Selected Abstracts from the 3rd European Hereditary Tumour Group Meeting (EHTG 2018)

Correspondence: C. Engel Aim The German HNPCC Consortium, founded in 1999, is a joint network of currently 14 clinical university centers, a reference pathology, and a central documentation facility aiming to provide structured interdisciplinary care and research for individuals suspected of having an inherited predisposition for colorectal cancer. In the past, the consortium has focused on Lynch Syndrome (LS) but aims to cover also the broad spectrum of other familial colorectal cancer entities. Method Families are ascertained based on the established Amsterdam and Bethesda criteria. Interdisciplinary care comprises genetic counseling, molecular pathological tumour analyses for mismatch repair deficiency, germline testing of predisposing genes, and structured intensified surveillance measures. Research goals are e.g. search for new disease causing genes, genotype-phenotype correlations and tumour risks, tumour immunology, and efficacy of intensified surveillance. Results The consortium has established a central research database, which is populated by the clinical centres using a web-based remote online data capture application based on standardized documentation. The scope of the retroand prospective data collection comprises fully structured pedigrees, familial tumour history, detailed results of diagnostics and results of surveillance. Conclusion Currently, approx. 8,800 individuals (patients, asymptomatic mutation carriers, relatives at risk) from 5,500 families are centrally registered (2,100 LS patients).

DNA isolated from leucocytes, buccal mucosa and urine will then be analyzed to see whether the variant can be identified in these tissues as well. Eligible patients will need to provide written consent before they are included.

Results
The main outcome will be prevalence of mosaic mutations in the above mentioned patient groups. Furthermore, mutation patterns and clinical phenotype will be recorded to study the mechanisms behind mosaicism and provide data to adapt surveillance guidelines. Conclusion A new study is starting this year at the LUMC in Leiden with the aim of further clarifying the prevalence, phenotype and clinical consequences of APC mosaicism. We invite attendees of the EHTG who have several cases that meet the selection criteria to contact us to discuss participation. We require tissue from multiple tumors from well described cases and can accommodate the NGS gene panel.

N4
Breast cancer pathology and stage are better predicted by risk stratification models including mammographic density and common genetic variants D. G. Evans 1, 2, 6, 7, 8  To better stratify breast cancer risks to enable more targeted early detection/prevention strategies particularly to balance the risks/benefits of population. Method Data from 9,362 women unaffected by breast cancer at study entry who provided a DNA sample for polygenic-risk-score (PRS) were analysed from the 57,902 women in the PROCAS study. The PRS score was analysed along with mammographic density (density residual-DR) and standard risk factors to assess future risk of breast cancer pathological type and hormonal receptor status Results For the 195 prospective breast cancers a predictor based on Tyrer-Cuzick/DR/PRS was informative for subsequent cancer overall and more so for stage 2+ cancers and calibrated (0.99) for predicting cancers across all risk groups. Although DR was most predictive for HER2+ and stage 2+ cancers it did not discriminate as well between poor prognosis cancers and extremely good prognosis cancers as Tyrer-Cuzick or the PRS, with the PRS providing the highest OR for post-prevalent stage 2+ cancers IQR OR=1.79 (95%CI:1.30-2.46). Conclusion A combined approach using Tyrer-Cuzick, mammographic density and a PRS provides accurate risk stratification not only overall but also for worse prognosis cancers. This provides support for reducing screening intervals in the high and increasing them in the low risk groups.

N5
Exogenous and endogenous associated factors to early onset colorectal cancer R. Zuppardo 1 , M. Di Leo 2 , A. Mannucci 1 , F. Azzolini 1 , D. Esposito 1 , L. Correspondence: R. Zuppardo Aim Early onset colorectal cancers (eoCRC < 50 years), is projected to increase by as much as 90% and 140%, respectively by 2030, and germline mutations appear to account for only about 20%. We investigate the role of exogenous and endogenous risk factors as associated factors in eoCRCs.

Method
Clinical, anamnestic and pathological data were retrieved on eoCRC patients from 06/2017 to 04/2018, and compared with a group of late onset CRC (loCRC) of the same period.

Results
We enrolled 33 eoCRCs and 48 loCRCs, mean age 40.7 +/-7.3 and 66.1 +/-9.8, respectively (p<0.001), prevalence of females (54.5% in eoCRCs and 52.1% in loCRCs). Diagnostic delay was higher in eoCRC group: 42.4% of eoCRCs diagnosed in the 6th months from symptoms onset versus 100% of loCRC patients (p<0.001). Lynch syndrome was more frequent in eoCRC (12%) than loCRC group (0%) p= 0.02. A statistically significant difference was found in alcohol habit, 66.7% of no-drinker in eoCRCs and 41.7% of loCRCs (p=0.04), and a trend through significance for no-smokers in eoCRCs. Conclusion CRC should be considered earlier for differential diagnosis in young patients. We confirmed alcohol as cofactor in development of eoCRC and we underlined that familiar history should be collected to identify mutations carriers. analysed by age and gender, deriving more precise risk and survival estimates to inform management.

Results
The validation cohort (N=425) provided 2,367 prospective observation years and confirmed previously reported cumulative risks for any cancer: 14% vs 18% at fifty years and 48% vs 53% at 70 years. The combined series of 841 carriers of class 4/5 path_MSH6 variants provided 5,205 prospective observation years. Cumulative risks at 75 years in males/females were: any cancer 42%/60%; colorectum 18%/20%; endometrium NA/41%; ovary NA/11%; stomach, duodenum, bileduct, pancreas 8%/4%; ureter,_kidney 2%/6%; bladder 8%/1%; prostate 9%/NA; breast NA/14%; brain 2%/1%. Ten-year crude survival following cancer was: colon 100%; rectum 86% and endometrium 90%. See www.PLSD.eu to calculate risks for individual patients by age and gender. Conclusion MSH6-associated Lynch syndrome has distinct characteristics with a high risk of endometrial cancer compared to other organs. The International Mismatch Repair Consortium (IMRC) was established to determine cancer risks by geographic region. Method Pedigree data for 6,436 Lynch syndrome families from 22 countries were submitted by researchers/clinicians throughout the world to the analysis team at the University of Melbourne. We estimated the cumulative risks (penetrance) by geographic region. We used a modified segregation analysis and adjusted for any ascertainment of families. Results Preliminary analysis suggest that for MLH1 mutations, the risk of colorectal cancer to age 70 is highest for carriers in Australasia (68% males, 55% females) and North America (61% males, 48% females) and lowest for carriers in South America (12% males, 10% females) and East Asia (20% males, 14% females). For MSH2, the patterns were similar, except for South America which had a high estimated average risk (82% males, 75% females). Conclusion Collection of MMR family data from many international sites has progressed well despite the challenges faces by sites to establish databases for epidemiological research with varying resources. Preliminary results suggest that cancer risks for people with Lynch syndrome differ by geographic region which is consistent with environmental modifiers for the disease and might justify region specific screening guidelines.

N10
Breast cancer risk in neurofibromatosis type 1 is a function of the type of NF1 gene mutation: a new genotype-phenotype correlation I. Frayling 1  Correspondence: I. Frayling Aim NF1 predisposes to breast cancer (BC), but no genotype-phenotype correlations have been described. Method Constitutional NF1 mutations in 78 NF1 patients with BC (NF1-BC) were compared to the NF1 LOVD (N=3432).

