Weroha, Saravut J.

Abstract Ovarian cancer is one of the most common gynecologic malignancies, with high-grade serous carcinoma (HGSC) as the most frequent and deadliest type of ovarian cancer (OC). With the recent advances in cancer epigenomic profiling, it has become clear that intra-tumor heterogeneity driven by genetic and epigenetic factors may be the basis of resistance, where current treatments could efficiently target sub-populations of cells while a sub-group may be resistant, or prone to develop resistance. With epigenomic profiling and single cell sequencing technology, we aim to characterize cell signatures that translate as prognostic markers and novel drug targets. We have mapped the epigenomic landscape of HGSC in 12 patient-derived xenografts (PDX), and their metastatic counterparts (circulating tumor cells and ascites present in the PDX models) using an assay for transposase-accessible chromatin (or ATAC-seq) and compared it with normal fallopian tube organoids to identify tumor and metastatic-specific regulatory elements. We also performed RNA-seq to validate potential transcriptional dependencies associated with resistance to chemotherapy. In order to identify potentially chemotherapy-resistant subpopulations and further validate the activity of some of the transcription factors that can be driving resistance, we performed single cell multiomic profiling using the 10x Multiome sequencing platform (scRNA+scATAC from the same nuclei) in HGSC patient samples and PDX. Finally, we performed immunohistochemistry of patient tissue slides for target markers and correlated it with overall survival in OC patients. As previously suggested, a subset of cancer clonal cells can drive treatment resistance and metastasis and result in poor patient outcomes. Data derived from PDX ATAC-seq showed a cancer stem cell signature driven by SOX2 in metastatic samples compared to the paired PDX primary tumor, suggesting a role for SOX2 in disease progression. Interestingly, SOX2 expression in epithelial cells in the Multiome analysis is more patient-specific than other stem cell markers, like CD133. We also observed that SOX2 expression negatively correlated with prognosis. These data, taken together, point out the promising role of SOX2 as a prognostic and targetable marker for HGSC. Citation Format: Leticia Assad Maia Sandoval, Wazim Mohammed Ismail, Amelia Mazzone, Amik Munankarmy, Jagneet Kaur, Tsering Tashi, Xiaonan Hou, Nagarajan Kannan, Ann Oberg, Scott H. Kaufmann, Saravut John Weroha, Alexandre Gaspar-Maia. Single cell multiomic profiling of high-grade serous carcinoma. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4724.

Abstract Ovarian cancer is the 7th most common cancer in women and the most fatal of all female reproductive cancers, with 5-year overall survival of 10-30%. Despite initial responses to first line platinum- and paclitaxel-based chemotherapies, >70% of ovarian cancers recur with increasingly resistant disease; and nearly all of these women die of their disease. Recently, the use of targeted therapies such as bevacizumab and PARP inhibitors have been shown to improve progression-free survival in a subset of patients. However, these agents have thus far failed to improve overall survival in patients without specific genetic signatures, highlighting the continued need for alternative therapies. Through an unbiased drug screen, we have identified lestaurtinib as a potent inhibitor of many sensitive and resistant ovarian cancer cell lines and patient derived organoid models. Lestaurtinib is a known tyrosine kinase inhibitor originally developed to block FLT3, a protein that is not expressed in any of the ovarian cancer cell lines tested. To elucidate the mechanisms by which lestaurtinib inhibits ovarian cancer cells, we have examined its impact on the phospho-proteome in parallel with RNAseq. Results of these studied have identified inhibition of the JAK/STAT signaling pathway as a top hit. Assessment of STAT1 and STAT3 has revealed that these transcription factors are constitutively activated in chemotherapy- and PARP inhibitor-resistant cell lines as reflected by high levels of Tyr701/705 and Ser727 phosphorylation, suggesting that induction of this pathway may drive resistance. Surprisingly, profound differences in the ability of various JAK/STAT inhibitors to block these specific phosphorylation events is observed, which correlates with their potency. Further, STAT1 and STAT3 ChIPseq experiments following cytokine-mediated activation of this pathway reveal that these transcription factors bind to genomic regions encoding genes involved in oncogenic and drug resistance pathways. Genetic knockdown of STAT1 and STAT3 via siRNA, or knockout via CRISPR/Cas9, results in significant growth inhibition of sensitive and resistant models of ovarian cancer, confirming their importance in maintaining cell viability. Finally, combining lestuartinib with standard-of-care cisplatin or olaparib (a PARP inhibitor) is shown to be synergistic, indicating that pharmacological inhibition of JAK/STAT signaling has the potential to counteract drug resistance. Ongoing studies are aimed at further understanding the role of JAK/STAT signaling in ovarian cancer, elucidating the mechanistic processes by which STAT1/3 mediate progression of this disease, and identifying the most effective pharmacological strategies to study in possible future clinical trials. Citation Format: Esther Rodman, Michael Emch, Elizabeth Bruinsma, Xiaonan Hou, Scott Kaufmann, Saravut J. Weroha, John Hawse. Investigation of the JAK/STAT signaling pathway in chemotherapy and PARP inhibitor resistant ovarian cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4862.

