic testing by multigene panel or LS-Jewish founder mutation panel.
Results
Overall, 92 patients were included (median age 35 years, range 23–45 years), of whom 79 (85.8%) underwent multigene panel genotyping, and 13 (14.2%) analysis for Jewish founder LS gene mutations. Altogether, 18 patients were identified with pathogenic mutations in actionable genes, including LS-associated genes in 6 (6.5%), BRCA2 in 2 (2.5%), GREM1 in 1(1.2%), and low-penetrance genes- APC I1307K and CHEK2- in 9 (11.4%) patients. Compared with non-LS patients, LS-carriers had a significantly higher median PREMM5 score (2.6 vs. 1.3; P = 0.04).
Conclusions
Young individuals diagnosed with adenomatous polyps should be offered genetic testing when fulfilling clinical guidelines for LS, but weight should also be given to adenoma characteristics in the PREMM5 score.
Lynch syndrome (LS), the most common hereditary colorectal cancer (CRC) syndrome, is clinically hallmarked by early age at CRC diagnosis, increased risk for extracolonic malignancies, and rapid adenoma-to-carcinoma sequence transition compared to sporadic CRC [
]. These features emphasize the importance of identifying cases at young age in order to offer an early surveillance strategy to facilitate cancer prevention and cascade screening. While microsatellite instability (MSI) and immunohistochemistry (IHC) for somatic loss of DNA mismatch repair (DNA-MMR) protein expression can help in identifying LS-associated tumors and patients who require evaluation [
], a definite diagnosis of LS is made by identifying germline mutations in the relevant genes, mostly DNA MMR genes. Current guidelines recommend universal tumor screening by MSI and/or IHC analysis in all CRC cases diagnosed by age 70 years, or fulfilling Bethesda (BT) guidelines or Amsterdam criteria (AC) at any age [
The diagnostic paradigm for LS analysis in young (≤ 45 yo) patients with pre-malignant adenomatous polyps is less clear. Since these adenomas, especially the more advanced (size ≥ 10 mm, villous histology, or harboring high-grade dysplasia) are a surrogate marker for CRC [
], patients could greatly benefit from being diagnosed as LS, as the surveillance scheme has proven efficient in prevention and early detection of CRC, and in lowering morbidity and mortality [
], similar analyses targeting benign adenomas in cases with an established LS gene mutation have only shown loss of IHC staining in 50%–84% of the lesions, depending on size and degree of dysplasia [
]. In a recent meta-analysis, characteristics of advanced adenomas significantly increased deficient MMR/MSI prevalence in LS patients from 60 to 65% to 81–88% [
]. Several studies that evaluated the yield of MSI/IHC screening of adenomas removed from young patients (< 50 yo) without known LS, reported low yield for detecting LS (0%−2.4%) [
]. However, these studies had several inherent limitations, such as substantial inclusion of patients with low-risk adenomas (LRA), small sample size, and no germline mutational analysis for LS. Since LS cannot be excluded on the basis of normal MMR staining in adenomas, the true percentage of young patients with polyps that represent genetically confirmed LS is unknown.
Table 1Primary (relevant) genes analyzed by a multigene hereditary cancer panel.
With recent advances in next-generation sequencing (NGS) technologies, multi-gene germline panel testing has emerged as an alternative strategy for diagnosis of hereditary cancer syndromes, in which both LS- and non-LS cancer susceptibility genes can be directly identified [
]. This study aimed to evaluate the yield of genetic testing in young (≤ 45 yo) individuals with adenomas, with either multi-gene panel or common mutation analysis, in diagnosing LS and other cancer predisposition conditions.
Patients and methods
Study population
This retrospective study was conducted in three medical centers in Israel-Sheba Medical Center (Ramat Gan), Rabin Medical Center (Petach-Tikwa), and Rambam Health Care Campus (Haifa). Electronic medical records were reviewed to identify all individuals’ ≤ 45 yo who underwent removal of at least one adenomatous polyp during 2015–2020, and subsequently genetic testing with either multigene panel or LS-Jewish founder mutation testing, with or without MMR analysis (MSI and\or IHC). Participants were recruited from attendees of cancer genetic/high-risk clinics in each participating medical center. Exclusion criteria included age < 18 yo, personal history of inflammatory bowel disease or CRC, and a known cancer-related hereditary gastrointestinal syndrome at the time of investigation. In order to avoid individuals with potential inherited polyposis syndrome, patients with multiple (≥ 10) adenomas were excluded.