Results
There are no gross relationships with mutation position. No cases were observed with large deletions (HR=0.10; 95%CI: 0.006-1.63; p=0.014, Fisher's exact (FE)) 64.3% of the 70 different mutations have p<0.05 (FE), while 74.3% are significant when adjusted for multiple comparisons (Benjamini-Hochberg p≤0.125). Two pairs of patients shared the same predicted effects on neurofibromin, but had different mutations at the DNA level. 6/14 (43%) of the missenses (MS) were located in the CSRD (p=0.093; FE). 10/11 (91%) of MS cases with known age of BC occurred <50y (p=0.041; FE). 18 had BRCA1/2 testing, revealing one BRCA2 mutation. Conclusion This demonstrates that certain heritable mutation types, and indeed certain specific mutations in NF1 confer different risks of BC. The observation that NF1 amplification does not always occur with, and can occur independently of ERBB2 amplification, supports the concept that BC risk in NF1 may be due to gain of function mutations. A prospective NF1-BC study needs to be established. Regardless of NF1 mutation status NF1-BC patients warrant testing of other BCpredisposing genes.
Correspondence: D. G. Evans Aim Pathogenic variants in BRCA1/BRCA2 are identified in~20% of families with multiple individuals with early-onset breast/ovarian cancer. Extensive searches for additional highly penetrant genes/alternative mutational mechanisms altering BRCA1/2 have not explained the missing heritability. For the first time, we report a dominantly inherited 5'UTR variant associated with epigenetic silencing of BRCA1 due to promoter hypermethylation in two families with breast/ovarian cancer. Method BRCA1 promoter methylation of ten CpG dinucleotides in breast/ovarian cancer families without germline BRCA1/2 pathogenic variants was assessed by pyrosequencing and clonal bisulfite sequencing. BRCA1 RNA/DNA sequencing from lymphocytes was undertaken to establish allelic expression and germline variants. Results BRCA1 promoter hypermethylation was identified in 2/49 families with multiple women affected with grade-3 breast/high-grade-serous ovarian cancer. Soma-wide BRCA1 promoter hypermethylation was confirmed in blood/buccal mucosa/hair follicles. Methylation levels were~50%, consistent with complete silencing of one allele. RNA sequencing revealed allelic BRCA1 expression loss in both families segregating with a novel heterozygous variant c.-107A>T in 5'UTR.

Conclusion
Our results indicate a novel mechanism for familial breast/ovarian cancer, caused by an in cis 5'UTR variant associated with epigenetic silencing of BRCA1 promoter. We propose methylation analyses are undertaken to establish the frequency of this mechanism in families affected by early onset breast/ovarian cancer without a BRCA1/2 pathogenic variant.

N13
Identification Of genetic variants in early-onset and familial cancers by targeted next generation sequencing M. To study the potential contribution of genes other than BRCA1/2, PTEN, TP53 and MMR to the biological and clinical characteristics of early-onset and familial cancers in Norwegian families.

Method
The Hereditary Cancer Biobank from the Norwegian Radium Hospital was used to identify early-onset families and individuals with a high risk of developing breast, gynecological and colorectal cancers. Forty-four cancer susceptibility genes were selected and analyzed by our in-house designed TruSeq amplicon-based assay for targeted sequencing. Protein-and RNA splicing-dedicated in silico analyses were performed for all variants of unknown significance (VUS). Variants predicted as the more likely to affect splicing were experimentally analyzed by a minigene assay (PMID: 29458332, 29371908, 28608266).

Conclusion
Our study provides new information on genetic loci that may affect the risk of developing cancer in these patients and their families, demonstrating that genes presently not routinely tested in molecular diagnostic settings may be important for capturing cancer predisposition in these families.

Aim
Pathogenic Alu element insertions are rarely reported because this type of insertions are undetectable with the classical screening methods. The aim of this work has been the identification and characterization of an Alu element insertion in a Spanish family with a history of breast/ovarian cancer. Method Molecular analysis was carried out using the BRCA MASTRDx (Multiplicom) and Massively Parallel Sequencing (Illumina). The Alu insertion was identified and characterized by fragments analysis, genotyping, PCR amplification and Sanger sequencing (ABI3130).

Results
We have identified and characterized a heterozygous pathogenic variant c.5007_5008ins174 located at the exon 11 of the BRCA2 gene in a patient with prostate cancer. The variant identified is a pathogenic Alu element insertion (AluYb8BRCA2) of about 174 bp long. Conclusion NGS has been incorporated into clinical genetic testing for hereditary cancer risk. NGS-based techniques and the standard bioinformatic pipelines, however, are unable to detect and precissely characterize ALU element insertions. In this work, we report, by using classical screening methods and bioinformatic programs, BLAST and Repeat-Masked, the identification of the AluYb8BRCA2 insertion in BRCA2 coding region. This insertion could generate a framesift resulting in the abrogation of BRCA2 protein function that has been associated with oxidative stress involved in carcinogenesis.

N18
Deciphering the contribution of recently proposed polyposis predisposing genes M. Terradas 1, 2 , P. M. Muñoz 1, 2 , S. Belhadj 1, 2 , G. Aiza 1, 2, 3 , M. Navarro 1 Correspondence: A. Ragunathan Aim Bilallelic loss-of-function germline mutations in the base excision repair gene NTHL1 result in an increased risk of colorectal polyps and different cancer types, resulting in the inclusion of this gene on many multi-gene cancer predisposition panels. However, the impact of heterozygous germline NTHL1 mutations on colorectal cancer (CRC) risk is unclear. Method 1953 CRC-affected individuals and 1207 controls from the Colon Cancer Family Registry Cohort were screened for coding single nucleotide and short indels variants in NTHL1 using a targeted multiplex PCR-based sequencing approach (Hi-PLEX). Variants were filtered on sequencing depth and allele proportions. Variants were predicted to be pathogenic if they were novel or rare (gnomAD < 0.05%), protein truncating variants or missense variants predicted to be deleterious (based on CADD>20 or REVEL>0.5).

Conclusion
The effect of heterozygous NTHL1 predicted pathogenic variants on CRC risk, if any, is not likely to be more than 1.5 fold. In Slovakia, the incidence of colorectal cancer is one of the highest worldwide and could be a result of higher incidence of cancer predisposing syndromes, such as Lynch syndrome. Novel large gene panel, exome or whole genome tests become less costly and widely available which allow detection of cancer predisposing genetic variants. In addition, novel non-invasive methods for tumor screening (liquid biopsy) become available as well. Screening for genetic cancer predisposing syndromes and accordingly adjusted cancer screening regimes with inclusion of liquid biopsy methods could be viable options but the implementation and social aspects need to be studied. Method A clinical exome test was carried out in 20yo male. Identification of variants relevant for secondary findings was carried out too. Identified variants were verified by Sanger sequencing. A family follow-up included clinical exome test as well as Sanger confirming tests. Written informed consent to publish was obtained from the patients involved in this study.

Results
We identified a rare potentially pathogenic variant in MSH6 gene in 20yo male as secondary finding. In addition a rare BRCA2 variant was also detected and confirmed. Despite family cancer history did not meet used criteria it was tested for these variants. A suspected case of metastatic breast cancer in the family was confirmed in 65yo female bearing the in 12/2017 and a suspected case od CRC was identified in 67yo male bearing the MSH6 variant. All family members adherred to required medical procedures after genetic testing. Conclusion Genomic tests and their wider availability with novel liquid biopsy methods offer novel cancer screening algorithm options. A case of family with two detected variants in both BRCA2 and MSH6 a secondary finding shows possible benefits but social aspects have to be considered for wider implementation. To create a national service collecting pseudonymised germline cancer-predisposing genotypes, and link these to the National Cancer Registration and Analysis Service for individuals with a prior or subsequent cancer diagnosis. Method NHS molecular genetics laboratories submit patient-level genotype data through a secure online portal. Unique patient demographics are pseudonymised using a one-way hash function that generates an irreversible pseudoID; additional identifiers are secondarily encrypted. The same hash function applied to cancer registration records, where patient identity is already known, enables linkage of the genotype data; decryption of additional identifiers is then possible. We can thus obtain accurate variant counts nationally, and identify those who develop cancer, without compromising patient privacy.

