Abstract Purpose Ovarian clear cell carcinoma (OCCC) is strongly associated with endometriosis and shows geographic variation in incidence. We investigated whether OCCC and adjacent endometriosis exhibit distinct transcriptional states and whether these patterns differ between United Kingdom (UK) and Japanese cohorts. Experimental Design We performed whole-transcriptome spatial profiling on specimens from 16 OCCC cases (8 UK, 8 Japan) in which tumor and endometriosis were both present. Gene expression was analyzed in tumor, endometriosis and stroma. ARID1A status was assessed by immunohistochemistry. Results Median age was 59 years (range 26–82). 13/16 cases (81.3%) had early-stage disease. Tissue compartment rather than cohort of origin was the dominant source of variation across endometriosis and tumor regions. Endometriosis was enriched for inflammatory and immune-related pathways compared to tumor, whilst there was greater representation of chromatin and protein–DNA complex assembly pathways in tumor regions. These patterns were conserved across both cohorts and after stratification by ARID1A status. Mesenchymal-associated gene expression scores also significantly differed across stroma, endometriosis and tumor with clear compartmental separation. Cell type deconvolution analyses showed clear compositional differences between stromal and epithelial disease compartments. Conclusions OCCC and coexisting endometriosis are transcriptionally distinct, with the dominant contrast being compartmental rather than geographic. ARID1A alone is unlikely to account for the principal spatial transcriptional states identified here. Further analyses will be required to ascertain whether these differences reflect genuine biological differences between OCCC and coexisting endometriosis or represent different stages of endometriosis-associated tumorigenesis. Translational Relevance Ovarian clear cell carcinoma often arises in association with endometriosis, yet the biological transition between these lesions remains poorly understood. Using spatial transcriptomics in matched tumor and adjacent endometriosis from Japanese and UK cohorts, we showed that endometriosis is characterized by inflammatory and antigen-presentation features, whereas tumor regions showed chromatin-organization and oncogenic transcriptional states. These patterns were largely maintained irrespective of ARID1A status and geographic background. In addition, spatial deconvolution suggested differences in local immune composition, with tumor regions showing relatively greater neutrophil- and T cell-associated signals. Together, our data suggest that OCCC and coexisting endometriosis share a spatially linked tissue context, but that tumor regions have distinct transcriptional profile and microenvironment that may be involved in the malignant transformation and inform interpretation of molecular classification in endometriosis-associated OCCC.

Objective: Adult-type granulosa cell tumors of the ovary (aGCTs) show ambiguous morphology and may be misdiagnosed as other tumors. Recently, heterozygous FOXL2 C402G mutations and TERT promoter C228T mutation have been reported as diagnostic and prognostic biomarkers of aGCTs. The objective of this study was to identify the characteristics of true aGCT cohort using these biomarkers in 72 aGCT samples. Methods: FOXL2 and TERT promoter mutational statuses of 64 primary and matched 8 recurrent aGCT samples were assessed. Non-aGCTs were excluded by the combination of FOXL2 mutational analysis and the pathological review. The characteristics and prognosis of molecularly/pathologically confirmed aGCTs (MP-aGCTs) were analyzed. Results: Of 18 FOXL2 wild-type (WT) tumors, 3 were excluded as they were of other histotype. None of 20 samples with the FOXL2 C402G mutation include other histotype. Clinical stage and age were prognostic factors for recurrence. Of the 61 MP-aGCTs, 46 harbored FOXL2 C402G mutation (44 heterozygous, 2 homozygous/hemizygous) and 15 had WT FOXL2. The presence of the FOXL2 mutation was associated with a worse prognosis. The mutational status of the TERT promoter in MP-aGCTs was 10 heterozygous and 51 WT. The TERT promoter mutation was highly identified in older patients and in larger tumors but had no prognostic impact. Conclusion: This is the first study to clearly demonstrate its practical application of FOXL2 in the diagnosis of aGCTs. Application of the molecular analysis to a large aGCT cohort is crucial for understanding true characteristics and establishing novel treatment strategy of this disease.

