T-cell Inflamed Gene Expression Profile Research Articles

Abstract CT207: Association of gene expression profiles and clinical outcomes in patients with locally advanced or recurrent/metastatic cutaneous squamous cell carcinoma treated with pembrolizumab in KEYNOTE-629

Abstract Background: Pembrolizumab (pembro) monotherapy demonstrated clinically meaningful antitumor activity for both locally advanced (LA) and recurrent or metastatic (R/M) cutaneous squamous cell carcinoma (cSCC) in the phase 2 KEYNOTE-629 trial (NCT03284424). cSCC has high immunogenicity driven by a high tumor mutational burden and it is of interest to determine if the 18-gene T-cell inflamed gene expression profile (TcellinfGEP) and a set of other signatures relevant to the tumor microenvironment are associated with response to treatment. This retrospective analysis aimed to evaluate the association between GEP signatures and clinical outcomes in patients (pts) treated with pembro in KEYNOTE-629. Methods: Pts with histologically confirmed LA or R/M cSCC, measurable disease per RECIST v1.1, ECOG PS 0 or 1 received pembro 200 mg IV Q3W ≤35 cycles. Association between TcellinfGEP by RNAseq with clinical response (objective response rate [ORR], progression-free survival [PFS], and overall survival [OS]) to pembro, and association between a set of additional consensus signatures with clinical response to pembro adjusted for TcellinfGEP were evaluated. Significance of GEP signatures was prespecified at 0.05 for 1-sided P values from logistic (ORR) and Cox proportional hazard regression (PFS, OS) adjusted for ECOG PS. Additional assessments included TcellinfGEP AUC or Harrell’s C-index in predicting response to pembro. Results: Of 159 pts, 143 (n = 52, LA cSCC; n = 91, R/M cSCC) had RNAseq samples available for analysis. Of 143 pts, 33.6 % had a TcellinfGEP <1st tertile and 66.4% had a TcellinfGEP ≥1st tertile. TcellinfGEP was significantly associated with improved ORR (P = 0.040) but not PFS (P = 0.087) or OS (P = 0.709). None of the 10 other consensus signatures were significantly associated with clinical outcomes after adjusting for TcellinfGEP and multiplicity (P > 0.05). AUC of TcellinfGEP was moderately predictive of ORR (0.61, 95% CI, 0.51-0.70). Harrell’s C-index of TcellinfGEP was moderately predictive of PFS (0.57, 95% CI 0.51-0.63), and duration of response (0.56, 95% CI 0.38-0.73), but not predictive of OS (0.50, 95% CI 0.42-0.57). Median PFS (95% CI) was 4.1 mo (1.4-9.8) for the TcellinfGEP <1st tertile subgroup and 8.5 mo (5.5-21.0) for TcellinfGEP ≥1st tertile subgroup. Median OS (95% CI) for the TcellinfGEP <1st tertile subgroup was 25.1 mo (11.4-NR) and 21.0 mo (15.8-29.8) for the TcellinfGEP ≥1st tertile subgroup. Conclusions: In this retrospective analysis of KEYNOTE-629, Tcellinf GEP was associated with tumor objective response to pembro, but was not clearly associated with longer survival in pts with LA or R/M cSCC treated with pembro. While definitive conclusions require more analyses, these findings provide additional support for TcellinfGEP as a predictive biomarker for cSCC. Citation Format: Åse Bratland, Jean-Jacques Grob, Eva Munoz Couselo, Laurent Mortier, Ralf Gutzmer, Osama Roshdy, Rene Mendoza Gonzalez, Jacob Schachter, Ana M. Arance Fernandez, Florent Grange, Nicolas Meyer, Abhishek J. Joshi, Salem Billan, Judong Shen, Razvan Cristescu, Michael Nebozhyn, Andrey Loboda, Jason Sparkowski, Burak Gumuscu, Jianda Yuan, Brett Hughes. Association of gene expression profiles and clinical outcomes in patients with locally advanced or recurrent/metastatic cutaneous squamous cell carcinoma treated with pembrolizumab in KEYNOTE-629 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr CT207.

