American Association For Cancer Research Research Articles

Abstract CT108: A phase 1b study of abemaciclib in combination with pembrolizumab for patients (pts) with hormone receptor positive (HR+), human epidermal growth factor receptor 2 negative (HER2-) metastatic breast cancer (mBC) (NCT02779751): Preliminary results

Abstract Background: Abemaciclib is an orally administered, selective small molecule cyclin-dependent kinase (CDK) 4 and 6 inhibitor, approved to treat HR+, HER2- mBC pts on a continuous twice daily dosing schedule as monotherapy or in combination with endocrine therapy (ET), an aromatase inhibitor as initial endocrine based therapy, or in combination with fulvestrant. In preclinical models, abemaciclib monotherapy induced intra-tumoral immune inflammation and had a synergistically enhanced anti-tumor efficacy when administered in combination with programmed cell death protein 1 blockade. Here we report the safety and preliminary anti-tumor activity of abemaciclib plus pembrolizumab plus anastrozole in HR+, HER2- locally advanced or mBC pts. Methods: This multicenter, nonrandomized, open-label, multi-cohort phase 1b study of abemaciclib plus pembrolizumab plus anastrozole enrolled a cohort of post-menopausal HR+, HER2- pts with locally advanced or mBC. No prior CDK4/6 inhibitor, chemotherapy, or ET for locoregional advanced or mBC were allowed. Patients received 150-mg abemaciclib orally every 12 hours plus pembrolizumab 200 mg intravenously on day 1 every 21 days plus anastrozole 1 mg orally every 24 hours. The primary objective was to characterize the safety of abemaciclib plus pembrolizumab plus anastrozole. Key secondary objectives included objective response rate (ORR), progression-free survival, and overall survival. Results: Out of 26 pts enrolled, 13 (50%) had received prior systemic anti-cancer therapy, including 12 (46%) who had received ET in the adjuvant setting. The majority of pts had visceral disease (65%); 54% had 3 or more metastatic sites. The nature of adverse events (AEs) observed for the triplet therapy was overall consistent with the known side effects of abemaciclib, pembrolizumab and anastrozole. Grade 3/4 AEs in > 2 pts included increased alanine aminotransferase and neutropenia (8 pts each, 31%) and increased aspartate aminotransferase (6 pts, 23%). In general, Grade 3/4 AEs were reversible following drug holds and corticosteroid treatment. There were 2 fatal events as a result of pneumonitis. Fifteen pts (58%) discontinued treatment including 9 pts (35%) due to an AE (7 transaminase elevations). Preliminary response assessment showed 5 pts had a confirmed partial response (19% ORR), and disease control rate (CR+PR+SD) was 77%. Clinical benefit rate (CR+PR+SD persisting for ≥6 months) was 27%. Conclusions: Abemaciclib plus pembrolizumab plus anastrozole demonstrated a numerically higher rate of transaminase elevations and pneumonitis than reported for the individual treatments and 2 experienced a fatal AE. At this time, evaluation of anti-cancer activity is premature. Citation Format: Hope S. Rugo, J. Thad Beck, Guy Jerusalem, Hans Wildiers, Peter Kabos, Michael Chisamore, Rhian McNaughton, Yanyun Chen, Anwar Hossain, Sara M. Tolaney. A phase 1b study of abemaciclib in combination with pembrolizumab for patients (pts) with hormone receptor positive (HR+), human epidermal growth factor receptor 2 negative (HER2-) metastatic breast cancer (mBC) (NCT02779751): Preliminary results [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT108.

Abstract 2173: CRISPR-Cas9 gene editing of inhibitory receptor signaling pathways improves CAR T cell activity against glioblastoma

Abstract Early clinical trials have demonstrated the therapeutic potential of chimeric antigen receptor (CAR) T cells against glioblastoma (GBM), but also highlighted the obstacles that need to be addressed to improve their anti-tumor activity. GBM is highly immunosuppressive and thus necessitates endowing adoptively transferred CAR T cells with resistance mechanisms to achieve sustained treatment responses. This is, in part, due to the effect of immune-inhibitory receptors, including PD1, CTLA4, LAG3 and BTLA, that antagonize the tumor-directed T cells. A common mechanism of T cell inhibition involves the recruitment of Src homology region 2-containing protein tyrosine phosphatase 2 (SHP2) to the immune synapse upon engagement of multiple co-inhibitory receptors. SHP2, in turn, de-phosphorylates key signaling proteins involved in T cell activation. We hypothesized that disrupting SHP2 expression will offset multiple co-inhibitory pathways, thereby improving the anti-tumor activity of CAR T cells. Using CRISPR/Cas9 gene editing, we disrupted the SHP2-encoding gene locus, PTPN11, in primary human T cells by generating and testing a library of single-guide RNA (sgRNA). We selected the top candidate that reproducibly reduced the SHP2 protein expression to <10%. A HER2-targeting second-generation CAR (HER2-CAR) was then expressed in SHP2 knock-out (SHP2KO) T cells via gamma retroviral vector transduction. SHP2KO CAR T cell manufacturing following a single round of stimulation was optimized to minimize ex vivo manipulation of the cells. The phenotypic profiles of SHP2KO and wild-type HER2-CAR T cells were comparable at baseline. Loss of SHP2 did not alter the CAR T cell viability, expansion or proliferative capacity in comparison to the wild-type CAR T cells. Mass cytometry (CyTOF)-based analysis of SHP2KO CAR T cells after encountering GBM cells demonstrated the preservation of the central memory (CCR7+CD45RA-) T cell subset. The majority of PD1+ SHP2KO CAR T cells co-expressed functional activation markers such as ICOS and CD107a. The PD1+ T cell subset was comprised of mostly T-betHi progenitor population and only a smaller percentage of EomesHi terminally exhausted population. Functionally, HER2-specific SHP2KO CAR T cells from different donors (n=4) exhibited improved anti-tumor activity that was sustained in long-term cytotoxicity assays. In an experimental orthotopic LN229-GBM mouse model, SHP2KO HER2-CAR T cells better controlled the growth of established xenografts and improved survival in comparison to wild-type HER2-CAR T cells. In conclusion, our results show that disrupting SHP2 expression using a CRISPR/Cas9 gene editing strategy could improve the CAR T cell function against GBM and warrants further clinical testing. Citation Format: Khaled Sanber, Zeid Nawas, Vita Salsman, Ahmed Gad, Pretty Rose Matthew, Daniel Landi, Ciaran Lee, Amel Sengal, Rikhia Chakraborty, Sujith Joseph, Ahmed Nabil, Meenakshi Hegde. CRISPR-Cas9 gene editing of inhibitory receptor signaling pathways improves CAR T cell activity against glioblastoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2173.

