10th Biennial Ovarian Cancer Research Research Articles

Abstract POSTER-TECH-1121: Rapid identification and targeting of chemotherapy-resistant tumor stem cell clones in ovarian cancer

Abstract Even though ovarian cancer is typically detected very late, advances in surgery and chemotherapy have resulted in the remarkable response in most patients such that cancer is essentially undetectable at six months. However, a frustratingly large number of these patients will see a return or “recurrence” of this disease over the following year against which chemotherapy is less effective. There is now increasing evidence that this recurrent disease is the result of a very small number of cancer cells in the original cancer that are somehow “different”. These differences allow these cancer cells to avoid death from the same chemotherapy drugs that kill the other tumor cells. As a result, these so-called “resistant” cells continue to grow in number that can appear as cancer 1-3 years later in the same patient. Thus oncologists know full-well how to kill 99.9% of ovarian cancer cells, and their most pressing problem in ovarian cancer today is understanding how to kill this very small number of resistant cells. To meet this challenge we have very recently developed an exciting and perhaps revolutionary technology that may permit us to identify these rare, resistant tumor cells very soon after the patient is diagnosed with ovarian cancer. In particular, we can grow these rare resistant cells as individual clones in a testtube so that we can employ very powerful techniques to determine at a molecular level how they are different and thereby identify drugs that specifically kill them before they have a chance to multiply and participate in the recurrence of ovarian cancer. Moreover, this information will be applied to develop a method of rapidly staining tissue sections of tumor to determine the number and type of resistant cells in the tumor before therapy is begun. Using this method, we hope to determine what changes in the resistant cells that allow them to circumvent chemotherapy, and identify drugs that specifically target these changes. We anticipate that given the frequency with which resistance appears in high-grade ovarian cancer patients, our strategy of rapidly identifying and understanding resistant cells in these patients will be relevant to all of these patients. Our long-term goal is to provide a rapid means of predicting resistance in a given patient prior to the onset of therapy and more effectively target resistant cells at the onset of treatment. The ultimate aim is to more effectively prevent the clinical relapse and provide a more lasting benefit and ultimate cure to the woman with ovarian cancer. Citation Format: Gang Ning, Yusuke Yamamoto, Ting Zhang, Xia Wang, Lane Wilson, Suzy Torti, Molly Brewer, Christopher Crum, Frank McKeon, Wa Xian. Rapid identification and targeting of chemotherapy-resistant tumor stem cell clones in ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-TECH-1121.

Abstract POSTER-BIOL-1329: The ovary is an alternative site of origin for high-grade serous ovarian cancer in mice

Abstract The tumor suppressor p53 is the most frequently mutated gene in high-grade serous carcinoma (HGSC), the most common and deadliest ovarian cancer. It remains unknown, however, whether p53 mutation can cause HGSC. To address this, we have incorporated a p53 activating mutation, p53R172H, into conditional Dicer-Pten double-knockout (DKO) mice, generating triple-mutant (TKO) mice. Like DKO mice, TKO mice develop metastatic HGSCs originating from the fallopian tube. Besides, even when fallopian tubes are removed in TKO mice, ovaries alone can spur their own metastatic HGSCs, indicating that p53 mutation can drive HGSC arising from the ovary. To further confirm the ability of p53 mutation to induce ovarian HGSC, we have generated p53R172H-Pten double-mutant (DMu) mice, one genetic control line for TKO mice. As anticipated, these DMu mice develop metastatic HGSCs from the ovary, which spread to the abdominal cavity before killing the mice, just like fallopian tube-deficient TKO mice. Our study therefore shows that the ovary, as well as the fallopian tube, can be a source of high-grade serous ovarian cancer. (Grant support: Marsha Rivkin Ovarian Cancer Challenge Grant; Ovarian Cancer Research Fund (OCRF); F32 NCI National Research Service Award (NRSA); and K99/R00 NCI Pathway to Independence Award) Citation Format: Jaeyeon Kim, Donna M. Coffey, Lang Ma, Martin M. Matzuk. The ovary is an alternative site of origin for high-grade serous ovarian cancer in mice [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1329.

Abstract POSTER-TECH-1103: Aspirin, NSAIDS, and ovarian cancer survival from electronic medical records

Abstract Background: Extensive literature indicates that aspirin and other non-steroidal anti-inflammatory drugs (NSAIDs) are associated with reduced cancer incidence and mortality. For ovarian cancer, a pooled analysis of 12 case-control studies recently demonstrated a significantly reduced risk for regular aspirin use and a similar but non-significant reduced risk for non-aspirin NSAID use. However, studies specifically evaluating NSAIDs and ovarian cancer survival are absent from the published literature. Purpose: To determine if aspirin and other NSAIDs are associated with ovarian cancer survival. Methods: Ovarian cancer cases were identified from de-identified electronic medical records (EMR) from the Vanderbilt University Medical Center. Ever use of aspirin or any NSAID was determined by MedEx, a natural language processing system designed to capture medication use from EMR. Linked tumor registry data was used to generate hazard ratios (HR) and confidence intervals (CI) for overall survival using Cox proportional hazards regression; covariates included age, race, stage of disease, histologic subtype, treatments received, and year of diagnosis. Results: The majority of 1,147 confirmed ovarian cancer cases were Caucasian (87.0%), had serous histologic subtypes (53.6%), and late stage (III/IV, 50.0%) or stage unknown (29.3%) disease. Ovarian cancer cases with any aspirin use (207, 18.1%) had significantly longer overall survival in both unadjusted (HR: 0.56, 95% CI: 0.45-0.70) and adjusted analyses (HR: 0.59, 95% CI: 0.46-0.74). Similarly, ovarian cancer cases with any non-aspirin NSAID use (248, 21.6%) had significantly longer overall survival in both unadjusted (HR: 0.60, 95% CI: 0.48-0.74) and adjusted analyses (HR: 0.69, 95% CI: 0.54-0.88). Associations between any aspirin or other NSAID use (360, 31.4%) and longer ovarian cancer survival did not significantly differ by stage of disease (P-interaction=0.650), histologic subtype (P-interaction=0.076), or treatment with a platinum-based agent (P-interaction=0.069). Conclusions: Indication of aspirin or non-aspirin NSAID use ascertained retrospectively from EMR was associated with improved overall ovarian cancer survival. Further evaluation, including timing and duration of use, are underway; replication of this association in additional study populations is needed. Citation Format: Alicia Beeghly-Fadiel, Gwendolyn Holman, Samantha P. Stansel, Gretchen Edwards, Edward Baeg, Hannah Berke, Ryan J. Delahanty, Dineo Khabele. Aspirin, NSAIDS, and ovarian cancer survival from electronic medical records [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-TECH-1103.