Evaluation of potential biomarkers for lenvatinib plus pembrolizumab among patients with advanced endometrial cancer: results from Study 111/KEYNOTE-146

BackgroundLenvatinib plus pembrolizumab demonstrated clinically meaningful benefit in patients with previously treated advanced endometrial carcinoma in Study 111/KEYNOTE-146 (NCT02501096). In these exploratory analyses from this study, we evaluated the associations...

Molecular determinants of clinical outcomes of pembrolizumab in recurrent ovarian cancer: Exploratory analysis of KEYNOTE-100

ObjectiveThis prespecified exploratory analysis evaluated the association of gene expression signatures, tumor mutational burden (TMB), and multiplex immunohistochemistry (mIHC) tumor microenvironment–associated cell phenotypes with clinical outcomes of pembrolizumab in advanced recurrent ovarian cancer (ROC) from the phase II KEYNOTE-100 study. MethodsPembrolizumab-treated patients with evaluable RNA-sequencing (n = 317), whole exome sequencing (n = 293), or select mIHC (n = 125) data were evaluated. The association between outcomes (objective response rate [ORR], progression-free survival [PFS], and overall survival [OS]) and gene expression signatures (T-cell–inflamed gene expression profile [TcellinfGEP] and 10 non-TcellinfGEP signatures), TMB, and prespecified mIHC cell phenotype densities as continuous variables was evaluated using logistic (ORR) and Cox proportional hazards regression (PFS; OS). One-sided p-values were calculated at prespecified α = 0.05 for TcellinfGEP, TMB, and mIHC cell phenotypes and at α = 0.10 for non-TcellinfGEP signatures; all but TcellinfGEP and TMB were adjusted for multiplicity. ResultsNo evidence of associations between ORR and key axes of gene expression was observed. Negative associations were observed between outcomes and TcellinfGEP-adjusted glycolysis (PFS, adjusted-p = 0.019; OS, adjusted-p = 0.085) and hypoxia (PFS, adjusted-p = 0.064) signatures. TMB as a continuous variable was not associated with outcomes (p > 0.05). Positive associations were observed between densities of myeloid cell phenotypes CD11c+ and CD11c+/MHCII−/CD163−/CD68− in the tumor compartment and ORR (adjusted-p = 0.025 and 0.013, respectively). ConclusionsThis exploratory analysis in advanced ROC did not find evidence for associations between gene expression signatures and outcomes of pembrolizumab. mIHC analysis suggests CD11c+ and CD11c+/MHCII−/CD163−/CD68− phenotypes representing myeloid cell populations may be associated with improved outcomes with pembrolizumab in advanced ROC.Clinical trial registration: ClinicalTrials.gov, NCT02674061.

763P Combined gene signature (T-cell inflamed gene expression profile – TcellinfGEP – & IMMETCOLS) identifies a metabolic rewiring program, key for immune evasion

The intrinsic and acquired resistance to immune checkpoint inhibitors (ICI) is a fact. We propose an immune-metabolic signature (IMMETCOLS sig.) that distributes patients (pts) across 11 tumors in 3 clusters (IMC1-3); IMC1 has more immune-suppressive score, despite higher PD1 and PD-L1 expression and IMC3, more metabolic flexibility in terms of glycolysis/mitochondrial respiration capacities. We hypothesize that our sig. in combination with TcellinfGEP and tumor mutational burden (TMB) can better identify pts for ICI.

Immune Gene Signature Expression Differs between African American and Caucasian Patients with Renal Cell Carcinoma

