Search

PURPOSE Advances in biological measurement technologies are enabling large-scale studies of patient cohorts across multiple omics platforms. Holistic analysis of these data can generate actionable insights for translational research and necessitate new approaches for data integration and mining. METHODS We present a novel approach for integrating data across platforms on the basis of the shared nearest neighbors algorithm and use it to create a network of multiplatform data from the immunogenomic profiling of non–small-cell lung cancer project. RESULTS Benchmarking demonstrates that the shared nearest neighbors-based network approach outperforms a traditional gene-gene network in capturing established interactions while providing new ones on the basis of the interplay between measurements from different platforms. When used to examine patient characteristics of interest, our approach provided signatures associated with and new leads related to recurrence and TP53 oncogenotype. CONCLUSION The network developed offers an unprecedented, holistic view into immunogenomic profiling of non–small-cell lung cancer, which can be explored through the accompanying interactive browser that we built.

PTEN loss in melanoma promotes resistance to T cell-mediated anti-tumor activity in the context of concurrent a selective inhibitor for mutant BRAF.

Supplementary Table S1. Patient characteristics of the single-agent anti-PD-1 cohort. Supplementary Table S2. List of primers used for gene expression analysis by real-time PCR. Supplementary Table S3. List of autophagy-related genes for overexpressing experiment.

Although immunotherapy has achieved impressive durable clinical responses, many cancers respond only temporarily or not at all to immunotherapy. To find novel, targetable mechanisms of resistance to immunotherapy, patient-derived melanoma cell lines were transduced with 576 open reading frames, or exposed to arrayed libraries of 850 bioactive compounds, prior to co-culture with autologous tumor-infiltrating lymphocytes (TILs). The synergy between the targets and TILs to induce apoptosis, and the mechanisms of inhibiting resistance to TILs were interrogated. Gene expression analyses were performed on tumor samples from patients undergoing immunotherapy for metastatic melanoma. Finally, the effect of inhibiting the top targets on the efficacy of immunotherapy was investigated in multiple preclinical models. Aurora kinase was identified as a mediator of melanoma cell resistance to T-cell-mediated cytotoxicity in both complementary screens. Aurora kinase inhibitors were validated to synergize with T-cell-mediated cytotoxicity in vitro. The Aurora kinase inhibition-mediated sensitivity to T-cell cytotoxicity was shown to be partially driven by p21-mediated induction of cellular senescence. The expression levels of Aurora kinase and related proteins were inversely correlated with immune infiltration, response to immunotherapy and survival in melanoma patients. Aurora kinase inhibition showed variable responses in combination with immunotherapy in vivo, suggesting its activity is modified by other factors in the tumor microenvironment. These data suggest that Aurora kinase inhibition enhances T-cell cytotoxicity in vitro and can potentiate antitumor immunity in vivo in some but not all settings. Further studies are required to determine the mechanism of primary resistance to this therapeutic intervention. Electronic supplementary material The online version of this article (10.1007/s00262-020-02748-9) contains supplementary material, which is available to authorized users.

BackgroundOur ongoing clinical trial for the treatment of melanoma with TGF-β-resistant tumor-infiltrating lymphocytes (TIL) [TGF-β dominant negative receptor 2 (TGFβDNR2) transduced-TIL] has yielded long-term responses in checkpoint refractory patients (NCT01955460). Building on this success, we sought to extend the impact of TGF-β–resistant TIL therapy to additional cancers while optimizing a non-viral alternative to transduction with a TGFβDNR2. Ovarian cancer (OvCa), which is characterized by an abundance of TGF-β, a dysfunctional immune infiltrate, and a paucity of novel treatment options, is an ideal candidate for TGF-β–resistant TIL therapy. Here, we present an optimized and clinically-scalable method for CRISPR/Cas9-mediated deletion of the TGF-β receptor (TGFBR2) in OvCa TIL.MethodsOvCa TIL were generated from tumor fragments1 and subjected to CRISPR-mediated knockout of TGFBR2 before going through a rapid expansion protocol. Resistance of TGFBR2-knockout TIL to TGF-β signaling was evaluated via quantification of downstream SMAD-2/-3 phosphorylation, global transcriptional changes upon TGF-β exposure, and cytokine release following TCR stimulation in the presence of TGF-β. The impact of CRISPR modification on TIL expansion and TCR clonal diversity was evaluated. Finally, the risk of off-target CRISPR activity throughout the genome was evaluated using Target Enriched GUIDE-seq (TEG-seq)2 followed by next generation sequencing (NGS) validation of putative off-target sites.ResultsUsing five TGFBR2-directed guide RNAs (gRNAs), we achieved gene disruption efficiencies ranging from 48%–90%, which correlated inversely with the degree of SMAD phosphorylation after TGF-β exposure (r=-0.9440, p=0.0158, n=4 donors) (figure 1A-C). TGF-β exposure induced a strong transcriptional response in wild-type TIL but had little to no effect on TGFBR2-knockout TIL (figure 2). TGFBR2-knockout TIL functioned well in the presence of exogenous TGF-β as evidenced by equally strong secretion of pro-inflammatory cytokines in the presence and absence of TGF-β (figure 3). CRISPR-modification did not hamper the ex vivo expansion efficiency nor the TCR clonal diversity of expanded OvCa TIL (figure 4). Using TEG-seq, we identified ≤5 low-probability off-target sites for gRNA-#3 and gRNA-#4, each of which were attributed to background sequencing artifacts upon further validation by NGS of specific amplicons (figure 5).Abstract 172 Figure 1(A) Genomic-level TGFBR2 knockout efficiency using 5 different gRNAs as evidenced by NGS of specific amplicons (n=1 TIL donor). (B) SMAD-2 and SMAD-3 phosphorylation in TGFBR2 knockout TIL vs. control TIL after 30 min exposure to TGF-β1. The left panel shows representative histograms of phospho-SMAD staining, and the right panel shows quantification of cells positive for phospho-SMAD-2/-3 after TGF-β exposure (n=4 TIL donors). The statistical significance of each experimental condition compared to the non-transfected control is shown. (C) Inverse correlation of TGFBR2 knockout efficiency and TGF-β-mediated SMAD phosphorylation.Abstract 172 Figure 2Top 100 differentially expressed genes in non-transfected (WT) TIL exposed to TGF-β. TGFBR2 knockout (KO) TIL display minimal gene expression changes upon TGF-β exposure (n=3 technical replicates).Abstract 172 Figure 3TIL were collected after 14 days of expansion and re-stimulated with 300 ng/mL plate-bound anti-CD3 in the presence of 3000 IU/mL IL2 and 10 ng/mL human TGF-β1 or vehicle. Cell culture supernatant was collected after 72 hrs of stimulation and assayed for the presence of 10 proinflammatory cytokines (IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and TNF-α). For TIL with intact TGFBR2 (non-transfected and Cas9 mock transfected TIL), the production of many pro-inflammatory cytokines decreased significantly in the presence of TGF-β. Conversely, TGFBR2 knockout TIL (generated using gRNA #3 or gRNA #4) retain cytokine secretion in the presence of TGF-β. IL-12p70 was below the limit of detection in this assay and is therefore not presented.Abstract 172 Figure 4(A) Control and CRISPR-modified OvCa TIL expand with equal efficiency during a 14-day rapid expansion protocol. Fold expansions ranging from 1000× - 3000× were observed across 4 independent patient samples. (B) The TCR clonal diversity of TIL after 14-day expansion was assessed by TCRB sequencing. Productive Simpson Clonality was equivalent in CRISPR-modified TIL compared to control TIL samples.Abstract 172 Figure 5TEG-seq revealed 3 putative off-target sites for gRNA #3 and 5 putative off-target sites for gRNA #4. The aligned sequences show similarities and differences between the gRNA sequence and the reference genome site. Dots represent exact matches in the reference genome compared to the gRNA sequence. Dashes represent missing bases, lower-case letters represent extra bases, and upper-case letters represent a base mismatch. Validation by NGS of specific amplicons confirmed the presence of TEG-seq Tag integration and large indels at the on-target cleavage sites for gRNA #3 and #4, indicating successful Cas9 editing and Tag integration in our experiment. NGS validation revealed that all putative low probability off-target sites were background artifacts as evidenced by the lack of Tag identification and lack of large indels.ConclusionsCRISPR/Cas9-mediated knockout of TGFBR2 is feasible and efficient in patient-derived OvCa TIL using clinically-scalable methods that yield little to no evidence of off-target activity. This study lays the groundwork for clinical translation of CRISPR-modified, TGF-β-resistant TIL for OvCa treatment, which will not only provide a novel immunotherapy for OvCa patients but also a platform for engineering more potent TIL therapies in general.ReferencesSakellariou-Thompson D, Forget MA, Hinchcliff E, Celestino J, Hwu P, Jazaeri AA, et al. Potential clinical application of tumor-infiltrating lymphocyte therapy for ovarian epithelial cancer prior or post-resistance to chemotherapy. Cancer Immunology, Immunotherapy: CII 2019;68(11):1747–57.Tang PZ, Ding B, Peng L, Mozhayskiy V, Potter J, Chesnut JD. TEG-seq: an ion torrent-adapted NGS workflow for in cellulo mapping of CRISPR specificity. Bio Techniques 2018;65(5):259–67.Ethics ApprovalAll procedures performed were in accordance with the 1975 Helsinki declaration. Ethical approval and tissue from surgical resections used to expand TIL were both obtained under protocols (PA16-0912 and LAB02-188) approved by the Institutional Review Board of The University of Texas MD Anderson Cancer Center. Written informed consent was obtained from all individual participants included in the study for their specimens and data to be used in research and for publication.