Abstract POSTER-BIOL-1346: Growth-suppressive effect of WRN RECQ helicase inactivation in breast and ovarian cancer cells

Abstract One important and high-yielding paradigm in the development of novel, targeted anticancer therapies is to identify genes that are not essential to normal cells but that have become critically important to particular cancer cells as they rely on these genes for survival. This concept, sometimes called addiction of cancer cells to specific genes or pathways can be exploited in the clinic because the function of the genes that cancer cells are addicted to can be disrupted using genetic or chemical manipulation, leading to selective killing of cancer cells while normal cells are minimally affected. For example, it was shown that homologous recombination-defective BRCA1-deficient cancer cells rely heavily on other repair pathways such as NHEJ, one of the pathways facilitated by PARP1. This finding has led to clinical trials designed to determine if PARP inhibitor drugs are an effective treatment for patients with BRCA1-deficient breast and ovarian cancer. We have previously discovered, via a high throughput screen, that simultaneous deficiency in BRCA1 and the WRN RECQ helicase resulted in additive growth-suppressive phenotype in human fibroblasts. WRN is a gene mutated in the Werner syndrome of premature aging, and it is a multifunctional exonuclease/helicase involved in maintenance of genomic stability. Congenital loss of WRN also predisposes to sarcomas and melanomas, though interestingly, not to the most common type of age-related cancer – carcinoma. Near-complete inactivation of WRN in human fibroblasts and common cell lines such as sarcoma-derived U2OS and HT1080, is typically non-lethal, though somewhat reduces cellular growth rate and fitness. Following our original observation, we were interested to determine whether loss of WRN is synthetically lethal with loss of BRCA1 in breast and ovarian cancer cell lines, and if loss of WRN restores sensitivity to platinum or impairs growth in platinum resistant clones derived from breast and ovarian cancer cell lines. Our results show that depletion of WRN in BRCA1-negative or positive breast or ovarian cancer cell lines dramatically suppresses their growth. Both in breast and ovarian cancer cells, at least some of this growth suppression could be attributed to the induction of RNAi-triggered cytotoxicity by WRN downregulation. In addition, WRN depletion also hypersensitized the HCC1937 breast cancer cell line to platinum, however this effect of WRN was attenuated in platinum-resistant clones derived in vitro from this line. Strikingly, non-cancerous mammary epithelial line MCF10a showed virtually no growth suppression by WRN depletion, suggesting that breast and ovarian cancer cells may be more reliant on WRN than normal cells of at least some of the relevant epithelial lineages — which is a therapeutically exploitable trait. We are currently following up on this observation using other non-cancerous epithelial cells, such as primary mammary and fallopian tube epithelial cells, as well as primary human keratinocytes. We will present these data as well as the update on reproducing WRN growth-suppressive phenotype using small molecule inhibitors of WRN. Significance: Our results highlight WRN as a potential novel target for development of breast and ovarian cancer therapies. Citation Format: Piri Welcsh, Keffy Kehrli, Paul Lazarchuk, Julia Sidorova. Growth-suppressive effect of WRN RECQ helicase inactivation in breast and ovarian cancer cells [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1346.

Abstract AS12: Integrating genomics, transcriptomics, and proteomics for the discovery of novel biomarkers to complement CA125 in ovarian cancer early detection

Abstract Recent work suggests that a 2-stage strategy incorporating serial measurements of the glycoprotein CA125 in patient serum and the Risk of Ovarian Cancer Algorithm (ROCA), followed by transvaginal ultrasound for women with high-risk scores, may have utility as a screening practice in the general population. However, CA125 only has a sensitivity of 50-60% for early stage disease, and is only present in approximately 80% of all ovarian cancers. Multiple marker panels have shown promise at increasing sensitivity and specificity over CA125 alone, but have not improved lead time for early cancer detection. Therefore, we have taken an unbiased in-depth approach to identify novel plasma biomarkers to complement CA125 in ovarian cancer early detection. Using high-resolution quantitative mass spectrometry, we have profiled the plasma proteome of preclinical patient samples, taken 1 year prior to CA125 elevation and diagnosis, and healthy controls from the Normal Risk Ovarian Screening (NROS) study. With this approach, we have identified over 30 biomarker candidates that are elevated in at least 50% of preclinical samples relative to controls in triplicate experiments, including cancer-specific post-translational modifications. As a means of preliminary validation of promising preclinical biomarkers, we have profiled the plasma proteome of 24 newly diagnosed, early stage ovarian cancer patients and 24 matched healthy controls. Further, half of these patients were selected with uninformative CA125 levels at diagnosis, allowing for the discovery of additional candidate biomarkers that may complement CA125. Correlation of preclinical and newly diagnosed patient plasma data with publicly available genomic and transcriptomic data, as well as proteomic data previously collected from ovarian cancer cell lines resulted in a prioritized list of candidate protein biomarkers and autoantibodies in patient plasma that may improve detection and lead time over CA125. For the most promising candidates, monoclonal antibodies and MagPlex assays are being optimized in order to determine their performance as useful biomarkers in additional patient sera. Multimarker panels will be tested in a training set for the development of a screening algorithm that will be evaluated in preclinical sera to determine the optimized set that improves detection and lead time over CA125. Citation Format: Clayton Boldt, Ph.D., Archana Simmons, Ph.D., Hong Wang, MS, Ph.D., Keith Baggerly, Ph.D., Robert Bast, Jr., M.D., Samir Hanash, M.D., Ph.D.. Integrating genomics, transcriptomics, and proteomics for the discovery of novel biomarkers to complement CA125 in ovarian cancer early detection [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS12.

Abstract POSTER-CTRL-1205: Neurocognitive evaluation of a cognitive training intervention in ovarian and gynecologic cancer survivors

