Merete Bjørnslett

Novel mutations of the suppressor gene PTEN in colorectal carcinomas stratified by microsatellite instability‐ and TP53 mutation‐ status

PTEN regulates cell homeostasis by inhibiting growth signals transduced through PI3‐kinases. The gene is mutated in several cancer types, but so far, only a limited number of mutations have been reported in colorectal cancer. In the present study, direct sequencing was used to analyze the whole coding region and exon‐intron boundaries of PTEN in a series of microsatellite stable (n=34) and microsatellite unstable (n=30) colorectal carcinomas with known TP53 mutation status. We detected 21 PTEN mutations in altogether 13 tumors (20%), including 19 mutations in the coding sequence and two in the exon‐intron boundaries. Sixteen of these alterations have not been previously reported in colorectal cancer. Furthermore, seven out of the 13 altered tumors harbored more than one mutation, potentially leading to loss of gene function. Finally, all PTEN mutations found were in tumors harboring wild‐type TP53. In conclusion, PTEN is mutated in a significant subgroup of colorectal carcinomas, and our findings further extend the previously small spectrum of reported PTEN changes. Additionally, it seems that mutations in PTEN and TP53 are mutually exclusive for this cancer type.

MDM2 promoter SNP344T>A (rs1196333) status does not affect cancer risk

The MDM2 proto-oncogene plays a key role in central cellular processes like growth control and apoptosis, and the gene locus is frequently amplified in sarcomas. Two polymorphisms located in the MDM2 promoter P2 have been shown to affect cancer risk. One of these polymorphisms (SNP309T>G; rs2279744) facilitates Sp1 transcription factor binding to the promoter and is associated with increased cancer risk. In contrast, SNP285G>C (rs117039649), located 24 bp upstream of rs2279744, and in complete linkage disequilibrium with the SNP309G allele, reduces Sp1 recruitment and lowers cancer risk. Thus, fine tuning of MDM2 expression has proven to be of significant importance with respect to tumorigenesis. We assessed the potential functional effects of a third MDM2 promoter P2 polymorphism (SNP344T>A; rs1196333) located on the SNP309T allele. While in silico analyses indicated SNP344A to modulate TFAP2A, SPIB and AP1 transcription factor binding, we found no effect of SNP344 status on MDM2 expression levels. Assessing the frequency of SNP344A in healthy Caucasians (n = 2,954) and patients suffering from ovarian (n = 1,927), breast (n = 1,271), endometrial (n = 895) or prostatic cancer (n = 641), we detected no significant difference in the distribution of this polymorphism between any of these cancer forms and healthy controls (6.1% in healthy controls, and 4.9%, 5.0%, 5.4% and 7.2% in the cancer groups, respectively). In conclusion, our findings provide no evidence indicating that SNP344A may affect MDM2 transcription or cancer risk.

t(14;22)(q32;q11) in non-Hodgkin lymphoma and myeloid leukaemia : molecular cytogenetic investigations

Two non‐Hodgkin lymphomas (NHL), one chronic lymphocytic leukaemia/small lymphocytic lymphoma and one diffuse large B‐cell lymphoma and three cases of myeloid leukaemia, two chronic (CML) and one acute (AML), showed, by G‐banding analysis, apparently identical chromosomal translocations t(14;22)(q32;q11), in three of the cases as the sole abnormality. Fluorescence in situ hybridisation (FISH) analysis with locus‐specific probes for ABL at 9q34 [bacterial artificial chromosomes (BACs) 835J22 and 1132H12], IGH at 14q32 [P1 artificial chromosome (PAC) 998D24] and IGL (PAC 1019H10) and BCR (BAC 74M14) at 22q11, as well as multicolour in situ hybridisation (M‐FISH) analyses were performed. A three‐way variant translocation of the classical t(9;22)(q34;q11), t(9;22;14)(q34;q11;q32), involving both BCR and ABL, was unravelled by the molecular cytogenetic investigations in the three myeloid leukaemia cases; a similar variant translocation has previously been reported in seven CML. The two cases of NHL (one NHL with a similar 14;22‐translocation has been reported previously) had no involvement of BCR or ABL, but instead the IGH and IGL genes were shown to be juxtaposed by the t(14;22)(q32;q11). How such a rearrangement with recombination of IGH and IGL might elicit a pathogenetic effect is completely unknown.

