Search

Accurate risk prediction of distant recurrence (DR) is crucial for personalized adjuvant systemic therapy of endometrial cancer (EC) stage I-III patients, as DR is associated with a 5-year overall survival of 10-20%. Risk stratification and treatment recommendation are currently based on histopathological and molecular markers, which is challenging due to high inter-observer variability and testing costs respectively. Deep Learning (DL) models can predict prognosis by identifying relevant visual features from H&E whole slide images (WSIs) at different resolutions without prior assumptions. Here, we developed and tested the first interpretable state-of-the-art DL model for WSI-based risk prediction of DR of stage I-III EC (DeREC) from the randomized PORTEC-1/-2/-3 trials and three clinical cohorts with long-term follow-up data. We used one representative H&E WSI each from 1761 EC patients, excluding those who received adjuvant chemotherapy as it lowers the risk of DR. We randomly sampled 20% as a held-out internal test set (N=353 with 62 events; 8.45 year median follow-up) and performed a 5-fold cross-validation on the training set (N=1408). WSIs were partitioned into 360 micron patches at 40x magnification. DeREC combined self-supervised representation learning of patches using a multi-resolution vision transformer and a WSI-level graph attention-based time-to-event prediction model. The model performance of correctly ranking patients by predicted risk scores and true time to DR was measured with the concordance-index and compared with a Cox’ Proportional Hazards (CPH) model fitted on histopathological variables (histotype, grade, lymphovascular space invasion, stage). Discriminative quality of the predicted risk groups was investigated with Kaplan-Meier analysis and the log-rank test. Most predictive patches by predicted risk groups were reviewed by an expert gynecopathologist for identification of prognostic morphological features. DeREC achieved a concordance-index of 0.764 [95%CI 0.754-0.773] on 5-fold cross validation and 0.757 on the test set, as compared to 0.704 [95%CI 0.662-0.746] with CPH. Predicted risk groups around quartiles 1 and 3 accurately stratified patients between low (N=89), intermediate (N=175), high (N=89) risk of DR (p Citation Format: Sarah Fremond, Sonali Andani, Jurriaan Barkey Wolf, Gitte Ørtoft, Estrid Høgdall, Jouke Dijkstra, Jan J. Jobsen, Ina M. Jürgenliemk-Schulz, Ludy CHW Lutgens, Melanie E. Powell, Naveena Singh, Linda R. Mileshkin, Helen J. Mackay, Alexandra Leary, Dionyssios Katsaros, Hans W. Nijman, Stephanie M. de Boer, Remi A. Nout, Vincent T.H.B.M Smit, Carien L. Creutzberg, Nanda Horeweg, Viktor H. Koelzer, Tjalling Bosse. Deep learning risk prediction model of distant recurrence from H&E endometrial cancer slides. [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 5695.

Background: Knowledge regarding the homologous recombination deficiency (HRD) status at diagnosis is essential to manage patients (pts) with advanced epithelial ovarian cancer (EOC). HRD is defined by either a BRCA1 or BRCA2 mutation, or genomic instability (GI). These genomic tests are performed on FFPE tumor samples and unfortunately non-contributive in 12-18% of cases. Advanced OC frequently presents with ascites which could provide an alternative source of Cell-free tumor DNA (cftDNA). Methods: We collected data from 53 patients undergoing primary (N=35) or interval (N=12) laparoscopy for suspected/confirmed EOC, or abdominal paracentesis at relapse (N=6). cftDNA was extracted from 20ml of ascites, or 20ml of peritoneal washings if no ascites was present. Intraperitoneal cfDNA was subjected to next generation sequencing (NGS) to confirm tumoral origin and to identify BRCA1/2 mutations, and GI was assessed by measuring large scale genomic alterations on shallow Whole Genome Sequencing (sWGS). Matching FFPE tumor samples underwent NGS and GI testing. Results: A total of 53 patient were included. Overall, intraperitoneal cfDNA was detectable in more than 90% (48/53) pts, and in 100% (35/35) of pts at diagnosis. Even among patients without visible ascites, peritoneal washings yielded cfDNA in 40% of cases (3/7). DNA quality (presence of mononucleosomal pics around 160bp and low contamination with high molecular DNA greater than 700bp) and quantity (median concentration: 3700 ng/ml; range 109 - 65 000 ng/ml) was excellent and median turn-around testing time was 21 days. Ascitic cfDNA was confirmed as tumoral in 98% of cases, largely due to detection of TP53 mutations (85,5%, 41/48), other variants in cftDNA included. Seven (15%) and 6 (12,5%) patients had BRCA1 and BRCA2 mutations, respectively. For 36 pts, NGS results on matching tumor tissue were available, and among those 97% (35/36) had concordant mutations detected in ascitic cftDNA and tumor tissue DNA. In 5 pts for whom tumor tissue-based NGS was non-contributive, NGS on cftDNA from ascites was informative and uncovered a BRCA2 mutation in one pt. We then performed GI testing on 18 cftDNA samples using sWGS and 100% of samples yielded contributive GI test results. Of these 18 pts, 42% (7/18) has failed GI testing on matching tumor sample. Conclusion: Genomic profiling on cftDNA from ascites is feasible, yields high quality and quantity tumor derived cfDNA. Importantly this approach has a fast testing turn-around time and only requires 20ml of ascites so that the majority of pts with advanced OC would be eligible to this approach. This cftDNA is suitable for HRD testing combining both BRCA mutation analysis and GI testing and should be done in every patient at primary laparoscopy Citation Format: Cyril Roussel-Simonin, Felix Blanc-Durand, Etienne Roueau, Audrey Le Formal, Laure Chardin, Elisa Yaniz, Sebastien Gouy, Amandine Maulard, Stéphanie Cherrier, Claire Sanson, Lea Mauny, Sophie Coterret, Francois Zaccarini, Roseline Tang, Alexandra Leary. Homologous recombination deficiency (HRD) testing on cell-free tumor DNA in ascites of patients with newly diagnosed advanced epithelial 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 3368.