2536 Background: TIL therapy has been used extensively in metastatic melanoma patients for many years, now with ongoing efforts to commercialize treatment. The efficacy of TIL outside of melanoma is largely unknown thus we designed and implemented a trial using TIL manufactured at a single academic center for treatment refractory metastatic colorectal (CRC), pancreas (PDAC) and ovarian (OVA) cancers. Methods: Patients with CRC, PDAC and OVA refractory to standard therapies with ECOG PS 0-1 and normal organ function were eligible for TIL harvest. Ex vivo TIL expansion and manufacturing was conducted at the MD Anderson TIL lab under conditions that included IL2 and 41BB stimulation (using urelumab). All patients received a lymphodepletion regimen consisting of cyclophosphamide 60mg/kg days -7 and -6 and fludarabine 25mg/m2 days -5 through day -1, followed by infusion of pooled ex-vivo expanded TIL. Patients received up to 6 doses of high dose IL-2 (600,000 IU/kg) after TIL infusion. The primary endpoint was evaluation of the objective response rate (ORR) using RECIST 1.1 criteria with secondary endpoints including disease control rate, duration of response, PFS, OS and safety. Results: A total of 17 patients underwent TIL harvest and 16 were treated on protocol; including 8 CRC, 5 PDAC and 3 OVA. Median age was 57.5 (range 33-70) and 50% were females. Median number of lines of prior therapy was 2 (range 1-8). Median number of TIL infused was 76 X 109 (range 20.3 x 109-150 x 109). Median doses of cyclophosphamide and fludarabine administered were 2 (range, 2-2) and 3 (range, 1-5), respectively. Median doses of IL-2 administered was 6 (range, 1-6). There were no responders. Best response included prolonged SD in a patient with PDAC lasting until 18 months. Grade 3 or higher toxicities attributable to therapy was seen in 14 subjects (87.5%; 95% CI: 61.7 – 98.4) with the majority of toxicities representing expected pancytopenia from lymphodepletion. Infusion product analysis showed the presence of effector memory cells with high expression of CD39 irrespective of tumor type. Early on-treatment biopsy of PDAC patient with prolonged SD showed presence of proliferating (KI67+) CD4+ and CD8+ TIL. Conclusions: Generation of TIL at a single academic center for CRC, PDAC and OVA is feasible and treatment is associated with no new safety signals. For these tumor types, further research is required to identify host factors associated with resistance to TIL therapy and optimize manufacturing processes to create more effective TIL cell therapy. Clinical trial information: NCT03610490.

Background and objectives Adoptive T-cell therapies (ACTs) using expansion of tumor-infiltrating lymphocyte (TIL) populations are of great interest for advanced malignancies, with promising response rates in trial settings. However, postoperative outcomes following pulmonary TIL harvest have not been widely documented, and surgeons may be hesitant to operate in the setting of widespread disease. Methods Patients who underwent pulmonary TIL harvest were identified, and postoperative outcomes were studied, including pulmonary, cardiovascular, infectious, and wound complications. Results 83 patients met inclusion criteria. Pulmonary TIL harvest was undertaken primarily via a thoracoscopy with a median operative blood loss and duration of 30 ml and 65 min, respectively. The median length of stay was 2 days. Postoperative events were rare, occurring in only five (6%) patients, including two discharged with a chest tube, one discharged with oxygen, one episode of urinary retention, and one blood transfusion. No reoperations occurred. The median time from TIL harvest to ACT infusion was 37 days. Conclusions Pulmonary TIL harvest is safe and feasible, without major postoperative events in our cohort. All patients were able to receive intended ACT infusion without delays. Therefore, thoracic surgeons should actively participate in ongoing ACT trials and aggressively seek to enroll patients on these protocols.

Background: The ImmunogenomiC prOfiling of Non-small cell lung cancer (NSCLC) Project (ICON) represents an ambitious undertaking to comprehensively characterize immuno-genomic diversity in NSCLC across diverse platforms. The depth and breadth of this cohort presented a unique opportunity to develop a specialized method for multi-platform data integration and exploration, which can be broadly applied to forthcoming large-scale patient profiling studies. Such a holistic approach can unlock insights for therapeutic targets, biomarkers, and treatment plans by providing a more complete view of phenomena driving disease pathogenesis and evolution. Purpose: We developed a novel shared nearest neighbors (SNN) approach to create an integrated network of ICON’s multi-platform data and identified collections of closely related measurements within the resulting network tied to noteworthy patient characteristics, including recurrence and oncogenotype. Methods: The ICON dataset is derived from tumor and normal lung tissue samples collected from 150 patients at time of resection as well as blood samples collected then and at intervals during the year following. Tissue samples underwent RNA-sequencing (RNA-seq), whole exome sequencing, T-cell receptor sequencing, multiplex immunofluorescence for immune cells, and reverse phase protein array profiling; flow cytometry for immune cells was performed on tissue and blood samples. From these data, the ICON data network was built using an integrative approach based on the SNN algorithm in which genes were linked on the basis of their shared top correlates in orthogonal datasets. Results: The ICON data network currently includes over 20,000 genes linked by over 500,000 connections derived from correlations between RNA-seq and orthogonal platforms. We captured established associations between cancer-related genes and examined these along with new ones in the network. To do so, we used the InfoMap algorithm to extract more interpretable sub-networks, termed modules, from the ICON data network. Single sample gene set enrichment scores for each module were used in multivariate analysis to highlight modules linked to clinical characteristics of interest. As an example, we found modules significantly tied to disease recurrence. The most notable of these was strongly associated with metabolic pathways, and other modules associated with platelets and ion channels were also identified. The metabolic pathway module is being explored as a prognostic biomarker, underscoring the opportunites enabled by mining the network. Conclusions: Through the framework developed, we identified modules in the ICON data network significantly associated with important patient characteristics like recurrence and oncogenotype. We are validating the gene sets identified as potential biomarkers and are developing an interactive application to facilitate further mining of the network. Taken together, our SNN network-building approach enables the integration and exploration of patient data from diverse platforms. Citation Format: Stephanie T. Schmidt, Neal Akhave, Alexandre Reuben, Tina Cascone, Jianhua Zhang, Jun Li, Junya Fujimoto, Lauren A. Byers, Beatriz Sanchez-Espiridion, Lixia Diao, Jing Wang, Lorenzo Federico, Marie-Andree Forget, Daniel J McGrail, Annikka Weissferdt, Shiaw-Yih Lin, Younghee Lee, Natalie Vokes, Carmen Behrens, Ignacio I. Wistuba, Andrew Futreal, Ara Vaporciyan, Boris Sepesi, John V. Heymach, Chantale Bernatchez, Cara Haymaker, Jianjun Zhang, Christopher A. Bristow, Timothy P. Heffernan, Marcelo V. Negrao, Don L. Gibbons. A shared nearest neighbors approach for integrated, multi-platform networks and its application to the exploration of multiomics data from early-stage non-small cell lung cancers [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2021 Oct 7-10. Philadelphia (PA): AACR; Mol Cancer Ther 2021;20(12 Suppl):Abstract nr P009.