Effect of the MDM2 promoter polymorphisms SNP309T>G and SNP285G>C on the risk of ovarian cancer in BRCA1 mutation carriers

BackgroundWhile BRCA mutation carriers possess a 20-40% lifetime risk of developing ovarian cancer, knowledge about genetic modifying factors influencing the phenotypic expression remains obscure. We explored the distribution of the MDM2 polymorphisms SNP309T>G and the recently discovered SNP285G>C in Norwegian patients with BRCA related ovarian cancer.Methods221 BRCA related ovarian cancer cases (BRCA1; n = 161 and BRCA2; n = 60) were tested for the MDM2 polymorphisms. Results were compared to healthy controls (n = 2,465).ResultsThe SNP309G allele was associated with elevated OR for ovarian cancer in BRCA1 mutation carriers (SNP309TG: OR 1.53; CI 1.07-2.19; p = 0.020; SNP309GG: OR 1.92; CI 1.19-3.10; p = 0.009; SNP309TG+GG combined: OR 1.61; CI 1.15-2.27; p = 0.005). In contrast, the SNP285C allele reduced risk of BRCA1 related ovarian cancer in carriers of the SNP309G allele (OR 0.50; CI 0.24-1.04; p = 0.057). Censoring individuals carrying the SNP285C/309G haplotype from the analysis elevated the OR related to the SNP309G allele (OR 1.73; CI 1.23-2.45; p = 0.002). The mean age at disease onset was 3.1 years earlier in carriers of SNP309TG+GG as compared to carriers of SNP309TT (p = 0.068). No such associations were found in BRCA2 related ovarian cancer.ConclusionsOur results indicate the SNP309G allele to increase and the SNP285C allele to reduce the risk of BRCA1 related ovarian cancer. If confirmed in independent studies, this finding may have implications to counseling and decision-making regarding risk reducing measures in BRCA1 mutation carriers.