Objectives: In patients with recurrent advanced endometrial carcinomas (EC), innovative treatments are needed after platinum chemotherapy. EC are characterized by frequent defective DNA repair, along with alterations of PI3K-AKT-mTor, and IGF1R pathways. In preclinical studies, the activity of PARP inhibitor is increased by the inhibition of PI3K-AKT pathway. Metformin inhibits the PI3K-AKT-mTor signaling pathway, and reduces IGF1 circulating levels. Metronomic cyclophosphamide may increase the activity of PARP inhibitors through DNA alkylating and anti-angiogenic effects. ENDOLA trial was meant to assess the safety and efficacy of the triplet olaparib (PARP 1-3 inhibitor) + metronomic cyclophosphamide + metformin in patients with advanced EC Methods: ENDOLA trial (NCT02755844) was an academic French prospective multicenter phase I/II open-label dose-escalation study meant to assess the safety, the Recommended Phase II trial Dose (RP2D), and the efficacy of olaparib combined to metronomic cyclophosphamide 50 mg PO QD and metformin 500 mg PO TID in patients with recurrent advanced EC. In the phase I part, olaparib was dose-escalated from PO 150 mg to 300 mg BID with a continual reassessment method (CRM), aiming at defining the RP2D. In the phase II part, the cohort was expanded to assess the efficacy of the triplet in terms of non-progression rate at 10 weeks (H0= 20%; H1= 50%; 1-sided α risk=0.01; power = 90%). Results: From Sept 2016 to Nov 2019, 35 patients enrolled, and 31 were assessable (median age: 69, IQR: 62-72; endometrioid: n=18 (51.4%); serous: n=11 (35.5%); carcinosarcoma: n=2 (6.5%)). In the phase I part (17 patients), 1 DLT (5.9%) was observed: grade 3 fatigue during cycle 1 at 150 mg. Dose-escalation was possible up to 300 mg BID, that was the RP2D. In the phase II part (14 patients), the overall response rate was 20.8%, and the disease-control rate was 66.6 %. The non-progression rate at 10 weeks was 61.5% (95% CI 42.8-80.2). The median PFS was 5.1 months, including 7.5 months (5.1-8.9) for endometrioid, and 4.3 months (2.7-5.8) for serous carcinomas. Treatment-related serious adverse-events (AEs) occurred in 3 patients (grade 3 fatigue; grade 4 lymphopenia; grade 4 lymphopenia & neutropenia). The most frequent grade 3-4 treatment-related AEs were lymphopenia (32.3%); anemia (16.1%); neutropenia (16.1%); and fatigue (12.9%). Conclusions: In heavily pre-treated patients with advanced EC, olaparib 300 mg BID could be safely combined to metformin 1500 mg/day and metronomic cyclophosphamide 50 mg/day. With a 61.5% non-progression rate at 10 weeks, the 1st-objective was reached. The 5.1 month median PFS (7.5 for endometrioid and 4.3 for serous) is promising compared to those observed with chemotherapy or immunotherapy-based combination in KEYNOTE-775 (Colombo et al. Proc ESMO 2021). Investigation in phase III is warranted. Citation Format: Benoit You, Alexandra Leary, Manuel Rodrigues, Philippe Follana, Cyril Abdeddaim, Florence Joly, Sylvie Bin, Laurent Villeneuve, Marine Alexandre, Florent Boutitie, Delphine Maucort-Boulch, Verane Schwiertz, Gilles Freyer. Safety and efficacy of olaparib combined to metronomic cyclophosphamide and metformin in recurrent advanced/metastatic endometrial cancer patients: ENDOLA trial [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 CT005.