Since genetic testing for LS diagnosis is not recommended or covered by the Israeli health basket, the option for genetic testing was primarily discussed with patients < 45 yo with high-risk adenomas (HRA), patients ≤ 30 yo with any adenoma, and patients at any age with adenoma and significant family history of malignancy, such as one first-degree relative (FDR) < 50 yo or at least two first- or second-degree relatives (SDR) at any age with LS-associated malignancies. HRA was defined as advanced adenoma [i.e., adenoma size ≥ 10 mm, tubulo-villous/villous (TVA/VA) histology, or those harboring high-grade dysplasia (HGD)] and ≥ 3 non-advanced adenomas (NAA's) [i.e., small tubular adenoma (TA) with low-grade dysplasia (LGD)], and LRA was defined as 1–2 NAA's.
MMR analysis from the adenoma was initially performed in some of the patients, but not used to exclude LS, and thus was complemented by a genetic panel. In all patients, no IHC staining for somatic loss of DNA- MMR proteins from a LS-associated malignancy of the patient's relative was available.
Genetic testing
Genetic testing was self-funded and carried out either by founder (common) Jewish LS mutations or inherited cancer multigene panels. Founder Ashkenazi Jews (AJ) mutations in the MMR genes include MSH2 c.1906G>C (p.Ala636Pro), MSH6 c.3984_3987dupGTCA (p.Leu1330ValfsTer12), and c.3959_3962delCAAG (p.Ala1320GlufsTer6) [
Thus, AJ and Ethiopian Jews were offered a two-stage genotyping strategy with initial resting of the predominant ethnically relevant mutations, and if negative, testing for multigene panel. Alternatively, patients were offered to directly undergo multigene panel, especially since 2017 when prices of multigene panels dramatically decreased. LS was assigned only for patients with a pathogenic or likely-pathogenic mutation in one of the MMR-genes.
Gene analysis
Founder mutation analysis was performed as reported in previous studies [
Variants were classified according to the standards and guidelines for sequence variant interpretation of the American College of Medical Genetics and Genomics (ACMG) [
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American college of medical genetics and genomics and the association for molecular pathology.
Genet Med Off J Am Coll Med Genet.2015; 17: 405-424
Pathogenic / Likely pathogenic variants were classified as having high penetrance (relative risk >4), moderate penetrance (relative risk 2–4), or low penetrance (relative risk ≤ 2); and sub classified as CRC-related or not well-established with CRC (Table 1).
Personal history of cancer - Breast cancer - Cervix - Brain (Medulloblastoma)
2 (2.1) 1 (1.1) 1 (1.1)
Family history (FDR) of cancer Any cancer - Colon cancer - Endometrial cancer - Urothelial cancer - Pancreatic cancer - Stomach cancer - Breast cancer - Other No family history
Data retrieved from the electronic medical records included demographic information, smoking habits, personal history of extra-colonic cancer, and complete family history of CRC or other cancer. Details regarding the polyp characteristics (number, size, and histology) were retrieved from colonoscopy and pathology reports. When more than one polyp was found, the more advanced was used for categorization. When both advanced polyps and ≥ 3 NAA's were found, they were both included in the calculations.
Estimates of the likelihood of identifying a germline mutation in MLH1, MSH2, MSH6, PMS2, and EPCAM were calculated for each subject using the PREMM 5 prediction model [
Categorical variables were described as frequencies and percentages, and Pearson Chi square test (or Fisher's exact test if > 20% cells had expected count < 5) were used to compare between age LS- and non-LS groups. Subject's ages and PREMM5 scores were described as continuous variables, and mean ± standard deviation (SD) or median and interquartile range (IQR) were reported and compared using Student's t-test. All statistical tests were two sided and p< 0.05 was considered as statistically significant. SPSS was used for all statistical analysis (IBM SPSS Statistics for Windows, ver. 25 (IBM corp., ARMNOK, NY, USA).