Abstract Purpose: Our prior work indicates that a group-based cognitive training intervention can improve memory and attention in cancer survivors. This study examined potential for training transfer on two cognitive tasks and corresponding changes in neural activation patterns as measured by fMRI prior to and following the cognitive training intervention. Methods: Seven gynecologic cancer (ovarian, breast, uterine) survivors (mean age: 64 years, mean time since primary treatment: 7.2 years) underwent a 7-week cognitive rehabilitation intervention delivered in group format. Participants were evaluated with a comprehensive cognitive battery and two experimental cognitive tasks prior to and following treatment. A word pair task required participants to learn pairs of words and later indicate whether words had been seen before (Item condition) or whether words had been paired together at learning (Relational condition). A verbal working memory task required participants to learn a set of 3, 5, or 7 letters and confirm or deny the presence of a probe letter. Participants also underwent fMRI scanning during word pair and verbal working memory tasks prior to and following treatment. Results: Participants showed significant improvement following intervention in accuracy in the Relational condition, in which strong associations between presented word pairs were required for successful performance, t(3)=5.21, p< .05. By contrast, accuracy did not increase significantly in the Item condition following intervention, t(3)=.48, p=.66). There was also a significant improvement in accuracy for the intermediate load (5-item) condition of the verbal working memory task following intervention, t(3)=4.17, p<.05. fMRI analyses revealed activation in a frontoparietal network during both the word pairs and verbal working memory tasks. Future analyses will investigate changes in recruitment of this network following intervention. Conclusions: Cancer survivors show evidence of specific transfer of memory skills following a group-based cognitive training intervention as demonstrated by improvement on the word memory task- relational item. In addition, there was a trend toward improvement on a working memory task. Neuroimaging results suggest frontoparietal network involvement for the cognitive task as well as some indication of a more efficient neural response following training. These results suggest that cognitive training may have a beneficial impact on both behavioral measures of cognition as well as neural activation markers. Citation Format: H. Gray, M. K. Askren Thomas, K. Anderson, D. David, A. Singh, K. Murphy, S. Willis, H. Mendoza, M.M. Cherrier. Neurocognitive evaluation of a cognitive training intervention in ovarian and gynecologic cancer survivors [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-CTRL-1205.

Abstract AS28: Phase Ib/II with expansion of patients at the mtd study of olaparib plus weekly (metronomic) carboplatin and paclitaxel in relapsed ovarian cancer

Abstract BACKGROUND: To establish the maximum-tolerated dose (MTD) and evaluate dose-limiting toxicities (DLTs) and response to therapy of combination therapy with carboplatin/paclitaxel and olaparib, an oral tablet inhibitor of poly ADP ribose polymerase (PARP), in advanced (Stage III or IV) relapsed ovarian cancer. METHODS: Eligibility required measurable disease, adequate organ function and ECOG performance status of ≤ 2. Patients had to have failed first line platinum containing chemotherapy. All patients were tested for BRCA 1 and 2 mutations. Patients received the metronomic therapy of paclitaxel 60mg/m2 IV and carboplatin AUC 2 IV weekly, 3 weeks out of 4 and increasing doses of olaparib until the maximum tolerated dose was obtained. Olaparib started at 50 mg bid administered orally for 3 consecutive days (D1-D3), every week for each cycle. Patients were assessed for toxicity and response according to the protocol. Patients received combination therapy until DLT or disease progression. RESULTS: The maximum tolerated dose was found to be olaparib 150 mg bid administered orally, three (D1-D3) consecutive days of each week of each cycle. Total number of patients enrolled in the phase 1b part of this study was 14. Median age was 58 (range 42-77). The median number of prior therapeutic regimens was 4 (range 3-8). Median number of cycles on study to date is 9.3 and the mean is 7.3. There were no grade 4 toxicities. The most common grade 3 toxicities were neutropenia (11), leukopenia (7), lymphopenia (4), anemia (4), fatigue (1) and MDS (1). There was no evidence of GI, cardiac, hepatic, pulmonary or dermatologic toxicities in any of these patients. Two patients had an allergic reaction prior to participating in the trial and once enrolled were desensitized to carboplatin for each study visit. One patient had a mild allergic reaction to the chemotherapy on trial, but did not require desensitization. 4 patients had a complete response (CR), 3 had partial response (PR), 3 had stable disease (SD), 2 had progressive disease (PD) and 2 were not evaluable. Of the 4 CR's three were BRCA mutant and 1 had no known BRCA mutation. Of the PD's one was BRCA mutant. CONCLUSION: Olaparib, an oral tablet can be safely administered with a weekly regimen of carboplatin and paclitaxel in heavily pretreated, ovarian cancer patients. This is early preliminary data. This study is ongoing with a planned total expansion of 40 additional patients at the MTD. Citation Format: Rivkin SE, Iriarte D, Sloan, H, Wiseman, C, Moon J,. Phase Ib/II with expansion of patients at the mtd study of olaparib plus weekly (metronomic) carboplatin and paclitaxel in relapsed ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS28.

Abstract POSTER-THER-1418: PD-1 mediated paralysis of ovarian cancer infiltrating dendritic cells

Abstract Purpose: It is well known that the immune system impacts the clinical course of ovarian cancer suggesting that immune-based approaches may effective in treating the disease. This has led to several clinical trials of novel treatments such as antibody, vaccine, and adoptive T cell therapy. PD-1:PD-L1 axis is a major immune regulatory pathway that blunts immune effectors in the tumor microenvironment. It is well known that high surface expression of PD-1 on tumor infiltrating T cells is a sign of their exhaustion. However, the complete mechanism behind the PD-1 regulation of T cell exhaustion is yet to be explored. Recently we identified that PD-1 is also expressed on ovarian tumor-infiltrating myeloid dendritic cells (DCs) that exhibit an immunosuppressive phenotype. Our goal in this study was to understand the mechanism by which PD-1 mediates the paralysis of ovarian cancer infiltrating DCs. Experimental procedures: PD-1+ DCs obtained from ID8 mouse model of ovarian cancer were used in this study. Data from this preclinical model was confirmed by using myeloid DCs obtained from ovarian cancer patient’s samples. Using standard flow cytometry, immunoassays (ELISA, multiplexed cytokine assay), PCR array, western blot and confocal microscopy, PD-1 mediated paralysis of ovarian cancer DCs was determined. Results: Our data shows that blockade of surface PD-1 on ovarian cancer infiltrating DCs results in their activation by, enhancing their surface co-stimulatory molecules expression, increasing their motility, enhancing their antigen presentation capacity, and increasing their production of immunostimulatory cytokines. Also, our results suggest that PD-1 expressed on tumor DCs mediates the suppression of NFkB which occurs through SHP-2 and IKK dependent mechanisms. Conclusion: Our study reveals that blockade of PD-1 on ovarian cancer infiltrating DCs results in the reversal of their paralysis to such an extent that it overrides their B7-H1 mediated suppression of T cells and this occurs through the reversal of PD-1 mediated tonic suppression of NFkB. Citation Format: Lavakumar Karyampudi James Krempski, Purushottam Lamichhane, Kimberly R. Kalli, Marshall D. Behrens, Doris M. Vargas, Lynn C Hartmann, Karen E Hedin, Ellen L. Goode, Keith L. Knutson. PD-1 mediated paralysis of ovarian cancer infiltrating dendritic cells [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-THER-1418.