High tumor mutational burden assessed in tissue biopsies (tTMB-H) or blood (bTMB-H) is associated with clinical efficacy in patients treated with immunotherapies. CheckMate 848 (NCT03668119) is a prospective phase 2 study of nivolumab (NIVO) with or without ipilimumab (IPI) in patients with advanced or metastatic solid tumors that are tTMB-H or bTMB-H (≥ 10 mutations/megabase) who were immunotherapy-naive and refractory to standard local therapies. The primary endpoint was objective response rate (ORR) in patients with tTMB-H or bTMB-H, assessed by FoundationOne® CDx-based and Clinical Trial Assays (Foundation Medicine), respectively. The study was not powered to compare NIVO + IPI vs NIVO. We present the interim and final analyses for the tTMB-H and bTMB-H cohorts, respectively (≥ 12 months follow-up, database lock June 2021). Of 1954 screened patients, 212 were randomized 2:1 to NIVO 240 mg Q2W + IPI 1 mg/kg Q6W or NIVO 480 mg Q4W for ≤ 24 months, and 201 (135 tTMB-H; 147 bTMB-H) were refractory to standard therapies. Of > 40 tumor types, colorectal (10.8%), small-cell lung (7.5%), breast (7.1%), and uterine (7.1%) were the most common. ORR and survival outcomes with NIVO + IPI were improved in patients with tTMB-H. The responses were independent of bTMB-H status in the tTMB-H cohort but improved with tTMB-H status in the bTMB-H cohort (Table). The safety profile of NIVO + IPI was manageable, and clinical outcomes with NIVO were comparable with previous studies. The impact of TMB cutoff, PD-L1 expression, and microsatellite instability were explored. In conclusion, NIVO + IPI demonstrated clinical efficacy with a manageable safety profile in patients with advanced or metastatic solid tumors that are tTMB-H or bTMB-H and refractory to standard therapies, with increased efficacy observed in patients with tTMB-H. NIVO + IPI tTMB-H cohort bTMB-H cohorta Patients, n (%)b,c 68 (32.1) 80 (37.7) Number of prior treatments, median (range) 2 (0–7) 2 (1–9) ORR, n (%)c, 95% CI 24 (35.3), 24.1–47.8 18 (22.5), 13.9–33.2 ORR in patients with bTMB-H by tTMBc: < 10 mut/Mb (n = 31), n (%), 95% CI NA 3 (9.7), 2.0–25.8 ≥ 10 mut/Mb (n = 39), n (%), 95% CI NA 13 (33.3), 19.1–50.2 ≥ 10 to < 16 mut/Mb (n = 18), n (%), 95% CI NA 3 (16.7), 3.6–41.4 ≥ 16 mut/Mb (n = 21), n (%), 95% CI NA 10 (47.6), 25.7–70.2 ORR in patients with tTMB-H by bTMBc: < 10 mut/Mb (n = 20), n (%), 95% CI 7 (35.0), 15.4–59.2 NA ≥ 10 mut/Mb (n = 43), n (%), 95% CI 16 (37.2), 23.0–53.3 NA ≥ 10 to < 16 mut/Mb (n = 12), n (%), 95% CI 3 (25.0), 5.5–57.2 NA ≥ 16 mut/Mb (n = 31), n (%), 95% CI 13 (41.9), 24.5–60.9 NA Percentage of responders (≥ 9 months) (95% CI) 91 (68–98) 88 (61–97) Median PFS, months (95% CI)c 4.1 (2.8–11.3) 2.8 (2.3–3.0) Median OS, months (95% CI)c 14.5 (7.7–NE) 8.5 (5.8–10.5) aThe bTMB cohort was randomized prior to December 20, 2019. bOut of 212 randomized patients; data presented in this table are from patients who were refractory to standard therapies. cMinimum follow-up 12 months. bTMB, blood tumor mutational burden; NA, not applicable; NE, not evaluable; PFS, progression-free survival; OS, overall survival; tTMB, tissue tumor mutational burden. Citation Format: Michael Schenker, Mauricio Burotto, Martin Richardet, Tudor Ciuleanu, Anthony Goncalves, Neeltje Steeghs, Patrick Schöffski, Paolo A. Ascierto, Michele Maio, Iwona Lugowska, Lorena Lupinacci, Alexandra Leary, Jean-Pierre Delord, Julieta Grasselli, David S. Tan, Jennifer E. Friedmann, Jacqueline Vuky, Marina Tschaika, Ruta Slepetis, Georgia D. Kollia, Misena Pacius, Ning Huang, Parul Doshi, Jonathan Baden, Massimo Di Nicola. CheckMate 848: A randomized, open-label, phase 2 study of nivolumab in combination with ipilimumab or nivolumab monotherapy in patients with advanced or metastatic solid tumors of high tumor mutational burden [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 CT022.

Background: Small cell ovarian carcinoma of the hypercalcemic type (SCCOHT) is an extremely rare and aggressive tumor that affects mainly young women (median age = 24 years). Prognosis is poor as most patients die within 2 years of diagnosis. Until recently, the literature describing the genomic profile of SCCOHT was scarce. In the last few months, four groups have now confirmed that SMARCA4 mutations are a frequent event in SCCOHT. Methods: We performed an integrated genomic analysis using WES, RNA-seq and CGH approaches to identify other recurrent genomic alteration and potential therapeutic targets. Candidate SNVs identified by WES were validated by Sanger and RNA-Seq. 8 frozen tumors were available, including 6 with matching normal control DNA. In addition 33 tumor samples were available as FFPE. Results: Despite their high grade and aggressive clinical course, SCCOHT tumors display genomic stability, low mutation load (0.53mut.Mb) and few SCNAs. However when present, genomic alterations were recurrent. Integrated WES and CGH analysis revealed a frequent and unusual 19p CN LOH (in 5 of 6 tumors). Although overall mutation rate was low, the mutations that did occur were novel and recurrent in 50%-80% of SCCOHT. In addition to SMARCA4 mutation (M+), we validated 10 novel recurrent M+ (including ABCA7, ANKRD24, CACTIN, EMR1, FBN3, FUT5, GRIN3B, KANK3, KRI1 and PLK5) in 50%-80% of tumors at high allelic frequencies (mean=0.87). Similarly, when present, amplifications were recurrent, common to a third of tumors: such as the mitochondrial redox enzymes (SHMT2, NDUFA4L2) or the transmembrane transporter LRP1 were amplified (Log2>2.3). No common fusion transcripts were identified. Since SCCOHT often display initial chemosensitivity, but rapid progressions, we investigated the differential genomic profiles of treatment naive versus chemotherapy treated tumors to uncover candidate resistance genes. The only alteration significantly enriched in post treatment tumors were M+ in the putative efflux pump, ABCA7. These M+ could not be identified even at low allele fractions in treatment naive tumors, but were identified in all 3 treated tumors with a mean allelic frequency of 0.83. In terms of therapeutic implications, modulating mitochondrial metabolism or targeting the membrane transporter LRP1 or the ABCA7 efflux pump could provide therapeutic venues in the future. However the most promising approaches for now may be targeting cell cycle checkpoint regulators or chromatin remodelling. M+ in the tumor suppressor PLK5 were confirmed in 79% of a cohort of 33 SCCOHT and many were predicted as damaging. Inactivation of PLK5 is associated with a loss of G2/M checkpoint regulation and cell harbouring PLK5 mutations were sensitive to PLK1 inhibitors. More interestingly, SMARCA4-M+ SCCOHT demonstrated complete loss of SMARCA4 expression as well as loss of the only other SWI/SNF catalytic subunit SMARCA2 suggesting that double SMARCA4/SMARCA2 negative SCCOHTs are characterized by a catalytically inactive SWI/SNF complex. This hypothesis was supported by changes in RNA expression levels in the main transcriptional targets for SWI/SNF (RHOA, RB and the E2F family). Importantly, the double SMARCA2/4 negative BIN-67 SCCOHT cell line displaying both a SMARCA4 M+ and loss of SMARCA2 expression was exquisitely sensitive to HDAC and MT inhibitors, while cell lines with either a SMARCB1-M+ or SMARCA4-M+ with retained SMARCA2 expression were not. These data make double SMARCA2/4 negative by IHC a potential predictive biomarker for histone modifying inhibitors in SCCOHT or potentially in other double SMARCA2/4 negative malignancies. Conclusions: We present the 1st integrated genomic analysis of SCCOHT. SCCOHT demonstrate a remarkably homogeneous genomic profile and potentially actionable alterations. In particular, dual null SMARCA2/SMARCA4 SCCOHTs could be selected for clinical trials with EZH2, HDAC or MT inhibitors. Citation Format: Aurélie Auguste, Marc Deloger, Joanna Cyrta, Audrey Le Formal, Catherine Richon, Nathalie Droin, Beatrice Brunn, Olivier Caron, Mojgan Devouassoux, Catherine Lhomme, Patricia Pautier, Catherine Genestie, Alexandra Leary. Small cell carcinoma of the ovary, hypercalcemic type (SCCOHT) beyond SMARCA4 mutations: A comprehensive genomic analysis. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A1-27.