White Blood Cell BRCA1 Promoter Methylation Status and Ovarian Cancer Risk

Women carrying germline BRCA1 mutations are at high risk for epithelial ovarian cancer, especially high-grade serous ovarian cancer (HGSOC) (1, 2). Thus, 4% to 10% of all women with ovarian cancer may carry BRCA1 germline mutations (3, 4). Recently, germline mutations in other genes, including PALB2, BRIP1, RAD51C, and RAD51D, all acting in the same DNA repair pathway as BRCA1 and BRCA2, have been associated with familial risk for ovarian and breast cancer (58). These findings are consistent with the hypothesis that disturbances in DNA repair by homologous recombination are important for cancer development in these organs. Promoter methylation represents an alternative mechanism of gene inactivation, and promoter methylation of the MLH1 mismatch repair gene, as well as BRCA1 methylation, has been observed in normal tissues in some families with a high risk for colorectal or breast cancer who do not have germline mutations in these genes (912). For breast and ovarian cancer risk, associations with BRCA1 methylation in white blood cells (WBCs) have not been consistent (10, 1316), and most studies have been small, with limited statistical power to detect any clear differences. We hypothesized that normal tissue BRCA1 methylation may be associated with an increased risk for ovarian cancer, with a particular propensity for HGSOC, analogous to observations for germline BRCA1 mutations. Here, we determined WBC BRCA1 promoter methylation status in a large casecontrol study and then attempted to replicate the findings in a similarly designed validation study. To determine whether normal tissue BRCA1 methylation may be established early in life, we also assessed WBC BRCA1 methylation in samples of newborn girls and healthy young women. Methods Study Design Overview We compared WBC BRCA1 promoter methylation status between patients with ovarian cancer and population control participants. The initial study was followed by a similarly designed casecontrol study to validate our results. In addition, we performed extensive sensitivity analyses to test the robustness of our findings. Initial Study White blood cell DNA was available from 934 patients with epithelial ovarian cancer treated at Oslo University Hospital, Norwegian Radium Hospital, between 1993 and 2011 (Figure 1, A). All samples collected in the biobank during that period were included; however, borderline ovarian tumors were excluded. Also, all patients had been tested for pathogenic BRCA1 or BRCA2 germline mutations, and patients carrying such mutation were excluded. Samples were collected before any systemic chemotherapy, with 583 samples collected from patients after surgery and 351 collected from patients before surgery or from those who did not have surgery. As a control, we used random samples of women from CONOR (Cohort of Norway), a large collection of population studies in Norway with similar questionnaire data, clinical measurements, and blood samples (17). The CONOR participants were recruited between 1994 and 2003 (for details, see Supplement Table 1). Thus, in the initial casecontrol study, 1698 women without cancer were frequency matched by 5-year age categories (at blood sampling) to the 934 patients with ovarian cancer. Among participants with successful BRCA1 methylation analysis, the patients (age range, 15 to 90 years; median, 62 years) were somewhat older than the control participants (range, 20 to 93 years; median, 57 years); therefore, we adjusted for age at blood sampling in the casecontrol comparisons. Figure 1. Flow chart of patients with ovarian cancer and healthy control participants included and successfully analyzed in the initial (A) and validation (B) studies. HGSOC = high-grade serous ovarian cancer; LGSOC = low-grade serous ovarian cancer; qPCR = quantitative polymerase chain reaction. Supplement. Supplementary Material and Methods To assess whether the percentage of methylated alleles would display a doserisk association with ovarian cancer, we analyzed methylation-positive samples by pyrosequencing. Validation Study We estimated sample size for the validation study on the basis of the strength of association found for the HGSOC group in our initial study (Supplement). Thus, we analyzed WBC DNA from 607 patients with ovarian cancer (274 from Oslo University Hospital and 333 from Haukeland University Hospital, Bergen), including 286 with HGSOC, and 1984 population control participants randomly selected from CONOR, with frequency matching in 5-year age increments. The Oslo patients with ovarian cancer had blood collected in 2011 to 2015, and the Bergen patients in 2001 to 2015 (Figure 1, B). Among the patients with ovarian cancer, 433 had blood collected before and 174 after surgery. All the Oslo patients had been tested for pathogenic BRCA1 and BRCA2 germline mutations, with negative results, whereas the Bergen patients had not been tested. As in the initial casecontrol study, samples were