Results
Pilot work has focused upon BRCA1 and BRCA2 genes; we are now commencing collection of colorectal cancer predisposition gene data. To date, ten laboratories have submitted BRCA1/2 data, covering a time period from 2001 onwards, and including~1300 different gene variants. Initial linkage to cancer registry records showed a 68% match rate.

Conclusion
This robust, secure system collects depersonalised but linkable genotypes on all individuals tested. Record-level linkage to the rich phenotype, treatment and outcome data in the national cancer registry provides allelic frequency and associated phenotype data, and facilitates variant interpretation.

Method
Stakeholders from Europe and North America worked together over a two-day workshop to achieve consensus on best practice. Stakeholders included patients, patient support groups, gynaecologists, clinical geneticists, medical oncologists, colorectal surgeons, gastroenterologists, histopathologists, genetic pathologists, health economists, epidemiologists, gynaecology nurse specialists and genetic counsellors.

Results
Guidance was developed in four key areas: 1) whether women with gynaecological cancer should be screened for Lynch syndrome and 2) how this should be done; 3) whether gynaecological surveillance was of value for women with Lynch syndrome; and 4) what preventive measures should be recommended for women with Lynch syndrome to reduce their gynaecological cancer risk.

Conclusion
The Manchester International Consensus Guideline provides comprehensive clinical guidance that can be referenced by both patients and clinicians so that women with Lynch syndrome can expect and receive appropriate standards of care. This group constitutes Lynch-like syndrome (LLS), estimated to represent 3% of colorectal cancer cases. Evidence suggests the majority of LLS cases are caused by somatic variants in the tumour. Colorectal and extracolonic cancer risks in LLS are increased in comparison to population risks. UK guidelines suggest 2 yearly colonoscopies for individuals with LLS and their first degree relatives; assuming there may be an unknown hereditary cause.

Method
We conducted a survey amongst clinicians practising in regional clinical genetics departments within the UK. We aimed to explore clinicians' understanding and management of LLS families. The survey was disseminated by the cancer lead clinician within each department, through a "surveymonkey" link.

Results
We received 44 responses from 19 centres. 40% of participants were aware of the definition of LLS while 27% would offer 2 yearly colonoscopies. There were variations in practice within and between departments.

Conclusion
These results emphasize the importance of increasing awareness of LLS, and contribute towards the need for clear management guidelines. Correspondence: M. Kloor Aim Lynch syndrome-associated cancers accumulate a high load of immunogenic frameshift peptide neoantigens as a consequence of DNA mismatch repair (MMR) deficiency. MMR-deficient cells can therefore be recognized by the immune system. We aimed to comprehensively characterize the immune phenotype of MSI cancers, accounting for somatic mutations inducing immune evasion and for immune cell infiltration.

Method
We combined the analysis of our own cohort of MSI cancers with mutation data of MSI cancers of the TCGA/DFCI cancer collections. Immune cell infiltration was quantified by immunohistochemistry, using antibodies specific for T cell subtypes, including CD3, FOXP3, and PD-1.

Results
72% of MMR-deficient colorectal cancers of the DFCI database harbored alterations affecting genes involved in HLA class I-mediated antigen presentation. The most common alterations were truncating mutations affecting the Beta2-microglobulin (B2M) gene. B2M mutations were related to a higher density of activated T cells infiltrating the tumor, and to a lower frequency of regulatory T cells in the tumor environment.

Conclusion
The extraordinarily high prevalence of immune evasion phenomena in MSI cancer most likely reflects life-long immune surveillance and suggests that most MSI pre-cancers are eliminated by the immune system if they fail to evade the immune attack. North West Thames Regional Genetics Service; 2 South East Thames Regional Genetics Service; 3

Results
We present the pathway as adopted at St Marks Hospital and the outcomes from the first year post implementation.

Conclusion
This pathway was effective at our hospital.

Method
Previously reported results were validated in an independent series of path_MLH1 carriers followed-up by colonoscopy. Combined results merging former and present series included only carriers with pathogenic class 4 or 5 variants listed in the InSiGHT database.

Conclusion
These findings provide a basis for recommending to introduce the investigation of these markers in biopsy specimens from LS patients, as a supportive tool for selecting the most appropriate management option in these patients (prophylactic hysterectomy vs surveillance)

N36
Back to back comparison of colonoscopy with virtual chromoendoscopy using third generation narrow band imaging system to chromoendoscopy with indigo carmine in Lynch syndrome patients E. Samaha 1 , C. Correspondence: E. Samaha Aim Colonoscopic screening with indigo carmine chromoendoscopy (ICC) in Lynch Syndrome (LS) patients improves adenoma detection rate and is widely used nowadays. Nevertheless, it is a time-and-moneyconsuming technique which requires a dedicated training. Narrow band imaging (NBI) is a well-known virtual chromoendoscopy technique that highlights superficial mucosal vessels and improves contrast for adenomas. We conducted a prospective multicenter study in a back-to-back fashion to compare 3rd generation NBI to ICC for detecting colonic adenomas in LS patients. Method One hundred and thirty eight patients underwent a double colonoscopy, first with NBI, followed by ICC, in a back-to-back fashion. All polyps detected in either pass were removed for histopathological analysis. The primary outcome measure was the number of patients with at least one adenoma after NBI compared with the number of patients with at least one adenoma after NBI and ICC. Proportions were compared with the paired exact test (McNemar's test). Continuous variables were compared with the Wilcoxon signed-rank test.

Conclusion
Neither colorectal, endometrial, ovarian nor urinary tract cancer was observed before 50 years of age. The point estimates for colorectal and endometrial cancers at age 75 were, however, higher than expected despite undergoing regular surveillance . The patients examined were mostly selected from cancer kindreds, and the late onset cancers might not necessarily have been caused by the path_PMS2 variants. Clinical guidelines for monoallelic path_PMS2 carriers should be revised. and median follow-up time 6 years (IQR 2-10 years). At first colonoscopy CRC was found in eight patients and during 916 follow-up colonoscopies in nine patients. No CRC was found in MSH6 or PMS2 mutation carriers. There were no significant differences in the number of colonoscopies with adenomas or advanced adenomas between the different gene mutation carrier groups. In MSH6 mutation carriers advanced neoplasia (advanced adenoma or colorectal carcinoma) was found after a longer follow-up time than in the other mutation carrier groups.

Conclusion
Since no CRC was found during follow-up in MSH6 mutation carriers and advanced neoplasia was found in shorter follow-up time in MLH1 and MSH2 mutation carriers, the colonoscopy interval in MSH6 mutation carriers might be less stringent than for MLH1 and MSH2 mutation carriers. Patient information: Age, gender, pathogenic variant, reporting centre, age and ICD9 diagnoses of all cancers (before, at or after inclusion), organs removed when. Events scored: All prospectively diagnosed cancers after inclusion by ICD9 code and age at diagnosis. Age at death.
Information not yet analysed: polyps removed, stage at colorectal cancer and time since last colonoscopy. For detailed protocol see https://ehtg.org/ Results Incidences of cancer by age, genetic variant and gender. Survival after cancer. Results of intervention (international guidelines).

Conclusion
The reports migrate knowledge on Lynch syndrome from expert opinions based mainly on retrospective studies to assumption-free empirical observations in carriers subjected to follow-up according to accepted clinical guidelines. The interactive website www.PLSD.eu returning risk for any single case when indicating age, gender and gene is referred to for clinical use by InSiGHT and others.

Method
Previously reported results were validated in an independent series of path_MSH2 carriers followed-up by colonoscopy. We combined results merging former and present series including only carriers with pathogenic class 4 or 5 variants listed in the InSiGHT database.