Abstract 206: The landscape of oncogenic gene fusions in Chinese colorectal cancer patients

Abstract Background: Oncogenic gene fusions have been reported in patients with colorectal cancer (CRC) as promising actionable targets, but recognition of rare molecular subgroups is a challenge for precision oncology. We aimed to comprehensively characterize the clinical, pathological, and molecular landscape of Chinese CRC patients with oncogenic gene fusions. Methods: A total of 1226 patients with CRC underwent genomic testing between Aug. 2017 and Sep. 2019 using a tissue-based next-generation sequencing (NGS) assay. All the coding exons of 450 cancer-related genes and selected introns were captured by the custom hybridization capture panel. Microsatellite instability (MSI) and tumor mutational burden (TMB) were assessed in all patients. Results: Oncogenic gene fusions were found in 1.6% (20/1226) of Chinese CRC tumors, which was higher than that in the western population (0.68%, R. Madison, ESMO 2018 Congress). The most frequently detected fusions were NTRK1 (N = 5, 25%), BRAF (N = 5, 25%), ALK (N = 3, 15%), NTRK3 (N = 2, 10%), and RET (N = 2, 10%). FGFR1, ROS1, and EGFR fusions each occurred in one patient. Compared with patients without fusions, those with fusions were older (median age 69 vs 59 years, respectively, P=0.009), had right-sided CRC (75.0% vs 27.4%, respectively, P<0.001), MSI-high (45.0% vs 8.7%, respectively, P<0.001), and a higher TMB (median TMB of 9.8 vs 5.6 muts/mb, respectively). All MSI-high patients with fusions were TMB-high with a median value of 73.5 muts/Mb (range 51.8-180.3). The most frequently co-mutated genes in fusion cases were TP53 (70%), APC (60%), KMT2D (50%), ACVR2A (45%), RNF43 (45%), and TGFBR2 (40%). KRAS, NRAS, BRAF, POLE, and mismatch repair gene mutations were all mutually exclusive with oncogenic gene fusions. One patient harboring an EGFR-SEPT14 fusion, who had been heavily treated for right colon cancer metastasis, achieved a partial response to erlotinib. Conclusions: This study revealed that oncogenic gene fusions were found in 1.6% of Chinese CRC patients by using a tissue-based NGS assay. As oncogenic gene fusions are mutually exclusive with common oncogenic mutations, using NGS to detect fusions as actionable targets or resistance mechanisms has notable implications for precision therapy. Citation Format: Zhenhai Lu, Rongfeng Song, Kunsong Li, Ying Cheng, Wenjing Wang, Weifeng Wang. The landscape of oncogenic gene fusions in Chinese colorectal cancer patients [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 206.

Abstract 3271: Different pathologic response rates between Australia and Europe in macroscopic stage III melanoma patients upon neoadjuvant ipilimumab plus nivolumab in the phase II OpACIN-neo trial

Abstract Background In the multicenter investigator-initiated OpACIN-neo trial, patients (pts) with macroscopic stage III melanoma were randomized (stratified by center) to three different dosing schemes of neoadjuvant ipilimumab (IPI) + nivolumab (NIVO). Two cycles IPI 1 mg/kg + NIVO 3 mg/kg was identified as the most favorable regimen with 20% grade 3-4 adverse events and a pathologic response rate (pRR) of 77%. After a median follow-up of 17.7 months, relapses were observed in 1/64 (2%) of the pts with a pathologic response, and in 13/21 (62%) of the non-responders. Post-hoc analyses according to continent of study inclusion were conducted to investigate potential differences between pts treated in Europe (EU) and in Australia (AUS). Methods We evaluated baseline patient characteristics, safety and efficacy in terms of pathologic response in pts treated in EU (n=48) and AUS (n=38). Mutational (mut) load of baseline biopsies was assessed using whole exome sequencing. Multivariate analyses were performed using the logistic regression method. Median follow-up was 18.2 months for EU pts and 16.6 months for AUS pts. Results Baseline characteristics (AUS vs EU) differed in age (median 60 vs 53 year [yr], p=0.017). There were numerically more male pts in the Australian cohort (65.8 vs 50.0%, p=0.142) and more pts with unknown primary melanoma (36.8 vs 20.8%, p=0.100). A numerical higher pRR was observed in AUS pts vs EU pts (84.2% vs 64.7%, OR 2.50, p=0.092). The pRR was significantly higher for older pts (OR per yr 1.059, p=0.003), males (83.7% vs 63.9%, OR 2.90, p=0.041), and pts with higher mut load (OR per mutation 1.002, p=0.014). Mut load was higher in pts with pathologic response (p=0.0013) and in AUS pts (p=0.0003). There was a positive correlation between age and mut load (R=0.26, p=0.043). Multivariate analysis including continent, age, gender and mut load revealed that only mut load was significantly associated with response (OR 1.002, p=0.037).The frequency of grade 3-5 toxicities was the same in pts <60 compared to pts >60 yr (42.3% vs 32.4%, p=0.353). Conclusion The numerical higher pRR in AUS vs EU melanoma pts upon neoadjuvant IPI + NIVO appears mostly driven by a higher mut load found in the melanomas of AUS pts. AUS pts were older and there was a positive correlation between age and mut load, indicating that the higher mut load in AUS pts might be explained by higher age. It remains to be elucidated if continental variance in sun exposure also contributed to the difference in mut load. The fact that older pts achieve a higher response rate in absence of increased toxicity rates indicates that older pts should not be withheld neoadjuvant IPI + NIVO. Citation Format: Irene L. Reijers, Elisa A. Rozeman, Alexander M. Menzies, Judith M. Versluis, Bart A. van de Wiel, Karolina Sikorska, Hanna Eriksson, Kerwin Shannon, Carolien Bierman, Harm van Tinteren, Maria Gonzalez, Andrew J. Spillane, Robyn P. Saw, Alexander C. van Akkooi, Richard A. Scolyer, Johan Hansson, Georgina V. Long, Christian U. Blank. Different pathologic response rates between Australia and Europe in macroscopic stage III melanoma patients upon neoadjuvant ipilimumab plus nivolumab in the phase II OpACIN-neo trial [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3271.