Results
Clinical characteristics
Overall, 92 eligible participants were included (Table 2), of whom 34 (36.9%) were male, with a median age of 35 (range 23–45) yo at polyp diagnosis. The most common indications for colonoscopy were family history of CRC/ polyps (31.5%), rectal bleeding, and abdominal pain (23.9% each). Four subjects had a personal history of cancer – breast (N = 2) at 32 and 33 yo, medulloblastoma (N = 1) at 27 yo, and cervix (N = 1) at 20 yo; 40 (43.5%) patients had a family history of one or more malignancies in FDR, most commonly CRC. Based on the reported family history, 4 (4.3%) families of patients fulfilled AC and 24 (26.0%) fulfilled Bethesda guidelines in at least one FDR/SDR. The median PREMM5 score was 1.4 (range 0–34.4), and 24/92 (26.0%) had PREMM5 score ≥ 2.5%, fulfilling NCCN guidelines for LS testing.
Twenty three (25.0%) patients had LRA at colonoscopy, of whom 12 (52.2%) had PREMM5 score ≥ 2.5% and another 5 (21.7%) were ≤ 30 yo. Thus 6 (26.1%) patients without PREMM5 score ≥ 2.5% or age < 30 yo were offered testing due to family history of cancer. Of the 69 (75.0%) patients with HRA, 4 (5.8%) had 3–9 NAA's only, 58 (84.0%) advanced polyp only, and 7 (10.2%) both. The most common components of advanced polyp were size ≥ 10 mm and villous histology, while high-grade dysplasia was less frequent. Twelve patients with HRA (17.4%) had PREMM5 score ≥ 2.5%.
Genetic analysis findings
Among participants, 79 (85.8%) patients were tested with multigene panel and 13 (14.2%) with founder LS mutations, either AJ (N = 12) or Ethiopian Jews (N = 1).
Altogether, 18 patients were identified with P\LP mutations in actionable genes; 6/92 (6.5%) were found to carry a pathogenic LS mutation with either the multigene or founder mutation panel, while 12/79 (15.1%) had non-LS mutations in a multigene panel (Fig. 1, Table 3). Of the LS mutations identified, there were 4 (4.3%) MSH6 mutations and 2 (2.2%) MSH2 mutations. Two of the MSH6 mutations were AJ founder (c.3984_3987dupGTCA) and one MSH2 mutation was the Ethiopian founder mutation (del exon 9–10). Of the non-LS mutations identified, 3 (3.8%) were in high-penetrance genes, including BRCA2 (N = 2) and GREM1 (N = 1), and 9 (11.3%) in low/moderate-penetrance genes, including CHEK2 (N = 3) and APC I1307K variant. None of these patients had double pathogenic mutations.
Fig. 1The proportion of pathogenic mutations identified with either a multigene panel and a founder mutation panel among 92 young individuals with adenomatous polyps.
Mismatch repair analysis by IHC was initially performed from the adenoma in seven patients with advanced adenoma (TVA-HGD). Six adenomas were MMR-proficient and one showed MLH1-PMS2 loss of staining and normal BRAF staining. None of the patients had mutations in multi-gene panel analysis.
Clinical findings in pathogenic mutation carriers
Of the 6 LS-carriers, 2 (33.3%) were male, 3 (50%) AJ, and 1 (16.6%) of Ethiopian descent (Table 3 and 4). The median age at polyp detection was 35 yo. Based on family history, 1 (16.7%) patient fulfilled AC, 3 (50%) Bethesda guidelines in FDR/SDR, and 5 (83.3%) had either PREMM5 score ≥ 2.5% or significant family history of CRC. Two (33.3%) patients with advanced polyps at 24 yo and 35 yo did meet NCCN criteria for LS testing. High-risk adenoma at colonoscopy was detected in 4 (66.7%) patients, 1 (16.7%) had NAA's, and 1 (16.7%) had NAA and an advanced duodenal polyp.
Table 4Comparison of clinical findings between Lynch carriers and non-Lynch patients.
The two patients with BRCA2 mutations were non-AJ, had NAA's on colonoscopy at 30 yo and 38 yo, and despite LS- and BRCA-associated malignancies in their families, did not fulfill NCCN criteria for genetic testing for LS or Hereditary Breast Ovarian Cancer (HBOC) syndrome. The patient with GREM1 40-kb dup had eight polyps detected during his colonoscopy with diverse pathology (tubular / villous adenomas and hyperplastic polyps) and no family history of cancer.