Background: Neoadjuvant chemotherapy (NAC) followed by interval debulking does not present inferior results to those of primary cytoreduction and offers the opportunity to evaluate chemo-sensitivity in vivo. Pathological response scores (CRS) have been shown to correlate with outcome with a complete or near-complete CRS (CRS3) predicting improved progression-free survival. Approximately 20% of ovarian cancers present BRCA1/2 mutations, which predict a better response to platinum salts. Our proposal is to determine the prevalence of BRCA mutations or other homologous recombination (HR) alterations among patients with a CRS3 after NAC. Methods: We performed a retrospective analysis of clinical, pathological, and sequencing data of patients who experienced a complete or near-complete response to platinum based NAC. For patients with wild-type (WT) sequencing and with tumor samples available, somatic analysis based on Next Generation Sequencing (NGS) with a large panel of genes (BRCA1, BRCA2, ATM, BARD1, BRIP1, CCNE1, CDK12, CHEK2, PALB2,RAD51C, RAD51D, TP53), including specific non-coding regulatory regions of BRCA1, BRCA2 and RAD51C, was also performed. We present preliminary data on the 1st cohort analysed at the time of abstract submission. Results: A total of 33 patients were identified who demonstrated CRS3 post-NAC. Tumors were in majority EC IIIc (57%) and grade II-III serous ovarian carcinoma (90%). Median overall survival (m OS) of the entire cohort was 55 months (IC95% 36-74). To date, germline and/or somatic analyses were available on 23 pts. The prevalence of pathogenic BRCA1/2 variants is higher than expected in this cohort of patients presenting a CRS3. In total, 7 of 23 patients (30%) had a germline (4) or somatic pathogenic (3) BRCA variant. In addition, among the 16 BRCA1/2 WT pts subjected to further HR NGS analysis, the following cases of alterations were identified: one ATM nonsense mutation (c.2465T>G; p.Leu822*), one pathogenic mutation of CDK12 gene (c.3G>A;p.Met1?), one likely pathogenic variant located on BRCT domain of BRCA1 (c.5165C>T p.Ser1722Phe), two 3'UTR BRCA2 variants of uncertain significance (VUS) (c.*14C>T; c.*72A>G), and one BRIP1 VUS (c.2932G>C, p.Gly978Arg). Further, one BRCA WT case presented an unexpected CCNE1 amplification. Conclusion: HGOC patients presenting with a pCR are enriched for BRCA germline and/or somatic BRCA mutations. In addition, among the small subset of BRCA WT pCR tumors, a deleterious ATM and CDK12 mutation were identified. This prevalence can be underestimated in a context of pCR since somatic screening is impossible on the debulking material. Somatic BRCA1/2 and complete HR sequencing on the initial biopsy is recommended to identify additional pts with exclusive somatic mutations. More complete data on this cohort will be presented. Citation Format: Etienne Rouleau, Elizabeth Santana dos Santos, Francilette Maela, Ludovic Lacroix, Aurélie Auguste, Patricia Pautier, Philippe Morice, Catherine Genestie, Alexandra Leary. Mutation analysis of ovarian carcinoma patients presenting optimal response to neoadjuvant chemotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3504.