Abstract Background: The kinases ataxia-telangiectasia-mutated-and-Rad3-related (ATR), checkpoint kinase 1 (CHK1) and WEE1 participate in the response to replication stress. Inhibitors of ATR, WEE1 and CHK1 are being tested for antineoplastic activity as monotherapies and in combination with chemotherapy. Although these agents impair cell cycle checkpoints, fork stabilization, origin firing and homologous recombination (HR), little is known about the apoptotic pathways that are engaged and the critical mediators activating cell death mechanisms. The present study utilizes high grade serous ovarian cancer (HGSOC) cell lines and patient derived xenografts (PDXs) to identify the primary mechanism of cell death in ovarian cancer and assess changes involved in the resistance setting after treatment with these inhibitors. Methods: A panel of HGSOC cell lines was examined for responses to ceralasertib (ATRi), prexasertib (CHK1i) or adavosertib (WEE1i) using colony forming assays, immunoblotting and assays for apoptosis. Mechanisms involved in acquired resistance and downstream changes in cell fate were evaluated by generating prexasertib-resistant HGSOC cell lines. Subsequent NGS analysis was performed in the sensitive and resistant pairs. Results: Irrespective of the homologous recombination (HR) status, ovarian cell lines and PDXs underwent cell death during prolonged exposure at clinically achievable concentrations. Treatment of ovarian cancer cell lines with ATRi, CHK1i or WEE1i activated the replication checkpoint, consistent with previously published reports. The subsequent apoptotic response involved PMAIP1 and BCL2L11 upregulation rather than the TNFa-induced death receptor-mediated apoptosis we recently described in acute leukemia (Cancer Res. PMID: 33414171). Additionally, ovarian lines selected for CHK1i resistance failed to die in response to either ATRi or WEE1i. Conclusions: Prexasertib, ceralasertib, and adavosertib exhibit monotherapy activity in both HR deficient and proficient HGSOC cell lines and PDX models. Unlike AML, death receptor mediated apoptosis was not observed; instead, the primary mechanism of cell killing in HGSOC lines involves activation of the mitochondrial apoptotic pathway. Cross-resistance of these DNA damage repair modulators is likely multifactorial and mainly involves diminished replication stress response. Citation Format: Annapoorna Venkatachalam, Kevin L. Peterson, Cristina Correia, Karen S. Flatten, Xianon Hou, Paula A. Schneider, Emily Balczewski, Cordelia McGehee, Rachel M. Hurley, Xue W. Meng, Chance Sine, Rameen Shah, Nicole Vincelettte, Husheng Ding, Hu Li, Saravut (John) Weroha, Scott H. Kaufmann. CHK1 inhibitor prexasertib induces NOXA-dependent apoptosis in ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1832.