Abstract 832: Construction and optimization of gene expression signatures for prediction of survival in two-arm clinical trials

Abstract Gene expression signatures for the prediction of differential survival of patients undergoing anti-cancer therapies are of great interest because they can be used to prospectively stratify patients entering new clinical trials, or to determine optimal treatment for patients in more routine clinical settings. Unlike prognostic signatures however, predictive signatures require training set data from clinical studies with at least two treatment arms, and the methodology for constructing and optimizing predictive signatures has been less prominently explored. Focusing on the “use case” of a two-arm clinical trial for metastatic colorectal cancer (CRC) patients treated with the anti-angiogenic molecule aflibercept, we present tools for the construction of signatures predictive of progression-free survival based on cross-validated multivariate Cox models. The work is organized around a specific approach called subtype correlation, which leverages a priori knowledge of molecular subtypes of tumors for a given indication. The tools and concepts presented here include the so-called differential log-hazard ratio, the survival scatter plot, the hazard ratio receiver operating characteristic, the area between curves and the patient selection matrix. In the CRC use case, the resulting signature stratifies the patient population into “sensitive” and “relatively-resistant” groups achieving a more than two-fold difference in the aflibercept-to-control hazard ratios across signature-defined patient groups. Through cross-validation and resampling the probability of generalization of the signature to similar CRC data sets is predicted to be high. The tools presented here should be of general use for building and using predictive multivariate signatures in oncology and in other therapeutic areas. Citation Format: Joachim Theilhaber, Marielle Chiron, Jennifer Dreymann, Donald Bergstrom, Jack Pollard. Construction and optimization of gene expression signatures for prediction of survival in two-arm clinical trials [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 832.

Abstract 1697: Signaling pathways involved in melanoma cell functions and dedifferentiation induced by inflammatory cytokines

Abstract INTRODUCTION: Melanoma-acquired resistance against T cell-based immunotherapy may be mediated by alterations of their responses to inflammatory cytokines. However, the cytokine signaling pathways in cancer cells remain poorly understood. METHODS: 8 melanoma cell lines were exposed to interferon gamma (IFNγ) and tumor necrosis factor alpha (TNFα) over a time course. Signaling molecules and their phosphorylated state were subsequently analyzed using mass cytometry. Briefly, cells were incubated with IFNγ (220 IU/mL), TNFα (1000 IU/mL), and both cytokines (220 IU/mL IFNγ + 200 IU/mL TNFα) for up to 48 hours. Staining was done with antibodies specific for (1) phosphorylated - STAT1(Y701), STAT3(Y707), AKT(S437), ERK1/2(T202/Y204), IKK(S176/180), NK-kbetap65 (S536), JNK (T183/T22), P38 (T180/Y182), Rb(S807/Y182), cMYC (S62), S6 (S235/S236), Histone3 (S28) and (2) non-phosphorylated - cPARP, β-catenin, CyclinB1, Ki67. Cells were read on a Helios Mass cytometer and de-barcoded using PREMESSA (R package). FlowJo was used to normalize time and eliminate calibration beads. Viable cell files were introduced to Cytobank where fold change (FC) with respect to time 0 of each protein was calculated with median intensity. The FC was exported to R for final analysis. RESULTS: All cell lines were sensitive to IFNγ/TNFα, with detectable (phosphorylated) signaling molecules. No synergistic effect was observed at any dose for IFNγ+TNFα. For the following data, numbers contained within parentheses represent the mean range of FC control vs treatment or mean standard error. pSTAT1 was significantly activated at all time points for IFNγ ([0.8, 1.0] vs [1.8, 3.6] p<0.02) and for IFNγ+TNFα ([1.9, 3.5] p<0.03). For TNFα it was activated at 48h ([3.1±0.9 vs 1.9±0.2] p=0.003). pJNK was significantly activated for IFNγ at time 0.25h and 0.5h ([0.9, 1.0] vs [1, 5] p<0.01); for TNFα at 0.25h and 1h ([0.9, 1.0] vs [1.5, 1.6] p<0.02); and for IFNγ+TNFα at 0.25h and 1h ([0.9, 1.0] vs 1.5 p<0.007). pNF-KBp65 was only significantly different for IFNγ+TNFα at 24h ([0.9±0.0 vs 11.6±0.2] p=0.025). The other time points followed a similar trend to activation. pSTAT3 was significantly activated for IFNγ at time 0.25h and 24h ([0.6, 0.9] vs [1.7, 2.1] p<0.04); for IFNγ+TNFα at time 0.25h and 24h ([0.6, 0.9] vs [1.9, 2.0] p<0.02); and for TNFα at 0.25h and 0.5h ([0.5, 0.6] vs [1.3, 1.5] p<0.02). pERK was significantly activated for IFNγ at 0.25h ([0.9±0.2 vs 1.6±0.3] p=0.04). β-catenin presented no difference in FC across cell lines. CONCLUSIONS: Our data show evidence for functional IFNγ/TNFα signaling in most melanoma cells. However, unexpected signaling was observed for some cell lines, possibly reflecting interpatient heterogeneity. Additional studies are necessary to understand the roles of cytokine signaling in cancer immunotherapy. Citation Format: Michael Cerniglia, Katie M. Campbell, Julianne Schwerdtfeger, Christophe Martignier, Connie B. Gilfillan, Karin Schaeuble, Gardenia Cheung-Lau, Yaroslav Teper, David Berger Maneiro, Miriam Guemes-Aragon, Alejandro J. Garcia, Daniel E. Speiser, Antoni Ribas, Begoña Comin-Anduix. Signaling pathways involved in melanoma cell functions and dedifferentiation induced by inflammatory cytokines [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1697.

Abstract 3854: Prevalence and prognostic role of FoxP3regulatory T lymphocytes in cancer: A tissue microarray study on >20,000 cancers

Abstract Regulatory FoxP3+ lymphocytes function as suppressors of T-cell activity. The clinical impact of high FoxP3+ cell density in cancers is not fully understood, as some studies have linked high FoxP3+ cell density to good prognosis and others to poor prognosis in tumor cohorts with associated clinical data. While some data suggest that these variable data are due to biological differences between tumor entities, it is also possible that methodological differences have caused these discrepancies. This study was undertaken to analyze the density of FoxP3+ cells in various different cancer types by employing standardized methods. Tissue microarrays and large sections made from >20,000 prostate, breast, colorectal, ovarian, pancreatic, bladder and stomach cancers were analyzed together with various normal and inflamed tissues by conventional brightfield FoxP3 immunohistochemistry. Samples were also analyzed by fluorescent multiplex immunohistochemistry to assess the fraction of Ki67+ FoxP3+ cells. Our results indeed suggested a variable role of FoxP3+ cells in different tumor types. High FoxP3+ density was linked to high Gleason grade (p=0.0003) and early biochemical recurrence (p<0.0001) in 16923 prostate cancers, but to low tumor stage (p=0.027) and prolonged survival (p=0.0029) in 1341 breast cancers, and to low tumor stage (p<0.0001) in 744 colorectal cancers. No significant associations were found to tumor phenotype in 549 ovarian, 574 pancreatic, 549 bladder and 346 stomach cancers. Multiplex fluorescence IHC analysis of FoxP3 and Ki67 revealed comparable fractions of proliferating FoxP3+ cells in healthy tissues (average 12.3%, range 5.8-18.5%) and inflammatory conditions (average 7.6%, range 2.6-17.2%). Interestingly, the rate of Ki67+FoxP3+ cells was markedly higher in 36 bladder cancers (average 14.2%, range 0-49.3%) suggesting active expansion of FOXP3+ cells in cancer. Our data demonstrate an inverse prognostic impact of the FoxP3+ cell density in prostate and breast cancers. The increased proliferation rate of immune-regulatory FoxP3+ cells in some bladder cancer is interesting in the light of the variable response of these tumors to immune checkpoint inhibitors. Citation Format: Tim Mandelkow, Elena Bady, Niclas Blessin, Claudia Hube-Magg, Ronald Simon, Guido Sauter, Christoph Fraune, Maximilian Lennartz, Katharin Möller, Sören Weidemann, Andreas Luebke, Doris Höflmayer. Prevalence and prognostic role of FoxP3regulatory T lymphocytes in cancer: A tissue microarray study on >20,000 cancers [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3854.