collected before chemotherapy began. The control group drawn from the CONOR study (17) did not overlap with that of the initial study. We adjusted for age at blood sampling by using a procedure similar to that of the initial study. Newborns and Healthy Young Persons To assess whether normal tissue BRCA1 methylation might be established early in life, we determined WBC methylation status in umbilical cord blood from a sample of newborn girls (n= 611) from MoBa (the Norwegian Mother and Child Cohort Study) (18). In addition, WBC methylation was determined in a group of healthy women aged 20 to 25 years (n= 292) selected from the CONOR study (17). These 2 groups were not part of the control groups used in the casecontrol comparisons (Supplement). Tumor and Normal Tissue BRCA1 Methylation We hypothesized that WBC BRCA1 methylation may be a surrogate marker of BRCA1 tissue methylation in general. If correct, one would expect that many women with ovarian cancer and positive BRCA1 methylation status would have BRCA1 methylation in their tumor tissue. To address these questions, we analyzed normal and ovarian cancer tissue samples from patients with and without positive WBC methylation status (Supplement). Methylation Status Among Persons Carrying BRCA1/2 Germline Mutations To explore a potential relationship between WBC BRCA1 methylation and BRCA1/2 germline mutation, we determined BRCA1 methylation status in 251 patients with ovarian cancer and germline BRCA1 mutations, 100 with BRCA2 mutations, and 15 with familial ovarian cancer who had negative results on BRCA1/2 mutation testing. Sensitivity Analysis Previous studies suggested that cancer burden influences WBC global DNA methylation patterns (19). To evaluate whether tumor load affects BRCA1 promoter methylation in blood, we compared BRCA1 methylation frequency between samples obtained before and after surgery; we also compared methylation status among FIGO (International Federation of Gynecology and Obstetrics) stages across the pooled groups of patients with ovarian cancer and those with HGSOC. We also assessed the potential effect of previous breast cancer treatment as well as of sample storage time. Further, we measured WBC BRCA1 methylation status in a separate cohort of 658 patients with ovarian cancer, from whom blood samples were collected at various time points after chemotherapy, when most patients had limited or no detectable residual disease. Previously, global DNA methylation assessments suggested that methylation of some CpG (cytosineguanine) nucleotide sequence sites across the genome might vary among different WBC fractions (20). Thus, we examined published data sets for potential variation among WBC fractions in adults (21) as well as newborns (18) to identify any variation with respect to methylation of the BRCA1 promoter CpGs of relevance to the present study (Supplement). Laboratory Analysis In brief, DNA was isolated from the samples and bisulfite converted (>99.5% validated conversion rate) before BRCA1 promoter methylation status was determined by quantitative polymerase chain reaction (qPCR). Two researchers (E.O.B. and L.M.), who were blinded to sample identity, independently scored each sample as methylation positive or negative. Further details, including assessment of individual CpG methylation across the promoter area, methylation quantification by pyrosequencing, and haplotype analysis, are given in the Supplement. Data Analysis We compared BRCA1 methylation status between patients with ovarian cancer and population control participants by using logistic regression models adjusted for age, and we reported odds ratios (ORs) from these models. In addition, we looked for heterogeneity by using a likelihood ratio test to compare results between the initial and validation studies. Because germline BRCA1 mutation carriers seem to have a propensity for HGSOC, we analyzed this subgroup separately. In a separate analysis pooling cases and controls from the 2 studies (initial and validation), we assessed the frequency of BRCA1 methylation within 10-year age groups by using logistic regression, stratified for the 2 studies. The patient and control samples from the initial study analyzed by pyrosequencing were dichotomized according to the median value of methylation, and ORs were calculated separately for each group (Supplement). We used chi-square tests to compare the proportions of persons with BRCA1 methylation between groups. The precision of the estimates of the proportion with BRCA1 methylation was presented with 95% CIs. Potential confounding by other unknown covariates was assessed by using the method described by VanderWeele (22) (Supplement). The statistical analyses were performed in Stata, version 12.1 (StataCorp), and SPSS, version 19 (IBM). The study was conducted according to the STREGA (Strengthening the Reporting of Genetic Association Studies) and GRIPS (Strengthening the Reporting of Genetic Risk Prediction Studies) statements (23, 24). Ethical Considerations