2651 Background: Etigilimab (etig), a humanized IgG1 monoclonal antibody, blocks TIGIT interaction with PVR (poliovirus receptor) and inhibits downstream signalling with target cell killing. Etig +/- nivolumab (nivo) showed acceptable safety and preliminary activity in a FIH Phase 1a/b study in solid tumors. ACTIVATE, an open-label Phase 1b/2 basket study is evaluating further efficacy, safety, tolerability and PK/PD of etig+nivo. This is a preliminary efficacy and safety analysis from ACTIVATE. Methods: Subjects with advanced/metastatic solid tumors without curative/standard of care therapies are given IV etig+nivo Q2W until disease progression or intolerable toxicity. Six open cohorts are enrolling in parallel: endometrial cancer (post-front-line platinum), endometrial cancer (2-3 prior lines), PD-L1+ checkpoint-inhibitor-naïve (CPI-n) cervical cancer, rare cancers (germ cell tumor (GCT), sarcoma, uveal melanoma), ovarian cancer (post-front-line platinum) and TMB-h/MSS, (TMB >10mut/mb) tumors. Tumor assessments are done every 8 weeks following baseline scan. Primary endpoint is investigator-assessed ORR by RECIST 1.1. Secondary endpoints include duration of response and safety. Results: Of 27 efficacy-evaluable subjects enrolled in the 6 open cohorts with minimum 1 staging scan or radiological/clinical progression at data cut-off of 02/10/22, 12 had clinical benefit with 1 complete response (CR), 2 partial responses (PRs) and 9 stable diseases (SDs) and 15 had radiological/clinical progressive disease (PD), with an overall response rate (ORR) of 11% and disease control rate (DCR) of 44%. Conclusions: Etig+nivo is safe and well tolerated with no new safety signals. Early efficacy was noted in cervical cancer (1CR, 1PR and 1SD) and uveal melanoma (3 SDs >20 weeks). Encouraging activity was also noted in ovarian cancer and post-CPI treated TMB-H/MSS NSCLC. These early data support continued evaluation of the combination of etig+nivo. Clinical activity was notable as below. Clinical trial information: NCT04761198. [Table: see text]

Abstract Background: Next-generation sequencing provides comprehensive information about individuals’ genetic makeup and is commonplace in oncology clinical practice. However, the utility of genetic information in clinical decision-making process has not been examined extensively from a real-world, data-driven perspective. Through mining real-world data (RWD) from clinical notes, we could extract patients’ genetic information and further associate treatment decisions with genetic information.Methods: We proposed a real-world evidence (RWE) study framework that incorporates context-based natural language processing (NLP) methods and data quality examination before final association analysis. The framework was demonstrated in a Foundation-tested women cancer cohort (N=196). Upon retrieval of patients’ genetic information using NLP system, we assessed the completeness of genetic data captured in unstructured clinical notes according to a genetic data-model. We examined the distribution of different topics regarding BRCA1/2 throughout patients’ treatment process, and then analyzed the association between BRCA1/2 mutation status and the discussion/prescription of targeted therapy. Results: We identified seven topics in clinical context of genetic mentions including: Information, Evaluation, Insurance, Order, Negative, Positive, and Variants of unknown significance (VUS). Our rule-based system achieved a precision of 0.87, recall of 0.93 and F-measure of 0.91. Our machine learning system achieved a precision of 0.901, recall of 0.899 and F-measure of 0.9 for four-topic classification and a precision of 0.833, recall of 0.823 and F-measure of 0.82 for seven-topic classification. We found in result-containing sentences, capture of BRCA1/2 mutation information was 75%, but detailed variant information (e.g. variant types) is largely missing. Using cleaned RWD, significant associations were found between BRCA1/2 positive mutation and targeted therapies.Conclusions: In conclusion, we demonstrated a framework to generate RWE using RWD from different clinical sources. Rule-based NLP system achieved the best performance for resolving contextual variability when extracting RWD from unstructured clinical notes. Data quality issues such as incompleteness and discrepancies exist thus manual data cleaning is needed before further analysis can be performed. Finally, we were able to use cleaned RWD to evaluate real-world utility of genetic information to initiate a prescription of targeted therapy.