Background Immunotherapy has increasingly become a staple in cancer treatment. However, substantial limitations in the durability of response highlight the need for more rational therapeutic combinations. The aim of this study is to investigate how to make tumor cells more sensitive to T-cell-based cancer immunotherapy. Methods Two pairs of melanoma patient-derived tumor cell lines and their autologous tumor-infiltrating lymphocytes were utilized in a high-throughput screen of 850 compounds to identify bioactive agents that could be used in combinatorial strategies to improve T-cell-mediated killing of tumor cells. RNAi, overexpression, and gene expression analyses were utilized to identify the mechanism underlying the effect of Topoisomerase I (Top1) inhibitors on T-cell-mediated killing. Using a syngeneic mouse model (n = 5 per group), the antitumor efficacy of the combination of a clinically relevant Top1 inhibitor, liposomal irinotecan (MM-398), with immune checkpoint inhibitors was also assessed. All statistical tests were two-sided. Results We found that Top1 inhibitors increased the sensitivity of patient-derived melanoma cell lines (n = 7) to T-cell-mediated cytotoxicity (P < .001, Dunnett’s test). This enhancement is mediated by TP53INP1, whose overexpression increased the susceptibility of melanoma cell lines to T-cell cytotoxicity (2549 cell line: P = .009, unpaired t test), whereas its knockdown impeded T-cell killing of Top1 inhibitor–treated melanoma cells (2549 cell line: P < .001, unpaired t test). In vivo, greater tumor control was achieved with MM-398 in combination with α-PD-L1 or α-PD1 (P < .001, Tukey’s test). Prolonged survival was also observed in tumor-bearing mice treated with MM-398 in combination with α-PD-L1 (P = .002, log-rank test) or α-PD1 (P = .008, log-rank test). Conclusions We demonstrated that Top1 inhibitors can improve the antitumor efficacy of cancer immunotherapy, thus providing the basis for developing novel strategies using Top1 inhibitors to augment the efficacy of immunotherapy.

Introduction: Differences in the host immune environment are thought to mediate heterogeneous treatment responses in non-small cell lung cancer (NSCLC). Unlike individual platform analyses, integrative analysis of multi-platform profiling allows for the discovery of novel interactions that expand our understanding of the disease. Utilizing the ImmunogenomiC prOfiling of NSCLC patient cohort (ICON), a prospective multi-omics protocol of operable early-stage NSCLC tumors with integrated immune, genomic, and clinical data, we hypothesized that multi-platform analyses would identify differences in the immune-genomic landscape that are associated with disease recurrence. Methods: Tumor and tumor-adjacent uninvolved lung was collected at resection; blood was collected before and after surgery. Tissue samples underwent WES, RNAseq, TCR sequencing (TCRseq), multiplex immunofluorescence (mIF), and RPPA profiling; tissue and blood (PBMC) samples were analyzed by flow cytometry. An integrated, inter-modality network was built using Spearman correlations between measurement pairs from different data modalities. Multivariate analysis was performed to adjust for stage and histology. Results: A total of 89 treatment-naïve patients with Stage 1-3 resected NSCLC (Squamous: 19; Non-squamous: 70) and 24 months of follow-up were analyzed (recurrence N = 24; no recurrence N = 65). The data network includes over 4,000 measurements linked by over 50,000 correlations. InfoMap, a community detection approach, was used to extract sub-network modules, which were used to contextualize the results of multivariate analysis. Tumors from patients with recurrence demonstrated decreased immune cell infiltration and activation including decreased cytotoxic CD8 T-cells (CD8+PD1+; fold-change (FC) = 0.898, p = 0.018; flow cytometry), decreased T-cell clonality (FC = 0.954, p = 0.017; TCRseq), and decreased tumor-associated macrophages (CD68+PD-L1+; FC = 0.426, p = 0.011; mIF). Furthermore, circulating CD8+ICOS+ activated T cells were decreased in patients with recurrence suggesting an impaired systemic anti-tumor immune response (FC = 0.552, p = 0.042; PBMC Flow). Finally, tumor-adjacent uninvolved lungs showed distinct T-cell phenotypes with accumulation of inactive CD8 T-cells (CD8+PD1-TIM3-) in patients with recurrence and increased populations of activated CD8 T-cells (CD8+PD1+) in patients without recurrence. Conclusion: Integrative multi-omic analysis suggests preserved anti-tumor immune surveillance in patients who are disease-free after 2 years from surgical resection with curative intent for treatment of NSCLC relative to patients with disease recurrence. Further analysis is ongoing to interrogate genomic and immune variables that are associated with disease recurrence. Citation Format: Neal Akhave, Stephanie Schmidt, Alexandre Reuben, Tina Cascone, Jianhua Zhang, Jun Li, Junya Fujimoto, Lauren A. Byers, Beatriz Sanchez-Espiridion, Lixia Diao, Jing Wang, Lorenzo Federico, Marie-Andree Forget, Daniel J. McGrail, Annikka Weissferdt, Shiaw-Yih Lin, Younghee Lee, Carmen Behrens, Ignacio I. Wistuba, Andrew Futreal, Ara Vaporciyan, Boris Sepesi, John V. Heymach, Chantale Bernatchez, Cara Haymaker, Jianjun Zhang, Christopher A. Bristow, Marcelo V. Negrao, Don L. Gibbons. Integrated multi-platform profiling of early-stage non-small cell lung cancer identifies relationship between disease recurrence and decreased native immune response in treatment-naïve resected NSCLC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 619.