Abstract POSTER-THER-1428: Effects of atypical PKC inhibitors on ovarian cancer proliferation and RNA levels

Abstract Ovarian cancers are highly lethal tumors which account for approximately four percent of all women’s cancers and are the fifth leading cause of cancer-related death among women. At diagnosis the majority of patients have metastatic disease and the long-term survival remains low. Certain ovarian cancers are highly lethal tumors due to the emergence of therapy-resistant ovarian cancer cells. Protein kinase C-iota (PKC-ι) has been shown to aid in the ability of cancer cells to resist drug-induced apoptosis. Recently, it has been reported that PKC-ι which is located in chromosome 3 at 3q26.2 is the most common genomic amplicon as identified by comparative genomic hybridization [Proc Natl Acad Sci U S A 102:12519–12524(2005)]. Additionally, an increase in PKC-ι DNA copy number was associated with decreased progression-free survival of ovarian cancer patients. Moreover, only PKC-ι gene amplification is highly correlated with protein overexpression, tumor size, lymph node metastasis and clinical stage out of four genes studied on the 3q26 amplification [Genes Chromosomes Cancer 47:127-136(2008)]. The focus of this research was to determine the in-vitro efficacy of the pan-atypical PKC inhibitor, 2-acetyl-1,3-cyclopentanedione (ACPD; Diabetes. 2014 Apr 4) and the novel PKC-iota inhibitor, ICA-1, (1H-imidazole-4-carboxamide, 5-amino-1-[2,3-dihydroxy-4-[(phosphonooxy) methyl]cyclopentyl]-,[1R-(1α, 2β, 3 β, 4 α) [The Inter. J. Biochem. & Cell Biol. 43:784-794(2011)] on HEY ovarian cancer cell proliferation and RNA concentration. In contrast to ACPD which inhibits both PKC-ι and PKC-zeta (PKC-ζ), ICA-1 specifically inhibits the activity of PKC-ι but not (PKC-ζ). The effects of ACPD and ICA-1on HEY proliferation was quantified by Trypan blue exclusion assay. RNA was isolated from treated and untreated cells using the Rneasy Mini Kit and protocol. Total RNA concentration was estimated using the NanoDrop1000 Spectrophotometer. Results showed that incubation of HEY ovarian cancer cells with ICA-1(1 μm) decreased proliferation by 41% compared to controls at 72 hours post-treatment. ICA-1(2 μM) reduced RNA levels by 17% at 72 hours post-treatment compared to controls. In contrast, ACPD (2 μM) inhibited proliferation by 82% compared to controls and RNA levels were reduced by 97% at 24 hours post-treatment. These results suggest the potential of ICA-1 and ACPD as chemotherapeutic agents. Citation Format: Minjel Shah, Christopher Apostolatos, Hercules Apostolatos, Mildred Acevedo-Duncan. Effects of atypical PKC inhibitors on ovarian cancer proliferation and RNA levels [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-THER-1428.

Abstract POSTER-BIOL-1306: DNA methylation changes during epithelial-to-mesenchymal transition in ovarian cancer cells

Abstract Epithelial-to-mesenchymal transition (EMT) is an important step in the process of metastasis and has been linked to cancer cell stemness. We hypothesize that aside from defined genetic changes associated with EMT, epigenetic mechanisms are involved and have not been fully elucidated. We determined DNA methylation in SKOV3 ovarian cancer (OC) cells during EMT induced by TGF-ß using the Infinium HumanMethylation450 array. Methylation levels of CpG sites were expressed as ß-values ranging from 0 (totally unmethylated) to 1 (totally methylated). Methylation of 468 and 390 sites changed (P<.01, ANOVA and FDR<0.05) after 48h and 120h of TGF-ß stimulation. The majority of changes (81%) reflected TGF-ß-induced increase in methylation. Methylation of 160 sites changed similarly at 48 and 120 h indicating that 34% of initial changes were sustained. The number of methylation changes decreased significantly after TGF-ß removal from the cell cultures, demonstrating that removal of TGF-ß stimulation reversed the methylation changes at most sites and restituted the baseline methylome. These results were consistent with those of the principal component and hierarchical clustering analyses. TGF-ß increased overall (average of all CpG sites) DNA methylation, total DNA methyltransferase activity, and amounts of DNMT1 protein and mRNA. The average methylation of CpG islands located in or near promoters increased in 193 genes at 24h and 494 at 120h, while no changes were observed after TGF-ß removal. Pathway analysis identified methylation changes in genes belonging to functional networks relevant to EMT and cancer progression, including: cellular movement, cell cycle, organ morphology, cellular development, and cell death and survival. Treatment of OC cells with the DNA methylation inhibitor SGI-110 prevented TGF-ß-induced EMT in association with an inhibition of the TGF-ß-induced downregulation of E-Cadherin. Results demonstrate that DNA methylation is a dynamic process involved in regulation of genes implicated in EMT and metastasis. Citation Format: Horacio Cardenas, Edyta Vieth, Jiyoon Lee, Yunlong Liu, Kenneth P. Nephew, Daniela Matei. DNA methylation changes during epithelial-to-mesenchymal transition in ovarian cancer cells [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1306.

Abstract POSTER-CTRL-1212: The impact of social isolation on ovarian cancer risk and survival

