Supplementary Methods, Figures 1-5. Supplementary Figure 1. Outline of cohort selection. Supplementary Figure 2. Example of whole-genome single-nucleotide polymorphism (SNP)array profile and genome alteration print (GAP). Supplementary Figure 3. Pre-treatment serum CA125 levels. Supplementary Figure 4. Progression-free and overall survival in patients with SBT-EOC and EOC without adjacent SBT. Supplementary Figure 5. Progression-free and overall survival of patients with Grade 3 SBTEOC and Grade 3 EOC without adjacent SBT

Supplementary Highlighted Methods. This document contains the revised supplementary methods and results with changes highlighted

Supplementary Tables 1-7. Supplementary Table 1. SBT-EOC cases used for molecular analyses Supplementary Table 2. Mutations Screened on the Sequenom/OncoMap3 Platform Supplementary Table 3. Primer sequences for the indicated exons of TP53, ERBB2 and NRAS Supplementary Table 4. Clinico-pathologic characteristics of the subset of SBT-EOC used for molecular analysis compared to the entire SBT-EOC cohort Supplementary Table 5A. Somatic mutations found in borderline and invasive ovarian tumors Supplementary Table 5B. Genes in which somatic mutations were found according to grade and co-existing serous borderline regions Supplementary Table 6. Genes differentially expressed between paired borderline and invasive regions of SBTEOC (p<0.01, not corrected for multiple comparison). Supplementary Table 7. Clinico-pathological characteristics of SBTEOC cases used for progesterone receptor (PGR) immunostaining

PDF file, 747K, Supplementary data: sequencing primers, integration of CDKN2A/2B copy number events with oncogenic mutations, tissue microarray IHC scoring, representative images of IHC staining, copy number segmentation.

Low-grade serous ovarian carcinoma (LGSOC) is associated with a poor response to existing chemotherapy, highlighting the need to perform comprehensive genomic analysis and identify new therapeutic vulnerabilities. The data presented here represent the largest genetic study of LGSOCs to date (n = 71), analysing 127 candidate genes derived from whole exome sequencing cohorts to generate mutation and copy-number variation data. Additionally, immunohistochemistry was performed on our LGSOC cohort assessing oestrogen receptor, progesterone receptor, TP53, and CDKN2A status. Targeted sequencing identified 47% of cases with mutations in key RAS/RAF pathway genes (KRAS, BRAF, and NRAS), as well as mutations in putative novel driver genes including USP9X (27%), MACF1 (11%), ARID1A (9%), NF2 (4%), DOT1L (6%), and ASH1L (4%). Immunohistochemistry evaluation revealed frequent oestrogen/progesterone receptor positivity (85%), along with CDKN2A protein loss (10%) and CDKN2A protein overexpression (6%), which were linked to shorter disease outcomes. Indeed, 90% of LGSOC samples harboured at least one potentially actionable alteration, which in 19/71 (27%) cases were predictive of clinical benefit from a standard treatment, either in another cancer's indication or in LGSOC specifically. In addition, we validated ubiquitin-specific protease 9X (USP9X), which is a chromosome X-linked substrate-specific deubiquitinase and tumour suppressor, as a relevant therapeutic target for LGSOC. Our comprehensive genomic study highlighted that there is an addiction to a limited number of unique 'driver' aberrations that could be translated into improved therapeutic paths. (c) 2020 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Objective Ovarian fibromas and adenofibromas are rare ovarian tumours. They are benign tumours composed of spindle-like stromal cells (pure fibroma) or a mixture of fibroblast and epithelial components (adenofibroma). We have previously shown that 40% of benign serous ovarian tumours are likely primary fibromas due to the neoplastic alterations being restricted to the stromal compartment of these tumours. We further explore this finding by comparing benign serous tumours to pure fibromas. Results Performing copy number aberration (CNA) analysis on the stromal component of 45 benign serous tumours and 8 pure fibromas, we have again shown that trisomy of chromosome 12 is the most common aberration in ovarian fibromas. CNAs were more frequent in the pure fibromas than the benign serous tumours (88% vs 33%), however pure fibromas more frequently harboured more than one CNA event compared with benign serous tumours. As these extra CNA events observed in the pure fibromas were unique to this subset our data indicates a unique tumour evolution. Gene expression analysis on the two cohorts was unable to show gene expression changes that differed based on tumour subtype. Exome analysis did not reveal any recurrently mutated genes.