T cell–mediated immunotherapies are promising cancer treatments. However, most patients still fail to respond to these therapies. The molecular determinants of immune resistance are poorly understood. We show that loss of PTEN in tumor cells in preclinical models of melanoma inhibits T cell–mediated tumor killing and decreases T-cell trafficking into tumors. In patients, PTEN loss correlates with decreased T-cell infiltration at tumor sites, reduced likelihood of successful T-cell expansion from resected tumors, and inferior outcomes with PD-1 inhibitor therapy. PTEN loss in tumor cells increased the expression of immunosuppressive cytokines, resulting in decreased T-cell infiltration in tumors, and inhibited autophagy, which decreased T cell–mediated cell death. Treatment with a selective PI3Kβ inhibitor improved the efficacy of both anti–PD-1 and anti–CTLA-4 antibodies in murine models. Together, these findings demonstrate that PTEN loss promotes immune resistance and support the rationale to explore combinations of immunotherapies and PI3K–AKT pathway inhibitors. Significance: This study adds to the growing evidence that oncogenic pathways in tumors can promote resistance to the antitumor immune response. As PTEN loss and PI3K–AKT pathway activation occur in multiple tumor types, the results support the rationale to further evaluate combinatorial strategies targeting the PI3K–AKT pathway to increase the efficacy of immunotherapy. Cancer Discov; 6(2); 202–16. ©2015 AACR. See related commentary by Rizvi and Chan, p. 128. This article is highlighted in the In This Issue feature, p. 109

The purpose of this study was to determine the role of BAP1 levels in cutaneous melanoma (CM). BAP1 is a tumor suppressor in which loss of heterozygosity (LOH) from mutation and copy number alteration is well described in germline and somatic cancers. Although BAP1 genomic alterations in CM are extremely rare (2% of 665 samples from 5 datasets), marked variability in BAP1 expression is observed in CM. We show that low nuclear BAP1 levels portend a significantly worse clinical outcome in stage III CM (n=37, log rank p ≤0.01 for both overall survival and progression-free survival). Gene Set Enrichment Analysis (GSEA) revealed low BAP1 expression to be most highly ranked with an increased epithelial–mesenchymal transition (EMT) gene expression profile in CM tumors (n=379, FDR q = 1.34E-26) and cell lines (n=53, FDR q = 2.86E-116). We identify the expression of ZEB1, a master regulator of EMT, to be significantly associated with low BAP1 expression in CM tumors and cell lines (p= 1.5E-04 and 3.3E-05, respectively). Analysis of the BAP1 promoter indicates three canonical ZEB1 binding sites. Functional experiments show ZEB1 to bind to the BAP1 promoter, and luciferase activity assays indicate that ZEB1 acts as a transcriptional suppressor of BAP1 expression with differential utilization of the promoter binding sites. Targeted reduction of endogenous ZEB1 caused increased BAP1 levels, while targeted reduction of BAP1 did not modulate ZEB1 levels, consistent with ZEB1 having a suppressive effect on BAP1. Phenotypically, targeted reduction of BAP1 in CM cells resulted in a switch from a more differentiated, melanocytic state, to a less differentiated, more migratory and invasive phenotype. Extinguishing melanocyte-specific BAP1 in mice with a BRAF V600E mutant genetic background resulted in the emergence of primary melanoma tumors, with a marked EMT gene expression profile, and resultant metastases. Given the phenotypic changes associated with BAP1 levels in our mouse and human studies, we then tested the effect of modulating BAP1 on BRAF targeted therapy. Exogenous expression of BAP1 sensitized BRAF inhibitor (vemurafenib)-resistant melanoma cells, while the targeted reduction of BAP1 desensitized BRAF inhibitor-sensitive melanoma cells. BRAF mutant/BAP1 loss mice failed to exhibit a marked response to vemurafenib treatment compared to control mice. These data implicate regulation of BAP1 to be a major mechanism that characterizes a highly malignant and treatment-resistant subset of tumors. Our study indicates that nongenomic reduction in BAP1 through ZEB1 transcriptional modulation may be a key factor in aggressive CM. This abstract is also being presented as Poster A30. Citation Format: Haifeng Zhu, Junna Oba, Xiaoxing Yu, Caitlin A. Creasy, Marie-Andrée Forget, Fernando Carapeto, Cara L. Haymaker, Chang-Jiun Wu, Tatiana V. Karpinets, Wei-Lien Wang, Michael T. Tetzlaff, Alexander J. Lazar, Gordon B. Mills, Amanda R. Moore, Yu Chen, Jianhua Zhang, Jeffrey E. Gershenwald, Jennifer A. Wargo, Chantale Bernatchez, Patrick Hwu, P. Andrew Futreal, Scott E. Woodman. Nongenomic BAP1 aberrancy drives highly aggressive cutaneous melanoma phenotype [abstract]. In: Proceedings of the AACR Special Conference on Melanoma: From Biology to Target; 2019 Jan 15-18; Houston, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(19 Suppl):Abstract nr PR03.

3027 Background: Metastatic UM is a rare cancer with poor response rates to systemic therapy. Adoptive transfer of patient-specific TIL may represent the best strategy for treatment. TIL are harvested from primary or metastatic tumors and initially expanded in culture with high dose IL-2 prior to undergoing rapid expansion protocol and therapeutic administration. Here, we report improved rates of initial expansion using a previously described TIL 3.0 method which utilizes dual agonistic antibodies to TCR and 4-1BB (Urelumab) for stimulation, respectively, with high dose IL-2, compared to the traditional method. Methods: Between 2006 and 2019, patients were consented for TIL harvest from either primary or metastatic UM tumors. Demographics, clinical features, and outcomes of the TIL initial expansion were collected. Success rates, number of cells expanded, and days in culture for the two methods were analyzed using partially overlapping samples t-tests and z-tests. Results: There were 85 harvests and expansions from 76 patients using the traditional method and 32 expansions from 30 patients using TIL 3.0. Initial TIL expansion was successful in 97% of TIL 3.0 harvests compared to 35% for the traditional method (p < 0.001). More TIL were expanded with TIL 3.0 compared to the traditional method (291.3 million cells vs. 88.6 million cells, p < 0.001), and fewer days were required in culture (18.5 vs. 29.0, p < 0.001). Both primary UM harvests and metastatic harvests were more successful with TIL 3.0 (90% vs. 12% for primary, p < 0.001, and 100% vs. 42% for metastatic, p < 0.001). Conclusions: Expansion of UM tumors via the TIL 3.0 method led to successful growth in 97% of harvests. Therapeutic administration to patients with TIL 3.0 is under active investigation. [Table: see text]

TPS2650 Background: Adoptive cell therapy (ACT) with tumor-infiltrating lymphocytes (TIL) has a long history of efficacy in metastatic melanoma, and is being increasingly considered across other solid tumors. Preclinical data generated at MD Anderson Cancer Center has demonstrated the ability to grow TIL from a variety of tumor types including various types of sarcomas, ovarian and pancreas cancers. We are testing the efficacy of TIL across multiple tumor types using two different manufacturing protocols. Methods: We are conducting two ongoing investigator initiated basket TIL therapy trials. The first (NCT03449108) includes cohorts with poorly differentiated soft tissue and bone sarcomas, osteosarcoma, and platinum resistant ovarian cancer. The TIL product used in this trial is an investigational cell product (LN-145, Iovance Biotherapeutics, Inc.). The second trial (NCT03610490) includes cohorts of osteosarcoma, platinum resistant ovarian cancer, and pancreatic cancer (who have progressed on, or received maximal benefit from, front-line therapy). For this trial, TIL are manufactured at MD Anderson Cancer Center using a protocol that includes the use of urelumab (an agonistic anti-CD137 antibody) combined with T cell receptor activation during TIL expansion. In both trials eligible subjects undergo tumor harvest using a surgical excisional biopsy of the tumor for TIL manufacturing, receive a modified cyclophosphamide and fludarabaine lymphodepletion regimen and up to six doses of IL-2 (600,000 IU/kg) following TIL infusion. No intervening therapy is allowed between tumor harvest and initiation of lymphodepletion. The primary endpoint for each cohort is ORR as assessed by investigators using RECIST 1.1 criteria. The Simon’s two stage design is used to monitor the efficacy of each cohort independently. In the first stage, 10 patients will be treated per cohort. If there is no confirmed response in these 10 evaluable patients, the cohort will be terminated. If the cohort moves forward to Stage II, an additional 8 patients will be treated leading to a total of 18 patients. Three or more responders out of 18 treated patients for the cohort will be considered clinically relevant to justify further investigation. Enrollment is ongoing in all cohorts in both trials. An accrual update will be provided at the annual meeting. Clinical trial information: NCT03449108, NCT03610490.