Abstract Introduction: In ovarian cancer mouse models, social isolation has been associated with increased circulating markers of angiogenesis, cell migration, and impaired immune response, as well as increasing tumor burden and aggressiveness. In ovarian cancer patients, social isolation has been associated with increased circulating levels of inflammatory and pro-angiogenic factors. However, social isolation has not been investigated in relation to ovarian cancer risk or survival in humans; we addressed these questions in the Nurses’ Health Study (NHS), a longitudinal cohort of 121,701 US-based nurses. Methods: The Berkman-Syme social network index (BSSNI), a validated measure of social isolation, was assessed every four years in the NHS, starting in 1992. The BSSNI is based on marital status, close friends and relatives, as well as attendance at religious services and other group activities. Women were categorized into four groups: socially isolated, moderately isolated, moderately integrated, and socially integrated. We assessed risk of ovarian cancer associated with the BSSNI and its component parts using Cox proportional hazards models, adjusted for age, duration of oral contraceptive use, family history of ovarian cancer, parity, menopause status, duration of menopausal hormone therapy (by type), and tubal ligation. We further investigated whether the timing of social isolation impacted the results, by evaluating social isolation measured 0-4, 4-8, and 8-12 years prior to cancer diagnosis. In addition, we used Cox proportional hazards models to evaluate whether pre-diagnosis social isolation affected ovarian cancer survival, adjusting as described above plus stage and histology. Results: In follow-up from 1992-2010, we identified 494 epithelial ovarian cancer cases among 53,839 women who completed the BSSNI. In our main analysis, socially isolated women had no increased ovarian cancer risk compared to socially integrated women (relative risk [RR]: 0.93; 95% confidence interval [CI]: 0.65-1.34), although case numbers were small (N=38). However, moderate isolation was associated with increased ovarian cancer risk (RR: 1.27; 95% CI: 1.00-1.61). When we assessed the components of the BSSNI, widowed women were at increased ovarian cancer risk (RR: 1.40; 95% CI: 1.12-1.75) compared to married women; none of the other BSSNI components were associated with ovarian cancer risk. In the analysis evaluating the role of timing of isolation, socially isolated women in the 8-12 years prior to diagnosis had increased ovarian cancer risk compared to socially integrated women (RR: 1.54; 95% CI: 1.02-2.33). Among the 494 ovarian cancer cases in this analysis, 313 women died due to ovarian cancer during follow-up. Neither the BSSNI nor its components were statistically significantly associated with ovarian cancer-specific survival; however, several of the associations were suggestive. For example, social isolation was not associated with survival (RR: 1.13; 95% CI: 0.41-3.13), but moderate isolation was suggestively associated (RR: 1.61; 95% CI: 0.87-3.00). Additionally, being widowed was associated with a 71% increased risk of ovarian cancer death (95% CI: 0.99-2.96). Conclusion: These data add to a growing body of evidence that psychosocial stress is important for ovarian cancer risk and progression. Our data suggest that the timing of social isolation is key, which may have implications for the biology of stress. Future studies should confirm these results and evaluate additional sources of psychosocial stress. Citation Format: Elizabeth M. Poole, Laura Kubzansky, Olivia I. Okereke, Shelley S. Tworoger. The impact of social isolation on ovarian cancer risk and survival [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-CTRL-1212.

Abstract POSTER-TECH-1112: Quantitate epithelial-mesenchymal transition in ovarian cancer

Abstract Purpose: Metastases account for more than 90% of cancer deaths and hence, inhibiting metastasis is an attractive therapeutic option. Accumulating evidences show that invasion and metastasis of ovarian carcinoma could be driven by Epithelial-Mesenchymal Transition (EMT), a developmental program being employed during carcinoma progression. As EMT is a reversible process, inhibiting or reversing EMT thus emerges as a potential anti-metastasis strategy. EMT is a spectrum consisting of the binary epithelial and mesenchymal states as well as transitional intermediate states. Therefore, reversing EMT from the mesenchymal state would result in the sliding along an EMT Spectrum. Currently, there is no quantitative tool available to assess the EMT spectrum. In this study, we describe a bioinformatics method for quantitating EMT. Experimental procedure: By using immunofluorescence staining of E-cadherin and N-cadherin, we defined the most epithelial and most mesenchymal cells in a panel of ovarian carcinoma cell lines. We derived an EMT signature from Affymetrix gene expression microarrays of these cell lines. Subsequently, with a Kolmogorov-Smirnov-based method, we computed EMT score that assesses the similarity of an ovarian carcinoma cell transcriptome with the derived EMT signature. This method scores ovarian carcinoma cell lines in the range of [-1.0,+1.0] which corresponded to the extreme epithelial and mesenchymal states respectively along the EMT spectrum. We thus applied the EMT scoring to estimate EMT status in ovarian carcinoma tissues. Results: The EMT signature derived included known EMT inducers and markers such as ZEB1, ZEB2, SNAI2 and VIM. Expressions of selected markers such as ZEB1, VIM and FN1 were confirmed by immunofluorescence staining. Ovarian carcinoma cell lines displaying a mesenchymal morphology were accurately reflected as having high EMT scores. Cell lines with high EMT scores were anoikis resistant. In addition, ovarian carcinoma tissues displaying high EMT scores had poorer overall survival than those of low EMT scores. Correlation analysis of EMT scores and pathways revealed that ovarian carcinoma with high EMT scores are associated with increased TGFβ, WNT, NOTCH, and SRC signaling pathways. This suggested that inhibitors targeting these pathways could potentially reverse EMT. We thus treated mesenchymal ovarian carcinoma cell lines SKOV3, HEY, DOV13 with a SRC inhibitor, AZD0530. We observed a partial reversal of EMT evident by decreased EMT scores (-0.11 to -0.17) in these cells. Conclusions: EMT scoring could reflect both the EMT status of a sample and the transitional changes along the EMT spectrum upon intervening with inhibitors. Thus, EMT scoring presents a promising tool to assess the effectiveness of anti-EMT or EMT reversing compounds. Citation Format: Tuan Zea Tan, Ruby Yun-Ju Huang, Qing Hao Miow, Jieru Ye, Meng Kang Wong, Jieying Amelia Lau, Seiichi Mori, Jean Paul Thiery. Quantitate epithelial-mesenchymal transition in ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-TECH-1112.

Abstract AS18: The somatic mutational landscape of endometriosis associated ovarian cancers and precursor lesions

Abstract Introduction: Clear Cell ovarian carcinomas (CCC) represent 10% of ovarian carcinomas, with outcomes for high-stage significantly worse than high-grade serous form. The complete mutational landscape, the molecular basis of the transformation of endometriosis, the putative precursor, and patterns of clonal evolution in CCC are not well understood. Methods: Whole genome sequencing and gene expression profiling was done to uncover somatic alterations and measure effects on transcriptional networks. Targeted deep sequencing of primary tumors, metastases and endometriosis was also performed and statistical modeling approaches were used to validate mutations, quantify clonal diversity, and trace patterns of selection. Results: Mutations in ARID1A (11/19) and PIK3CA (8/19) were by far the most frequent aberrations seen. Three other SWI/SNF components also showed somatic alteration: two in non-ARID1A mutant cases and a truncating mutation of ARID1B in an ARID1A-null case. Amongst 24 significant “candidate drivers” impacting expression, five “cancer genes” have been previously described: PIK3CA, ARID1A, CTNNB1, TP53, and PPP2R1A. We observed no association between PIK3CA or ARID1A status with disease stage, genomic instability, or mutation load. Deep sequencing data suggested multiple clones in every case. In cases with matching precursor lesions, we observed multiple mutations in at least one such lesion. In precursor lesions where tumor-matched somatic mutations were observed, ARID1A and PIK3CA mutations were also always present, if observed in the matched tumor. Conclusions: ARID1A and PIK3CA mutations appear as early and histo-type defining events in CCC. Pattern of endometriosis transformation can be associated with somatic mutations in all cases, including histologically “Atypical” and non-atypical endometriosis. Finally, patterns of mutational conservation across the series of precursor lesions may present an opportunity for early screening of endometriosis tissues as an indicator of transformation potential. Citation Format: Michael S Anglesio, Ali Bashashati, Yikan Wang, Gavin Ha, Janine Senz, Winnie Yang, Steve E Kalloger, Leah M Prentice, Satoshi Yanagida, Clara Salamanca, Galina Soukhatcheva, Anthony Kazernis, Hector Chang, Anne-Marie Mes-Mason, Aikou Okamoto, Marco A Marra, Blake Gilks, Sohrab P Shah, David G Huntsman. The somatic mutational landscape of endometriosis associated ovarian cancers and precursor lesions [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS18.