Abstract 5189: Neoadjuvant and adjuvant anti-PD-1 based combination immunotherapy with a novel AXL inhibitor, SKI-G-801 in syngeneic tumor model: A combined analysis of immune profiling

Abstract Introduction: Neoadjuvant therapy suggests the systemic treatment of cancer prior to surgical therapy. The main purpose of neoadjuvant therapy is to improve surgical outcomes in patients for whom a primary surgical approach is difficult to practice. Recently, Anti-PD-1 has been applied to neoadjuvant and adjuvant therapy, but the anti-cancer responses are not sufficient to prevent tumor recurrence. According to our previous study, a novel AXL inhibitor, SKI-G-801 strongly inhibits metastasis of syngeneic tumor with immune responses. The primary goal of this study is to overcome limitation of anti-PD-1 based neoadjuvant and adjuvant therapy with SKI-G-801. Methods: 4T1 cancer bearing mouse models were established for neoadjuvant or adjuvant treatment. Mice were treated with single anti-PD-1 (BIW), SKI-G-801 (QD) or a combination for 6 days, followed by the surgical dissection of tumor tissue in a subcutaneous xenograft model. In five mice per group, their survival rates were monitored for 50 days. We used flow cytometry and immunohistochemistry for comprehensive immune profiling and tumor-infiltrating lymphocytes (TILs) from three mice per group surgically removed tumor tissues. Three tumors per group were stained, and CD3-positive cells per unit area (mm2) were analyzed using Vectra Polaris. Results: According to results, the median survival of control group (only surgical operation without treatment) was 16 days (N=5), whereas, the longest survival group was neoadjuvant-combination (median survival = 37 days). Two out of five mice were survived until end of experiment (D-50). Most of the groups had similar outcomes excluding neoadjuvant-combination group. The median survival of adjuvant-anti-PD-1 group was 30 days, and adjuvant-SKI-G-801 and neoadjuvant-anti-PD-1 group were 25 days. The Neoadjuvant-SKI-G-801 was recorded 23.5 days. Statistically, survival rate of neoadjuvant-combination therapy was significantly higher than of neoadjuvant-anti-PD-1 and neoadjuvant-SKI-G-801 (log-rank test, p <0.01).The reason for neoadjuvant therapy is to improve immune responses before surgical operation. We investigated the effects of SKI-G-801 or anti-PD-1 alone and combination therapy on neoadjuvant treatment in the tumor microenvironment. IHC and flow cytometry were performed to elucidate tumor infiltrating immune cell populations with the tumor obtained by surgery. In IHC result, CD3+ cells population was significantly increased in SKI-G-801 alone and combination of neoadjuvant therapy compared to control group (p<0.05, p < 0.001, respectively), however, anti-PD-1 group was not significant. IFN-γ+ Tc cells, CD3+CD44+ memory Tc, and PD-1+Tc cells were significantly infiltrated into the tumor in the combination of neoadjuvant therapy (p < 0.01). Conclusion: This study confirmed that a potential combination with aPD-1 and SKI-G-801 applies to neoadjuvant therapy, as evidenced by increased T cell infiltration and function. Our findings provide a rationale for further clinical investigations. Keywords: Axl inhibitor, Neoadjuvant therapy, Adjuvant therapy, Immunotherapy, Anti-PD-1 Citation Format: Ha Ni Jo, Wongeun Lee, Chun-Bong Synn, Hee Kyu Lee, Jae-Hwan Kim, Yeongseon Byeon, Sung Eun Kim, Ji Min Lee, Do Hee Kim, Jung-Ho Kim, Beung-Chul Ahn, Min Hee Hong, Hye Ryun Kim, Kyoung-Ho Pyo, Byoung Chul Cho. Neoadjuvant and adjuvant anti-PD-1 based combination immunotherapy with a novel AXL inhibitor, SKI-G-801 in syngeneic tumor model: A combined analysis of immune profiling [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5189.

Abstract 1394: Gene expression profile in BCBL-1 cell line following activation of KSHV

Abstract Patients with HIV/AIDS are at a 3,400 times higher risk of developing Kaposi's sarcoma when compared to healthy populations. Our hypothesis is that the upregulation of viral associated proteins is causing an increase in DNA damage, allowing for the higher incidence rate of Kaposi's sarcoma in infected populations. The direct mechanisms leading to this increased risk is unknown, but the viral infection with Kaposi's sarcoma-associated herpesvirus (KSHV) is an apparent cause of this risk. The cell is under attack of endogenous threats including oxidative stress, causing DNA damage and leading to mutations, but the viral proteins may interfere with the repair mechanisms. In previous experiments we looked at DNA repair dynamics following X-ray exposure in the two different stages and determined there is an inefficiency in the DNA repair process in the cells during active production of viral particles. This project aimed to compare the transcriptomes of the BCBL-1 cell line during latent and acute phases of KSHV production. Our results are showing that genes involved in genome stability are differently expressed during the lytic stage. Comparison of these two transcriptomes allows for identification of which proteins are being upregulated during active viral production. This may be of importance for the HIV/AIDS cancer patients because it shows the mechanisms behind genome stability in KSHV-infected cells. 1) Research reported in this publication was supported by the National Institute Of General Medical Sciences of the National Institutes of Health under Award Numbers UL1GM118991, TL4GM118992, or RL5GM118990. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. 2) NASA EPSCoR, NNH15ZHA003C Citation Format: Samantha L. Haines, Sabrina R. Bishop, Viktorija Podlutskaya, Andrej Podlutsky. Gene expression profile in BCBL-1 cell line following activation of KSHV [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1394.