Mucinous ovarian carcinoma (MOC) is a unique subtype of ovarian cancer with an uncertain etiology, including whether it genuinely arises at the ovary or is metastatic disease from other organs. In addition, the molecular drivers of invasive progression, high-grade and metastatic disease are poorly defined. We perform genetic analysis of MOC across all histological grades, including benign and borderline mucinous ovarian tumors, and compare these to tumors from other potential extra-ovarian sites of origin. Here we show that MOC is distinct from tumors from other sites and supports a progressive model of evolution from borderline precursors to high-grade invasive MOC. Key drivers of progression identified are TP53 mutation and copy number aberrations, including a notable amplicon on 9p13. High copy number aberration burden is associated with worse prognosis in MOC. Our data conclusively demonstrate that MOC arise from benign and borderline precursors at the ovary and are not extra-ovarian metastases.

While the genomics of high-grade serous carcinoma are well-studied in large international consortia, the less common subtypes have been neglected. We have sought to rectify this gap by analyzing international collections of low-grade serous and mucinous ovarian carcinomas and their putative benign and borderline precursors. Exome sequencing and copy number analysis of low-grade serous carcinomas (n=9) and serous borderline tumours (n=13) and targeted sequencing and copy number analysis in additional carcinomas (n=10) and borderline tumours (n=44) identified recurrent mutations in novel drivers such as EIF1AX and USP9X, as well as the known drivers KRAS, BRAF and NRAS. Copy number changes including 9p and 1p losses were significantly associated with progression from borderline to carcinoma. Exome and targeted sequencing analysis of mucinous carcinomas (GAMuT study) found a surprisingly high proportion (~50%) with TP53 mutations, and mutations in new drivers like RNF43 and ELF3. Despite similarities in early RAS/RAF pathway oncogenic drivers and CDKN2A disruption, the genetics of these two subtypes are otherwise distinct, suggesting differing etiologies and selective pressures. We also present here the first whole-genome sequencing analysis of a high-grade mucinous ovarian carcinoma collected from multiple sites at autopsy (CASCADE study). The patient, aged just 41 when diagnosed with Stage I mucinous ovarian carcinoma, had a 26-month progression-free interval, including normal CA125 and CA19-9 measurements at 21 months. The primary tumor was mostly borderline in appearance, with only a small focus of carcinoma. At autopsy, the carcinoma was widespread in the body, and whole-genome sequencing data was obtained from deposits in the omentum, iliac lymph node, para-aortic lymph node and upper diaphragm. These data were compared to the primary ovarian tumor and nine other sites sampled at autopsy. Citation Format: Kylie L. Gorringe, Matthew Wakefield, Sally M. Hunter, Georgina L. Ryland, Dane Cheasley, Michael S. Anglesio, Michael Christie, Raghwa Sharma, Antill Yoland, Simone M. Rowley, Jason Li, Blake Gilks, Prue E. Allan, Andrew N. Stephens, Sumi Ananda, Jan Pyman, Martin Koebel, Jessica McAlpine, Charlie Gourley, David G. Huntsman, Anna deFazio, David DL Bowtell, Ian G. Campbell, Clare Scott. Genomics analyses of less common epithelial ovarian cancer subtypes. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: Exploiting Vulnerabilities; Oct 17-20, 2015; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(2 Suppl):Abstract nr B08.

Objective: A family history of prostate cancer (PC) is a well-recognized high-risk factor for the development of clinically significant PC. To date, traditional linkage and association studies have identified only a limited number of genes and specific gene variants that account for only a small proportion of PC risk. To identify novel PC predisposition genes we performed whole-exome sequencing of PC-affected men from families with a significant history of PC.Methods and materials: Exome sequencing was performed on 5 PC-affected men from 3 families where there were multiple cases of PCs and where diagnostic testing returned a negative result for BRCA1 and BRCA2 mutations. Genotyping was performed for all potentially predisposing variants detected within each family on the affected and unaffected male participants.Results: Essential splice site, missense, and stop-lost variants were filtered against a recently published candidate gene list. A total of 19 truncating variants and 17 missense variants were identified for genotyping in all prostate-affected and unaffected male participants. hi all, 3 missense variants, PCTP, MCRS1, and ATRIP, demonstrated complete segregation and 1 missense variant, PARP2, demonstrated partial segregation with PC. hi addition, 3 truncating variants, CYP3A43, DOK3, and PLEKHH3, demonstrated complete segregation and 3 truncation mutations, HEATR5B, GPR124, and HKR1, demonstrated partial segregation with PC. No segregating variants between the 3 families were shared.Conclusions: In all, 10 truncating or missense variants showed either complete or partial segregation with PC in the relevant families. CYP3A43 and PARP2 variants have been shown to occur in other familial PCs and our findings add to the contribution that these variants potentially have in the risk and development of PC in BRCAX cases. (C) 2016 Elsevier Inc. All rights reserved.