Conclusion
The PLSD and InSiGHT databases are complementary: PLSD reports prospectively observed average risks and survival in carriers of variants determined to be pathogenic by InSiGHT. Aim Screening for small-bowel cancer (SBC) is not yet included in surveillance guidelines for LS. In 2016 Mallorca group advised may be appropriate in MSH2 and MLH1 mutation carriers, after 40 years. Aim of the study was to determine SBC incidence in asymptomatic LS patients by means of video capsule endoscopy (VCE). Method Two prospective VCE databases were retrospectively reviewed to identify consecutive asymptomatic LS patients, compared with a group of patients who underwent VCE for obscure gastrointestinal bleeding (OBS). Results 25 LS patients and 280 OBS patients were enrolled by two Italian centers. In 91.5%, caecal visualization was achieved. SBC was detected in two LS patients and three OBS patients (p=0.06). The two groups have a significant statistically different mean age (SD): 41.3 ys± 14.0 ys in LS group and 62.9 ys ± 17.2 ys in OBS group. Besides SBC, LS patients and OBS patients have statistically significant difference in incidence of vascular lesion, angiectasia and minute polyps.

Conclusion
The prevalence of SBC in asymptomatic patients with LS was 8% vs 1.1%. Although the incidence of SMC did not reach statistical significance difference, a trend through statistically significant difference was observed and this suggests further multicentric studies are needed.

N43
Hide and seek with hereditary cancer: testing the effectiveness and cost-effectiveness of implementation approaches for translating Lynch syndrome evidence into practice N. Taylor  Correspondence: N. Taylor Aim Evidence indicates that hospitals face infrastructural, psychosocial and environmental barriers to detecting Lynch Syndrome (LS) patients. In Australia, less than half of all high-risk colorectal cancer (CRC) patients are being referred for LS genetic testing. This study aims to compare the effectiveness and cost-effectiveness of two implementation approaches for increasing the proportion of CRC patients with risk-appropriate completion of the LS testing and referral pathway. Method This randomised controlled trial will test the Theoretical Domains Framework Implementation approach against a non-theory-based implementation approach in eight large Australian hospitals. Site based healthcare professionals will be trained to lead the following process: 1) Baseline audits, 2) Form Implementation Teams, 3) Identify practice change behaviours, 4) Identify/confirm barriers to change, 5) Generate intervention strategies, 6) Support intervention implementation, 7) Evaluate practice/culture change. Theoretical and non-theoretical components are distinguished in 4-5.

Results
Progress to date of baseline data analysis will be presented. Plans for the analysis of health and economic outcomes of each implementation approach to be estimated using "POLICY1-Lynch" will be provided. Conclusion This will be a world first study to compare theory-based and nontheory based approaches to evidence translation in healthcare, and to incorporate these findings into existing microsimulation models to accurately assess implementation cost-effectiveness. The aim of this study was to assess genetic and clinical characteristics of Slovenian Lynch syndrome (LS) families, as such evaluation has not yet been performed for our population.

Method
We analyzed the results of genetic testing performed in 2008-2018 for probands fulfilling LS testing criteria. LS spectrum cancers identified in confirmed, obligate and assumed carriers of mismatch repair (MMR) gene mutations were verified in Slovenian cancer registry and analyzed according to site, age of onset, and genes involved.

Conclusion
We had very few referrals for LS testing in the 10-year period analyzed considering its prevalence in the population. LS is therefore likely to be drastically underdiagnosed in Slovenia. Screening of all colorectal and possibly endometrial cancers with immunohistochemical test should be performed in order to systematically identify LS families and offer them adequate treatment and familial risk assessment in the future.   Excluding samples with BRAF mutation or MLH1 promoter hypermetilation (MLH1-Hy), deficient-IHC cases were addressed to germline MMR gene testing.

Conclusion
We support the systematic evaluation of MMR proteins in colorectal and gynecological cancers to select patients with LS. MSI could be useful to manage borderline-IHC cases.

Method
We analysed the MMR genes in frozen tumor tissue samples by NGS for 28 LLS patients. When a tumoral variant was found, we performed a targeted re-examination of the germline NGS results with lower detection rates and targeted Sanger analysis in normal adjacent tissue DNA and lymphocytes DNA from offspring when available.

Results
Eight patients had double somatic hits in their tumors. Two patients had a germline de novo mosaic variant of MSH2 with low variant allele frequency (9% and less than 2%). Those variants were missed by NGS analysis in lymphocytes DNA. Their identification in tumors allowed a targeted NGS reanalysis. In both cases, these variants were found to be heterozygous in one of the offspring.

Conclusion
These mosaic cases confirm that identification of the mechanism that causes tumors in LLS is crucial for genetic counselling and guiding screening of patients and their relatives To further develop the concept of a cancer-preventive vaccine in Lynch syndrome, we aimed to establish a preclinical mouse model. Method A systematic database search was performed to identify coding microsatellites (cMS) and potential neoantigens in Lynch syndrome mice (Msh2flox/floxVpC+/+). After mutation analysis of murine tumors, most promising FSP neoantigens were evaluated for immunogenicity by ELISpot after vaccination of C57BL/6 mice.

Results
Four FSP neoantigens derived from common cMS mutations in the genes Nacad, Maz, Xirp1, and Senp6 elicited strong antigen-specific cellular and humoral immune responses. Based on the cMS mutation data, a vaccine with these four FSP neoantigens is predicted to cover about 75% of cancers in Lynch mice.

Conclusion
We have identified four immunogenic FSP neoantigens derived from commonly mutated cMS in murine Lynch syndrome colorectal cancers. This allows evaluating the concept of cancer-preventive neoantigen vaccines in mouse models of Lynch syndrome, including longitudinal monitoring of immune responses and combination with different adjuvants and chemoprevention approaches. Correspondence: A. Ahadova Aim For distinguishing Lynch syndrome patients from sporadic microsatellite unstable (MSI) patients, BRAF V600E testing has become one of the most important tools. In order to analyze the discriminatory power of BRAF mutations in different age groups, we looked at the age distribution of BRAF mutations in MSI colorectal cancers. Method Age at diagnosis and BRAF mutation status were retrieved for unselected series of MSI colorectal cancers (n=151) from publicly available databases (DFCI) and the DACHS cohort.

Results
The prevalence of BRAF V600E mutations in MSI cancers strongly increased with age at diagnosis, with 87% of BRAF mutations occurring after the age of 65. There was no patient with a BRAF mutation under the age of 50, and the youngest patient with a BRAF mutation was 52 year old. Conclusion Our data demonstrate that BRAF mutation testing to exclude Lynch syndrome has very limited value in patients younger than 50, as the likelihood of detecting BRAF mutation in a patient under 50 is close to 0%. Reports of BRAF mutations in 1-2% of cancers from proven Lynch syndrome mutation carriers call into question the role of BRAF mutations as a bona-fide exclusion marker for Lynch syndrome. Correspondence: A. Ballhausen Aim Lynch syndrome cancers are caused by DNA mismatch repair (MMR) deficiency. MMR deficiency leads to microsatellite instability (MSI) and to a high mutational load. Insertion/deletion mutations (indels) of coding microsatellites are drivers of MSI cancer development and responsible for the accumulation of immunogenic frameshift neoantigens. Next-generation sequencing has a limited sensitivity for detecting such indels. We have developed a novel tool to provide a high-resolution map of the MSI cancer mutation and neoantigen landscape.

Method
The 'qMSI' algorithm processes fragment length analysis data, removing stutter band artifacts using a linear matrix. QMSI allows the quantification of the true allele frequency of mutations and the distinction of different mutation types that give rise to distinct frameshift neoantigens.