PARP inhibitors have changed the management of advanced high-grade epithelial ovarian cancer (EOC), especially homologous recombinant (HR)-deficient advanced high-grade EOC. However, the effect of PARP inhibitors on HR-proficient (HRP) EOC is limited. Thus, new therapeutic strategy for HRP EOC is desired. In recent clinical study, the combination of PARP inhibitors with anti-angiogenic agents improved therapeutic efficacy, even in HRP cases. These data suggested that anti-angiogenic agents might potentiate the response to PARP inhibitors in EOC cells. Here, we demonstrated that anti-angiogenic agents, bevacizumab and cediranib, increased the sensitivity of olaparib in HRP EOC cells by suppressing HR activity. Most of the γ-H2AX foci were co-localized with RAD51 foci in control cells. However, most of the RAD51 were decreased in the bevacizumab-treated cells. RNA sequencing showed that bevacizumab decreased the expression of CRY1 under DNA damage stress. CRY1 is one of the transcriptional coregulators associated with circadian rhythm and has recently been reported to regulate the expression of genes required for HR in cancer cells. We found that the anti-angiogenic agents suppressed the increase of CRY1 expression by inhibiting VEGF/VEGFR/PI3K pathway. The suppression of CRY1 expression resulted in decrease of HR activity. In addition, CRY1 inhibition also sensitized EOC cells to olaparib. These data suggested that anti-angiogenic agents and CRY1 inhibitors will be the promising candidate in the combination therapy with PARP inhibitors in HR-proficient EOC.

Logistic regression classifiers' normalized coefficient and confidence interval

ObjectiveAn effective treatment strategy for epithelial ovarian cancer (EOC) with homologous recombination (HR)-proficient (HRP) phenotype has not been established, although poly (ADP-ribose) polymerase inhibitors (PARPi) impact the disease course with HR-deficient (HRD) phenotype. Here, we aimed to clarify the cellular effects of paclitaxel (PTX) on the DNA damage response and the therapeutic application of PTX with PARPi in HRP ovarian cancer.MethodsTwo models with different PTX dosing schedules were established in HRP ovarian cancer OVISE cells. Growth inhibition and HR activity were analyzed in these models with or without PARPi. BRCA1 phosphorylation status was examined in OVISE cells by inhibiting CDK1, which was reduced by PTX treatment. CDK1 expression was evaluated in EOC patients treated with PTX-based neoadjuvant chemotherapy.ResultsPTX suppressed CDK1 expression resulting in impaired BRCA1 phosphorylation in OVISE cells. The reduced CDK1 activity by PTX could decrease HR activity in response to DNA damage and therefore increase the sensitivity to PARPi. Immunohistochemistry showed that CDK1 expression was attenuated in samples collected after PTX-based chemotherapy compared to those collected before chemotherapy. The decrease in CDK1 expression was greater with dose-dense PTX schedule than with the conventional PTX schedule.ConculsionsPTX could act synergistically with PARPi in HRP ovarian cancer cells, suggesting that the combination of PTX with PARPi may be a novel treatment strategy extending the utility of PARPi to EOC. Our findings provide cules for future translational clinical trials evaluating the efficacy of PTX in combination with PARPi in HRP ovarian cancer.

Diagnostic accuracy of clinicopathological factors using Machine Learning Algorithms

Pretreatment peripheral blood tests of 334 patients with epithelial ovarian cancer and 101 patients with benign ovarian tumor

Explanation and evaluation of the random forest (RF) classifier. (A) Schematic illustration of the classification of samples by the RF. (B, C) Representative classification trees from the discrimination between malignant and benign tumors. These trees are only representations out of 4,000 trees constructed in the RF classifier. The final class is determined as a result of voting by all 4,000 trees. (D, E) The highest accuracy of prediction (D) and the AUC (E) using different numbers of samples in RF classification between malignant and benign tumors. The mean accuracy and AUC with these 95% confidence intervals were presented for 10 independent sets of randomly selected data of 20%, 40%, 60%, and 80% of patients from the training and test cohorts.