Background Low-grade fibromatosis-like spindle cell carcinomas are very rare breast carcinomas comprising <0.5% of all breast cancers. They demonstrate immunohistochemical (IHC) features of basal-like/metaplastic breast carcinomas, but the underlying molecular characteristics are unknown. We hypothesised that, as with IHC similarities, there may be common genomic alterations between spindle cell and basal-like/metaplastic carcinomas. Methods and results Genomic mutational profile and genomic copy number aberration (CNA) analyses were performed on three cases of this unusual entity, and findings were compared with that reported for basal-like/metaplastic breast carcinomas. Copy number analyses by molecular inversion probe assays of the three spindle cell carcinoma samples revealed little overall genomic CNAs with only minor changes identified (fraction of the genome altered; 1.3%–6.4%), but with a common 9p21.3 loss in 2 out of 3 samples, with CDKN2A (p16) being a likely candidate. No areas of commonality were identified in an in silico analysis compared with publically available basal-like/metaplastic carcinoma copy number data. Conclusions These tumours are characterised by low genomic instability, and share no CNAs with other metaplastic carcinomas. These findings favour this entity being a unique group genotype and belie their apparent homogeneous morphology and phenotype.

BACKGROUND:Mucinous ovarian tumors are an unusual group of rare neoplasms with an apparently clear progression from benign to borderline to carcinoma, yet with a controversial cell of origin in the ovarian surface epithelium. They are thought to be molecularly distinct from other ovarian tumors but there have been no exome-level sequencing studies performed to date.METHODS:To understand the genetic etiology of mucinous ovarian tumors and assess the presence of novel therapeutic targets or pathways, we undertook exome sequencing of 24 tumors encompassing benign (5), borderline (8) and carcinoma (11) histologies and also assessed a validation cohort of 58 tumors for specific gene regions including exons 4-9 of TP53.RESULTS:The predominant mutational signature was of C>T transitions in a NpCpG context, indicative of deamination of methyl-cytosines. As well as mutations in known drivers (KRAS, BRAF and CDKN2A), we identified a high percentage of carcinomas with TP53 mutations (52 %), and recurrent mutations in RNF43, ELF3, GNAS, ERBB3 and KLF5.CONCLUSIONS:The diversity of mutational targets suggests multiple routes to tumorigenesis in this heterogeneous group of tumors that is generally distinct from other ovarian subtypes.

Low grade serous ovarian tumours are a rare and under-characterised histological subtype of epithelial ovarian tumours, with little known of the molecular drivers and facilitators of tumorigenesis beyond classic oncogenic RAS/RAF mutations. With a move towards targeted therapies due to the chemoresistant nature of this subtype, it is pertinent to more fully characterise the genetic events driving this tumour type, some of which may influence response to therapy and/or development of drug resistance. We performed genome-wide high-resolution genomic copy number analysis (Affymetrix SNP6.0) and mutation hotspot screening (KRAS, BRAF, NRAS, HRAS, ERBB2 and TP53) to compare a large cohort of ovarian serous borderline tumours (SBTs, n = 57) with low grade serous carcinomas (LGSCs, n = 19). Whole exome sequencing was performed for 13 SBTs, nine LGSCs and one mixed low/high grade carcinoma. Copy number aberrations were detected in 61% (35/57) of SBTs, compared to 100% (19/19) of LGSCs. Oncogenic RAS/RAF/ERBB2 mutations were detected in 82.5% (47/57) of SBTs compared to 63% (12/19) of LGSCs, with NRAS mutations detected only in LGSC. Some copy number aberrations appeared to be enriched in LGSC, most significantly loss of 9p and homozygous deletions of the CDKN2A/2B locus. Exome sequencing identified BRAF, KRAS, NRAS, USP9X and EIF1AX as the most frequently mutated genes. We have identified markers of progression from borderline to LGSC and novel drivers of LGSC. USP9X and EIF1AX have both been linked to regulation of mTOR, suggesting that mTOR inhibitors may be a key companion treatment for targeted therapy trials of MEK and RAF inhibitors.