Results
Using qMSI for 40 target genes in MSI colorectal cancers (n=139) we demonstrate that most indels in MSI cancer are single-nucleotide deletions (77%) followed by two-nucleotide deletions and singlenucleotide insertions (21%). Neoantigen-inducing mutations were surprisingly similar across different MSI cancers.

Conclusion
The qMSI algorithm is a powerful tool to identify driver genes and mutational neoantigens in MSI cancer. The identification of shared, recurrent neoantigen-inducing mutations indicates that a vaccine for tumor prevention in Lynch syndrome is highly promising. The onset of mismatch repair (MMR) deficiency in Lynch syndromeassociated tumors has been discussed to be a late event of pathogenesis. Since the time point of MMR deficiency onset and its consequences have a direct impact on the selection of suitable therapeutic and preventive measures, we aimed to reconstruct the sequence of mutational events in Lynch syndrome cancers. Method MMR protein expression and mutational signatures were analyzed to address the time point of MMR deficiency in Lynch syndrome adenomas and carcinomas from public databases and our own cohort. Results 77% of Lynch syndrome adenomas (n=640) were MMR-deficient. Mutational signatures of MMR deficiency were reflected in canonical CRC gene mutations, demonstrating that 100% of KRAS and more than 60% of APC mutations likely occurred after the onset of MMR deficiency. A substantial proportion of Lynch syndrome-associated colorectal cancers lacked evidence of polypous growth. These tumors showed a distinct molecular pattern enriched for TP53 and CTNNB1 mutations.

Conclusion
There is more than one pathway of CRC development in Lynch syndrome. MMR deficiency commonly occurs early during Lynch colorectal carcinogenesis. Non-polypous cancers developing from MMR-deficient crypts may be missed by colonoscopy, strengthening the need for additional primary prevention measures in Lynch syndrome.

Results
The data consists of 9719 family members and 838 proven mutation carriers. MSH6-mutation carriers with CRC or EC (n=306) were included in the study, accounting for 219 CRCs and 122 ECs. Of the tested tumors, discordant staining for MSH6 was reported in 10 out of 68 CRCs (14.7%) and 3 out of 26 ECs (11,5%). Six out of 62 CRCs (9.7%) and 5 out of 25 (20.0%) ECs appeared to be microsatellite stable. Fifteen germline MSH6 mutation carriers also displayed negative staining for MSH2 in addition to negative MSH6 staining, but did not harbor a germline MSH2 mutation. Conclusion Germline MSH6 mutation carriers can be missed using reflex IHC MMR testing as is currently standard in most western countries. MSH6 germline or tumor DNA analysis -preferably as part of a larger gene panel -should therefore be considered, especially in patients fulfilling the Bethesda criteria.  Correspondence: B. Desouza Background UK guidelines recommend that all newly diagnosed colorectal cancers (CRCs) be screened for mismatch repair deficiency (MMR-D) that may be indicative of Lynch syndrome (LS). Current diagnostic approaches, will fail to detect MLH1 promoter hypermethylation or a germline mutation in approximately 60% of suspected LS cases. In most cases the diagnosis of LS can be excluded by somatic sequencing through the demonstration of double somatic mismatch repair (MMR) mutations.

Method
We have used our clinical data from over 1100 families to model costs for different diagnostic strategies for LS that integrate germline and somatic testing. Outcomes were correlated to family history category of either revised Bethesda guidelines or modified Amsterdam criteria.

Results
Modelling shows that for Bethesda families, performing concurrent germline and somatic testing would be more cost-effective than sequential germline and somatic testing (£523 vs. £940 per proband). For Amsterdam families, however, performing sequential testing would be more cost-effective than concurrent testing (£617 vs. £1256 per proband). Conclusion LS diagnostic strategies for CRC cases could be accelerated and simplified by concurrent germline and somatic testing. Moreover, our data suggests that this approach is more cost-effective than sequential testing in Bethesda families.

Method
We used a microsimulation to model costs of DNA mismatch repair gene mutation testing for five target population subgroups: i) incident colorectal cancers (CRCs) diagnosed under age 50; ii) under age 70; iii) at any age; iv) unaffected people aged 20-50 years; and v) unaffected people aged 20-80 years. For the incident CRC subgroups, three strategies were considered: multi-gene panel testing; immunohistochemistry (IHC) followed by a multi-gene panel test; and IHC followed by MLH1 methylation testing and a multi-gene panel test. For the strategies targeting the general population (no CRC), only multi-gene panel testing was considered. Results IHC followed by panel testing yielded the lowest cost per mutation carrier identified at AU$2,529, AU$6,331, and AU$11,182 for the approaches targeting incident CRCs under age 50, 70 and any age, respectively. For the general population approaches, testing unaffected people aged 20-50 years was the cheapest option (AU$112,282 per carrier identified). Testing incident CRCs under age 50 identified the highest number of carriers (11,774 per 100,000 probands). Conclusion Testing incident CRC cases under age 50 years appears as the most effective and cheapest strategy to identify LS mutation carriers.  ) and a mean of 3.7 followup years (range 1-9.6). Conclusion Several research projects and publications have been implemented, generating knowledge of MMR variants in these populations to bring additional awareness to medical professionals and public health leaders. Participation in PLSD and international collaborations have been initiated to support the implementation of genetic testing and research in most of the countries of Latin America.

Method
This is a multicenter nation-wide study (25 Spanish hospitals). Patients were included when CRC tumors showed immunochemical loss of MSH2, MSH6, PMS2 or loss of MLH1 with BRAF-WT and/or no MLH1 methylation and absence of pathogenic mutation in these genes.

Results
Our study included 160 LLS patients. Loss of MLH1/PMS2 was found in 48% of CRC, loss of MSH2/MSH6 in 25%, loss of MSH6/PMS2 in 2%, isolated loss of MSH6, PMS2 and MSH2 was found in 11%, 9% and 2% respectively. In 3% of patients no gene loss of expression was found. 5 patients (3%) developed CRC during the follow up time since diagnosis, (median time of 7 years (SD 3.95)); 20 patients (12.5%) had personal history of non-CRC, and only 5 (3%) patients had LS-related cancer history.

Conclusion
In this LLS cohort, the largest until now, there are no clinical, molecular or pathological characteristics that could help distinguish between probably sporadic and hereditary patients. These results support the need of homogeneous follow-up for this group of patients.

Aim
Hereditary non-polyposis colorectal cancer defines the development of colorectal cancer within the spectrum of presentation of Lynch syndrome. A major characteristic of CRC in Lynch individuals is the failure to metastasize despite the large tumor size. Herein we present a case of metastatic CRC in a patient with a pathogenic MSH2 / MSH 6 mutation.

Method
A 68 year old Caucasian male patient with a history of right nephrectomy 25 years after a uretheral cancer. Mismatch repair analysis confirmed MSI-H for MSH2 / MSH6. The patient now presented with a rectal cancer and to date he had not been recommended genetic testing. He underwent an anterior rectal resection with a protective loop ileostomy in December 2017 for colorectal cancer of the rectosigmoidal junction (pT3N0pV0pL0G2R0). An abdominal wall mass was found 10 months after surgery at the former ileostomy site during follow-up, which was completely excised with negative margins. Five months later, computed tomographic scans of the abdomen suspected recurrent metastasis including a peritoneal mass. Surgical exploration was performed.

Results
The abdominal wall mass was completely removed with negative margins. Equally, limited peritonectomy was performed during the second exploratory laparotomy. Histopathology confirmed the presence of metastasis from a colorectal cancer with loss of MSH 2 / MSH 6 proteins on immunohistochemistry. The patient was discussed at the interdisciplinary oncologic board after which adjuvant checkpoint inhibitor therapy was recommended However, the insurance was not willing to pay for this treatment.