Background: Human cancers are associated with defects in DNA damage repair (DDR). Evidence suggests a key role for poly(ADP-ribose) polymerases (PARPs) in repairing certain DNA damage lesions. Olaparib (a PARP inhibitor) is approved for pts with germline BRCA1 and BRCA2-mutated breast and ovarian cancers, and pts with or without BRCA mutation in recurrent ovarian cancer. Mutations in HRR genes (HRRm; beyond BRCA1 or BRCA2) and HRD may sensitize tumors to respond to PARP inhibitors. Olaparib has also shown favorable ORRs in HRRm/HRD-positive (HRD+) cancers, including prostate and ovarian. Thus, this study (NCT03742895) aims to evaluate the use of HRRm/HRD+ as biomarkers of tumor-agnostic response to olaparib monotherapy. Methods: Eligibility for this open-label, Phase II study requires pts ≥18 years of age with a previously treated, histologically/cytologically confirmed, HRRm/HRD+ advanced solid tumor; who have failed, are intolerant to, or ineligible for all available standard of care therapies; with no disease progression during prior platinum-based treatment. At baseline, pts should also have measurable disease per RECIST v1.1 or Prostate Cancer Working Group (PCWG)-modified RECIST v1.1 (for pts with prostate cancer) without CNS metastases, ECOG PS of 0 or 1, and no persistent toxicities (CTCAE grade >2) from prior cancer therapy. Pts may have received any number of prior regimens. Newly obtained or archival tumor samples will be centrally evaluated using the Lynparza HRR Assay (under development at Foundation Medicine, Inc). Pts will be grouped into 2 cohorts: cohort 1 will include pts with any solid tumor type with a BRCA1 or BRCA2 mutation (excluding breast and ovarian cancers); cohort 2 will include pts whose tumor (any type) is BRCA1 or BRCA2 wild-type and HRRm and/or HRD+ as determined by loss of heterozygosity score. Pts will receive olaparib 300 mg orally BID until documented disease progression, unacceptable toxicity, or study withdrawal. Tumor imaging will be performed at baseline, with response assessments Q8W in the first year and Q12W thereafter, per RECIST v1.1 by BICR or PCWG-modified RECIST v1.1, both modified to follow ≤5 target lesions per organ (10 total). AEs will be graded using NCI CTCAE v4.0. The primary endpoint is ORR; the point estimate and 95% CI calculated using the Clopper-Pearson method will be provided. Summary statistics for key secondary efficacy endpoints of duration of response, OS, and PFS will be provided using the Kaplan-Meier method. Safety endpoints will be summarized. Approximately 370 pts will be enrolled (~84 in cohort 1 [BRCA1 and BRCA2 mutated]; ~174 HRRm and ~112 HRD+ in cohort 2 [BRCA1 and BRCA2 wild-type]). Enrollment is ongoing in 5 countries beginning December 12, 2018. Citation Format: David Hyman, Andrew Hendifar, Hyun Cheol Chung, Michele Maio, Alexandra Leary, Iben Spanggaard, Joon Rhee, Matt Marton, Menghui Chen, Suba Krishnan, Ronnie Shapira. Olaparib in patients (pts) with previously treated, homologous recombination repair mutation (HRRm) or homologous recombination deficiency (HRD)-positive advanced cancer: Phase II LYNK-002 study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr CT229.

Rationale: ER+/HER2+ BC may be resistant to endocrine therapies, but sensitive to HER2 inhibitors such as lapatinib. There is evidence that lapatinib may reactivate ERα expression1,2 raising the possibility that lapatinib may sensitize BC to subsequent endocrine therapy. Whether targeting the ER and HER2 in combination or in sequence in ER+/HER2+ BC may enhance anti-tumour effect is unknown. Aims: (1) To determine in ER+/HER2+ aromatase-expressing breast cancer models in vitro and in vivo.whether the combination of lapatinib with letrozole or tamoxifen is synergistic, and (2) to investigate whether using lapatinib and letrozole in sequence rather than in combination may provide a superior strategy. Methods: ER+/HER2+ BT474 cells were stably transfected with aromatase (BTArom). Combination growth studies, immunoblots and gene expression assays were conducted in BTArom cells. BTArom xenograft studies were performed over 18 weeks under androstenedione support and mice randomized to daily treatment with vehicle, lapatinib, letrozole, or lapatinib and letrozole (L+L). In addition, mice in the lapatinib alone arm, were randomized at 6 weeks to continued lapatinib, vs. switch to letrozole, vs switch to L+L. Results: Formal combination studies in aromatase-transfected BT474 (BTArom) cells demonstrated synergy between lapatinib and letrozole (combination index, CI=0.4, p0.05). However for the group of mice treated with lapatinib for 6 weeks followed by the combination of lapatinib and letrozole for 12 weeks, tumour volume increased only 2.2-fold, compared to 4.5-fold for those continuing on lapatinib, 5.7-fold for those switching to letrozole and 5.2-fold for those on lapatinib and letrozole from the start. Conclusion: We have demonstrated that HER2 targeted therapy enhanced responsiveness to aromatase inhibition in an ER+/HER2+ breast cancer model in vitro and in vivo. In contrast, the interaction between lapatinib and 4-OH-tamoxifen was antagonistic. Lapatinib increased ER expression and function, and the most effective tumour control was achieved with lapatinib alone followed by the combination of lapatinib and letrozole. Maximizing benefit from ER and HER2 targeted therapies in ER+/HER2+ BC may involve sequential rather than concurrent treatment to exploit lapatinib-induced ER reactivation. 1 Leary et al, Clin Canc Res 2010 2Xia et al, PNAS 2006 Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P3-14-03.