9543 Background: TIL adoptive cell transfer (ACT) therapy can produce durable responses for MM pts although efficacy appears lower in the era of checkpoint inhibitors. Toxicities from HD-IL2, including sepsis physiology, limits widespread use of this regimen. Suppression of transferred TIL by either tumor cells or the tumor microenvironment could limit TIL responses. Pembro is known to promote T cell activation, thus, we evaluated the efficacy and safety of TIL with pembro with HD-IL2 versus LD-IL2. Methods: Pts with MM who had tumor harvested and cryopreserved TIL at MD Anderson with PS 0-1 and normal organ function were eligible. All pts received a standard LD regimen consisting of cyclophosphamide and fludarabine, followed by infusion of pooled ex-vivo expanded TIL and either HD-IL2 (Arm 1: 720,000 IU/kg IV q 8 hrs up to 15 doses) or LD-IL2 (Arm 2: 2 million IU SC for 14 d). Pts received pembro 200mg IV starting 21 d post T cell infusion every 3 wks for up to 2 yrs. Pts were randomized 1:1 based on stage and LDH. Paired blood and tumor biopsies were obtained prior to LD, prior to first and second dose of pembro and at time of progression. Results: A total of 36 pts were planned to enroll (18 in each arm); however, the protocol met pre-specified futility boundaries in Arm 1 which prompted early closure after treatment of 14 pts (7 in each Arm). Median age was 50 yrs, 6 were female, 8 had cutaneous melanoma, 2 mucosal, 2 uveal and 2 unknown primary. 86% were stage M1c, 14% M1D, 50% had LDH elevation. Median lines of prior therapy were 3 (range 1-6), including prior anti PD-1 in 13 pts. Best overall response was 1 PR (for 10 mos), 2 SD, 3 PD, 1 NE in Arm 1; 1 PR (ongoing over 36 mos), 1 SD, 5 PD in Arm 2. With median follow up of 9.2 mos, PFS was 3.9 mos for Arm 1 and 2.1 mos for Arm 2 (p = 0.99). Median OS was 9.7 mos for Arm 1 and 8.8 mos for Arm 2 (p = 0.71). Toxicity was similar in both Arms but with lower rates of grade 3 febrile neutropenia (57% vs. 71%) and shorter hospital stay (median 16 vs. 18 d) in Arm 2 vs. Arm 1. Conclusions: In a heavily treated pt population, TIL with pembro achieved low response rates. Use of LD-IL2 did not diminish efficacy and may be better tolerated than HD-IL2 for TIL ACT. Correlative studies are ongoing to determine mechanisms of treatment response and failure. Clinical trial information: NCT02500576.

Immune checkpoint inhibitors (ICI) produce durable responses in some melanoma patients, but many patients derive no clinical benefit. The molecular underpinnings of ICI resistance involve intricate cell-cell interactions that are yet elusive. To systematically map the interactions between malignant and immune cells in the tumor ecosystem, we applied single-cell RNA sequencing to 31 human melanoma tumors, profiling thousands of malignant, immune, and stromal cells. We identified a transcriptional program in malignanT-cells that is strongly associated with T-cell exclusion and immunotherapy resistance. Using highly multiplexed in situ imaging we first demonstrated that this program characterizes malignanT-cells in “cold” niches. Next, we showed that the program predicts clinical responses to ICI according to multiple independent validation cohorts, including a new cohort that we obtained from 112 melanoma patients treated with anti-PD-1 therapy. We then identified CDK4/6 as master regulators of this resistance program, and found that CDK4/6 inhibitors repress the program and shift melanoma cells into a senescence-associated secretory phenotype. Lastly, we showed that CDK4/6-inhibition leads to a substantial reduction in melanoma tumor outgrowth in a B16 mouse model when given in combination with immunotherapy. Taken together, our study provides a high-resolution landscape of ICI-resistant cell states, identifies clinically predictive signatures, and forms a basis for the development of novel therapeutic strategies that could overcome immunotherapy resistance. Citation Format: Livnat Jerby, Parin Shah, Michael S. Cuoco, Christopher Rodman, Mei-Ju Su, Johannes M. Melms, Rachel Leeson, Abhay Kanodia, Shaolin Mei, Jia-Ren Lin, Shu Wang, Bokang Rabasha, David Liu, Gao Zhang, Claire Margolais, Orr Ashenberg, Patrick A. Ott, Elizabeth I. Buchbinder, Riz Haq, Stephen Hodi, Genevieve M. Boland, Ryan J. Sullivan, Dennie Frederick, Benchun Miao, Tabea Moll, Keith Flaherty, Meenhard Herlyn, Russell S. Jenkins, Rohit Thummalapalli, Monika S. Kowalczyk, Israel Canadas, Bastian Schilling, Adam N.R Cartwright, Adrienne M. Luoma, Shruti Malu, Patrick Hwu, Chantale Bernatchez, Marie-Andree Forget, David A. Barbie, Alex K. Shalek, Itay Tirosh, Peter K. Sorger, Kai W. Wucherpfennig, Eliezer M. Van Allen, Dirk Schadendorf, Bruce E. Johnson, Asaf Rotem, Orit Rosenblatt-Rozen, Levi A. Garraway, Charles H. Yoon, Benjamin Izar, Aviv Regev. Single-cell RNA-sequencing of metastatic melanoma identifies a cancer cell-intrinsic program associated with immune checkpoint inhibitor resistance [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A082.

A hallmark of melanoma is its inherent immunogenicity through both innate and adaptive immune mechanisms. However, a number of factors inhibit these immune responses through intrinsic mechanisms in tumor cells or adaptive resistance triggered by the immune response via negative feedback. Understanding how these processes are balanced in context of pathways regulating T-cell activation, migration, and differentiation, and T-cell dysfunction in tumors has become a critical area of research. This chapter describes the immunoregulatory mechanisms in the melanoma tumor microenvironment and how positive and negative signaling elements can be harnessed to facilitate enhanced anti-tumor immune responses through immunotherapy.