Conclusion
Histopathologic features including loss of MSH 2 / MSH6 protein expression on immunohistochemistry in both the primary tumor as well as the metastatic lesions confirms the presence of metastatic seeding. This provides evidence for a metastasis of CRC in a patient with Lynch syndrome and disapproves the currently accepted nonmetastatic theory. We conclude, that we cannot rely on the theory and are mandated to adhere to all principles of oncological surgery and also of stringent follow-up.

Method
We performed whole exome sequencing in a cohort of 27 LLS patients. We performed an analysis to identify rare likely pathogenic variants that could be predisposing to cancer. Only high-quality called variants, present with a population frequency <2.10-5 were included. Based on the fact that the mutations in the MMR genes could be passenger mutations that drive further instability, a targeted analysis including a comprehensive list of DNA repair genes was also included. We also performed tumor exome analysis from the matching samples to search for somatic hits.

Results
We identified 4 LLS patients with rare germinal variants in the following genes: AXIN1, PIWIL3, CD109, RECQL5 and GEN1. No somatic second hit was found in any of these genes. 2/8 cases where we could evaluate somatic events had a somatic mutation in one MMR gene and 1 showed LOH of the other copy. One tumor had a single mutation in a MMR gene and in one case I did not identify any somatic alterations.

Conclusion
Based on these results we hypothesize that there is a group of patients with predisposition to CRC due to a germinal variant in one allele that triggers genomic inestability. But there is also another group of patients where it could be due to a biallelic somatic mutation in MMR genes

Method
Initial attempts through the HVP sourced variants from laboratories by streamlining with Laboratory Information Management Systems. Subsequently, a grant was awarded from the New South Wales Cancer Council to build a database of pathogenic (Class 4 and 5) variants identified in the cancer genes through the familial cancer clinics (FCCs); a collaboration across the clinics (ICCon) was formed to facilitate this.

Results
The ICCon database holds information about MMR gene pathogenic variants in adult carriers as follows: MLH1 124 (90 unique), MSH2 121 (94), MSH6 68 (50), PMS2 36 (25); totalling 349 (259). Ten discordant interpretations between clinics and/or InSiGHT's classifications were resolved as part of the ICCon process. Importantly, clinical and other data to assist VUSs was accessible from the FCCs. Conclusion Sourcing variants via the FCCs has proved feasible. The ICCon database has contributed to variant interpretation internationally, including InSiGHT's Variant Interpretation Committee and, in part, the PLSD. ICCon is working to achieve governance around transforming the variant database to a national registry, to permit changes in counselling, and clinical management, such as when new information emerges through contemporary experience or research. Reason for heterogeneity of cancer risk within specific variants in each gene is unknown.

Method
We estimated colorectal cancer risk for MSH2 c.942+3A>T variant using 234 families from the International Mismatch Repair Consortium. Age-specific cumulative risks (penetrance) and 95% confidence intervals were estimated using a modified segregation analysis with appropriate ascertainment conditioning and allowing for risk to vary between families by fitting a polygenic effect.

Results
The estimated average cumulative risks to age 70 years (95% confidence intervals), were 56% (38%-78%) for males carriers and 45% (28%-67%) for female carriers. However, the lifetime risks for different people were estimated to vary widely about these average risks (p=0.001). For carriers of this specific variant, 26% of males and 16% of females had colorectal cancer risk less than 20%; and 24% of males and 37% of females had risk greater than 70%.

Conclusion
Even for a specific variant in a DNA mismatch repair gene, there is a wide range of colorectal cancer risks. This is consistent with the existence of strong modifiers of risk, that if known, could be used to provide personalized risk of colorectal cancer for Lynch syndrome.

Method
Clinical and pathological data were retrieved during a single-session visit in gastroenterology and genetics from 2016 to 2017. FPC underwent either pancreatic endoscopic ultrasound (eUS) or magnetic resonance (MR) and Next Generation Sequencing analysis. Results 57FPC were evaluated; 17 had a personal diagnosis of PC. 29(50,9%) had ≥2 relatives affected, of whom ≥1 was a first-degree relative (FDR); 11(37,9%) had PC. 11(19,3%) had ≥3 relatives affected (1 had PC). 6(10,5%) had Lynch Syndrome with ≥1 FDR (1 had PC). 2(3,5%) had hereditary pancreatitis and 9(15,8%) BRCA1/2 mutation with 1 FDR affected (5 had PC). 17(29,8%) were genetically confirmed: 6 LS (35,3%), 2 PRSS1 (11,8%), 6 BRCA2 (35,3%), 1 BRCA1 (5,9%), 2 PALB2 (11,8%). 8 showed a Variant of Unknown Significance (VUS). 21(36,8%) underwent eUS, revealing 8 PC, 3 intraductal mucinous neoplasias, 1 pseudopapillary lesion. 16(28,1%) underwent MR, revealing 7 CP, 1 IPMNs, and 3 cystadenoma. Conclusion A multidisciplinary approach enriches the proportion of patients with genetically confirmed FPC from 5-10% to about 30% of all FPC. Correspondence: T. Gemoll Aim Familial adenomatous polyposis (FAP) is an autosomal dominant inherited disease with a germline mutation of the APC gene. In spite of this specific genetic alteration early diagnosis in young patients without polyposis onset and lack of family history can be difficult and finally letal. Thus, additional sensitive diagnostics are required. We aimed at identifying and validating a protein expression signature in macroscopically unaffected colon mucosa that allows identifying genetic carriers of the FAP-syndrome. Method Protein profiling by 2-D gel electrophoresis was performed on samples obtained from 15 different patients (FAP, n=8; sporadic colorectal cancer, n=7). Analysis was performed for normal mucosa, adenoma, and carcinoma while comparing FAP-associated tissue with the sporadic counterpart. Analysis aimed at identifying proteins that were expressed in FAP tissue but not in the corresponding sporadic tissue, comparing particularly FAP associated normal mucosa versus sporadic normal mucosa. Target validation was performed by Western and by immunohistochemistry on clinical samples (n=189), respectively. Results A total of 47 proteins were present in all macroscopically unaffected FAP mucosa specimens but absent in sporadic normal mucosa. Comparing FAP polyps with sporadic colonic polyps revealed 49 polypeptides being present in FAP samples but absent in all sporadic polyps. Comparing three FAP carcinomas with seven sporadic colorectal carcinomas yielded 66 proteins with absence/ presence expression pattern. CSTF2T and ACTB were validated by Western Blot and immunohistochemistry in unaffected colon mucosa of FAP patients.

Conclusion
The data obtained demonstrate specific differences of FAP and sporadic colorectal disease on the protein expression level and could help to identify patients with FAP disease already in macroscopically "normal" colorectal mucosa. The Danish HNPCC register was established in 1991 as a private research register, later developing into a national database financed within the National Public Health care System. Epidemiological, clinical, and genomic data generated all over the country on 6.297 CRC families hereof 443 Lynch families are registered. Initially paper-based reports were sent to and typed into the database. Later a model for electronic exchange of data between laboratories, departments and the register in an EC co-funded project to prevent cancer by optimizing screening, digitization of data transport and combining genotype-phenotype information, sufficiently usable and generic to be implemented in other countries were developed. As medical data are heterogeneous, focuses were on integration, development of classification systems and communication standards. Identified gaps and status of usability will be presented.

Results
Data in the HNPCC register belongs to the financing Capital Region and the multidisciplinary scientific societies providing data. To achieve commitment and ownership representatives off all parties are invited into the Scientific Board and Steering Comity of the register, where rules for ownership and data delivery are decided.

Conclusion
The Danish HNPCC-register is national and comprehensive, and researchers can request data via the Scientific Board.