The anti-Müllerian hormone (AMH) belongs to the TGF-β family and plays a key role during fetal sexual development. It has been shown that AMH inhibits proliferation and induces apoptosis of AMH type II receptor (AMHRII) -positive tumor cells, especially progenitor cancer cells. Moreover, various reports have described the expression of AMHRII in human gynecologic cancers including ovarian tumors. Based on qRT-PCR test results, which were confirmed by a specific In-Situ Hybridization (ISH) assay, AMHRII mRNA could be detected only in normal ovary, testis and adrenal gland tissue, and at lower levels in pancreas and stomach. However, when a tissue microarray of solid tumor samples was tested with ISH assay, AMHRII mRNA was surprisingly detected in several primary cancer tissues including melanoma, lung, head and neck, colorectal and breast cancers. For example, a large ISH study involving 58 colorectal cancer samples demonstrated AMHRII expression in 55% of cases. AMHRII protein expression was confirmed by ImmunoHistoChemistry (IHC) in gynecological cancer samples including 179 out of 210 (85%) Granulosa Cell Tumor samples and 55 out of 80 (69 %) Epithelial Ovarian Cancers. In patients with gynecological cancers pre-screened by IHC for inclusion in a Phase Ia/Ib clinical trial of GM102, an anti-AMHRII monoclonal antibody, AMHRII positive membrane expression was detected in 116 out of 166 (70%) archived tumor tissues. Using the same IHC protocol, over 1000 frozen samples covering 16 different cancer types were tested, which demonstrated AMHRII expression in over 50% of hepatocarcinoma, colorectal, lung and renal cancer samples. No AMHRII expression was observed in 75 neuroendocrine lung tumor samples nor in 18 non-Hodgkin lymphoma samples tested. On a small panel of frozen ovarian (11) and colorectal (9) cancer samples using an immunofluorescence assay with an AlexaFluor488-conjugated GM102 antibody, AMHRII expression was confirmed on the matching fresh tissue samples, previously detected by IHC in their fixed counterpart samples. Preliminary FACS analysis of fresh ovarian and colorectal tumor samples have demonstrated comparable expression levels with mean values of 50,000 and 40,000 AMHRII receptors per cell, respectively. In conclusion, the results presented above, using a variety of detection methods and systems, have consistently shown that AMHRII is expressed in many different histological types of solid tumors. These results suggest that this embryonic receptor could be a suitable target for selective antitumor agents such as GM102. Citation Format: Jean-Marc M. Barret, Didier Meseure, Guillaume Bataillon, Noora Andersson, Aurélie Auguste, Olivier Dubreuil, Anniina Färkkilä, Emeline Perrial, Celine Bossard, Emily Loison, Anne Jarry, André Nicolas, Fanny Lemée, Solenne Gaillard, Mehdi Lahmar, Mikko Anttonen, Gabriel Champenois, Charles Dumontet, Isabelle Ray-Coquard, Alexandra Leary, Isabelle Tabah-Fisch, Anne Vincent-Salomon, Jean-François F. Prost. Anti-Müllerian hormone type II receptor (AMHRII) found expressed in human non-gynecological solid tumors, suggesting potential broader applications for anti-AMHRII-based therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 774.

Background: Onapristone is a type I PR antagonist, which prevents PR-induced DNA transcription; its anti-cancer activity is well-documented. The reported T1/2 is 2-4h, so an extended-release (ER) tablet was designed to mitigate Cmax spikes, which may be involved in the liver function test (LFT) elevations seen with an immediate-release (IR) onapristone. Materials and methods: The expansion cohort of this multi-center, open-label, randomized, parallel-group, 2-stage Ph1 study (NCT02052128) is ongoing in endometrioid cancers. Female pts ≥18 years with PR-expressing tumors (including endometrial, ovarian, breast) were eligible. The primary endpoint was to recommend a Ph2 dose of ER onapristone (RP2D), with a 57-day DLT observation period; secondary endpoints include: safety, efficacy, and real-time PK. Pts received onapristone ER 10, 20, 30, 40 or 50 mg BID, or IR tablets 100 mg QD until progressive disease or intolerability in Stage 1. Results: 52 pts are enrolled (by 2 December 2014). Validated PK data are available for 35 pts. Onapristone AUC and Cmax are dose-proportional across all dose levels including 100mg IR (Table 1), with coefficients of determination (r2) of 0.76 and 0.79, respectively. The bioavailability of onapristone ER vs IR is high. Steady state is consistently attained at approximately 8 days (200 h), with T1/2 at 8-12 h, longer than previously published, with no evidence of onapristone accumulation through day 57. Table 1.Onapristone PK parameters following first doseOnapristone formERERERERERIRDose (mg)10 bid20 bid30 bid40 bid50 bid100 qdn666566Mean AUC (μg/L*h)51761765015450174403161067980CV%47.7202.722.943.070.555.1Min348029411200038091515033330Median4654860815040207302593062270Max952440000214302667062500125000Mean Cmax (μg/L)226.8676.3767.3641.614594296CV%46.0140.115.693.646.862.8Min922206762914591556Median218403732.547414953726Max4031466988134424927507 Conclusions: This study reveals a longer than anticipated onapristone T1/2, which nevertheless supports using an ER formulation. AUC and Cmax data are dose proportional and allow dosing flexibility. Protracted exposure minimizing Cmax spikes is best achieved with a twice-daily, ER formulation. There were no LFT elevations in the absence of liver metastases in 30 patients exposed to the ER formulation. The RP2D is 50 mg bid based on safety and PK, to be further explored in the endometrioid cohort. Citation Format: Francois Lokiec, Antoine Italiano, Andrea Varga, Jacques Bonneterre, Mario Campone, Alexandra Leary, Keyvan Rezai, Marie-Paule Sablin, Alice Bexon, Stefan Proniuk, Erard Gilles, Joseph Bisaha, Alexander Zukiwski, Paul Cottu. Onapristone in patients (pts) with progesterone receptor (PR)-expressing cancers: PK results from part 1 of a randomized, parallel-dose phase 1 study. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4512. doi:10.1158/1538-7445.AM2015-4512