Background: The immunocheckpoint represents a group of membrane proteins expressing on immune cells, composed of multiple co-inhibitory and co-stimulatory pathways. They are involved in eliminating unwanted substances while maintaining self-tolerance, which play a crucial role in immunomodulation. Malignant cells (MCs), on the other hand, often tends to take advantage of immunocheckpoints, for instance, by activating inhibitory checkpoint pathways in situ, to evade immune responses. Therefore, cancer immunotherapies, which target key immune checkpoints to restore effective anti-tumor immune function, emerge as a revolutionary strategy during recent years. The aim of this study was analyze a full range of checkpoint available to investigate their role in the microenvironment and their frequent co-expression in MCs and tumor infiltrated lymphocytes (TILs) in non-small cell lung carcinomas. Material and Methods: 4 µm-thick histology sections obtained from formalin-fixed paraffin-embedded (FFPE) tissues from a TMA containing 256 NSCLCs (adenocarcinomas, ADC=156; squamous cell carcinomas, SCC=100); were used for multiplex immunofluorescence (mIF) analyses. Nine immune markers including PD-L1, B7-H3, B7-H4, IDO-1, ICOS, OX-40, LAG3, TIM3, VISTA, CD3 and CD20 was optimized for immunofluorescence. Two mIF immunocheckpoints panels were composed included seven antibodies as follows: Panel 1: AE1/AE3, PD-L1, B7-H3, B7-H4, IDO1, VISTA, and CD3; and Panel 2: AE1/AE3, ICOS, LAG3, TIM3, OX40, CD3, and CD20. For quantitative image analysis of mIF markers, the stained TMA slides were scanned using the Polaris™ multispectral microscope (Perkin Elmer) and analyzed using the inForm™ 2.3.1 software. Results: In this preliminary report using the median density in MCs and TILs expressing those markers, we identified overall higher expression of immunocheckpoints in SCC than ADC. The higher density co-expression by MCs (AE/AE3+) was B7-H3 + IDO-1+ following by B7-H3+PD-L1+ expression when compared with the other markers (P~0.001) in the panel 1. Furthermore, we observed higher densities of TILs (CD3+) expressing IDO-1+ as well as B7-H3+ and high densities of TILs (CD3+) co-expressing B7-H3+PD-L1+ when compared the other (P~0.001) markers in the same panel. In panel 2, we observed higher densities of TILs (CD3+) expression ICOS+ than the other markers (P~0.001). Conclusions: Here, we describe a mIF platform with computational image processing workflows, enabling simultaneous evaluation of 12 biomarkers to FFPE tissue section. These data highlight digital image processing pipelines to the community for examining immune complexity of immunocheckpoints expression in MCs and TILs to improve biomarker discovery and assessment. Citation Format: Edwin Roger Parra, Mei Jiang, Jiexin Zhang, Carmen Behrens, Jack Lee, John Victor Heymach, Marie-Andrée Forget, Cara Haymaker, Chantale Bernatchez, Cesar Moran, Jianjun Zhang, Don Gibbons, Ignacio Wistuba. Immunofluorescence profiling of co-expression of multiple immune checkpoints in malignant and tumor infiltrated lymphocytes in non-small cell lung carcinomas using image analysis system [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 3633.

161 Background: Multiple clinical studies have shown that adoptive cell transfer (ACT) of autologous tumor-infiltrating lymphocytes (TIL) is remarkably effective in melanoma patients. Non-small cell lung cancer (NSCLC) shares similarities with melanoma in terms of mutational burden and sensitivity to immune checkpoint inhibitors. We therefore sought to test whether TIL ACT may represent a viable option for the treatment of NSCLC patients. We utilized tissue collected from patients enrolled on the prospective ImmunogenomiC prOfiling of Non-small cell lung cancer (ICON) study. Methods: TIL and tissue-infiltrating lymphocytes were expanded ex vivo from 97 freshly resected early-stage localized NSCLC tumors and 39 matched uninvolved lung tissues. Growth and functional characteristics of TIL were assessed via flow cytometry, TIL-tumor reactivity assays, and analysis of TCRβ sequencing data. Results: NSCLC showed an increased proportion of CD3+ lymphocytes within the tumor-infiltrating leukocyte component as compared to matched normal lung tissue. The TIL compartment included a suppressed CD8+ T cell subset expressing significantly higher levels of PD-1 and lacking cytolytic potential compared to T cells expanded from normal tissue. TIL contained a higher proportion of proliferating (Ki67+) CD8+CD103+ tissue-resident memory (TRM) cells expressing activation markers such as CTLA4, LAG3, PD1 and ICOS, and increased CD4+ Tregs. Despite a highly immunosuppressive environment, TIL expansion was achieved with a success rate of 68% (n = 97) but appeared hindered in patients undergoing neoadjuvant chemotherapy treatment prior to surgery (56.2%, n = 16 vs 72.5% success rate in therapy-naïve patients). In addition, expansion efficiency (number expanded and time of culture) of TIL and matched lung residing lymphocytes were significantly associated (r = 0.379, p = 0.017, n = 39). Importantly, expanded CD8+ TIL products were oligoclonal and showed reactivity toward autologous tumors. Conclusions: Although NSCLC TIL are functionally inhibited in vivo they can be successfully expanded ex vivo and demonstrate recognition of autologous tumor cells. These data suggest that TIL can potentially be used for adoptive T cell-based immunotherapy in NSCLC.

95 Background: Immunotherapy has become an effective cancer therapy, particularly in the case of checkpoint blockade and adoptive T-cell therapy (ACT). ACT exploits the presence of tumor-infiltrating lymphocytes (TIL) by exponentially expanding their numbers ex vivo and re-infusing them into the patient in an autologous setting. With the effectiveness of TIL therapy already well established in multiple phase II studies in melanoma, there is a push to translate it to other cancers in need of improved therapies. Colorectal cancer (CRC) is a cancer where the presence of TIL has been strongly correlated with increased survival. Metastatic CRC (mCRC) has a poor outcome with median overall survival of less than 3 years. At present anti-PD1 therapy is only active in the small subset of mCRC patients (4%) that are MSI-high. We sought to evaluate the ability to generate and characterize TIL from patients with mCRC to provide a rationale for future TIL therapy in this disease. Methods: To assess the feasibility of utilizing TIL ACT for this patient population, we characterized the immune infiltrate of mCRC, TIL growth from tumor fragments (n = 24), as well as the TIL repertoire (n = 4) and anti-tumor potential in samples with available autologous tumor target. Results: Flow cytometry analysis detected a CD4-rich T-cell infiltration at the tumor site as well as a scarce yet activated CD8+ TIL population. The outgrowth of CD8+ TIL from tumor fragments in media containing IL-2 was potentiated by the addition of an agonistic 4-1BB antibody (Urelumab, BMS). Additionally, the use of a 4-1BB mAb improved the likelihood of growing TIL (63% [14/24] to 83% [20/24]) and increased the total yield of TIL grown. T-cell receptor sequencing showed enrichment in the tumor of CD8+ T-cell clones shared between the blood and tumor, suggesting selective expansion at the tumor site. Using IFNγ ELISPOT, CD8+ TIL from one patient with an MSS tumor were found to be reactive against a peptide derived from mutated TP53R116W restricted to HLA-B*39:01. Conclusions: It is possible to expand CD8+ T cells from microsatellite stable (MSS) mCRC that are able to target tumor antigens. Although preliminary, the initial data suggest the feasibility of TIL therapy for mCRC.

5528 Background: Human papillomaviruses (HPVs) play a major role in development of cervical cancer, and HPV oncoproteins are being targeted by immunotherapies. Although these treatments show promising results in the clinic, many patients do not benefit or the durability is limited. Methods: To explore the landscape of neoantigens in cervix cancer, we predicted all possible mutated neopeptides using exome and RNA sequencing data of a large number of tumors in two independent data sets (n = 194 and 79), and analyzed whether mutation and neoantigen load correlate with expression of genes involved in antigen presentation, infiltrating immune cell type markers, and HPV oncoprotein-associated genes. Normal tissue expression analyses were also performed using data from the Genotype-Tissue Expression project (n = 8,555). Results: We found that potentially immunogenic and targetable neoantigens can be predicted for almost all cervix tumors, including recurrent neoantigens from mutations of known oncogenic driver genes (KRAS G12D, MAPK1 E322K, PIK3CA E545K, PIK3CA E542K, ERBB2 S310F, and ERBB3 V104M). Furthermore, we found that expression of HPV-associated “master regulator” genes is associated with mutation, neoantigen, and HPV load, and also with expression of several important markers in the immune microenvironment. Notably, the OVOL1 master regulator positively correlated with mutation and neoantigen load, and also with PD-L1 and TGFB1 expressions. Furthermore, ENO1 over-expression in cervical cancer was associated with high HPV load, and ENO1 level also positively correlated with PD-L1 and TGFB1 expressions, suggesting that it may be a potential target in cervical cancer. Conclusions: For most of the cervix tumors we report predicted neoantigens, and we also identified recurrent neopeptides derived from mutations in known oncogenes. We have also identified statistically significant associations between neoantigen availability, antigen presentation, cytotoxic T-cell function, suppressive mechanisms, and expression of HPV master regulators to help guide immunotherapies of cervical cancer.