BACKGROUND: Predictive immune signatures such as the T-effector, the 26-gene “Renal 101 Immuno signature” and the 18-gene T-cell inflamed gene expression profile were developed in clinical trials enrolling predominantly Caucasians and there is a dearth of literature comparing tumor biology between African American (AA) and Caucasian patients. OBJECTIVE: To compare the immune gene signature expression in AA (n = 55) and Caucasian (n = 457) patients. METHODS: Raw gene expression count data were downloaded from the TCGA KIRC dataset and tumor samples from “white” and “black or AA” patients were selected. The gene expression values of the immune signatures were VST-transformed normalized counts and compared between the groups. RESULTS: There were 457 Caucasian and 55 AA patients in the TCGA. The immune gene expression in all three signatures was significantly lower in AA patients compared to Caucasians (p < 0.05). We validated our findings in an independent dataset using Nanostring Immune Profile Panel. Since the majority of AA tumors in TCGA were stage I (71%), we compared gene expression between stage I AA tumors (n = 39) with stage I Caucasian tumors (n = 220). Once again, the immune gene expression was significantly lower in AA patients compared to Caucasians (p < 0.05), indicating differences in tumor biology between the races. CONCLUSIONS: Low expression of predictive immune gene signatures in AA compared to Caucasian patients indicates a possible difference in the biology of their tumors. Future studies are needed to validate our findings in other datasets and to study the predictive role of these signatures in AA patients.

Abstract 810: HRD is inversely correlated with MSI and identifies immunologically cold tumors in most cancer types

Abstract Homologous recombination deficiency (HRD) leads to DNA double strand breaks and can be exploited by the use of PARP inhibitors to induce synthetic lethality of cancer cells. Over the last years, the clinical utility of this therapeutic concept was successfully demonstrated in ovarian, breast, prostate and pancreatic cancer. Currently, the role of HRD is investigated in trials testing immune checkpoint blockers alone or in combination with PARP inhibitors. But the relationship between HRD and immune cell contexture in cancer is incompletely understood. Here, we analyzed the association of HRD with tumor mutational burden (TMB), immune cell composition and gene expression in 10,000 tumors of 32 solid cancer types from the TCGA project. For each of the tumors, the HRD sum score (HRDsum) was calculated from allele-specific copy numbers (derived from genotyping data) and the TMB was calculated from the missense mutation calls (derived from WES data). HRDsum and tumor mutational burden (TMB) correlated positively pan-cancer (R=0.42) and within most cancer types. By contrast, HRD was absent in ultra-hypermutated (TMB >= 100 mut/Mb) tumors. Ultra-hypermutation was typically associated with microsatellite instability (MSI) or POLE/POLD1 mutation. Significant positive correlation of the HRDsum and immune cell infiltration was observed pan-cancer, but only within a few cancer types. Significant positive correlation of HRDsum and the T-cell inflamed gene expression profile was observed only in 7 of 32 cancer types: in breast cancer, ovarian cancer, low grade glioma, testicular germ cell tumors and three kinds of kidney cancer. A functional genomics analysis was carried out by correlating genome-wide gene expression data (derived from RNA-Seq) with HRDsum. The resulting lists of significantly correlating genes were analyzed for enrichment of 50 hallmark gene sets (catalog H, MSigDB). We detected simultaneous enrichment of two proliferation-related categories, E2F_TARGETS and G2M_checkpoint, for 13 of the 32 cancer types (ACC, BLCA, BRCA, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, PRAD, SARC, THYM). The study shows that HRD is associated with immunological activation of the tumor microenvironment only in a minority of cancer types. Our data support further studies of immune activating therapies in combination with immune checkpoint blockade in the not intrinsically activated cancer types and advocate to combine different genomic and transcriptomic biomarkers (including HRD and TMB) for comprehensive molecular diagnostics and therapy guidance. Citation Format: Jan Budczies, Klaus Kluck, Susanne Beck, Iordanis Ouralidis, Michael Allgäuer, Michael Menzel, Eugen Rempel, Daniel Kazdal, Lukas Perkhofer, Alexander Kleger, Peter Schirmacher, Thomas Seufferlein, Albrecht Stenzinger. HRD is inversely correlated with MSI and identifies immunologically cold tumors in most cancer types [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 810.

Phase II study of KN046 in patients with thymic carcinoma who failed immune checkpoint inhibitors.

