Publications

Abstract Sarcomas are malignant soft tissue and bone tumours affecting adults, adolescents and children. They represent a morphologically heterogeneous class of tumours and some entities lack defining histopathological features. Therefore, the diagnosis of sarcomas is burdened with a high inter-observer variability and misclassification rate. Here, we demonstrate classification of soft tissue and bone tumours using a machine learning classifier algorithm based on array-generated DNA methylation data. This sarcoma classifier is trained using a dataset of 1077 methylation profiles from comprehensively pre-characterized cases comprising 62 tumour methylation classes constituting a broad range of soft tissue and bone sarcoma subtypes across the entire age spectrum. The performance is validated in a cohort of 428 sarcomatous tumours, of which 322 cases were classified by the sarcoma classifier. Our results demonstrate the potential of the DNA methylation-based sarcoma classification for research and future diagnostic applications.

<p>Whole exome sequencing data, including single nucleotide variants and copy number variants.</p>

Supplementary Figures 1-4 from Chromosomal Instability Correlates with Genome-wide DNA Demethylation in Human Primary Colorectal Cancers

<p>Genes included in the targeted next-generation sequencing panel</p>

<p>Supplementary Figure 5. Biochemical and cellular assays to determine KAT6B activity</p>

<p>Fisher exact test for each mutated pathway, comparing responders vs. non-responders at diagnosis</p>

<p>Comparative histological analysis of original non-treated orthoxenografts and their pairs with acquired resistance to CDDP after five cycles of treatment in mice</p>

TPS1114 Background: Identification of targeted therapies for advanced triple negative breast cancer (TNBC) remains as an important clinical challenge. TNBC frequently shows BRCA dysfunction and 80% of germline BRCA1 mutation (gBRCAm) carriers with BC diagnosis are TNBC patients (pts). In addition to germline mutations, epigenetic silencing by aberrant methylation of BRCA1/2 promoters can be responsible for a dysfunctional BRCA protein. Methylation of BRCA1 promoter occurs in 15-57% of TNBCs. Interestingly, BRCA1-methylated sporadic breast tumors display pathologic features and gene expression profiles similar to those of gBRCAm carriers, a phenotype called “BRCAness”. Olaparib (O) is an oral poly (ADP-ribose) polymerase (PARP) inhibitor approved by the FDA for treatment of gBRCA-mutated HER2-negative metastatic BC. Tutt et al. have shown antitumor activity in gBRCA-mutated advanced TNBC (54% Objective Response Rate [ORR]). We report an ongoing phase II clinical trial to analyze the O efficacy in advanced TNBC pts with BRCA1/2 promoter methylation (NCT03205761). Methods: Eligible pts are advanced TNBC cases with ≥ 1 prior treatment for advanced disease, without germline BRCA1/2 mutations and centrally confirmed somatic BRCA1/2 promoter methylation in metastatic lesions. Pts receive O 300 mg b.i.d. orally. Primary objective is to analyze O efficacy in terms of ORR according to RECIST 1.1 by investigator assessment. Secondary objectives include safety, other efficacy endpoints, and biomarker analyses. Sequential tumor and blood samples are collected to a) explore changes in methylation status, b) correlate BRCA-methylation status both in blood vs. tumor and in primary vs. metastatic counterpart lesions, and c) correlate methylation status with BRCA1/2 expression and outcome. Thirty-four evaluable pts are required based on an optimal two-stage Simon model (α error = 0.05, power = 80%, dropout rate = 10%). Recruitment started in October 2017, with 17 pts screened for methylation status, 3 pts enrolled by 31th January 2018, and 14 active sites out of 16 participants. Primary endpoint analysis is planned for Q2 2020. Clinical trial information: NCT03205761.