C77G in PTPRC (CD45) is no risk allele for ovarian cancer, but associated with less aggressive disease

The pan lymphocyte marker CD45 exists in various isoforms arising from alternative splicing of the exons 4, 5 and 6. While naïve T cells express CD45RA translated from an mRNA containing exon 4, exons 4–6 are spliced out to encode the shorter CD45R0 in antigen-experienced effector/memory T cells. The SNP C77G (rs17612648) is located in exon 4 and blocks the exon’s differential splicing from the pre-mRNA, enforcing expression of CD45RA. Several studies have linked C77G to autoimmune diseases but lack of validation in other cohorts has left its role elusive. An incidental finding in an ovarian cancer patient cohort from West Norway (Bergen region, n = 312), suggested that the frequency of C77G was higher among ovarian cancer patients than in healthy Norwegians (n = 1,357) (3.0% vs. 1.8% allele frequency). However, this finding could not be validated in a larger patient cohort from South-East Norway (Oslo region, n = 1,198) with 1.2% allele frequency. Hence, C77G is not associated with ovarian cancer in the Norwegian population. However, its frequency was increased in patients with FIGO stage II, endometrioid histology or an age at diagnosis of 60 years or older indicating a possible association with a less aggressive cancer type.

MDM2 promoter polymorphism del1518 (rs3730485) and its impact on endometrial and ovarian cancer risk

BackgroundThe del1518 (rs3730485) polymorphism is an in/del variant in the MDM2 promoter P1. The variant is in complete linkage disequilibrium with MDM2 SNP309 (rs2279744) and has previously been found associated with an increased risk of colon cancer. In this study we assessed the impact of MDM2 del1518 on risk of ovarian and endometrial cancer.MethodsHere, we genotyped del1518 in two large hospital-based series of patients diagnosed with ovarian (n = 1,385) or endometrial (n = 1,404) cancer and performed risk estimations as compared to the genotype distribution among 1,872 healthy female controls.ResultsIn overall analysis we observed no association between del1518 and risk of either ovarian or endometrial cancer. However, stratifying according to SNP309 status, we found the del1518 variant to be associated with a reduced risk of endometrial cancer among individuals carrying the SNP309TT genotype both in the dominant (OR = 0.64; 95% CI = 0.45 – 0.90) and the recessive model (OR = 0.80; 95% CI = 0.65 – 1.00). No such association was observed for ovarian cancer risk.ConclusionWe found the MDM2 del1518 del variant to be associated with reduced risk of endometrial cancer among individuals carrying the MDM2 SNP309TT genotype.

Abstract 4613: MDM4 SNP 34091 (rs4245739) effect on risk of breast, colon, lung, prostate, endometrial and ovarian cancer

Background MDM4 enhances MDM29s E3 ligase activity causing ubiquitin-proteasome-dependent degradation of p53. Thus, elevated levels of both MDM4 and MDM2 may result in p53 inactivation and elevated cancer risk. A single nucleotide polymorphism in the MDM4 3′ UTR (SNP34091A>C; rs4245739) has been found to exert biological function as the SNP34091C allele creates a putative target site for hsa-miR-191 leading to specific hsa-miR-191 down-regulation of MDM4. Further, the SNP34091AA genotype is associated with increased risk for both recurrence and tumor related death in estrogen negative ovarian cancer patients. In the present study we assessed the potential effect of MDM4 SNP34091A>C on cancer risk in six major cancer forms. Materials and methods We analyzed 1717 breast-, 1331 lung-, 1531 colon-, 2501 prostate cancer cases and 3747 healthy controls form the same population based study (CONOR) as well as 1404 endometrial- and 1699 ovarian cancer cases from hospital based studies. All samples were genotyped for MDM4 SNP34091A>C using custom Light-SNiP assay on a LightCycler 480 II instrument. Potential associations between MDM4 SNP34091 and cancer risk were estimated by calculating Odds Ratio (OR) with 95% confidence intervals (CI). All statistical analyses were performed using the IBM SPSS 19 software. Results We observed no significant association between MDM4 SNP34091A>C status and cancer risk in any of the analyzed cancer forms. Interestingly, stratifying the ovarian cancer samples according to grade and histology, we observed a reduced risk of high grade serous ovarian cancer in patients harboring the MDM4 SNP34091AA genotype (OR = 0.80; 95% CI = 0.65 - 0.98). Stratifying according to MDM2 SNP309 status we found the MDM4 SNP34091A allele to be associated with increased risk for breast cancer (OR = 2.10; 95% CI = 1.08 - 4.10) in patients carrying the SNP309 GG genotype. Conclusions The data presented here indicate that the effect of MDM4 SNP34091 on cancer risk may be tissue specific and that there may be cooperative effects with MDM2 SNP309. Citation Format: Liv B. Gansmo, Merete Bjornslett, Anne Dorum, Helga Salvesen, Pal Romundstad, Kristian Hveem, Lars Vatten, Per Eystein Lonning, Stian Knappskog. MDM4 SNP 34091 (rs4245739) effect on risk of breast, colon, lung, prostate, endometrial and ovarian cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4613. doi:10.1158/1538-7445.AM2015-4613

Abstract 2833: Mutation analysis of cancer drivers and DNA repair genes in chemosensitive versus resistant ovarian cancers