Abstract Background : Next- generation sequencing provides comprehensive information about individuals’ genetic makeup and is commonplace in oncology clinical practice. However, the utility of genetic information in clinical decision-making process has not been examined extensively from a real-world, data-driven perspective. Through mining real-world data (RWD) from clinical notes, we could extract patients’ genetic information and further associate treatment decisions with genetic information. Methods : We proposed a real-world evidence (RWE) study framework that incorporates context-based natural language processing (NLP) methods and data quality examination before final association analysis. The framework was demonstrated in a Foundation-tested women cancer cohort (N=196). Upon retrieval of patients’ genetic information using NLP system, we assessed the completeness of genetic data captured in unstructured clinical notes according to a genetic data-model. We examined the distribution of different topics regarding BRCA1/2 throughout patients’ treatment process, and then analyzed the association between BRCA1/2 mutation status and the discussion/prescription of targeted therapy . Results: We identified seven topics in clinical context of genetic mentions including: Information, Evaluation, Insurance, Order, Negative, Positive, and Variants of unknown significance (VUS) . O ur rule-based system achieved a precision of 0.87, recall of 0.93 and F- measure of 0.91. Our machine learning system achieved a precision of 0.901, recall of 0.899 and F- measure of 0.9 for four-topic classification and a precision of 0.833, recall of 0.823 and F- measure of 0.82 for seven-topic classification. We found in result-containing sentences, capture of BRCA1/2 mutation information was 75%, but detailed variant information (e.g. variant types) is largely missing. Using cleaned RWD, significant associations were found between BRCA1/2 positive mutation and targeted therapies. Conclusions : In conclusion, we demonstrated a framework to generate RWE using RWD from different clinical sources. Rule-based NLP system achieved the best performance for resolving contextual variability when extracting RWD from unstructured clinical notes . Data quality issues such as incompleteness and discrepancies exist thus manual data cleaning is needed before further analysis can be performed. Finally, we were able to use cleaned RWD to evaluate real-world utility of genetic information to initiate a prescription of targeted therapy.

Abstract Introduction: Acquired resistance to platinum remains a major clinical problem in ovarian cancer since systemic therapies have limited efficacy in this setting. There is growing evidence that the tumor microenvironment (TME) contributes to such resistance, as well as the pathogenesis of the disease. The study of the TME is limited by technical challenges of separating the epithelial from the stromal component. To overcome this limitation, we performed RNA sequencing (RNA-Seq) for three pairs of sensitive and induced resistant patient-derived xenograft models since the epithelial transcripts originate from human tissue and the stroma component is murine. Methods: To determine which TME genes might be contributing to chemotherapy resistance, three platinum-sensitive ovarian cancer PDX lines were treated repeatedly with carboplatin/paclitaxel until resistance occurred. Primary platinum-sensitive and induced platinum resistant tumors were characterized histologically by H&E and Masson Trichrome staining. The three pairs of sensitive and resistant PDX models were then subjected to RNA-Seq. Raw sequencing data were processed through Xenome to classify the reads into epithelial carcinoma (human) or TME (mouse) through Mayo Clinic’s internal RNA-Seq pipeline for transcriptome quantification. Differential expression analysis between induced resistant and sensitive models was carried out for both carcinoma and TME components using edgeR paired analysis. Gene set enrichment analysis was used to identify overrepresented biologic functions for differentially expressed genes between the pairs. Top significantly varied, differentially expressed genes between the pairs were validated using species-specific qPCR. Results: Histologic analysis of the matched sensitive/resistant pairs revealed a marked increase of the stromal component in the resistant tumors evident by H&E and Masson trichrome staining. A total number of 95 and 312 genes were differentially expressed in the epithelial and TME component, respectively (FDR<0.05; absolute log2 fold-change >1.5). The stemness marker SOX2 was significantly upregulated in the human component of the resistant tumors as confirmed by qPCR. Genes involved in stroma remodeling were upregulated in the “murine” component of the resistant tumors, and upregulation was confirmed by qPCR (normalized to a murine housekeeper gene). Gene set enrichment analysis showed that the upregulated genes in TME of resistant models were enriched for cytoskeleton related functions, while the downregulated genes were enriched for extracellular matrix-related functions. Conclusion: A bioinformatics strategy able to map human and murine reads separately may identify resistance mechanisms attributed to the two tumor subcompartments and could lead to strategies to potentially target each separately or concurrently. Citation Format: Valentina Zanfagnin, Yuanhang Liu, Pritha Chanana, Xiaonan Pritha, Beverly Long, Joe Pathoulas, Brad Evans, Jaime Davila, Chen Wang, Saravut John Weroha. Transcriptional profiling of tumor stroma using ovarian cancer PDX models with induced platinum-resistance [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research; 2019 Sep 13-16, 2019; Atlanta, GA. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(13_Suppl):Abstract nr B18.