138 Background: Adoptive cell therapy (ACT) of autologous tumor-infiltrating lymphocytes (TIL) can produce long lasting treatment responses in patients (pts) with metastatic melanoma. In our initial report of 31 treated pts, we demonstrated overall response rate (ORR) of 48%. Due to the evolution of treatment options for metastatic melanoma with heavy reliance on immunomodulatory therapies, we sought to re-evaluate responses in the era of checkpoint blockade. Methods: Pts receive treatment consisting of 7 days of lymphodepleting chemotherapy consisting of cyclophosphamide and fludarabine. High dose interleukin-2 (720,000 IU/kg IV q 8 hrs up to 15 doses) was infused after TIL infusion. Responses were assessed by imaging per irRC. Results: A total of 74 pts have been treated in our initial TIL study with an updated ORR assessment of 43%. Stratification of pts according to their checkpoint blockade immune-modulator therapy prior to TIL ACT, demonstrated that prior Ipilimumab (Ipi) decreased ORR to 33% compared to 51% in checkpoint naïve pts and decreased overall survival (OS) post-TIL therapy (median 7.7 vs 24.6 months respectively). There were too few pts to assess the impact of anti-PD1 as a single agent. A multiparametric analysis revealed that LDH levels at the time of therapy mainly influences OS and ORR to TIL but still could not invalidate the impact of Ipi prior to TIL ACT. The durability of the response was also found to be different between the 2 groups (30% for Ipi prior and 50% No Ipi prior). This new reality also impacted previously reported parameters that correlated with ORR to TIL ACT such as the expression of B-and-T lymphocyte attenuator (BTLA) on CD8+ TIL. Interestingly, assessment of mutation load (ML) of the tumors prior to TIL ACT showed that the check point naïve group display a high ML ( > 100) within their tumor whereas some patients who had Ipi prior to TIL have a low ML ( < 100). Conclusions: Our preliminary analysis shows that pre-treatment with Ipi decreases ORR to TIL ACT. Understanding how prior ipi modifies TIL, the tumor and microenvironment will help to define the full extent of the impact and how to best treat Ipi refractory pts with TIL. Clinical trial information: NCT00338377.

Manipulation of T-cell co-inhibitory molecules, such as CTLA-4, PD-1, and BTLA has recently moved to the forefront of cancer immunotherapy. Although these molecules serve as inhibitors of T-cell activation, they are also biomarkers for activated T cells and may in fact have positive immune

Meeting abstracts The ability of the immune system to treat established tumors is very well recognized. Immunotherapeutic approaches such as Adoptive T cell therapy and anti-CTLA-4 antibody treatment can result in long-term responses in some patients with metastatic melanoma. However, the majority

Meeting abstracts Adoptive cell therapy (ACT) with expanded autologous tumor infiltrating lymphocytes (TIL) has emerged to be a powerful salvage therapy for metastatic melanoma with response rates up to 50%. However, the generation of the TIL ACT product is technically challenging with current

T cell-mediated immunotherapies are promising cancer treatments. However, most patients still fail to respond to these therapies. The molecular determinants of immune resistance are poorly understood. Here, we interrogated the role of loss of expression of the tumor suppressor, PTEN, in immune resistance. In preclinical studies, we found that silencing PTEN in tumor cells inhibited T cell-mediated tumor killing and decreased T cell trafficking into tumors. In clinical studies, we observed that tumors with loss of PTEN had significantly less CD8+ T cell infiltration than PTEN-present tumors. In addition, 26% of melanomas that did not yield successful TIL growth demonstrated PTEN loss, which was more frequent than was observed in tumors that yielded successful TIL growth (11%). We further validated the association between reduced number and impaired function of TIL with PTEN loss using another independent cohort, TCGA dataset for SKCM. More importantly, we analyzed clinical outcomes of metastatic melanoma patients treated with the FDA-approved anti-PD-1 antibodies. Our analysis demonstrates that a greater reduction in tumor burden was achieved by PD-1 blockade in PTEN present patients, when compared with PTEN absent patients. To decipher the factors mediating the immunosuppressive effects of PTEN loss, we determined the expression profiles of tumor cells with or without PTEN expression. Our results indicated that PTEN loss increased the production of immunosuppressive factors, including CCL2 and VEGF. Anti-VEGF blocking antibody improved anti-tumor activity of transferred tumor-reactive T cells and enhanced tumor infiltration of transferred T cells in PTEN-silenced tumors. These results suggest that loss of PTEN can facilitate the resistance of T cell-mediated immune responses by increasing the expression of immunosuppressive factors. Given that PTEN loss results in activation of the PI3K pathway, we evaluated the efficacy of immunotherapy in combination with a selective PI3Kβinhibitor to treat spontaneously developed BRAF mutant, PTEN null melanomas in genetically engineered mouse models. Our result showed that the combination of PI3Kβ inhibitor and anti-PD-1 significantly delayed tumor growth in tumor-bearing mice. Mice treated with this combination had a median survival time of 28 days, which is longer than the survival time of mice treated with either therapy. Increased numbers of T cells at tumor sites were found in mice receiving the combination therapy compared with mice receiving either agent alone. Taken together, our results demonstrate that PTEN loss contributes to the generation of immunosuppressive tumor microenvironment. Notably, this study provides the first direct clinical evidence to support the association between PTEN loss and poor clinical outcome in immunotherapy treated patients. In addition, our study indicates that inhibition of the PI3K-AKT pathway can improve the efficacy of immunotherapy in cancer. Citation Format: Weiyi Peng, Jie Qing Chen, Chengwen Liu, Shruti Malu, Caitlin Creasy, Michael Tetzlaff, Chunyu Xu, Jodi McKenzie, Chunlei Zhang, Xiaoxuan Liang, Leila Williams, Wanleng Deng, Guo Chen, Rina Mbofung, Alexander Lazar, Carlos Torres-Cabala, Zachary Cooper, Pei-Ling Chen, Trang Tieu, Stefani Spranger, Xiaoxing Yu, Chantale Bernatchez, Marie-Andree Forget, Cara Haymaker, Rodabe Amaria, Jennifer McQuade, Isabella Glitza, Tina Cascone, Haiyan Li, Lawrence Kwong, Timothy Heffernan, Jianhua Hu, Roland Bassett, Marcus Bosenberg, Scott Woodman, Willem Overwijk, Gregory Lizée, Jason Roszik, Thomas Gajewski, Jennifer Wargo, Jeffrey Gershenwald, Laszlo Radvanyi, Michael Davies, Patrick Hwu. Loss of PTEN promotes resistance to T cell-mediated immunotherapy. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4363.