TPS8607 Background: Thymic carcinomas are the most aggressive form of thymic epithelial tumors. They are often not operable and are more resistant to chemotherapy than thymomas. Thymic carcinoma is sensitive to pembrolizumab. However, most patients who respond to pembrolizumab eventually recur. Recently, molecules that combine PD(L)1 and CTLA-4 have been developed for solid tumor patients, with the hope that targeted therapy will be more effective than standard of care. KN046 is a bi-specific antibody against PD-L1 and CTLA-4 with a much higher affinity of the anti-PD-L1 portion and a weaker affinity for anti-CTLA-4, potentially leading to less autoimmune disorders and toxicities. We developed a Phase II study to test the hypothesis that dual PD-L1 and CTLA-4 inhibition with KN046 may represent a safe and tolerable option for patients with advanced thymic carcinoma who have progressed on prior treatment with immune checkpoint inhibitors. Methods: Key eligibility criteria include thymic carcinoma with progression after treatment with an immune checkpoint inhibitor with no limit to prior lines of therapy, adequate organ function and performance status. History of prior or current autoimmune disorders are not allowed and history of baseline positive anti-acetylcholine receptor (AChR) autoantibody are not allowed. KN046 will be administered intravenously at 5 mg/kg every 2 weeks until progression or excessive toxicity for up to 2 years. A cycle is defined as 2 treatments (28 days). The primary objective is to evaluate the antitumor activity of KN046 in patients with thymic carcinoma as measured by overall response rate defined by RECIST 1.1 criteria. The secondary objectives are to assess the safety and tolerability of KN046 including safety as measured by the number of adverse events (CTCAE 5.0), duration of response (RECIST 1.1) from first documented response to the date of first documented disease progression, progression-free survival, and overall survival. Exploratory objectives include the association of biomarkers (PD-L1 expression, tumor immune microenvironment determined by multiplex IHC, tumor mutational burden, T-cell inflamed gene expression profile) and clinical efficacy parameters. We will also characterize the safety laboratory results (AChR autoantibodies and creatinine kinase) and the occurrence of adverse events of interest. Simon’s two-stage design will be used. The null hypothesis that the true response rate is 5% will be tested against a one-sided alternative of target response rate ≥20%. In the first stage, 10 patients will be accrued. If there are no responses in the first stage, then the study will be stopped. Otherwise, 19 additional patients will be accrued for a total of 29 patients. The null hypothesis will be rejected if ≥4 responses are observed in 29 patients, with a type 1 error rate of 0.05 and power of 80%. The study was activated at Weill Cornell Medicine in December 2021. Clinical trial information: NCT04925947.

Immune Biomarkers in Metastatic Castration-resistant Prostate Cancer

BackgroundMetastatic castration-resistant prostate cancer (mCRPC) is a heterogeneous disease in which molecular stratification is needed to improve clinical outcomes. The identification of predictive biomarkers can have a major impact on the care of these patients, but the availability of metastatic tissue samples for research in this setting is limited. ObjectiveTo study the prevalence of immune biomarkers of potential clinical utility to immunotherapy in mCRPC and to determine their association with overall survival (OS). Design, setting, and participantsFrom 100 patients, mCRPC biopsies were assayed by whole exome sequencing, targeted next-generation sequencing, RNA sequencing, tumor mutational burden, T-cell–inflamed gene expression profile (TcellinfGEP) score (Nanostring), and immunohistochemistry for programmed cell death 1 ligand 1 (PD-L1), ataxia-telangiectasia mutated (ATM), phosphatase and tensin homolog (PTEN), SRY homology box 2 (SOX2), and the presence of neuroendocrine features. Outcome measurements and statistical analysisThe phi coefficient determined correlations between biomarkers of interest. OS was assessed using Kaplan-Meier curves and adjusted hazard ratios (aHRs) from Cox regression. Results and limitationsPD-L1 and SOX2 protein expression was detected by immunohistochemistry (combined positive score ≥1 and >5% cells, respectively) in 24 (33%) and 27 (27%) mCRPC biopsies, respectively; 23 (26%) mCRPC biopsies had high TcellinfGEP scores (>–0.318). PD-L1 protein expression and TcellinfGEP scores were positively correlated (phi 0.63 [0.45; 0.76]). PD-L1 protein expression (aHR: 1.90 [1.05; 3.45]), high TcellinfGEP score (aHR: 1.86 [1.04; 3.31]), and SOX2 expression (aHR: 2.09 [1.20; 3.64]) were associated with worse OS. ConclusionsPD-L1, TcellinfGEP score, and SOX2 are prognostic of outcome from the mCRPC setting. If validated, predictive biomarker studies incorporating survival endpoints need to take these findings into consideration. Patient summaryThis study presents an analysis of immune biomarkers in biopsies from patients with metastatic prostate cancer. We describe tumor alterations that predict prognosis that can impact future studies.