Abstract Objectives: Cognitive computing has the potential to improve efficiency and accuracy of clinical trial enrollment using artificial intelligence. The cognitive computing clinical trial matching (CTM) system used for this study utilizes natural language processing to derive patient and tumor attributes from structured and unstructured electronic medical record (EMR) data. The attributes are matched to complex eligibility criteria in trial protocols. This CTM system has been implemented in our gynecologic oncology practice, for which there is lower representation of ethnic minorities, elderly, and uninsured in National Cancer Institute-sponsored trials. Methods: A clinical research coordinator (CRC) used the CTM system to screen patients for potential clinical trials one day prior to their clinic visit. Trial matches were shared with gynecologic oncology clinicians to raise awareness for study opportunities and assist in treatment decisions. Clinicians were surveyed regarding their experience with the CTM system prepared matches. The identical patients were evaluated by a clinician using the traditional manual screening method. Results: Seventeen patients with new diagnosis, recent resection, or restaging scans were screened for 41 potential gynecologic, phase I, and supportive care clinical trials. Trial screening by the CRC using the CTM system resulted in a total of 119 matched trials (mean: 7 trials/patient) compared to the clinician-generated list of 271 matched trials (mean: 16 trials/patient). The CRC using the CTM system spent an average of 18 minutes/patient (range: 7 to 40 minutes) compared to the clinician average of 22 minutes/patient (range: 7 to 37 minutes). A survey of 8 gynecologic oncology clinicians reported they were willing to spend an average of 8 minutes/patient (range: 1 to 15 minutes) to screen patients for trial eligibility. On independent review, discrepancies occurred when the clinician identified trials that the CRC excluded due to inclusion/exclusion criteria or were unknowingly closed to accrual. In addition, the CRC identified trials that were inadvertently missing from the clinician’s list of active protocols. Consistent with these observations, 100% of surveyed clinicians (8 of 8) agreed that “The CTM system has helped to include and exclude trials prior to discussion with patients.” Conclusions: Use of the CTM system by a CRC was superior to a clinician alone for the accuracy and efficiency of screening clinical trial eligibility. This process improved the identification of trials that a clinician might otherwise miss and the exclusion of trials that were not available or appropriate for patients. The CTM system can reduce the burden of clinicians and research staff to screen patients. Additional research is needed to determine if the process can improve clinical trial enrollment in gynecologic cancers. Citation Format: Thanh P. Ho, Jane M. Helgeson, Angela L. Andring, Jennifer C. Reed, Venessa L. Boyle, Nigar Sofiyeva, Konstantinos Leventakos, Tufia C. Haddad, Andrea E. Wahner Hendrickson, Saravut (John) Weroha. Implementation of cognitive computing to match clinical trials in gynecologic cancers: A single-institution experience [abstract]. In: Proceedings of the AACR Special Conference on Advancing Precision Medicine Drug Development: Incorporation of Real-World Data and Other Novel Strategies; Jan 9-12, 2020; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(12_Suppl_1):Abstract nr 01.