Background: The mechanisms underlying resistance to chemotherapy in ovarian cancer are incompletely understood. Identifying genetic alterations associated with treatment response is decisive in the determination of which patients may benefit from adjuvant chemotherapy. Methods: Biopsies were collected from twenty patients diagnosed with ovarian cancer who were subjected to post-operative taxane- and platinum-containing chemotherapy. Patients were selected for genetic analyses based on response to chemotherapy, determined as time to relapse (10 sensitive and 10 resistant patients). A panel of 620 genes, including known cancer driver genes, as well as genes involved in DNA repair were analysed by massively parallel sequencing. Alignment and mutation calling was performed using MiSeq Reporter, with further manual filtering of variants to exclude common SNPs. Validation of low quality mutation calls was done by Sanger sequencing. Results: A median of 6 genes (range: 3 - 45) per patient was found to harbour non-synonymous mutations. Among previously identified driver genes in ovarian cancer, we found mutations in TP53, BRCA1, CDK12, NF1 and CSMD3. These mutations were more common among patients with more advanced disease and higher grade. For example, TP53 mutations were found in 10 out of 12 patients with high grade, stage 3c or 4 disease, and in 2 out of 5 with lower stage and/or grade. One patient was found to have a tumor potentially of a hyper-mutator phenotype with 49 mutations in 45 genes identified within our gene panel. With respect to treatment efficacy, 73 and 40 genes were found to be mutated exclusively in patients with a good and poor response to treatment, respectively. Conclusion: We describe the profile of mutations in cancer driver genes and DNA repair genes among patients suffering from ovarian cancer according to treatment response. Citation Format: Einar Birkeland, Rakel Blaalid, Merete Bjornslett, Anne Dorum, Per Eystein Lonning, Stian Knappskog. Mutation analysis of cancer drivers and DNA repair genes in chemosensitive versus resistant ovarian cancers. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2833. doi:10.1158/1538-7445.AM2014-2833

Effect of WBC BRCA1 promoter methylation on ovarian cancer risk.

5029 Background: Recently, some studies have reported BRCA1 WBC hypermethylation to be associated with increased risk of breast cancer. We assessed WBC BRCA1 promoter methylations in ovarian (OC) and breast cancer (BC) patients and healthy controls. In an OC subgroup harboring WBC BRCA1 methylation, we confirmed promoter methylation status in normal tissue. METHODS WBC DNA from 899 OC patients, 425 BC patients and 719 healthy controls were analyzed for BRCA1 promoter methylation by methylation specific PCR. In addition, we analyzed WBC DNA from 256 OC and 393 BC patients in blood samples drawn prior to diagnosis in a population-based study. We also analyzed 24 BC patients with a family history (BRCAPRO scores > 80%; Manchester score >40) without BRCA1/2 mutations. Paraffin-embedded normal tissue from 5 OC patients harboring WBC BRCA1 methylation was analyzed for BRCA1 methylation status. Finally, to determine potential risk factors in cis, we investigated BRCA1 haplotype status in a sub-cohort of 10 individuals with WBC methylated BRCA1and 13 controls. RESULTS We detected WBC BRCA1 promoter hypermethylation in 2.4% of healthy controls and 3.1% of BC patients (all BC individuals). No difference in BRCA1 methylation incidence was recorded between BC patients with blood samples drawn before (4.1%) or after (2.1%) diagnosis. In contrast, we detected WBC BRCA1 methylation in 10.3% of OC patients having blood samples drawn at diagnosis. Among OC patients with blood sampling 0-13 years (median 4.6 y) prior to diagnosis, BRCA1 promoter methylation was detected in 6.6%. BRCA1 promoter methylation was associated with a non-significant elevated risk of BC (HR 1.302; 95% CI 0.697-2.431) but a significantly increased risk of OC in the cohort of patients with blood samples drawn at time of (OR 4.765; CI 2.814-8.069) or prior to (OR 2.937; CI 1.476-5.845) diagnosis. We confirmed BRCA1 promoter methylation in normal tissue from all 5 individuals analyzed, excluding WBC promoter methylation being due to circulating DNA contamination. No association between BRCA1methylation and promoter haplotype was found. CONCLUSIONS WBC BRCA1 promoter methylation is associated with increased risk of ovarian cancer.This finding has clinical as well as biological implications..