Abstract AS01: Molecular genetic approaches to understanding ovarian cancer

Abstract Ovarian cancer remains a challenging condition to prevent, diagnose and treat. We have conducted several different studies that have attempted to advance knowledge in one or more of these areas. These projects are briefly summarized below. 1) Sequencing of samples from a BRCA1 mutation carrier with high grade serous carcinoma We performed whole exome sequencing (WES) on the blood, primary tumor, omental metastasis and recurrence following therapy with carboplatin and paclitaxel, from a patient carrying a BRCA1 p.S1841R mutation. We observed loss of heterozygosity in the BRCA1 mutation in the primary and subsequent tumors, and somatic mutations in TP53 and NF1 were identified, suggesting their role along with BRCA1 driving the tumor development. We were able to detect large chromosomal rearrangements using WES. A large deletion was present in the three tumors in the regions containing BRCA1, TP53, and NF1 mutations, and an amplification in the regions containing MYC. We did not observe the emergence of new mutations among tumors from diagnosis to relapse after chemotherapy. Mutations present in the primary tumor were the main drivers of metastases and chemotherapy resistance. 2) Identification of DICER1 mutations in non -epithelial ovarian cancer We used Sanger sequencing to sequence the RNase IIIa and IIIb domains of DICER1 in 154 gonadal tumours from 135 females and 19 males, and 43 extra-gonadal germ cell tumors (GCTs) from 26 females and 17 males. We identified heterozygous non-synonymous mutations in the RNase IIIb domain of DICER1 in 14/197 non-epithelial tumours (7.1%). Mutations were found in 9/28 SCSTs (32%), 5/118 gonadal GCTs (4.2%), 0/43 extra-gonadal GCTs and 0/8 miscellaneous tumours. The 14 mutations affected only five residues: E1705, D1709, E1788, D1810 and E1813. In all five patients where matched and constitutional DNA was available, the mutations were somatic. There were no mutations found in the RNase IIIa domain. More than half (8/15) of Sertoli-Leydig cell tumours (SLCTs) harbour DICER1 mutations in the RNase IIIb domain, while mutations are rarely found in GCTs. 3) Characterization of small cell carcinoma of the ovary, hypercalcemic type (SCCOHT) We sequenced the exomes of six individuals from three families with SCCOHT. After discovering segregating deleterious germline mutations in SMARCA4 in all three families, we tested DNA from a fourth affected family, which also carried a segregating SMARCA4 germline mutation. All the familial tumors sequenced harbored either a somatic mutation or loss of the wild-type allele. Immunohistochemical analysis of these cases and additional familial and non-familial cases showed loss of SMARCA4 (BRG1) protein in 38 of 40 tumors overall. Sequencing of cases with available DNA identified at least one germline or somatic deleterious SMARCA4 mutation in 30 of 32 cases. Additionally, the SCCOHT cell line BIN-67 had biallelic deleterious mutations in SMARCA4. Our findings identify alterations in SMARCA4 as the major cause of SCCOHT, which could lead to improvements in genetic counseling and new treatment approaches. Citation Format: Leora Witkowski, Jian Carrot-Zhang, Jacek Majewski, William D. Foulkes. Molecular genetic approaches to understanding ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS01.

Abstract AS04: Epigenomic characterization of gene regulatory networks in human ovarian cancer stem cells

Abstract Ovarian cancer is the fifth leading cause of cancer mortality among women in the United States. Despite aggressive surgery and chemotherapy, most patients relapse and develop drug-resistant tumors, leading to the <40% chance of 5 year survival in ovarian cancer patients. Evidence suggests that a small subpopulation of cells in the tumor, ovarian cancer stem cells (OvCSCs), are responsible for tumor relapse and drug resistance. OvCSCs have enhanced tumorigenicity, inherent chemoresistance, and stem-like properties such as the ability to self-renew or divide asymmetrically. These latter characteristics give CSCs an analogous role in tumors that healthy stem cells have in developing organs (adult stem cells) or entire organisms (embryonic stem cells).The goal of this study is to determine how the cancer stem cell state is regulated by annotating the gene regulatory network of OvCSCs. In doing so, we will uncover commonalities that exist between OvCSCs and embryonic or progenitor stem cells. The epigenome provides the structural framework for cell-type specific gene regulation. Chromatin modifications that uniquely mark gene regulatory elements, like promoters and enhancers, are easily identified throughout the genome using ChIP-seq. ChIP-seq performed in pluripotent stem cells has revealed many epigenetic features that distinguish them from differentiated cells. In an effort to better understand CSC gene regulatory networks, we are generating comprehensive transcriptomic maps and globally identifying enhancers marked by histone 3, lysine 4 monomethylation (via RNAseq and H3K4me1 ChIP-seq respectively) in OvCSCs and their daughter tumor cells. We are also comparing OvCSCs to previously generated data in human embryonic stem cells (hESCs) to identify shared stemness traits. OvCSCs were acquired from patient biopsies and detailed characterization identified a CD133+ CSC population within the tumor. We are using a mouse xenograft model to expand tumors in vivo. Harvested tumors are depleted of blood lineage cells and CD133+ cells (CSCs) are isolated from the remaining cells in the tumor. Our analysis of RNA-seq and enhancer maps have revealed OvCSC signatures that are absent in tumor cells but present in embryonic stem cells. Many important embryonic stem cell transcription factors are expressed in OvCSCs and have almost identical chromatin enhancer profiles. We identified over 7000 H1 hESC enhancers that have at least 95% overlap with OvCSC enhancers implying a shared regulatory network in these cell types. Motif analysis of nucleosome-free regions within enhancers reveals enrichment for many ESC transcription factors that are also expressed in OvCSCs. Our integrative analysis indicates that OvCSCs share common features of the ESC regulatory network and is beginning to provide novel insight on why CSCs have stem cell properties. We are continuing to build maps for additional histone modifications and genome-wide DNA methylation in OvCSCs to gain a comprehensive view of how the epigenome regulates the cancer stem cell state. Citation Format: Stephanie L. Battle, Antti Larjo, Joling Liao, Harri Lähdesmäki, Andre Lieber, R. David Hawkins. Epigenomic characterization of gene regulatory networks in human ovarian cancer stem cells [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS04.