Abstract Background: Next-generation sequencing provides comprehensive information about individuals’ genetic makeup and is commonplace in oncology clinical practice. However, the utility of genetic information in clinical decision-making process has not been examined extensively from a real-world, data-driven perspective. Through mining real-world data (RWD) from clinical notes, we could extract patients’ genetic information and further associate treatment decisions with genetic information.Methods: We proposed a real-world evidence (RWE) study framework that incorporates context-based natural language processing (NLP) methods and data quality examination before final association analysis. The framework was demonstrated on a Foundation-tested women cancer cohort (N=196). Upon retrieval of patients’ genetic information using NLP system, we assessed completeness of genetic data captured in unstructured clinical notes according a genetic data-model. We examined the distribution of different topics regarding BRCA1/2 throughout patients’ treatment process, and then analyzed the association between BRCA1/2 mutation status and the discussion/prescription of targeted therapy. Results: We identified seven topics in clinical context of genetic mentions including: Information, Evaluation, Insurance, Order, Negative, Positive, and Variants of unknown significance (VUS). Our rule-based system achieved a precision of 0.87, recall of 0.93 and F-measure of 0.91. Our machine learning system achieved a precision of 0.901, recall of 0.899 and F-measure of 0.9 for four-topic classification and a precision of 0.833, recall of 0.823 and F-measure of 0.82 for seven-topic classification. We found in result-containing sentences, capture of BRCA1/2 mutation information was 75%, but detailed variant information (e.g. variant types) is largely missing. Using cleaned RWD, significant associations were found between BRCA1/2 positive mutation and targeted therapies.Conclusions: In conclusion, we demonstrated a framework to generate RWE using RWD from different clinical sources. Rule-based NLP system achieved the best performance for resolving contextual variability when extracting RWD from unstructured clinical notes. Data quality issue such as incompleteness and discrepancies exist thus manual data cleaning is needed before further analysis can be performed. Finally, we were able to use cleaned RWD to evaluate real-world utility of genetic information to initiate prescription of targeted therapy.

e18097 Background: AB160 is a 160 nm nano-immunoconjugate consisting of nab-paclitaxel (NP) nanoparticles non-covalently coated with bevacizumab (BEV) for targeted delivery into tissues expressing high levels of vascular-endothelial growth factor (VEGF). Both taxanes (GOG-0129C, GOG-0126R, GOG-0127V) and BEV (GOG-0229E, AURELIA, GOG-0227C) have demonstrated clinical activity in previously treated metastatic endometrial cancer (EC), ovarian cancer (OC), and cervical cancer (CC), respectively. Methods: A 3+3 phase I trial was conducted in patients with EC, platinum-resistant OC and CC who had prior systemic treatment for metastatic disease to determine the maximum tolerated dose of AB160 administered intravenously on days 1, 8 and 15 of a 28-day cycle. The starting dose level (DL1) was NP at 125 mg/m2 with BEV at 50 mg/m2. There were 2 higher dose levels: DL2 (NP at 150 mg/m2 with BEV at 60 mg/m2) and DL3 (NP at 175 mg/m2 with BEV at 70 mg/m2). Dose limiting toxicities (DLT) included grade (G) 4 neutropenia or anemia, PLT < 25,000, serum creatinine ≥2 times baseline, G2-4 neurologic toxicity or G3-4 non-hematologic toxicities. Disease evaluations were conducted after every 2 treatment cycles using RECIST criteria. Patients were treated until disease progression or intolerability. Samples were collected for pharmacokinetic (PK) studies. Results: Nine women 41 – 74 years of age (median 57) have enrolled (5 with EC and 4 with OC); data are available for the first 8. No DLTs have been observed among the 3 women enrolled on DL1, 3 women on DL2, and 2 women on DL3. All 3 patients on DL3 continue on treatment. The other 6 patients have discontinued due to adverse reactions (3), progression (2), and patient choice (1). The median number of cycles administered is 6 (4-14). The most common severe (G3/4) toxicities include neutropenia (37.5%) and leukopenia (25%). There have been 5 partial responses (62.5%): 1 on DL1 and 2 each on DL2 and DL3. PK evaluation is pending. Conclusions: AB160 therapy is safe and demonstrates promising clinical activity in patients with previously treated metastatic gynecologic malignancies. Further clinical testing is being pursued in this patient population. Clinical trial information: NCT02020707.