Abstract 2631: Deep learning-based analysis of tissue segmentation in histopathology images of colorectal cancer

Abstract Objective: Stromal elements in the tumor microenvironment (TME) impact prognosis and response to therapy, creating a barrier to absorption and penetration of therapeutic drugs and modulating immune cell infiltration. Few models exist to analyze the spatial distribution and mapping of stroma and cancer lesions composing the complex TME in association with mRNA levels via high throughput deep learning algorithm. Methods: The histopathology images (H&E stain) and mRNA-seq of 625 CRC patients (pts) were obtained from the Cancer Genome Atlas (TCGA). A deep learning-based model of tissue classification enabling segmentation of tumor (malignant glands composed of adenocarcinoma cells) and stroma in histopathology images was established using published data sets [1-2]. We defined the following parameters based on the pixel count of stroma and tumor: the stroma-to-tumor ratio (STR) defined as the pixel count ratio of stroma to stroma plus tumor; STRx, STR within the pixel distance x from the tumor; tumor-tissue-ratio (TTR), a pixel count ratio of tumor to stroma plus tumor. The expressivity of genes enriched in cellular and structural components of TME and its correlation with STR and TTR were analyzed. Results: The profiling results demonstrate that histopathology images of consensus molecular subtype (CMS) 4 of colorectal cancer have statistically significantly larger STR, STR10, STR20, STR30, STR40, and STR50 than those in other CMS groups (p<0.0001); TTR, smaller in CMS 4 than in other CMS groups (p<0.0001). Out of two patient groups clustered via k-means clustering (high and low STR), the group having higher STR has significantly decreased expression of genes related to antigen presentation, natural killer cells, and plasmacytoid dendritic cells, as well as significantly elevated expression of genes associated with epithelial-mesenchymal transition, desmoplastic reaction, fibroblastic cytokines, and angiogenesis (p<0.05) (refer to presentation, regarding the list of genes). Conclusion: Our analysis of histopathology images via the deep learning-based algorithm enables automated segmentation of stroma and malignant lesions. Parameters representing spatial and architectural distribution of stroma and cancer lesions are shown to have close correlation with varying expression of genes associated with immune system activation and suppression in the tumor microenvironment. [1] Kather et al. Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study. PLoS Medicine. 2019 [2] Kumar et al. A Multi-organ nucleus segmentation challenge. IEEE Trans Med Imaging. 2019 Citation Format: Yeongwon Kim, Kyungdoc Kim, Jeonghyuk Park, Hyunho Park, Kyu-Hwan Jung, Sunyoung S. Lee. Deep learning-based analysis of tissue segmentation in histopathology images of colorectal cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2631.

Abstract 6055: Discovery of epitope-selective anti-CD25 targeting therapeutic antibodies for effective Treg cell depletion in cancer using a novel AI/ML based platform

Abstract Therapeutic antibodies have contributed significantly to better clinical outcomes across multiple tumor types. However, there is still unmet clinical need in cancers with an immunosuppressive and stroma rich tumor microenvironment (TME). One key limitation in early antibody discovery is immunodominance - an evolutionary bias toward specific epitopes that steer antibody selection away from potentially high-efficacy epitopes. Our integrated computational and laboratory approach to antibody discovery enabled us to build focused epitope embodiments of structural epitopes. These epitope embodiments, called Meso-scale Engineered Molecules (MEMs), are used in antibody selection to steer hits toward intended epitopes. CD25 (IL2Rα) is highly expressed on regulatory T-cells (Tregs) that drive immunosuppression in the TME. Depletion of Tregs can restore anti-tumor T effector (Teff) function in the TME, by increasing the Teff/Treg ratio, especially when combined with checkpoint-inhibitors. Using our epitope-selective antibody discovery platform, we built MEMs that embody eight CD25 structural epitopes outside the CD25:IL-2 interface. Through in vitro selection, several high-affinity (median KD = 25 nM) anti-CD25 antibodies from each of the eight intended CD25 epitopes were identified and confirmed to be on epitope by cross-blocking assays and in-epitope alanine mutations. Several conventional- and epitope-selective anti-CD25 clones were converted to human IgG1 and tested in in vitro assays. As a result, these antibodies bound specifically to CD25+ cells, preserve IL-2 signaling via pSTAT5 assays and elicited potent ADCC activity using human effector cells, superior to benchmark antibodies. These studies show that our integrated computational and laboratory platform improves the antibody discovery process, producing clones with biological function and improved therapeutic efficacy. Citation Format: Phung Gip, Bing Li, Angeles Estelles, Hanako Daino-Laizure, Oleg Gurinovich, Piotr Zalicki, Kevin Hauser, Alex Taguchi, Christian Barrett, Andrew Morin, Jordan Willis, Mohan Srinivasan, Isaac Bright, Matt Greving. Discovery of epitope-selective anti-CD25 targeting therapeutic antibodies for effective Treg cell depletion in cancer using a novel AI/ML based platform [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6055.

Abstract 1535: Identification of novel epithelial-mesenchymal transition inducers and their clinical significance in colorectal cancer

Abstract Purpose. Cancer tissues contain a substantial amount of stromal tissues, and transcriptome derived from bulk samples originates from a mixture of the cancer epithelium and stroma. Gene signature of epithelial-mesenchymal transition (EMT) cells is similar to that of stromal cells; therefore, gene signature that we considered to be related EMT might be only contamination of stroma. Here, we identify novel EMT inducers by analysis of independent gene expression profiles in the colorectal cancer epithelium and stroma. Materials and Methods. Gene expression data of pure epithelial tissue was obtained using laser microdissection technique from 13 colorectal cancer tissues. Next, the ratio of each gene expression in the stroma to in the epithelium and correlation of each gene expression with a target gene derived from The Cancer Genome Atlas (TCGA) is plotted in two dimensions, and correlation coefficients are calculated. We regard genes with high correlation coefficients but highly expressed in the epithelium as candidates of EMT inducers. We manipulate the expression of some candidate genes in colorectal cancer cell lines to investigate the alterations of EMT markers and migration ability in vitro. Results. We extract 91 new putative EMT-related genes in silico. They are highly expressed in similar cell lines to known EMT markers in a public database. Furthermore, some candidates have reported having a relation to EMT, and that indicates the validity of this procedure. The knockdown of three candidate genes upregulates epithelial gene expression and downregulates mesenchymal gene expression. The knockdown of one of the three genes reduces cell migration. In contrast, overexpression of the gene downregulates epithelial gene expression and leads to cell migration. Conclusion. We identify novel EMT inducers successfully using purified gene expression data. They could be new prognostic markers and therapeutic targets of colorectal cancer. Citation Format: Satoshi Ishikawa, Naohiro Nishida, Shiki Fujino, Takayuki Ogino, Hidekazu Takahashi, Norikatsu Miyoshi, Mamoru Uemura, Chu Matsuda, Tsunekazu Mizusima, Hirofumi Yamamoto, Koshi Mimori, Masaki Mori, Yuichiro Doki, Hidedoshi Eguchi. Identification of novel epithelial-mesenchymal transition inducers and their clinical significance in colorectal cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1535.