Ovarian clear cell carcinoma (OCCC) is a subtype of epithelial ovarian cancer (EOC) that is associated with elevated interleukin-6 (IL-6) expression, resistance to chemotherapy, and increased mortality. Although bevacizumab (Bev) is a widely used anti-angiogenic agent for EOC, the efficacy of Bev and the role of IL-6 in modulating angiogenesis in OCCC are unknown. We performed tube formation assays using human umbilical vein endothelial cells (HUVEC) cultured in OCCC cell-conditioned medium and using cells directly co-cultured with OCCC cells. We observed that IL-6 inhibition significantly mitigated the ability of Bev to impede tube formation in both cases. Furthermore, IL-6 blockade disrupted the anti-angiogenic efficacy of Bev and its concomitant anti-tumor activity. In addition, IL-6 inhibition resulted in a significant increase in angiopoietin-1 (Ang1) secretion and decreased vascular endothelial growth factor (VEGF) expression. Clinical specimens also exhibited this reciprocal relationship between IL-6 and Ang1 expression. Finally, depletion of Ang1 abrogated the effects of IL-6 inhibition on Bev activity, demonstrating that IL-6 supports the anti-angiogenic activity of Bev by suppressing Ang1 expression and promoting dependence on VEGF for angiogenesis. Altogether, our data suggest that OCCC tumors with high IL-6 levels are candidates for Bev therapy.

AbstractPurpose: We aimed to develop an ovarian cancer–specific predictive framework for clinical stage, histotype, residual tumor burden, and prognosis using machine learning methods based on multiple biomarkers. Experimental Design: Overall, 334 patients with epithelial ovarian cancer (EOC) and 101 patients with benign ovarian tumors were randomly assigned to “training” and “test” cohorts. Seven supervised machine learning classifiers, including Gradient Boosting Machine (GBM), Support Vector Machine, Random Forest (RF), Conditional RF (CRF), Naïve Bayes, Neural Network, and Elastic Net, were used to derive diagnostic and prognostic information from 32 parameters commonly available from pretreatment peripheral blood tests and age. Results: Machine learning techniques were superior to conventional regression-based analyses in predicting multiple clinical parameters pertaining to EOC. Ensemble methods combining weak decision trees, such as GBM, RF, and CRF, showed the best performance in EOC prediction. The values for the highest accuracy and area under the ROC curve (AUC) for segregating EOC from benign ovarian tumors with RF were 92.4% and 0.968, respectively. The highest accuracy and AUC for predicting clinical stages with RF were 69.0% and 0.760, respectively. High-grade serous and mucinous histotypes of EOC could be preoperatively predicted with RF. An ordinal RF classifier could distinguish complete resection from others. Unsupervised clustering analysis identified subgroups among early-stage EOC patients with significantly worse survival. Conclusions: Machine learning systems can provide critical diagnostic and prognostic prediction for patients with EOC before initial intervention, and the use of predictive algorithms may facilitate personalized treatment options through pretreatment stratification of patients.

Accumulating evidence has indicated that microRNAs play a critical role in the pathogenesis of human cancers. microRNA-34a (miR-34a) has been shown to be a key regulator of tumor suppression by targeting several cancer-related signals, including the interleukin-6 receptor (IL-6R)/Signal Transducers and Activator of Transcription 3 (STAT3) signaling pathway. Previously, we determined that miR-34a expression was frequently reduced in high-grade serous carcinoma (HGSC), the major subtype of epithelial ovarian cancer (EOC). Considering that the IL-6R/STAT3 signaling pathway is upregulated and believed to be a potential therapeutic target in EOC, we investigated the biological significance of reduced miR-34a expression in HGSC with regard to IL-6R signaling. Additionally, we evaluated the viability of miR-34a as a therapeutic application for HGSC both in vitro and in vivo. Accordingly, we found that the ectopic expression of miR-34a significantly reduced tumor proliferation and invasion through downregulation of IL-6R expression, suggesting that reduced miR-34a expression might play an important role in the malignant potential of HGSC through upregulation of the IL-6R/STAT3 signaling pathway. Moreover, we demonstrated that replacement of miR-34a reduced tumorigenicity of HGSC in vivo. Therefore, this study may provide the rationale for miR-34a replacement as a promising therapeutic strategy for HGSC.