Literature review Results
Colonic varices is a rare entity and in majority of cases results from portal vein hypertension. It is even rarer when these lesions develop without an underlying hepatic or portal vein disease, termed, idiopathic colonic varices with less than 40 cases reported in the literature. Familial idiopathic colonic varices have also been described, where more than one family member is affected. These lesions could present an incidental finding, however, many cases presenting with lower GI bleeding were recognised in the literature, but no case was reported with a colonic tumor. Hereby we report a case of a 24 years old gentleman who presented with a history of acute abdominal pain and anemia. CT and Colonoscopy showed evidence of hepatic flexure mass, proved to be adenocarcinoma on histological examination, with an incidental finding of pancolonic varices. The patient has two relatives with pancolonic varices on colonoscopy, but no history of colonic tumors. He underwent right hemicolectomy with uneventful recovery. To our knowledge, this is the first case reported with coexistence of extensive idiopathic colonic varices and colonic tumor.

Method
Ten adults undergoing predictive genetic testing for cancer predisposition syndromes were included between January and March 2017. Demographic information, personality traits, psychological distress, behaviour in some daily activities and medical resources use were collected before testing and two months after results disclosure.

Results
High pre-and post-test psychological distress was associated to low education levels, having psychopathological history, pursuing testing for offspring, and being recruited at ICO (p<0.05). It was also associated with high negative affect, detachment, psychoticism and novelty seeking, and low reward dependence, self-directiveness, cooperativeness, and persistence (p<0.05). High post-test distress was also associated with having pre-test psychological distress (p<0.05). It would be important to know our counselees' personality because it gives us the opportunity to know who to offer more support and how to personalize genetic counselling.

Conclusion
Our results suggest that there are some personality traits which can influence psychological distress in individuals undergoing predictive genetic testing. Further studies need to be performed in order to extrapolate these results to this particular population.

Aim
Microsatellite instability (MSI) is reflective of a deficient mismatch repair system (dMMR) and occurs in 15% of all colorectal carcinomas (CRC). This most frequently occurs due to sporadic or constitutional mutations in mismatch repair genes. Mismatch repair (MMR) status is often identified by immunohistochemistry (IHC) for mismatch repair proteins (MMRPs) on CRC resection specimens. IHC testing performed on endoscopic biopsy material may be as reliable as that performed on resected specimens. We aimed to evaluate the reliability of MMR IHC staining on preoperative CRC endoscopic biopsies. Correspondence: P. Janega Aim The immune system plays crucial role in the development of the neoplastic diseases. Colorectal carcinoma is one of the most frequent oncological diseases with high mortality rate also in Slovak republic. Its development is the result of environmental, genetic and epigenetic changes accumulation leading to neoplastic transformation. Tumor-specific mutations manifest by neoantigens activating the immune system. The aim of the work was to evaluate the antitumor immune microenvironment in association of tumor grading. Method Archival surgical specimens of CRC were evaluated and graded according to the WHO criteria. Immunohistochemically detected CD4, CD8 and CD68 positive cells were evaluated morphometrically and expressed as % of the evaluated area.

Results
Neoplastic as well as the surrounding tissues were infiltrated by the three cell types in unchanged ratios, with predomonation of CD68+ histiocytes. With the increasing grade there was significant decrease of CD4+ and CD68+ cells and a clear decrease of CD8+ cells at the edge of significance, of infiltration of the tumor tissue. Changes in the peritumoral tissue infiltration were not significant.

Conclusion
Our findings support the idea of tumor suppressing activity of the anti tumor immunological response and that it plays an important role in progression of the neoplasm. Supported by the APVV-14-0318 grant. Correspondence: E. Meuser Aim Duodenal polyposis and cancer are important yet poorly understood causes of morbidity and mortality in FAP and MAP patients. We aimed to characterise the genomic and transcriptomic signatures associated with duodenal adenomas from patients with FAP and MAP, to better understand duodenal tumourigenesis in these hereditary disorders. Method A series of 67 samples from patients with a genetically confirmed diagnosis of FAP or MAP were subjected to whole transcriptome sequencing, consisting of 44 duodenal adenomas (FAP n=29, MAP n= 15) and 23 duodenal normal mucosa (FAP n=15, MAP n=8). Outcomes were compared to exome sequencing data from 50 duodenal adenomas (FAP n=25, MAP n=25).

Results
We found distinct gene expression profiles in FAP and MAP duodenal adenomas which were absent from the respective normal mucosa. MAP adenomas harboured aberrations in RAS signalling and immune system stimulation, whilst evidence for dysregulation of prostanoid synthesis and NOTCH signalling were found in FAP adenomas. Whole exome analysis revealed that MAP duodenal adenomas carried more somatic mutations than FAP (p=0.0226). Recurrently mutated genes in duodenal adenomas included known drivers (APC, KRAS) and additional potential duodenal-specific tumour initiators.

Conclusion
The identification of commonly deregulated pathways contributes to our understanding of duodenal tumourigenesis in the context of FAP and MAP.

N83
Surveillance recommendations for first-degree relatives of patients with unexplained multiple colorectal adenomas: a nationwide survey of UK regional genetic services B. Desouza, A. Elniel, N. Jakharia-Shah, G. Norbury, A. Kulkarni, D. Ruddy, V. Tripathi, A. Shaw , L. Izatt Guy's Regional Genetics Service Hereditary Cancer in Clinical Practice 2019, 17(Suppl 2):N83 Correspondence: B. Desouza Background Patients with multiple colorectal adenomas (MCRA; 10-100 adenomas cumulatively) without a known genetic cause are increasingly being diagnosed in the UK. Germline monoallelic APC or biallelic MUTYH mutations are not identified in the majority of patients. Possible explanations include; APC mosaicism, cryptic mutations, mutations in other polyposis genes, and polygenic inheritance. Some guidelines have recommended regular colorectal surveillance for first-degree relatives of this patient group, but currently there is no national UK guidance.

Method
We conducted a national survey of UK regional genetic services to explore management practices for first-degree relatives of patients with MCRA without a known genetic cause. A web based-survey was sent by email to the cancer genetic lead clinicians at the 24 regional genetics services. The survey was primarily designed to assess surveillance recommendations for first-degree relatives of MRCA patients, and to determine whether recommendations varied according to the total number of adenomas and age of onset. Testing criteria and genetic investigations were also assessed for patients with MCRA.

Results
National survey results are presented.

Conclusion
The survey aims to highlight variation in the management of this patient group and their first-degree relatives in the UK.

Results
We found 11 germline mutations (8 biallelic and 3 monoallelic) in MutYH gene among 93 patients with 4-100 polyps and 2 mutations (1 biallelic and 1 heterozygous) in 19 patients with 100 and more colorectal polyps. We don't found heterozygous mutations among 150 healthy controls Conclusion Frequency of germline mutations in MutYH gene among Russian patients with 4-99 and more than 100 colorectal polyps was 11,8% and 10,5% respectively Aim Blood selenium (Se) levels associated with significantly lower risk of cancers has been identified in Polish females from families with hereditary breast cancers (HBC). For BRCA1 mutation carriers: 70-89 μg/l at age <50 yrs (OR~12) and 95-120 μg/l at age ≥50 yrs (OR~4); for females without detected BRCA1 mutation: 98-108 μg/l (OR~5). The main goal of SELINA is validation of hypothesis that optimization of Se level can decrease the risk of malignancies. Method 7000 females (including 1200 BRCA1 carriers) from families with HBC and deficiency or excess of Se are qualified to one of the arms: "placebo", observational, supplement (Sodium Selenite) or diet modification. Blood Se level will be measured and optimized during 5 yrs.