Abstract 1047: Combinatorial benefit without synergy in recent clinical trials of immune checkpoint inhibitors

Abstract IMPORTANCE Immune Checkpoint Inhibitors (ICIs), frequently combined with other therapies, have become the standard of care for the treatment of several cancers, and hundreds of clinical trials evaluate the efficacy of novel combinations with ICI. However, it is unclear whether combination therapies truly enhance ICI activity or simply provide multiple “shots on target” for a response to occur. OBJECTIVE To assess whether ICIs in combination with other therapies result in ‘additive' or ‘synergistic' anti-tumor activity. The null hypothesis is independent drug action, where each patient's response to a combination of therapies is equal to that patient's best single-agent response. DATA SOURCES We analyzed twelve randomized-controlled clinical trials testing combinations of ICIs with other drug classes in patients with advanced cancers, including melanoma and lung, breast, gastric, kidney, and head and neck cancers. For each analysis, the clinical activity of each constituent therapy of the combination was available in the respective study population. METHODS We digitized Progression Free Survival curves for combination therapies and their constituents, and computed PFS expected for the null hypothesis of independent action by randomly sampling single-agent PFS durations (with correlation ρ=0.3±0.2 to account for cross-resistance), and assigning each simulated patient the longer of the two sampled PFS values. For each combination therapy, simulated and clinically observed PFS distributions were compared. RESULTS None of the analyzed clinical trials of ICIs in combination therapies exhibited PFS surpassing the null hypothesis of independent drug action. This suggest that there is no additive or synergistic effect in patients receiving combination therapies with ICIs. Rather, these combinations provided benefits that are largely predictable from the response probabilities of each constituent therapy. CONCLUSIONS Combining ICIs with other cancer therapies provides predictable and clinically meaningful benefit by bet-hedging: providing patients with more chances of a response to any one therapy. This implies that similar benefits may in principle result from sequential treatments or biomarker-stratified choice of treatment. This insight may be most valuable in cases where toxicity prevents full-dose combinations. Citation Format: Adam C. Palmer, Peter K. Sorger, Benjamin Izar. Combinatorial benefit without synergy in recent clinical trials of immune checkpoint inhibitors [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1047.

Abstract 6309: Development of protein conformational array ELISA for p53 and SHP2 allosteric regulator screening and analysis

Abstract P53 and SHP2 plays important role during cancer development. For p53, more than 50% of cancers were shown to have gene mutation in this tumor suppressor, however how these gene mutations impact the p53 protein conformation and function have not be studies in much detail. A Protein Conformational Array ELISA was developed with 17 polyclonal antibodies covering the whole p53 molecule, by measuring the linear epitope distribution on the surface of the p53 protein, the conformational (Higher Order Structure or HOS) can be described quantitively. Initial studies have shown that in lung cancer cells, the level of p53 is significantly increased and more interestingly, the p53 HOS has changed at its N-terminal as compared with the wild type. For SHP2, a protein phosphatase, a 27-antibody panel was developed to characterize the HOS changes. Because of the generic nature of the phosphate binding site, it has been very challenging to develop phosphatase inhibitors, however studies have shown that allosteric regulators can be identified with excellent anti-tumor activities. Currently, some of the allosteric regulators identified are being tested in the clinic. The SHP2 Conformational Array ELISA will provide a robust platform for the screening and testing of potential allosteric regulators with its molecular resolution and high-throughput capability. This presentation will describe the development of the two conformational arrays and their applications in the study of p53 and SHP2 from different cancer cell lines and the characterization of the allosteric regulators identified. Citation Format: Xing Wang. Development of protein conformational array ELISA for p53 and SHP2 allosteric regulator screening and analysis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6309.

Abstract 1942: AGX73 is an effective treatment for hepatocellular carcinoma (HCC) by selective inhibition of immunokinases that play crucial roles in the crosstalk between cancer and immune cells in the tumor microenvironment

Abstract HCC is the second leading cause of cancer-related death and the fifth-most-common cancer worldwide. Unlike almost all other tumor types, the vast majority (90%) of HCCs develop due to underlying chronic inflammation, the induction of fibrosis and/or subsequent cirrhosis. In established disease, persistent upregulation of inflammatory signals in the tumor microenvironment leads to suppression of antitumor immune response by various mechanisms such as reprogramming of tumor-associated macrophages (TAMs) and loss of critical immune cell subsets (e.g. tumor-infiltrating lymphocytes). We identified 6 key immunokinases (AXL, DDR1, DDR2, FMS, KIT and MER) that are frequently amplified in HCC and are significant predictive markers of poor outcome (TCGA). We design small molecule compounds that potently and selectively inhibit these kinases. AGX73, optimized for drug-like properties, is an orally bioavailable compound currently undergoing IND evaluation as potential clinical candidate for first-in-human trial. In in vitro assay, carried out at high ATP concentration (100 μM), AGX73 inhibited the activity of recombinant human AXL, DDR1, DDR2, FMS, KIT and MER with IC50 values of 30, 12, 0.63, 96, 54, and 40 nM, respectively, but was highly selective against a panel of 374 kinases. In an in vitro primary macrophage assays, AGX73 promoted M1 macrophage polarization and increased the secretion of IL-6 and TNFα. In contrast, AGX73 suppressed the activation of M2 macrophages and decreased the secretion of IL-10. Importantly, AGX73 prevented the reprogramming of human LPS/IFN-γ polarized M1 macrophages back to M2 macrophages. In contrast, AGX73 had no effect on the reprogramming of IL-4/IL-13 polarized M2 macrophages back to M1. In vivo, in the MC38 colon carcinoma model, AGX73 treatment reduced the number of TAMs, from 46 to 22% of the total CD45+ cells in the tumors. Moreover, the M1/M2 ratio was favorably increased, from 1.3 to 1.8. AGX73 demonstrated robust antitumor activity in multiple (11) patient-derived xenograft (PDX) models of HCC, with tumor growth inhibition (TGI) ranging 62-105% (Median = 84%) and was effective at multiple dose levels, from 30-150 mg/kg. In combination with anti-PD1 (10 mg/kg) in the MC38 model, AGX73 produced synergistic activity at multiple dose levels. In nonclinical ADME studies, AGX73 demonstrated favorable drug-like properties: including good bioavailability (F>50% in mouse, rat and monkey) and systemic exposure following oral administration in multiple species, acceptable metabolic stability, lack of significant drug-drug interaction liabilities (transporters, CYP inhibition, CYP induction) and excellent cardiac safety profile. IND toxicologic studies are currently in progress. Citation Format: Francis Y.F. Lee, Wenlian Wu, Zhiqiang Yang, John Tan. AGX73 is an effective treatment for hepatocellular carcinoma (HCC) by selective inhibition of immunokinases that play crucial roles in the crosstalk between cancer and immune cells in the tumor microenvironment [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1942.