Background: Onapristone is a type I PR antagonist, which prevents PR-induced DNA transcription. Onapristone anti-cancer activity is well documented. An extended-release (ER) tablet formulation of onapristone was designed to address the liver function test (LFT) elevations seen with immediate-release (IR) onapristone. A phase 1 study with onapristone in patients with tumors expressing PR is underway. Objectives included determining the PK profile of ER onapristone using a PPK approach. Materials and methods: This is an ongoing multi-center, open-label, randomized, parallel-group, 2-stage ph1 study. Female pts ≥18 yrs with tumors expressing PR are eligible. The Stage 1 primary endpoint is the recommended ph2 dose of ER onapristone; secondary endpoints include: safety, efficacy, and PK. Pts received onapristone ER 10, 20, 30, 40 or 50 mg BID, or onapristone IR tablets 100 mg QD until progressive disease or intolerability. PK blood samples from 8 time points were collected over 12 h post-dose Day 1 for the ER and 9 blood samples over 24 h post-dose for the IR formulation. Onapristone plasma concentrations were measured using validated UPLC with tandem mass spectrometry detection (range 1-250 ng/mL). Monolix V4.1 was used to calculate absorption constant (Ka); apparent clearance (CL/F); inter-compartmental clearance (Q); apparent distribution volume (V1/F), 2nd compartment distribution volume (V2) and bioavailability (F) of ER vs IR. Results: Stage 1 is complete. 42 pts have validated PK data. A 2-compartment open model adequately described the total onapristone time-concentration curve with linear elimination. Results are in Table 1. Table 1.Estimated PK parameters for onapristone in pts with PR-expressing cancers (n = 42)ParameterValueRelative standard error (%)Ka0.191 h-114CL/F1.51 L/h20Q3.11 L/h25V1/F5.41 L25V241.1 L45F60%20 Conclusions: The PPK modeling described the plasma onapristone time-concentration curves well. A central volume equivalent to the circulating blood volume and a large volume of the deep compartment suggest a large tissue diffusion. PPK/PD modeling to explore safety and efficacy is ongoing, with no overt PK/safety relationship detected. Citation Format: Keyvan Rezai, Paul Cottu, Samuel Huguet, Mario Campone, Antoine Italiano, Andrea Varga, Jacques Bonneterre, Alexandra Leary, Marie-Paule Sablin, Stefan Proniuk, Alice Bexon, Erard Gilles, Joseph Bisaha, Alexander Zukiwski, Francois Lokiec. Population pharmacokinetic (PPK) modeling of onapristone in patients (pts) with progesterone receptor (PR)-expressing cancers. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4523. doi:10.1158/1538-7445.AM2015-4523

Aim: To identify pretreatment biomarker predictors of Ki67 response to lapatinib in women with HER2− primary breast cancer. Background: Lapatinib is an EGFR/HER2 inhibitor. Its clinical use is restricted to HER2 overexpressing disease. The MAPLE (Molecular Antiproliferative Predictors of Lapatinib's Effects) presurgical window of opportunity study of lapatinib vs placebo was conducted in women with HER2-amplified (HER2+) or HER2− primary disease. Ki67 (primary end-point) was reduced by a geomean 46% (95%CI 23–63%, p = 0.002) and 27% (95%CI 8–42%, p = 0.008) in HER2+ and HER2− disease, respectively (Leary et al, AACR 2012). We have now assessed whether predictive biomarkers of the antiproliferative response in HER2− disease could be identified. Methods: 121 primary breast cancer patients were randomized (3:1) to 14 days of 1500mg/d lapatinib or placebo before surgery. Biopsies were taken before treatment and at surgery. Ki67 responders were defined as having a >/=50% reduction in Ki67 compared to baseline (Ellis, P et al, Breast Cancer Res Treat 1998, 48, 107). ER, PgR, HER2, EGFR, pAKT, pERK1/2 (nuclear and cytoplasmic), stathmin and apoptosis (TUNEL) were assessed by IHC (+FISH for HER2[all cases]) and scored visually by continuous methods. HER2, HER3, epiregulin (epir), amphiregulin (amphir) and neuregulin (neur) were assessed by qrtPCR. Results: Three of the 121 patients were excluded because of inadequate biopsy material. Ninety-one of the remaining 118 patients received lapatinib: 7/19 (37%) HER2+ cases and 10/72 (14%) HER2− cases were Ki67 responders. Thus while the proportion of Ki67 responders was higher for HER2+ disease there was a similar or higher absolute number of responders with HER2− disease. All of the following relates to patients with HER2− disease. None of the pretreatment levels of ER, PgR, pAKT, pERK1/2, EGFR, epir, amphir or neur were associated with Ki67 response (p > 0.20). However, HER3 (p = 0.01) and HER2 (p = 0.06) mRNA levels were associated with greater Ki67 response. There was a tendency for Ki67 response to be greater with lower baseline Ki67 (p = 0.07). Multivariate analysis showed only HER3 mRNA levels to be independently significant. HER2 and HER3 mRNA levels were highly correlated (rho = 0.67, p < 0.001), a relationship confirmed in 2 other datasets (Wang et al, Breast Cancer Res, 2011, 13, R92; Dunbier et al, submitted). All Ki67 responders were above the median for both HER3 and HER2 expression. Conclusions: Lapatinib is antiproliferative in a subgroup of HER2− tumours. This exploratory analysis indicates that they are characterized by high HER3 expression. The possible importance of high HER2:HER3 heterodimers in predicting this response is supported by the relationship between HER2 and HER3 expression. Further exploration of lapatinib is merited in HER2− cases with high HER3 expression. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD07-07.