Accumulating evidence has indicated that microRNAs play a critical role in the pathogenesis of human cancers. microRNA-34a (miR-34a) has been shown to be a key regulator of tumor suppression by targeting several cancer-related signals, including the interleukin-6 receptor (IL-6R)/Signal Transducers and Activator of Transcription 3 (STAT3) signaling pathway. Previously, we determined that miR-34a expression was frequently reduced in high-grade serous carcinoma (HGSC), the major subtype of epithelial ovarian cancer (EOC). Considering that the IL-6R/STAT3 signaling pathway is upregulated and believed to be a potential therapeutic target in EOC, we investigated the biological significance of reduced miR-34a expression in HGSC with regard to IL6R signaling. Additionally, we evaluated the viability of miR-34a as a therapeutic application for HGSC both in vitro and in vivo. Accordingly, we found that the ectopic expression of miR-34a significantly reduced tumor proliferation and invasion through downregulation of IL-6R expression, suggesting that reduced miR-34a expression might play an important role in the malignant potential of HGSC through upregulation of the IL-6R/STAT3 signaling pathway. Moreover, we demonstrated that replacement of miR-34a reduced tumorigenicity of HGSC in vivo. Therefore, this study may provide the rationale for miR-34a replacement as a promising therapeutic strategy for HGSC.

Ovarian clear cell carcinoma (OCCC) exhibits distinct phenotypes, such as resistance to chemotherapy, poor prognosis and an association with endometriosis. Biomarkers and imaging techniques currently in use are not sufficient for reliable diagnosis of this tumor or prediction of therapeutic response. It has recently been reported that analysis of somatic mutations in cell-free circulating DNA (cfDNA) released from tumor tissues can be useful for tumor diagnosis. In the present study, we attempted to detect mutations in PIK3CA and KRAS in cfDNA from OCCC patients using droplet digital PCR (ddPCR). Here we show that we were able to specifically detect PIK3CA-H1047R and KRAS-G12D in cfDNA from OCCC patients and monitor their response to therapy. Furthermore, we found that by cleaving wild-type PIK3CA using the CRISPR/Cas9 system, we were able to improve the sensitivity of the ddPCR method and detect cfDNA harboring PIK3CA-H1047R. Our results suggest that detection of mutations in cfDNA by ddPCR would be useful for the diagnosis of OCCC, and for predicting its recurrence.

Objective Ovarian clear cell carcinoma (OCCC) frequently presents at an early stage. In stage I OCCC, the prognosis differs according to substage. In particular, predictive biomarkers and new treatment strategies are needed for stage IC2/IC3 disease. We investigated tumor biology and prognostic factors for stage I OCCC from a clinicopathological perspective, including the expression of ARID1A and IL-6, which are considered critical for OCCC carcinogenesis. Methods A retrospective cohort study of 192 patients with stage I OCCC treated at a single institution was performed. We calculated overall survival (OS) with respect to 12 clinicopathological parameters that included the unique and diverse histological features of OCCC. Results The estimated 5-year OS rate in patients with all stage I OCCC was 88.9% during a median of 91 months of follow-up. The multivariate analysis indicated that substage classification and IL-6 expression status were associated with poor OS (p = 0.010 and p = 0.027, respectively). Loss of ARID1A expression had no impact on survival; however, it was associated with substage (p = 0.001), capsule rupture status (p = 0.011), and ascites cytology (p = 0.016). No clear association was found between ARID1A and IL-6 expressions. Histological findings, including the presence of endometriosis, adenofibroma, architectural pattern, and tumor cell type, showed no prognostic effects. Conclusions Both substage classification and IL-6 expression status may be independent prognostic factors in stage I OCCC. Therefore, IL-6 molecular stratification may be crucial in optimizing therapeutic strategies for early stage OCCC to improve survival.