Tripartite motif containing 27 (TRIM27) is highly expressed in lung cancer, including non-small-cell lung cancer (NSCLC). Here, we profiled DNA methylation of lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) tumours from 613 early-stage NSCLC patients and evaluated associations between CpG methylation of TRIM27 and overall survival. Significant CpG probes were confirmed in 617 samples from The Cancer Genome Atlas. The methylation of the CpG probe cg05293407TRIM27 was significantly associated with overall survival in patients with LUSC (HR = 1.65, 95% CI: 1.30-2.09, P = 4.52 × 10-5), but not in patients with LUAD (HR = 1.08, 95% CI: 0.87-1.33, P = 0.493). As incidence of LUSC is associated with higher smoking intensity compared to LUAD, we investigated whether smoking intensity impacted on the prognostic effect of cg05293407TRIM27 methylation in NSCLC. LUSC patients had a higher average pack-year of smoking (37.49LUAD vs 54.79LUSC, P = 1.03 × 10-19) and included a higher proportion of current smokers than LUAD patients (28.24%LUAD vs 34.09%LUSC, P = 0.037). cg05293407TRIM27 was significantly associated with overall survival only in NSCLC patients with medium-high pack-year of smoking (HR = 1.58, 95% CI: 1.26-1.96, P = 5.25 × 10-5). We conclude that cg05293407TRIM27 methylation is a potential predictor of LUSC prognosis, and smoking intensity may impact on its prognostic value across the various types of NSCLC.

B‐cell translocation gene 2 ( BTG2 ) is a tumour suppressor protein known to be downregulated in several types of cancer. In this study, we investigated a potential role for BTG2 in early‐stage non‐small cell lung cancer (NSCLC) survival. We analysed BTG2 methylation data from 1230 early‐stage NSCLC patients from five international cohorts, as well as gene expression data from 3038 lung cancer cases from multiple cohorts. Three CpG probes (cg01798157, cg06373167, cg23371584) that detected BTG2 hypermethylation in tumour tissues were associated with lower overall survival. The prognostic model based on methylation could distinguish patient survival in the four cohorts [hazard ratio (HR) range, 1.51–2.21] and the independent validation set (HR = 1.85). In the expression analysis, BTG2 expression was positively correlated with survival in each cohort (HR range, 0.28–0.68), which we confirmed with meta‐analysis (HR = 0.61, 95% CI 0.54–0.68). The three CpG probes were all negatively correlated with BTG2 expression. Importantly, an integrative model of BTG2 methylation, expression and clinical information showed better predictive ability in the training set and validation set. In conclusion, the methylation and integrated prognostic signatures based on BTG2 are stable and reliable biomarkers for early‐stage NSCLC. They may have new applications for appropriate clinical adjuvant trials and personalized treatments in the future.

Abstract Background. BRCA1 mutated breast tumor cells are defective in DNA repair by homologous recombination and therefore especially sensitive to treatment with DNA double-strand break (DSB) inducing agents, such as alkylators or PARP inhibitors. However, such tumors can eventually develop therapy resistance. Understanding therapy resistance mechanisms may help in designing treatment strategies to overcome therapy resistance. Methods. We have developed BRCA1-deficient triple-negative breast cancer (TNBC) xenograft models by implantating fresh human breast tumor pieces and subsequent serial passaging. These models show epigenetic loss of BRCA1 due to promoter hypermethylation or genetic inactivation of BRCA1 due to a frameshift mutation (c.2210delC) resulting in a premature stop codon (p.Thr737LeufsX15). We have used these BRCA1-deficient TNBC models to study response and acquisition of resistance to alkylating therapy (cisplatin, melphalan) and the clinical PARP inhibitor olaparib. Results. Treated tumors were sensitive to the alkylators cisplatin or melphalan or the PARP inhibitor olaparib, in some cases resulting in complete remission with no palpable tumor left. However, relapses did occur in most cases and repeated treatment of recurrent tumors eventually led to aquired resistance. Since restoration of BRCA1 function has been suggested as a mechanism of therapy resistance (Swisher et al., Cancer Res 2008; 68: 2581), we determined BRCA1 expression in therapy-sensitive and -resistant tumors by Western blot analysis. While no full length BRCA1 protein could be detected in the therapy-sensitive tumors, expression of full length BRCA1 protein was found in the majority of cisplatin resistant and olaparib resistant tumors. BRCA1 reexpression in the therapy-resistant BRCA1-c.2210delC tumors was caused by genetic restoration of the reading frame due to reversion to the wildtype sequence or additional deletions near the c.2210delC mutation. BRCA1-re-expression in the therapy-resistant TNBC xenografts with epigenetic loss of BRCA1 was caused either by loss of BRCA1 promoter methylation or by complex rearrangements at the BRCA1 locus. Conclusion. Although BRCA1-deficient TNBC xenografts are initially very sensitive to alkylating agents and olaparib, resistance to treatment develops in almost all treated tumors. This acquired resistance is frequently associated with re-expression of BRCA1 due to secondary mutations or loss of promoter methylation. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-06-02.