Abstract 5795: Long-term diabetes risk among ovarian cancer survivors in a population-based cohort study

Abstract Objective. In the United States, ovarian cancer is the eleventh most common female cancer and the fifth most common cause of cancer death among women. Although incidence and death rates have been declining over the last 10 years, the 5-year survival rate is 47.6%. Diabetes was the seventh leading cause of death in the United States and is associated with an increased mortality rate among cancer patients. While previous studies have examined the association between diabetes and the risk of ovarian cancer, there are very few studies investigating the risk of diabetes after an ovarian cancer diagnosis. We previously reported increased risks of type I and II diabetes among endometrial cancer survivors, but there is no previous study, to our knowledge, that examined the risk of diabetes among ovarian cancer survivors. The aim of this study was to investigate long-term diabetes risk among ovarian cancer survivors compared with a general population of women. Methods. Ovarian cancer patients diagnosed from 1996 to 2012 in the Utah Cancer Registry were matched to up to 5 cancer-free women on birth state and birth year using the Utah Population Database. Diabetes diagnoses were identified with ICD-9 codes from statewide ambulatory surgery records, statewide inpatient databases and electronic medical records. We estimated hazard ratios with Cox Proportional Hazards models to: (1) investigate risk for diabetes after cancer diagnosis among ovarian cancer survivors compared with the general population, and (2) examine risk factors for type II diabetes among ovarian cancer survivors. Results. A total 1,520 ovarian cancer patients and 5,709 women from the general population were included in the final cohort. A slightly higher proportion of ovarian cancer patients were obese at baseline (16.3%) than the general population cohort (14.9%). Ovarian cancer survivors had a higher risk of type II diabetes in the first year after cancer diagnosis (HR=3.09, 95%CI=2.09, 4.56) and overall (HR=1.52, 95%CI=1.23, 1.88) compared with the general population. However, associations between ovarian cancer and type II diabetes were not observed in the later follow up periods. High BMI was a significant risk factor for type II diabetes among ovarian cancer survivors, with a dose-response relation (overweight; HR=1.81, 95%CI=1.21, 2.71, obese; HR=2.48, 95%CI=1.61, 3.83, P-trend <0.001). Clinical factors such as cancer treatment, stage, and histology did not contribute to an increased risk of type II diabetes among ovarian cancer patients. Other factors such as race and baseline comorbidities were also not associated with type II diabetes risk among ovarian cancer patients. Conclusion. Ovarian cancer survivors had an increased risk of type II diabetes compared with women in the general population. Risk was elevated only within the first year of cancer diagnosis, thus surveillance bias cannot be ruled out. Our findings suggest that the increased risk of type II diabetes does not persist beyond the first year after cancer diagnosis. Citation Format: Seungmin Kim, Kerry Rowe, John Snyder, Alison Fraser, Ken Smith, Vikrant G. Deshmukh, Michael Newman, Kimberley Herget, Chun-Pin Chang, Dominik Ose, Mary C. Playdon, David Gaffney, Mia Hashibe. Long-term diabetes risk among ovarian cancer survivors in a population-based cohort study [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5795.

Abstract CT267: A phase 1b/2, open-label study of bintrafusp alfa with chemotherapy in patients with stage IV non-small cell lung cancer

Abstract Background: The combinations of pembrolizumab plus chemotherapy (CT) and atezolizumab plus bevacizumab and CT have been approved for first-line therapy in patients with advanced non-small cell lung cancer (NSCLC) regardless of PD-L1 expression; however, there is still a significant unmet need for this patient population who may not respond to immune checkpoint treatment. Bintrafusp alfa (M7824) is a first-in-class bifunctional fusion protein composed of the extracellular domain of the TGF-βRII receptor designed to function as a TGF-β “trap” fused to a human IgG1 mAb blocking PD-L1. Promising antitumor activity and a manageable safety profile were observed with bintrafusp alfa in multiple cohorts of a phase 1 study (NCT02517398) that enrolled patients with advanced, pretreated NSCLC who experienced disease progression either after platinum-based CT or anti-PD-(L)1 monotherapy. This study is evaluating the safety and efficacy of bintrafusp alfa in combination with CT in patients with stage IV NSCLC. Trial Design: This is a phase 1b/2, open-label, 4-cohort study (NCT03840915) evaluating bintrafusp alfa in combination with CT in patients with stage IV NSCLC. All patients will receive bintrafusp alfa 2400 mg Q3W intravenously and either cisplatin or carboplatin + pemetrexed (cohort A), carboplatin + nab-paclitaxel or paclitaxel (cohort B), cisplatin or carboplatin + gemcitabine (cohort C), or docetaxel (cohort D) Q3W for 4 cycles, followed by bintrafusp alfa maintenance (monotherapy or combination with pemetrexed [cohort A]) for up to 31 cycles or until disease progression, unacceptable toxicity, or death. Patients must be adults with histologically confirmed diagnosis of stage IV nonsquamous or squamous NSCLC (nonsquamous for cohort A), adequate organ function, ECOG PS ≤1, life expectancy ≥3 months, and measurable disease based on RECIST 1.1. Patients in cohorts A-C must not have received prior systemic therapy, and patients in cohort D must have disease progression on previous anti-PD-(L)1 therapy. Patients with tumors with actionable mutations (for which targeted therapy is locally approved), mixed SCLC and NSCLC, active CNS metastases, active autoimmune disease, known severe hypersensitivity, or interstitial lung disease are not eligible. The primary endpoint of this study is to assess the safety of bintrafusp alfa in combination with CT. Planned enrollment is 64 patients (32 patients among the 4 cohorts during the safety phase, and an additional 32 patients in cohort A for expansion phase). Citation Format: Fabrice Barlesi, Sandrine Hiret, Christian Rolfo, Italia Grenga, Andre Köenig, Yulia Vugmeyster, Leora Horn. A phase 1b/2, open-label study of bintrafusp alfa with chemotherapy in patients with stage IV non-small cell lung cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT267.