Melanoma is the major cause of skin cancer-related deaths. However, the recent emergence of immune checkpoint inhibitors such as Ipilimumab (anti-CTLA-4) and Nivolumab (anti-PD-1), and adoptive cell therapy, using tumor infiltrating lymphocytes (TILs), has improved clinical outcomes and produced response rates of approximately 50%. Despite the progress made in melanoma immunotherapy, there is a large cohort of patients for which these treatments are currently not applicable. They are either inherently resistant to or acquire resistance to immunotherapy. As such, understanding mechanisms of resistance to immune response and ascertaining efficacious therapy combinations is important in overcoming immunoresistance and developing novel treatment regimens. As proof of principle, preliminary work using a unique protocol for an in vitro T-cell mediated cytotoxicity screen was established. In this assay, intracellular staining of active caspase-3 in tumor cells, a marker of apoptosis, is measured by flow cytometry in a 96-well format. Using patient-derived tumor and TIL pairs, a screen of 850 compounds was performed to find candidate drugs that could modulate the sensitivity of tumor cells to T- cell mediated killing. Initial screens identified classes of compounds that inhibit Heat shock protein 90 (Hsp90), Aurora Kinase and Topoisomerase I, and studies to understand their mechanisms of action are currently underway. However, as this screening process is very time-intensive, we are currently increasing the efficiency and feasibility of this screen by miniaturizing this workflow. This work presents the development of a screening method that is performed in a 384-well format, with a greater drug concentration range, and utilizing robotics to streamline and increase output. With this, we can specifically identify novel targeted and immunotherapy combinations that may be used in the clinic to treat melanoma. Additionally, by understanding the signaling pathways and molecular factors that regulate tumor response to T cell mediated killing, we can translate these findings into making immune-based cancer therapies applicable to other tumor types. This project aims to discover novel compounds that work synergistically with T-cell mediated tumor cell cytotoxicity, understand their mechanism of action and how they contribute to resistance and as a result identify innovative immune-based therapies that can be translated into clinical use. Citation Format: Leila Williams, Shruti Malu, Jodi McKenzie, Rina Mbofung, Marie-Andree Forget, Chantale Bernatchez, Patrick Hwu. Identification of novel targeted and immunotherapy combinations by a high throughput assay of T cell-mediated cytotoxicity. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4010.

TPS9594Background: Improving T cell trafficking to tumors may help augment anti-tumor responses in ACT using tumor infiltrating lymphocytes (TILs). mm produces the chemokines CXCL1 and CXCL8, which...

Background: Allogeneic hematopoietic cell transplantation (HCT) can be curative in a subset of patients with advanced lymphoid malignancies but relapse remains a major reason for treatment failure. Donor-derived, non-specific lymphocyte infusions (DLI) can confer an immune anti-malignancy effect but can be complicated by graft-versus-host-disease (GVHD). Chimeric antigen receptor (CAR)-modified T cells directed toward CD19 have demonstrated dramatic efficacy in patients with refractory ALL and NHL. However, responses are often associated with life-threatening cytokine release syndrome. Aim: We hypothesized that infusing CAR-modified, CD19-specific T-cells after HCT as a directed DLI would be associated with a low rate of GVHD, better disease control, and a less severe cytokine release syndrome since administered in a minimal disease state. Methods: We employed a non-viral gene transfer using the Sleeping Beauty (SB) transposon/transposase system to stably express a CD19-specific CAR (designated CD19RCD28 that activates via CD3z & CD28) in donor-derived T cells for patients with advanced CD19+ lymphoid malignancies. T-cells were electroporated using a Nucleofector device to synchronously introduce two DNA plasmids coding for SB transposon (CD19RCD28) and hyperactive SB transposase (SB11). T-cells stably expressing the CAR were retrieved over 28 days of co-culture by recursive additions of g-irradiated activating and propagating cells (AaPC) in presence of soluble recombinant interleukin (IL)-2 and IL-21. The AaPC were derived from K562 cells and genetically modified to co-express CD19 as well as the co-stimulatory molecules CD86, CD137L, and a membrane-bound version of IL-15. Results: To date, we have successfully treated 21 patients with median age 36 years (range 21-62) with advanced CD19+ ALL (n=18) or NHL (n=3); 10 patients had active disease at time of HCT. Donor-derived CAR+ T cells (HLA-matched sibling n=10; 1 Ag mismatched sibling n=1; haplo family n=8; cord blood n=2) were infused at a median 64 days (range 42-91 days) following HCT to prevent disease progression. Transplant preparative regimens were myeloablative, busulfan-based (n=10) or reduced intensity, fludarabine-based (n=11). All patients were maintained on GVHD prophylaxis at time of CAR T-cell infusion with tacrolimus, plus mycophenolate mofeteil for cord, plus post-HCT cyclophosphamide for haplo donors. The starting CAR+ T-cell dose was 106 (n=7), escalated to 107 (n=6), 5x107 (n=5), and currently at 108 (n=3) modified T cells/m2 (based on recipient body surface area). Patients have not demonstrated any acute or late toxicity to CAR+ T cell infusions. Three patients developed acute grades 2-4 GVHD (liver n=1, upper GI n=1, skin=1) which was within the expected range after allogeneic HCT alone. Of note, the rate of CMV reactivation after CAR T cell infusion was 24% vs. 41 % previously reported for our patients without CAR T cell infusion (Wilhelm et al. J Oncol Parm Practice, 2014, 20:257). Nineteen patients have had at least 30 days follow-up post CAR T-cell infusion and are evaluable for disease progression. Forty-eight percent of patients (n=10) remain alive and in complete remission (CR) at median 5.2 months (range 0-21.3 months) following CAR T cell infusion. Importantly, among 8 patients who received haplo-HCT and CAR, 7 remain in remission at median 4.2 months. Conclusion: We demonstrate that infusing donor-derived CD19-specific CAR+ T cells, using the SB and AaPC platform, in the adjuvant HCT setting as pre-emptive DLI may provide an effective and safe approach for maintaining remission in patients at high risk for relapse. Graft-vs-host disease did not appear increased by administration of the donor derived CAR-T cells. Furthermore, the add-back of allogeneic T cells appears to have contributed to immune reconstitution and control of opportunistic viral infection. Disclosures Huls: Intrexon and Ziopharm: Employment, Equity Ownership. Singh:Intrexon and Ziopharm: Equity Ownership, Patents & Royalties. Olivares:Intrexon and Ziopharm: Equity Ownership, Patents & Royalties. Su:Ziopharm and Intrexon: Employment. Figliola:Intrexon and Ziopharm: Equity Ownership, Patents & Royalties. Kumar:Ziopharm and Intrexon: Equity Ownership. Jena:Ziopharm Oncology: Equity Ownership, Patents & Royalties: Potential roylaties (Patent submitted); Intrexon: Equity Ownership, Patents & Royalties: Potential royalties (Patent submitted). Ang:Intrexon and Ziopharm: Equity Ownership. Lee:Intrexon: Equity Ownership; Cyto-Sen: Equity Ownership; Ziopharm: Equity Ownership.

Meeting abstracts Immunotherapy has become an effective cancer therapy, particularly in the case of checkpoint blockade and adoptive T cell therapy (ACT). ACT exploits the presence of tumor-infiltrating lymphocytes (TIL) by exponentially expanding their numbers ex vivo and re-infusing them into the

Meeting abstracts Adoptive T cell therapy using tumor-infiltrating lymphocytes (TIL) for patients with metastatic melanoma has been shown to have a 45-50% response rate as demonstrated by multiple institutions worldwide. However, understanding how long infused T cells persist in patients remains

Meeting abstracts Despite recent advances in the treatment of metastatic melanoma through targeted and immunotherapy, the majority of patients do not achieve a durable response. Research efforts to better understand responses are underway, and numerous molecular mechanisms of resistance to targeted