Background: Intraductal carcinoma of the prostate (IDC-P) is a distinct clinicopathologic entity associated with aggressive prostate cancer (PCa). PCa patients carrying a breast cancer 2, early onset (BRCA2) germline mutation exhibit highly aggressive tumours with poor prognosis.Objective: To investigate the presence and implications of IDC-P in men with a strong family history of PCa who either carry a BRCA2 pathogenic mutation or do not carry the mutation (BRCAX).Design, setting, and participants: Patient-derived xenografts (PDXs) were generated from three germline BRCA2 mutation carriers and one BRCAX patient. Specimens were examined for histologic evidence of IDC-P. Whole-genome copy number analysis (WG-CNA) was performed on IDC-P from a primary and a matched PDX specimen.Outcome measurements and statistical analysis: The incidence of IDC-P and association with overall survival for BRCA2 and BRCAX patients were determined using KaplanMeier analysis.Results and limitations: PDXs from BRCA2 tumours showed increased incidence of IDC-P compared with sporadic PCa (p = 0.015). WG-CNA confirmed that the genetic profile of IDC-P from a matched (primary and PDX) BRCA2 tumour was similar. The incidence of IDC-P was significantly increased in BRCA2 carriers (42%, n = 33, p = 0.004) but not in BRCAX patients (25.8%, n = 62, p = 0.102) when both groups were compared with sporadic cases (9%, n = 32). BRCA2 carriers and BRCAX patients with IDC-P had significantly worse overall and PCa-specific survival compared with BRCA2 carriers and BRCAX patients without IDC-P (hazard ratio [HR]: 16.9, p = 0.0064 and HR: 3.57, p = 0.0086, respectively).Conclusions: PDXs revealed IDC-P in patients with germline BRCA2 mutations or BRCAX classification, identifying aggressive tumours with poor survival even when the stage and grade of cancer at diagnosis were similar. Further studies of the prognostic significance of IDC-P in sporadic PCa are warranted.Patient summary: Intraductal carcinoma of the prostate is common in patients with familial prostate cancer and is associated with poor outcomes. This finding affects genetic counselling and identifies patients in whom earlier multimodality treatment may be required. (C) 2014 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Ductal carcinoma in situ (DCIS) is a non-obligate precursor of invasive breast cancer and a frequent mammographic finding requiring treatment. Up to 25% of DCIS can recur and half of recurrences are invasive, but there are no reliable biomarkers for recurrence. We hypothesised that copy number aberrations could predict likelihood of recurrence. We analysed a cohort of pure DCIS cases treated only with wide local excision for genome-wide copy number and loss of heterozygosity using Affymetrix OncoScan MIP arrays. Cases included those without recurrence within 7 years (n=25) and with recurrence between 1 and 5 years after diagnosis (n=15). Pure DCIS were broadly similar in copy number changes compared with invasive breast cancer, with the consistent exception of a greater frequency of ERBB2 amplification in DCIS. There were no significant differences in age or ER status between the cases with a recurrence vs those without. Overall, the DCIS cases with recurrence had more copy number events than the DCIS without recurrence. The increased copy number appeared non-random with several genomic regions showing an increase in frequency in recurrent cases, including 20q gain, ERBB2 amplification and 15q loss. Copy number changes may provide prognostic information for DCIS recurrence, but validation in additional cohorts is required.

BACKGROUND:Using whole exome sequencing to predict aberrations in tumours is a cost effective alternative to whole genome sequencing, however is predominantly used for variant detection and infrequently utilised for detection of somatic copy number variation.RESULTS:We propose a new method to infer copy number and genotypes using whole exome data from paired tumour/normal samples. Our algorithm uses two Hidden Markov Models to predict copy number and genotypes and computationally resolves polyploidy/aneuploidy, normal cell contamination and signal baseline shift. Our method makes explicit detection on chromosome arm level events, which are commonly found in tumour samples. The methods are combined into a package named ADTEx (Aberration Detection in Tumour Exome). We applied our algorithm to a cohort of 17 in-house generated and 18 TCGA paired ovarian cancer/normal exomes and evaluated the performance by comparing against the copy number variations and genotypes predicted using Affymetrix SNP 6.0 data of the same samples. Further, we carried out a comparison study to show that ADTEx outperformed its competitors in terms of precision and F-measure.CONCLUSIONS:Our proposed method, ADTEx, uses both depth of coverage ratios and B allele frequencies calculated from whole exome sequencing data, to predict copy number variations along with their genotypes. ADTEx is implemented as a user friendly software package using Python and R statistical language. Source code and sample data are freely available under GNU license (GPLv3) at http://adtex.sourceforge.net/.

Targeted resequencing by massively parallel sequencing has become an effective and affordable way to survey small to large portions of the genome for genetic variation. Despite the rapid development in open source software for analysis of such data, the practical implementation of these tools through construction of sequencing analysis pipelines still remains a challenging and laborious activity, and a major hurdle for many small research and clinical laboratories. We developed TREVA (Targeted REsequencing Virtual Appliance), making pre-built pipelines immediately available as a virtual appliance. Based on virtual machine technologies, TREVA is a solution for rapid and efficient deployment of complex bioinformatics pipelines to laboratories of all sizes, enabling reproducible results. The analyses that are supported in TREVA include: somatic and germline single-nucleotide and insertion/deletion variant calling, copy number analysis, and cohort-based analyses such as pathway and significantly mutated genes analyses. TREVA is flexible and easy to use, and can be customised by Linux-based extensions if required. TREVA can also be deployed on the cloud (cloud computing), enabling instant access without investment overheads for additional hardware. TREVA is available at http://bioinformatics.petermac.org/treva/.