Abstract 2518: Function of microRNA-10a in cholangiocarcinoma

Abstract Cholangiocarcinoma, a primary liver cancer of the bile ducts, effects 6,000 patients yearly and incidence has been growing consistently over the past 15 years. Current therapies for cholangiocarcinoma are not curative and patient survival is less than 1 year after diagnosis. Strides toward new treatment rest on better understanding tumor biology. The expression of microRNAs in cholangiocarcinoma cells was analyzed by RNAseq after knockdown of fibroblast growth factor receptor 4 (FGFR4) to identify potential microRNAs regulated by FGFR4. Caspase activity was used to evaluate miR-10a roles in anti-apoptotic signaling. Cell scratch assays were used to test cell migration and cell counting was used to determine proliferation. Quantitative real time PCR was used to measure miR-10a levels in cholangiocarcinoma cell lines Mz-ChA-1, HUCCT-1 and KMCH after FGFR4 knockdown with siRNA. Potential target proteins of miR-10a were evaluated by western blot. Preliminary data showed dozens of microRNAs (miRs) in cholangiocarcinoma cells were modulated by FGFR4 knockdown. Treatment with siRNA to FGFR4 caused increased sensitivity to TRAIL-induced apoptosis, and also reduced miR-10a expression in Mz-ChA-1, HuCCT-1, and KMCH cell lines. Importantly, apoptosis sensitization was reversed by restoring miR-10a expression in cells. miR-10a helped regulate both cell proliferation and cell migration in cholangiocarcinoma cells. When miR-10a was increased in cholangiocarcinoma cells, migration and proliferation increased. When miR-10a was decreased these effects also decreased. Protein targets of miR-10a included the MAP Kinase-related proteins MLK-1 & TAK-1. Other protein targets such as PHLDA1 & AKTIP were also identified. Notably, microRNA binding prediction programs did not include MLK-1 as a target, though our sequence analysis shows multiple stable binding sites. MLK-1 showed the most dramatic response to miR-10a. We investigated the role miR-10a plays for cholangiocarcinoma and the mechanism through which this pathway is regulated. Data supports miR-10a as a key player in cell survival, proliferation and migration. We have identified a set of proteins that are regulated by miR-10a in cholangiocarcinoma cells. Looking forward, identification of the function of these proteins will help elucidate how miR-10a functions in cholangiocarcinoma progression. Citation Format: Matthieu Spriet, Cody Wehrkamp, Brandon Henslee, Ashley Mohr, Steve Kachman, Justin Mott. Function of microRNA-10a in cholangiocarcinoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2518.

Abstract 4306: Retrospective analysis of docetaxel in combination with ramucirumab for previously treated non-small cell lung cancer patients

Abstract BACKGROUND: Current clinical trials demonstrated that combination regimens comprising chemotherapy and immunotherapy lead to better patient outcomes compared to chemotherapy alone as the first line of treatment for non-small cell lung cancer (NSCLC). In addition, the combination therapy of docetaxel (Doc) and ramucirumab (Ram) was considered one of the standard treatments for advanced or relapsed NSCLC patients. However, little is known about the therapeutic responders of this combination therapy among previously treated NSCLC patients. In the present study, we aimed to identify predictive factors for therapeutic response, including programmed death-ligand 1 (PD-L1) expression in tumors, for Doc treatment in combination with Ram. METHODS: We retrospectively analyzed a total of 135 advanced or relapsed NSCLC patients who were refractory to platinum-based chemotherapy at eleven institutions in Japan between July 2016 and November 2018. RESULTS: Our observations showed that PD-L1 expression in tumors is not associated with the efficacy of combined therapy of Doc and Ram in previously treated NSCLC patients. Analysis of the patient clinical profiles indicated that prior treatment with immune checkpoint inhibitors (ICIs) is a reliable predictor for the good progression-free survival (PFS) to this combination therapy (P=0.041). CONCLUSIONS: Our retrospective study indicated that combination regimens comprising chemotherapy and ICIs followed by Doc and Ram could be an optimal therapeutic option for NSCLC patients regardless of the PD-L1 status of tumors. Further investigations are required to strengthen clinical evidence demonstrating the effectiveness of the combination therapy of Doc plus Ram in previously treated NSCLC patients. Citation Format: Rui Kitadai, Akihiro Yoshimura, Tadaaki Yamada, Yusuke Okuma, Takayuki Takeda, Takanori Kanematsu, Junji Uchino, Yasuhiko Nishioka, Koichi Takayama. Retrospective analysis of docetaxel in combination with ramucirumab for previously treated non-small cell lung cancer patients [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4306.

Abstract CT074: Pre-existing ESR1 mutations in early-stage primary breast cancer predict failure of endocrine therapy and poor survival

Abstract Background: More than three-quarters of primary breast cancers are positive for estrogen receptor alpha (ER; encoded by the gene ESR1), the most important factor for directing anti-estrogenic endocrine therapy. Although mutation in ESR1 is known as an acquired mechanism of resistance to endocrine therapy (ET), found in 12-55% of metastatic breast cancers treated previously with ET, the impact of ESR1 mutation on therapy response in primary breast cancer is unclear. Patients and methods: In this study we analyzed 3217 real-world and population-based early-stage primary breast cancers (within the SCAN-B study, ClinicalTrials.gov NCT02306096). Tissues were sampled from initial diagnosis prior to any treatment and analyzed for the presence of ESR1 mutations using RNA sequencing. Mutations were verified by SAGAsafe droplet digital PCR. Results: We identified ESR1 resistance mutations in 30 cases (0.9%), of which 29 were ER-positive (1.1%). In ER-positive disease, presence of ESR1 mutation was significantly associated to poor relapse-free survival (RFS) and overall survival (OS) (p=0.011 and p=0.019, respectively), and moreover predicted poor RFS and OS within the patient group that received ET (p=0.007 and p=0.010, respectively). Conclusions: These results indicate that ESR1 mutations at diagnosis of untreated primary breast cancer are rare, however we confirm for the first time that such early mutations predict eventual resistance to standard hormone therapy in the adjuvant setting. If replicated, tumor ESR1 screening may be considered in ER-positive primary breast cancer and, in mutated cases, ER-degraders such as fulvestrant or other therapeutic options may be considered as more appropriate. Citation Format: Malin Dahlgren, Anthony M. George, Christian Brueffer, Sergii Gladchuk, Yilun Chen, Johan Vallon-Christersson, Cecilia Hegardt, Jari Häkkinen, Lisa Rydén, Martin Malmberg, Christer Larsson, Sofia K. Gruvberger-Saal, Anna Ehinger, Niklas Loman, Åke Borg, Lao H. Saal. Pre-existing ESR1 mutations in early-stage primary breast cancer predict failure of endocrine therapy and poor survival [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT074.