Results
Recruitment will be closed in 2018. Conclusion SELINA is the first trial aimed to decrease the risk of cancers by active control of blood selenium levels . Interested scientists are welcome for collaboration. The nationwide Lynch Syndrome Registry of Finland (LSRFi) was founded in 1982 to organize endoscopic surveillance for high-risk families with colorectal cancer (CRC). To date, there are 298 families with confirmed pathogenic variants of mismatch repair (MMR) genes. Currently, LSRFi organises genetic counselling and predictive testing in research setting and co-ordinates endoscopic surveillance that takes place mostly in centralized public hospitals. Colonoscopy surveillance is offered from 25 years onwards, with 3-year interval for those with no prior cancer. LSRFi has access to national healthcare registries, such as registry for causes of death, parish registries and Finnish cancer registry. About 3,000 individuals have undergone genetic testing, so far. In May 2018, there were total of 1,416 path_MMR carriers; 1,044 path_MLH1 (74%), 246 path_MSH2 (17%), 123 path_MSH6 (9%) and 3 path_PMS2 (0.2%). The mean age for live carriers was 53 years for path_MLH1, 53 years for path_MSH2, 60 years for path_MSH6 and 48 years for path_PMS2. In 2015, about two thirds of eligible children (age >18 years) of verified path_MMR carriers had undergone predictive testing. Adherence to offered surveillance is high, well over 90%. CRC incidence, stage and survival do not differ from other countries compared to independent prospective datasets in Europe. Patients are informed about their inclusion into the registry, which generally contained data on family history, clinical information, age at onset and results of DNA testing or tumour screening in the diagnosis of LS. Written informed consent was obtained from all patients during genetic counselling sessions. Results From our registry, 61 suspected families fulfilled AMS criteria or Bethesda guidelines. Seventeen families (28%) had MMR deficiency and underwent genetic MMR testing. Path_ MLH1 variants was identified in 3 (21%) families, path_ MSH2/EPCAM variants in 11 (72%) families and path_ PMS2 variants in 1 family (7%). LS carriers have been identified with a mean age of 37.5 years (range 18-57) and a mean of 13 follow-up years.

Conclusion
The path_MSH2 variants are the most frequently identified in our registry and we provides support to set or improve LS genetic testing in South America. In addition, despite the small number of our registry, we described patients with a young age of onset and/or a positive family history of LS-associated cancers without an identified path_MMR variant, and may suggest the involvement of pathogenic variants in as yet undiscovered genes.
which generally contained data on family history, clinical information, age at onset and results of DNA testing or tumor screening. Written informed consent was obtained from all patients during genetic counseling sessions.

Results
In our registry, we have an overall record of 221 suspected families (with 533 registered individuals), 107 are Lynch syndrome suspected families, 98 familial adenomatous polyposis, 11 Peutz-Jeghgers syndrome, 2 juvenil polyposis, 1 Cowden syndrome and 2 hyperplastic polyposis. In total, 88 families present a mutation or variant of uncertain significant in APC (41), MUTYH (3), MLH1 (21), MSH2 (7), MSH6 (1), PMS2 (3), EPCAM (2), STK11 (8), PTEN (1) and SMAD4 (1) genes. In those families with pathogenic or likely pathogenic mutations, we have studied 386 relatives, of which, 223 are carriers and 163 are no carriers. All families have received clinical recommendations based on the National Comprehensive Cancer Network (NCCN) guidelines. Interestingly, 25 mutations have not yet been described in other studies, clearly demonstrating the relevance of evaluating different racial/ethnic populations like ours, which include an admixture of Amerindian and European -mainly Spanish -populations. Conclusion Our work shows the success to integrate multidisciplinary professionals as coloproctologists, PhD in biological sciences (genetic counselor), nurses, medical doctors from various disciplines, and the constant support of a psycho-oncologist. We would like to highlight our last challenge, a pioneering initiative in Latin America, which consisted in the creation of a Course of genetic counseling in hereditary cancer aimed for health care professionals belonging to oncology units.

Introduction
Since 1996, the Uruguayan Collaborative Group (UCG), a nonprofit organization is devoted to the registry, diagnosis, management and investigation of hereditary cancer. UGC is integrated by a multidisciplinary team of experts and represents in the country, a reference center for genetic counselling and risk assessment. Objective: To present an updated Uruguayan mutation catalog for gastrointestinal (GI) hereditary cancer susceptibility Methodology The UCG registry is integrated by 1536 non-related families. 548 families (35%) are defined as GI-high risk population following the National Comprehensive Cancer Network 2018 guidelines. These families were classified as: Amsterdam I-II, Bethesda, Li Fraumeni, Peutz Jeghers, Familial Adenomatous Polyposis, MUTYH-Associated Polyposis, or Serrated polyposis syndrome. Selected probands for genetic testing signed informed consent prior to obtain saliva or blood samples. Several DNA-analysis techniques were used over these 22 years, from Sanger sequencing alone (until 2010), Next Generation Sequencing of a group of genes and large rearrangements detection methods, to nowadays, panels of 30 genes.

Results
At present a total of 234 (43%) GI-high risk, non-related probands were tested and 63 families were diagnosed. We found 49 different mutations, classified according to ACMG as "Pathogenic" and distributed among the following genes: MLH1 (9), MSH2(11), PMS2(3), MSH6 (3), EPCAM(1), APC (11), STK11(2), NF1(1), FAN1(1), RAD51(1), SDHB(1), BMPR1A(1), MUTYH biallelic (3). A family carried a mutation class 4 (likely Pathogenic) in MLH1. In nine probands with a characteristic hereditary colon cancer phenotype, only MUTYH monoallelic mutations were found. An increasing number of variant of uncertain significance were found. Conclusion A research period of 22 years has unveiled the mutational spectrum of GI-high risk cancer of the Uruguayan population, allowing a broader vision regarding hereditary cancer profile in an understudied population. In spite of the large gene selection, only a few were involved in cancer predisposition. Lynch Syndrome, as expected, was the most frequent diagnosis, but with a relatively low pathogenic variant presentation.

Introduction
Mutations in BRCA1 or BRCA2 genes are considered the most prevalent cause of hereditary breast and ovarian cancer syndrome (HBOC), although other genes also explain this kind of affection. Since 2014, the Uruguayan Collaborative Group (UCG), a nonprofit organization is devoted to the registry, diagnosis, management and research of hereditary cancer, has been recruiting high-risk family groups with HBOC.
Objective: To report about pathogenic variants in BRCA and non-BRCA genes detected in Uruguayan high-risk for HBOC population. Methodology From the UCG registry, 592 non-related are defined as HBOC-high risk population following the National Comprehensive Cancer Network 2018 guidelines. Selected probands for genetic testing signed informed consent prior to obtain saliva or blood samples. Different approaches for searching gene mutations have been employed. At first, Next Generation Sequencing of BRCA1 and BRCA2, then large rearrangements detection methods were used, and lately multigene panels have been employed. Results 330 (56%) HBOC-high risk, non-related probands were tested, 56 were found positives and 49 different pathogenic variants identified. BRCA1-2 accounted for 31 (66%) pathogenic mutations (14 BRCA1 and 17 BRCA2) while mutations in non-BRCA genes were: PALB2(3) ATM(1) CHEK2(3) BARD1(3), TP53(6), CHD1(1), NBN(1).

Conclusion
Even though only HBOC high risk probands were selected, a relatively high proportion of non-BRCA genes presented with pathogenic variants. Although multigene panels can give unexpected and uninformative results, when used with thoughtfulness, they can be a valuable tool capable of diagnose beyond the traditional boundaries of BRCA genes. Despite technological improvements, a high number of families with no molecular diagnosis still remains. Since the role of constitutive epimutations in cancer development can be underestimated, future approaches will include a methylation screening.