Aberrant DNA methylation is recognized as being a common feature of human neoplasia.CpG island hypermethylation and global genomic hypomethylation occur simultaneously in the cancer cell. However, very little is known about the interindividual inherited susceptibility to these epigenetic processes. To address this matter, we have genotyped in 233 cancer patients (with colorectal, breast, or lung tumors), four germ-line variants in three key genes involved in the metabolism of the methyl group, methylene-tetrahydrofolate reductase, methionine synthase, and cystathionine beta-synthase, and analyzed their association with DNA methylation parameters. The epigenetic features analyzed were the 5-methylcytosine content in the genome of the tumors and their normal counterparts, and the presence of CpG island hypermethylation of tumor suppressor genes (p16(INK4a), p14(ARF), hMLH1, MGMT, APC, LKB1, DAPK, GSTP1, BRCA1, RAR beta 2, CDH1, and RASSF1). Two positive associations were found. First, carriers of genotypes containing the methylene-tetrahydrofolate reductase 677T allele show constitutive low levels of 5-methylcytosine in their genomes (P = 0.002), and tumors in these patients do not achieve severe degrees of global hypomethylation (P = 0.047). Second, tumors occurring in homozygous carriers of the methionine synthase 2756G allele show a lower number of hypermethylated CpG islands of tumor suppressor genes (P = 0.029). The existence of these associations may provide another example of the interplay between genetic and epigenetic factors in the cancer cell.

Abstract Fatty acids (FA) modify DNA methylation in vitro , but limited information is available on whether corresponding associations exist in vivo and reflect any short-term effect of the diet. Associations between global DNA methylation and FAs were sought in blood from lactating infants (LI; n = 49) and adult males (AMM; n = 12) equally distributed across the three conventional BMI classes. AMM provided multiple samples at 2-hour intervals during 8 hours after either a single Western diet-representative meal (post-prandial samples) or no meal (fasting samples). Lipid/glucose profile, HDAC4 promoter and PDK4 5’UTR methylation were determined in AMM. Multiple regression analysis revealed that global (in LI) and both global and PDK4 -specific DNA methylation (in AMM) were positively associated with eicosapentaenoic and arachidonic acid. HDAC4 methylation was inversely associated with arachidonic acid post-prandially in AMM. Global DNA methylation did not show any defined within-day pattern that would suggest a short-term response to the diet. Nonetheless, global DNA methylation was higher in normal weight subjects both post-prandially and in fasting and coincided with higher polyunsaturated relative to monounsaturated and saturated FAs. We show for the first time strong associations of DNA methylation with specific FAs in two human cohorts of distinct age, diet and postnatal development stage.

Age and ApoE status are among the most established non-modifiable risk factors to developing Alzheimer's disease (AD). It is also well known that human brain structure changes at both macro and microscopical level with healthy aging. The ApoE-ε4 allele is thought to impact with aging-related mechanisms and carriers are at increased risk of developing AD. AD patients carrying the e4 allele show increased pathological hallmarks and earlier age of onset. We present here results on microstructural differences found between cognitively healthy subjects with increased risk for AD as a function of their ApoE genotype using diffusion-weighted (DW) imaging. T1 and DW images were acquired from 514 subjects. ApoE genotyping was performed in all participants. We first applied denoising on both T1 and DWI using adaptive non-local means algorithm, then correction for eddy currents and of gradient directions. Diffusion models were estimated and mean diffusivity (MD) maps generated. DW B0 maps were warped to T1 space by coregistering white matter masks from both image types. We finally normalized MD maps from T1 space the MNI space. A multiple regression analysis was performed over MD maps to determine effects of genotype, age, and the interaction between both factors after accounting for gender and educational attainment. Images were also analyzed with Voxel-Based Morphometry (VBM). Our dataset included 245 non-carriers, 204 heterozygotes and 65 homozygotes for the ApoE-ε4 allele. We detected significant bilateral effects of genotype in the basal forebrain, in the pons as well as in orbitofrontal cortex. We also find some significant differences in posterior bilateral periventricular areas and in left corona radiata. In these regions carriers of the ApoE-ε4 allele showed decreased MD compared to non-carriers. These regions were not found in the VBM results.