All patients with APC I1307K (c.3920T>A) variant were AJ, with a median age of 39 yo at polyp detection; 5/6 (83.3%) patients had advanced polyp, and only 1 (16.7%) had PREMM5 score ≥ 2.5%. Likewise, all three patients with CHEK2 mutations were AJ, 2 (66%) of which had advanced polyp, and the third NAA but with PREMM5 score ≥ 2.5%.
None of the four patients with personal history of cancer, including the two breast cancer patients diagnosed at early age, had pathogenic mutations.
Compared to non-LS patients, the 6 LS-carriers had a significantly higher median PREMM5 score (2.6 vs. 1.3; P = 0.04) and proportion of patients with PREMM5 ≥ 2.5% (66.7% vs. 23.2%;P = 0.19) (Table 4). Demographics, AJ origin as well as polyp characteristics (proportion of LRA and HLA), were not different between the two groups.
Discussion
In the current study, germline mutations in LS associated genes were noted in 6.5% of patients ≤ 45 yo with adenomatous colonic polyps. Although most (95.7%) of our patients did not fulfill clinical criteria for LS, LS-carriers had significantly more family history of cancer than non-LS carriers, reflected by PREMM5 score.
Clinically genotyping for LS associated mutations in patients with colorectal adenomas diagnosed at ≤ 45 yo is justified due to the well-established value of adhering to an early detection scheme in LS, and as these adenomas (especially the more advanced) are known to be surrogate markers for CRC [
]. In addition, there is a high probability that advanced lesions at this age could be developed to cancer, thus fulfilling BT guidelines for clinical diagnosis of LS. Several studies evaluated the yield of MSI/ IHC analysis for defining LS in young patients with colorectal adenomas with no additional LS suggestive features. Velayos et al. evaluated 40 adenomas from 34 patients < 40 yo, and none demonstrated MSI or loss of MMR staining [
] of young adenoma cases, both had clinical histories suggestive of LS. In a recent meta-analysis, deficient MMR was present in 2.8% of adenomas from unselected patients, but most of the studies were not limited to young age group [
], 3/125 (2.4%) patients <45 yo with advanced adenoma screened for by IHC and MSI analysis were diagnosed with LS, of whom two fulfilled AC or BT guidelines. The higher rate of LS diagnosed herein (6.5%) can be explained by the performance of mutation analysis in our study rather than IHC, which is known to have a sensitivity of 50–84% in adenomatous polyps from LS carriers. In addition, higher proportion of patients with HRA (75% in this study), and possibly family history of LS-associated malignancies, as reflected by the reported PREMM5 scores, can explain the higher yield for LS detection.
Our findings support those studies in concluding that reflex testing for LS all young individuals with adenoma is likely not cost-effective, and selecting patients based on family history or prediction models for LS might be a more effective strategy. Despite the accelerated adenoma-carcinoma sequence in LS and their tendency to develop advanced neoplasia in young age [
], neither polyp characteristics nor age of polyp detection were significantly different between LS- and non-LS carriers. However, this could be related to selection bias, since genetic testing was originally offered to either very young (age ≤ 30 yo) patients or individuals with advanced polyps, so most patients with NAA's were probably not tested. Moreover, since 2/6 LS carriers did not fulfill the clinical criteria for LS testing, both with advanced polyps at the age of ≤ 35 yo, we think that young age of onset and adenoma characteristics (i.e., advanced adenoma), and not only malignancy-related features, should be taken into consideration and incorporated in the PREMM5 score. This issue should particularly be considered in high-prevalent MSH6 mutation populations, like AJ's [
]. Hence, PREMM5 score has limited value to identify those carriers. Indeed, 4/6 of LS-carriers in this study were MSH6 mutation carriers.
In the course of multigene panel testing, additional mutations in other high penetrance genes were noted. Notably, two carriers of a non-Jewish founder mutation in the BRCA2 were detected, both in non AJ's. The rate of BRCA2 mutations in our cohort (2/79, 2.5%) is substantially higher than the expected prevalence in the general non-AJ population (0.22%–0.33%) [
] findings that BRCA1/2 mutations were the most common non-LS mutations among patients with suspected LS undergoing multigene panel testing, suggesting that a subset of BRCA1/2 mutation carriers may have atypical phenotype that can mimic LS or predispose individuals to colon cancer.