Abstract POSTER-THER-1404: in vivo synergism between chemotherapy and the 3C23K monoclonal antibody directed against the anti-müllerian hormone type II receptor in ovarian cancer patient-derived xenografts

Abstract Purpose of the study: Müllerian inhibiting substance, also called anti-Müllerian hormone (AMH), belongs to the TGF-β family and plays a key role during sexual development. It has been shown that AMH inhibits proliferation and induces apoptosis of AMH type II receptor-positive tumor cells, and particularly progenitor cancer cells. Moreover, various reports have underlined expression of AMHRII in human gynecologic cancers including ovarian tumors. Based on these considerations, a humanized glyco-engineered monoclonal antibody (3C23K) capable of increasing effector cells engagement has been developed and tested in AMHR2 transfected xenografted cell lines. In order to document its activity in naturalistic models, we have initiated a preclinical study to assess AMHRII expression in a panel of ovarian cancer patient-derived xenografts (PDXs) and evaluate in vivo efficacy of 3C23K. Experimental procedures: AMHRII expression was studied in 14 ovarian cancers PDXs, using both flow cytometry (FC) and immunofluorescence (IF) analyses performed with alexaFluor488-conjugated 3C23K antibody on 300µm slices of xenograft fragments. In vivo efficacy of 3C23K alone (10 or 20 mg/kg twice a week, IP) or in combination with standard chemotherapy (CT) (carboplatin 66 mg/kg IP days 1-22 + paclitaxel 30 mg/kg IP days 1-22) has been planned to be evaluated in 4 AMHRII-positive ovarian cancers PDXs. One undifferentiated highly proliferative serous ovarian adenocarcinoma PDX, the OV54 model, has already been treated. On these 4 selected models, expression and quantification of AMHRII was or will be performed before and after treatments, as well as tumor-associated macrophages (TAMs) infiltrate. Summary of the data: Expression studies showed that over the 14 PDX tested, 6 (43%) were positive with at least 25% of tumor cells being positive, among these 3 expressed more than 50% of positive cells. This expression was confirmed by ex vivo IF with a profile comparable to that observed with fresh primary human ovarian tumors resected from stage IV patients. In vivo experiments (OV54 model) showed an intermediate efficacy of the 3C23K antibody administered alone, as well as a high and significant increase of CT antitumor effect. Synergistic effects were characterized by 50% of complete response (CR) when 3C23K (20mg/kg/inj.) was added to CT vs 10% CR with CT alone. The cumulative number of days under CR was 110 with 3C23K + CT vs 7 on CT alone. Impact on survival was consistent with a 90% survival at D37 with 3C23K + CT vs 60% survival on CT alone. We did not observe modification of AMHRII expression in OV54 tumors after 3C23K administration. Stromal FC analyses showed a high TAMs infiltrate of M2 type compatible with induction of Antibody Dependent Cell Phagocytosis by 3C23K. Conclusions: Using a panel of 14 ovarian cancer PDXs, we confirmed the expression of AMHRII in human ovarian cancers. Targeting this receptor by a specific monoclonal antibody, i.e. 3C23K, dramatically increases the efficacy of a standard chemotherapy. Our results therefore suggest that such combination could be further developed in ovarian cancer patients whose prognosis remains dismal. Citation Format: Fariba Némati, Jean-Marc Barret, Gérald Massonnet, Houcine Bougherara, Marie-Aude Le Frère-Belda, Jean-François Prost, Didier Decaudin. in vivo synergism between chemotherapy and the 3C23K monoclonal antibody directed against the anti-müllerian hormone type II receptor in ovarian cancer patient-derived xenografts [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-THER-1404.

Abstract POSTER-TECH-1131: Next generation protein multiplexing

Abstract While a variety of technologies exist to measure proteins, such as ELISA’s, Mass Spec, and 2D Gel’s, the most promising for screening multiple proteins are biochip or microarray based technologies. However, even with biochip devices, measuring specific interactions between multiple protein combinations remains problematic. Proteins are complex and fragile bio-molecules and often interact in complex and unpredictable ways with other proteins and/or protein analysis equipment, causing non-specific (false) signals. Furthermore, many proteins of interest to the life science industry are present in samples at very different concentrations, limiting which proteins can be screened in a single multiplex test and sample dilution. To address these problems, Inanovate has developed a new multiplexed protein quantification technology called Longitudinal Assay Screening (LAS), combining high sensitivity confocal imaging and microfluidics alongside protein based microarrays. Inanovate has completed testing and benchmarking of the first platform integrating LAS technology across a range of demonstration assays. Instead of depending on a 96-well micro-titer plate, the new LAS platform utilizes a glass slide based protein microarray and microfluidics for dispensing and binding samples and detection antibodies. In its most basic form, the protein microarray is composed of capture antibodies for the proteins being measured, as well as positive and negative quality control features for ensuring sample to sample, run to run, lot to lot, and user to user consistency. Conceptually very similar to a real-time PCR reaction, the LAS platform iteratively flows small volumes of sample and labeled detection antibodies across the protein microarray (housed on Inanovate’s fluidic cartridges) and fluorescently measures the formation of the sandwich between capture antibody, analyte of interest, and detection antibody in real-time. Due to the time-resolved nature of the assay, the resulting data is a rate of reaction, instead of a simple final fluorescent reading. This ‘rate of reaction’ based analysis helps deliver the following core advantages of LAS technology, each of which has been demonstrated through the development and validation of a five-plex assay consisting of CRP, IL-6, IL-1a, IL-8 and IL-1b. 1. Large detection range: LAS enables the accurate quantitation of protein concentrations across a 7 log range in a single multiplex test. This eliminates the need for serial dilutions, making multiplexing faster, cheaper and helping preserve precious samples. 2. Multiplexing flexibility: Due to its large detection range, LAS allows users to run virtually any assay of interest in one test, enabling the development of biologically relevant multiplexes. 3. Improved accuracy: LAS produces and analyses real-time kinetic data on protein interactions (rate of reaction data), improving identification and discrimination of background and nonspecific signals, delivering more accurate quantitation at low analyte concentrations. LAS technology holds the potential to become to proteomics what PCR was to genomics. It offers a new approach to multiplexed protein screening that helps address many of the problems presently affecting the utility of biochips in protein biomarker discovery, validation and clinical screening applications. The Poster will both summarize the technical components of the new LAS based platform and present data from the five-plex demonstration work plus recent work on serology assays. Citation Format: Sloan D, Nelson J, Ure D, Votaw G, Stevens E. Next generation protein multiplexing [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-TECH-1131.