We studied the whole-genome point mutation and structural variation patterns of 133 tumors (59 high-grade serous (HGSC), 35 clear cell (CCOC), 29 endometrioid (ENOC), and 10 adult granulosa cell (GCT)) as a substrate for class discovery in ovarian cancer. Ab initio clustering of integrated point mutation and structural variation signatures identified seven subgroups both between and within histotypes. Prevalence of foldback inversions identified a prognostically significant HGSC group associated with inferior survival. This finding was recapitulated in two independent cohorts (n = 576 cases), transcending BRCA1 and BRCA2 mutation and gene expression features of HGSC. CCOC cancers grouped according to APOBEC deamination (26%) and age-related mutational signatures (40%). ENOCs were divided by cases with microsatellite instability (28%), with a distinct mismatch-repair mutation signature. Taken together, our work establishes the potency of the somatic genome, reflective of diverse DNA repair deficiencies, to stratify ovarian cancers into distinct biological strata within the major histotypes.

Abstract Background: Ovarian carcinoma is comprised of distinct histological subtypes with different etiology, molecular, genomic and clinical attributes. Patterns of genomic diversity and different treatment responses differentiate each ovarian cancer histotype. The relative patterns of both mutational, copy number and structural variation have not been studied with relation to each disease phenotype. We hypothesized that global genomic architectures will stratify ovarian cancer patients and reveal different treatment response groups. Methods: Whole genome sequencing was performed on 133 ovarian tumors, including 123 carcinomas (59 high-grade serous (HGSC), 35 clear cell (CCOC), 29 endometrioid (ENOC)) and 10 granulosa cell tumours (GCT). Profiles of copy number aberrations, loss of heterozygosity (LOH), mutations (SNVs and INDELs) and structural variations were assessed. Mutational characteristics including mutation signatures derived from tri-nucleotide substitution patterns together with genomic structural characteristics, such as the relative proportion of rearrangement types, reflective of specific DNA repair processes were calculated for each patient. Results: Integrative clustering of the 133 patients according to their mutation and structural signatures resulted in seven distinct subgroups of patients. LOH and the homologous recombination deficiency mutation signature mainly distinguished HGSC cases from non-serous histotypes. HGSC cases were further clustered into two main subgroups. One subgroup (n = 23, 39%) showed a high prevalence of foldback inversions with homology size >5bp, while the other group (n = 25, 42%) was enriched in tandem duplications and deletions, and associated with microhomology (<3bp). Survival analysis revealed that the foldback inversion group associated with poor overall and progression-free survival (logrank p-value = 0.016 and 0.015). CCOC cases were characterized by tandem duplications (Median = 39%, p-value <0.001). The mutation signatures further identified two main subgroups of CCOC; one (n = 10, 29%) showing prevalence of kataegis events typically associated with an APOBEC mutational signature, and the other (n = 17, 49%) characterized by an age-related signature. Enrichment of a mis-match repair defect signature identified a microsatellite instable subgroup of ENOC (n = 8). A signature related to breast cancers uniquely identified GCT cases. Conclusion: Our results suggest that mutational and chromosomal structural variant signatures (rearrangement and copy number profiles) constitute new and defining features of ovarian carcinoma that relate to different DNA repair mechanisms. Our results provide insight into divergent etiologies within histotypes and suggest a novel structure on which to base treatment. Citation Format: Yikan Wang, Ali Bashashati, Michael S. Anglesio, Dawn Cochrane, Diljot Grewal, Hugo Horlings, Anthony Karnezis, Anne-Marie Mes-Masson, Aikou Okamoto, Satoshi Yanagida, Nozomu Yanaihara, Misato Saito, Blake Gilks, Jessica McAlpine, Samuel Aparicio, David Huntsman, Sohrab Shah. Genomic consequences of aberrant DNA repair stratify ovarian cancer histotypes. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-324.