One patient had a GREM1 pathogenic mutation, which is a 40-kb tandem duplication spanning at the region upstream of the GREM1 gene. This mutation has been reported to segregate in several AJ families with hereditary mixed polyposis syndrome (HMPS) and was present in 40 affected individuals within these families, while absent in 50 of the unaffected family members [
Hereditary mixed polyposis syndrome is caused by a 40-kb upstream duplication that leads to increased and ectopic expression of the BMP antagonist GREM1.
]. The syndrome is associated with adult-onset predisposition to various types of colon polyps, rapid progression to advanced polyps and subsequent colon cancer [
]. However, the lifetime cancer risks are currently not well established.
One of the downsides of multigene panel testing is the identification of low/moderate cancer susceptibility genes with questionable clinical value. The most commonly detected low risk variant in the current study was the heterozygous variant [c.3920T>A, p.I1307K, in the APC gene. This variant is associated with a moderately increased CRC risk (OR= 2.17) in AJs, and to a lesser extent in non-AJs (OR=1.36). This variant is found in ~ 6%–10% of AJ's and ~ 3% of non- AJ individuals, and up to 28% in AJs with a family history of CRC [
]. The rates of this variant in the current study [6/79 (7.6%) of all participants and 6/42 (14.2%) in AJs] is most probably attributed to simply detecting their background population prevalence. In practice, the need to genotype for this variant in young AJs with polyps is equivocal, since CRC is not diagnosed at an earlier age in carriers of this variant and there were no consistent reports of its association with early diagnosed colonic polyps [
] to perform colonoscopy from the age 40 yo at intervals of 5 years, regardless of negative results in surveillance colonoscopies.
Likewise, germline mutations in the cell cycle checkpoint kinase 2 (CHEK2) genes are associated with increased risk of breast cancer and CRC, though heterogeneity may exist based on type of pathogenic mutation [
] thus recommend similar management strategies as described for carriers of the APC I1307K mutation.
The foremost limitations of this study are the small cohort and potential for selection bias in choosing patients for genetic testing. As noted, although primarily discussed with patients ≤ 45 yo with HRA or ≤ 30 yo with any adenoma (93.5% of our study population), the option for genetic testing was available for all patients with adenomas since it was self-funded. If more patients with significant family history of CRC were selected or chose to do genetic testing, this could have produced a higher-than expected rate of detection of mutations. However, only one-third of our study population had a family history of CRC and a quarter had PREMM5 score ≥ 2.5%, confirming that clinical parameters rather than family history were primarily used for selection. Another limitation is the lack of performance of MSI/IHC analysis on the majority of adenomas before genetic testing, hence we could not determine the added value of multigene panel. Future research is needed in this population to define the yield of multigene panel testing after normal MSI/IHC analysis.
In conclusion, despite the higher than the previously reported rate of identified LS patients, it seems reasonable to conclude that routine screening with genetic testing for LS in individuals with adenoma diagnosed at ≤ 45 yo should primarily be recommended in the presence of features suggestive of inherited predisposition to cancer. However, adenoma characteristics, specifically features of advanced adenoma, should also be taken into consideration to be integrated in the PREMM5 score. Additionally, panel testing can potentially diagnose some “incidental” high-penetrance germline mutations that have profound clinical implications on the patient in terms of cancer prevention, while detecting other low/moderate- penetrance gene mutations, that while posing a clinical dilemma regarding their significance, have some clinical yield. In the evolving era of multigene panel testing, these clinical issues will become increasingly more common for practicing clinicians.
Declaration of Competing Interest
None
Acknowledgments
We thank Dr. Debby Mir (Rambam Health Care Campus, Haifa, Israel) for writing assistance. The authors state that LI is the guarantor of this article; LI, LRG, HE performed the research; GY, LRG, FD and FE C collected and analysed the data; KL, GBR, KI, LZ and KU contributed to the design of the study and manuscript review.
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American college of medical genetics and genomics and the association for molecular pathology.
Genet Med Off J Am Coll Med Genet.2015; 17: 405-424
Hereditary mixed polyposis syndrome is caused by a 40-kb upstream duplication that leads to increased and ectopic expression of the BMP antagonist GREM1.
00305-4&id=gr1.jpg){kind=link}