Abstract POSTER-BIOL-1335: Threonyl-tRNA synthetase overexpression correlates with angiogenesis and progression of human ovarian cancer

Abstract Ovarian tumors create a dynamic microenvironment that promotes angiogenesis and reduces immune responses. There is growing evidence that tRNA synthetases have a role in cell signaling separate from their function in protein synthesis. Our research has revealed that threonyl-tRNA synthetase (TARS) has an extracellular angiogenic activity. Exploration of protein and mRNA databases further showed a connection between TARS expression and ovarian cancer. The objective of this study was to test the hypothesis that TARS is secreted by ovarian cancer cells and that its expression in clinical samples correlates with angiogenic markers and ovarian cancer progression. TARS secretion by SKOV3 human ovarian cancer cells was induced by hypoxia or tumor necrosis factor-α (TNF-α) as detected by ELISA and Western blot of culture media (p<0.001). For the clinical study, TARS, vascular endothelial growth factor (VEGF) and PECAM were assessed by immunohistochemistry in serial sections of paraffin embedded ovarian tissues from 70 patients diagnosed with epithelial ovarian cancer and 12 control patients. Serum samples from 31 tissue-matched patients were analyzed by ELISA for TARS, CA125, and tumor necrosis factor-α (TNF-α). There was a strong association between the tumor expression of TARS and advancing stage of epithelial ovarian cancer (p<0.001). TARS expression and localization were also correlated with VEGF (p<0.001). A significant proportion of samples included heavy TARS staining of infiltrating leukocytes which also correlated with stage (p=0.017). Patient serum TARS was related to tumor TARS and angiogenic markers, but did not achieve significance with respect to stage. Multivariate Cox proportional hazard models revealed a surprising inverse relationship between TARS expression and mortality risk in late stage disease (p=0.062). Together these results show that TARS is secreted in response to cell stress and that its expression is increased in epithelial ovarian cancer according to stage and angiogenic progression. These findings and the association of TARS with disease survival provide clinical validation that TARS is associated with angiogenesis in ovarian cancer. These results encourage further study of TARS as a regulator of the tumor microenvironment and possible target for diagnosis and/or treatment in ovarian cancer. Citation Format: Theresa L. Wellman, Midori Eckenstein, Cheung Wong, Mercedes Rincon, Takamaru Ashikaga, Sharon L. Mount, Christopher S. Francklyn, Karen M. Lounsbury. Threonyl-tRNA synthetase overexpression correlates with angiogenesis and progression of human ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1335.

Abstract AS06: Development and characterization of an oviduct-specific model of high-grade serous carcinoma

Abstract Recent studies suggest the most common and lethal type of “ovarian” cancer, high-grade serous carcinoma (HGSC), often arises from fallopian tube epithelium rather than the ovarian surface epithelium. Their intraperitoneal location and microscopic size make it very difficult to screen for, or study the biology of early tubal lesions in humans, as most are discovered incidentally in women with hereditary predisposition to ovarian cancer who elect prophylactic salpingo-oophorectomy for cancer risk-reduction. Although most cases of pelvic HGSC probably originate in the fallopian tube, roughly one-third lack evidence of tubal origin. These cases may arise from ectopic tubal-type epithelium (endosalpingiosis) present in the ovary, peritoneum, or at other sites. Mouse models that closely mimic the genetics and biology of human HGSCs may be useful for clarifying how tubal and non-tubal pelvic HGSCs develop and progress. We have developed transgenic (Ovgp1-iCreERT2) mice allowing conditional (tamoxifen [TAM]-inducible) expression of Cre recombinase exclusively in the oviductal epithelium, using a single transgene. Using double transgenic Ovgp1-iCreERT2;R26LSL-eYFP mice in which TAM treatment activates eYFP reporter protein expression in oviductal epithelial cells, we have shown that tubal-type inclusion glands (endosalpingiosis) expressing both eYFP and OVGP1 are present in a subset of murine ovaries following treatment with TAM. The findings suggest that oviductal epithelium can detach from the oviduct and traffic to the ovaries, where it implants through breaks in the OSE and maintains its tubal-type differentiation. Preliminary data also suggest that the frequency of ovarian endosalpingiosis increases over time and by superovulating the mice. The latter finding raises the possibility that at least some of the protective effects of high parity and oral contraceptive use on human ovarian cancer risk is related to reduced likelihood of acquiring endosalpingiosis from which “ovarian” HGSCs may eventually arise. We have further shown that Ovgp1-iCreERT2;Trp53fl/fl;Brca1fl/fl mice develop oviductal lesions identical to human serous tubal intraepithelial carcinomas (STICs) after TAM treatment and have identified additional genetic alterations that cooperate with Trp53 and Brca1 inactivation in oviductal HGSC pathogenesis in our model system. Murine models of HGSC that recapitulate their human tumor counterparts provide excellent in vivo systems with which to define the cellular and molecular events associated with HGSC development and progression. Citation Format: Yali Zhai, Rong Wu, Tom C. Hu, Eric R. Fearon, Kathleen R. Cho. Development and characterization of an oviduct-specific model of high-grade serous carcinoma [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS06.

Abstract POSTER-BIOL-1351: Molecular characterization of cyclin-dependent kinase 1 pathway in newly established epithelial ovarian cancer cell lines

Abstract Background: Epithelial ovarian cancer (EOC) is a heterogeneous gynecologic tumor. To better understand the biological pathways involved in the development of EOC, we performed mRNA expression profiling on newly established EOC cell lines. Methods: Five newly established EOC cell lines were characterized by transcript profiling. Microarray analyses were validated using quantitative real-time PCR (RT-PCT) and immunoblotting. Furthermore, biological networks of the differentially expressed genes were predicted using the STRING database and literature review. Results: A total of 239 genes represented by 24,957 probes were identified from microarray analysis, of which 98 and 141 genes are up- (> 2 fold, p < 0.05) or down-regulated (< 3 fold, p < 0.05), respectively, in EOC cell line compare to human ovarian surface epithelial cell lines. We identified a subset of EOC characterized by the numerous genes related to cell cycle and apoptosis, along with cyclin-dependent kinase 1 (CDK1). A total of 24 genes were validated by quantitative RT-PCR, and expressional reliabilities of 4 CDK1 pathway-related genes were confirmed by immunoblotting. Further biological network analysis revealed that the CDK1 pathway might be essential for EOC tumorigenesis. Conclusions: These data reveal that the CDK1 pathway is important to EOC development and progression, providing new insight into the biology of EOC. Citation Format: Wookyeom Yang, Hanbyoul Cho, Kim Sol, Jae-Hoon Kim. Molecular characterization of cyclin-dependent kinase 1 pathway in newly established epithelial ovarian cancer cell lines [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr POSTER-BIOL-1351.