Previous studies have identified microRNA (miRNA) involvement in human cancers. This study aimed to elucidate potential clinical and biological associations of ovarian cancer-related miRNA gene expression profiles in high-grade serous carcinoma (HGSC) and ovarian clear cell carcinoma (OCCC). Accordingly, we investigated 27 patients with ovarian cancer (12 HGSC and 15 OCCC cases) using quantitative real-time reverse transcription polymerase chain reaction to determine the cancer-related miRNA expressions. Gene Cluster 3.0 was used for hierarchical clustering analysis, and differentially expressed miRNAs between HGSC and OCCC were identified by the class comparison analysis using BRB-Array Tools. An unsupervised hierarchical clustering analysis identified two distinct miRNA expression clusters, with histological subtype-related significant differences in the associations between clusters and clinicopathological features. A comparison of miRNA expression in HGSCs and OCCCs identified five miRNAs (miR-132, miR-9, miR-126, miR-34a, and miR-21), with OCCCs demonstrating a statistically higher expression. Further investigation of the biological significance of miR-9 overexpression in OCCC revealed that miR-9 inhibition reduced the cell invasion ability and upregulated E-cadherin expression. Using a luciferase reporter assay, we further demonstrated the direct binding of miR-9 to E-cadherin. Global cancer-related miRNA expression analysis identified statistically unique profiles that could discriminate ovarian cancer histotypes. In OCCC, miR-9 overexpression may affect pathogenesis by targeting E-cadherin, thereby inducing an epithelial-mesenchymal transition. Therefore, miR-9 may be a promising therapeutic target strategy for OCCC.

Among epithelial ovarian cancers, clear cell carcinoma of the ovary (CCC) has unique clinical and molecular characteristics that include chemoresistance resulting in poor prognosis. It was shown that CCC recently was characterized by specific upregulation of the IL‐6/IL‐6R‐signal transducer and activator of transcription 3 (Stat3) signaling pathway. In this study, we aim to clarify whether IL‐6/IL‐6R mediated signaling pathway could have clinical relations with CCC and to evaluate inhibitory effects of the pathway on CCC carcinogenesis. A total of 84 CCC cases collected from primary surgical specimens were evaluated by the immunohistochemical analysis for IL‐6R and phosphorylated Stat3 (pStat3), and we found that high IL‐6R expression correlated with poor patient survival both by the univariate and multivariate analyses, suggesting that IL‐6/IL‐6R signaling pathway could be implicated in the progression of CCC. We further investigated the effects of IL‐6/IL‐6R mediated signaling pathway inhibition either by IL‐6R small interfering RNA (siRNA) approach or humanized anti‐human IL‐6R antibody (tocilizumab) in CCC. Inhibition of endogenous IL‐6R including tocilizumab in CCC cells did reduce cell invasion ability and restored their response to cytotoxic reagent. These data suggest that IL‐6/IL‐6R signaling pathway could act on CCC cells to enhance invasion and chemoresistance and, therefore, targeting IL‐6/IL‐6R mediated signaling pathway could be a promising therapeutic strategy for CCC. © 2015 Wiley Periodicals, Inc.

To construct a profile of therapeutically actionable gene alterations in the major histological types of ovarian cancer, 72 Japanese patients with surgically resected ovarian cancers were selected from an original cohort consisting of 267 patients who had not received pre-treatment before surgery. Somatic mutations and copy number alterations at 740 hotspots in 46 cancer-related genes were detected by deep sequencing of genomic DNAs obtained from snap-frozen tumor tissues using a next generation sequencer. The alterations were verified by Sanger sequencing and quantitative genomic PCR. Mutations and/or copy number aberrations which will make tumors respond to molecular targeting drugs were detected in nine genes of 35/72 (48.6%) patients; PIK3CA (25.0%), KRAS (13.9%), ERBB2 (4.3%), PTEN (2.8%), RB1 (2.8%), CDKN2A (2.8%), AKT1 (1.4%), CTNNB1 (1.4%) and NRAS (1.4%). These mutations tended to occur in a mutually exclusive manner. Non-serous histological type tumors showed such actionable gene alterations frequently (32/47; 68.1%). Therefore, ovarian cancers, particularly of non-serous types, frequently carry gene aberrations that link to therapy using molecular targeting drugs.