Not all patients (pts) with HER2 positive (+) breast cancer (BC) benefit from treatment (tx) with EGFR/HER2 tyrosine kinase inhibitor lapatinib (L) & it is not known whether HER2 over-expression or amplification are obligate requirements for L response. Assessment of in vivo biomarkers of drug activity, e.g. change in Ki67, after short-term pre-surgical tx may elucidate the differential molecular profile of sensitive vs. resistant tumors. Aims: i) Determine whether L induces anti-proliferative effects in HER2+ & HER2 negative (-) BC & ii) identify biomarkers of sensitivity/resistance to anti-proliferative effects of L. Methods: Women with newly diagnosed BC were randomized (3:1) to receive 10-14 days of pre-operative L (1500mg/day) or placebo (P) in a multicentre phase II trial (ISRCTN68509377). Paired core biopsies before & after tx were analyzed for Ki67, TUNEL, HER2, EGFR, ER, PgR, pAkt, pErk, & stathmin (candidate marker of PI3K/Akt activation) by IHC +/− FISH. HER2, HER3 & ligands EREG, AREG & NRG1 mRNA was measured by RT-PCR. The primary endpoint was change in Ki67. HER2+ was defined as 3+ or 2+ by IHC and FISH+. Differences in geometric means were assessed by Mann-Whitney & correlations by Spearman rank. Results: 121 pts (L=94, P=27) were randomized between 12/2007-04/2011, 70% were ER+/PgR+, 13% ER+/PgR- & 17% ER-; 22% were HER2+ (including one 2+/FISH+) & 26% were EGFR+. L pts reported significantly more frequent G1-2 (64%) or G3 (6%) rash, & G1 diarrhea (56%); the only other G3 toxicity was infection in 1 pt. There were no delays in surgery. Paired samples containing tumor were obtained for 98% (118/121) of randomized pts. Ki67 fell significantly in the L group (relative reduction in post- vs. pre-tx samples = -31%, 95%CI: -44 to -16; p Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-222. doi:1538-7445.AM2012-LB-222

Background: Approximately 80% of human breast carcinomas present as estrogen receptor α-positive (ER+ve) disease and ER status is a critical factor in treatment decision-making. However, ER+ve tumors are heterogeneous in both their molecular profiles and response to therapy. Up to 50% of patients derive little or no clinical benefit from treatment and many exhibit de novo or intrinsic resistance to hormonal therapy. Aim: To investigate factors associated with the level of expression of the ER gene that may be potential determinants of the biology of ER+ve tumors. Methods: Pre- and 2wk post-anastrozole tumor biopsies were obtained from 104 postmenopausal women with stage I to IIIB ER+ early breast cancer. RNA extracted from biopsies was analyzed on Illumina 48K microarrays and quantitative trait analysis by Spearman correlation was used to identify genes whose expression in tumors correlated significantly with ESR1. Results: Expression of three previously uncharacterized open reading frames (ORFs) located immediately upstream of ESR1 on chromosome 6q21, RMND1, C6ORF97 and C6ORF211, were highly correlated with ESR1 (Rs=0.672, 0.637 and 0.546 respectively, p Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2916.

Abstract #3031 Background Growth factor (GF) signaling is frequently upregulated in hormone resistant breast cancer (BC) and provides the rationale for using the dual EGFR/HER2 inhibitor, lapatinib (LAP) in an effort to overcome endocrine resistance. Hormone resistant BC can also vary in its degree of continued ER dependence. We previously demonstrated potent synergy between LAP and TAM in cell models of acquired endocrine resistance to tamoxifen (TAMR cells), and to estrogen deprivation (LTED cells)1. This study aimed to characterize the phenotype of these two models with a focus on the ER pathway, and to determine the biological basis for the observed synergy between LAP and TAM. Methods LTED cells and TAMR cells were both derived from hormone-sensitive MCF7 parental cells as previously described1. The differential effects of LAP, TAM, estrogen deprivation (ED) or the combination were assessed by proliferation assays, ERE reporter assays, quantitative PCR and western blot analyses. Results In LTEDs: Increased ER expression and ERE transactivation were seen in LTEDs compared to MCF7 cells, as well as continued sensitivity to the ER downregulator, fulvestrant (FUL), indicating that in LTEDs, ER activity is not only enhanced but is also a significant driver of growth. In LTEDs, LAP alone suppressed ER activity as measured by both ERE transactivation and ER phosphorylation at Ser118; the most effective suppression of ER activity was achieved with a combination of LAP+TAM/ED where the effects were additive. In TAMRs: In contrast, TAMRs demonstrated decreased ER activity (low basal ERE transactivation and loss of PgR) and ER independent proliferation (FUL had minimal growth inhibitory effects despite ER downregulation). In TAMRs, neither LAP nor TAM had significant antiproliferative effects on their own: LAP alone decreased pMAPK/pAkt but increased basal ERE transactivation and transcription of the ER-responsive gene pS2. TAM alone inhibited ERE activation but increased HER2 gene and protein expression. Inhibiting both HER2 and ER pathways with LAP+TAM resulted in synergistic growth inhibition. Conclusions In LTED cells with enhanced ER function attributable in part to GF receptor activation, the synergistic antiproliferative effects from LAP and TAM or ED may be largely due to enhanced inhibition of ER signaling. In contrast in TAMR cells with suppressed ER activity, LAP alone appears to enhance ER signaling, while TAM alone inhibits ER activity but increases HER2 expression. In TAMRs, the combination of TAM+LAP is required to interrupt both ER and GF signaling. These results suggest that the basis for synergy between LAP and endocrine therapy in hormone resistant BC may vary depending on the nature of the cross-talk between the ER and GF pathways. 1SABCS, 2006 Abs 303. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 3031.