Integrative genomic analysis reveals low T-cell infiltration as the primary feature of tobacco use in HPV-positive oropharyngeal cancer

SummaryAlthough tobacco use is an independent adverse prognostic feature in HPV(+) oropharyngeal squamous cell carcinoma (OPSCC), the biologic features associated with tobacco use have not been systematically investigated. We characterized genomic and immunologic features associated with tobacco use through whole exome sequencing, mRNA hybridization, and immunohistochemical staining in 47 HPV(+) OPSCC tumors. Low expression of transcripts in a T cell-inflamed gene expression profile (TGEP) was associated with tobacco use at diagnosis and lower overall and disease-free survival. Tobacco use was associated with an increased proportion of T > C substitutions and a lower proportion of expected mutational signatures, but not with increases in mutational burden or recurrent oncogenic mutations. Our findings suggest that rather than increased mutational burden, tobacco’s primary and clinically relevant association in HPV(+) OPSCC is immunosuppression of the tumor immune microenvironment. Quantitative assays of T cell infiltration merit further study as prognostic markers in HPV(+) OPSCC.

Reverse Translating Molecular Determinants of Anti-Programmed Death 1 Immunotherapy Response in Mouse Syngeneic Tumor Models.

Targeting the programmed death 1/programmed death ligand 1 (PD-1/PD-L1) pathway with immunotherapy has revolutionized the treatment of many cancers. Somatic tumor mutational burden (TMB) and T-cell-inflamed gene expression profile (GEP) are clinically validated pan-tumor genomic biomarkers that can predict responsiveness to anti-PD-1/PD-L1 monotherapy in many tumor types. We analyzed the association between these biomarkers and the efficacy of PD-1 inhibitor in 11 commonly used preclinical syngeneic tumor mouse models using murinized rat anti-mouse PD-1 DX400 antibody muDX400, a surrogate for pembrolizumab. Response to muDX400 treatment was broadly classified into three categories: highly responsive, partially responsive, and intrinsically resistant to therapy. Molecular and cellular profiling validated differences in immune cell infiltration and activation in the tumor microenvironment of muDX400-responsive tumors. Baseline and on-treatment genomic analysis showed an association between TMB, murine T-cell-inflamed gene expression profile (murine-GEP), and response to muDX400 treatment. We extended our analysis to investigate a canonical set of cancer and immune biology-related gene signatures, including signatures of angiogenesis, myeloid-derived suppressor cells, and stromal/epithelial-to-mesenchymal transition/TGFβ biology previously shown to be inversely associated with the clinical efficacy of immune checkpoint blockade. Finally, we evaluated the association between murine-GEP and preclinical efficacy with standard-of-care chemotherapy or antiangiogenic agents that previously demonstrated promising clinical activity, in combination with muDX400. Our profiling studies begin to elucidate the underlying biological mechanisms of response and resistance to PD-1/PD-L1 blockade represented by these models, thereby providing insight into which models are most appropriate for the evaluation of orthogonal combination strategies.

Association between biomarkers and clinical outcomes of lenvatinib + pembrolizumab in advanced renal cell carcinoma (RCC): Results from Study 111/KEYNOTE-146.

375 Background: In the Study 111/KEYNOTE-146 trial (NCT02501096; N=147), lenvatinib (lenva) + pembrolizumab (pembro) showed encouraging antitumor activity and a manageable safety profile in treatment-naive (n=23) or previously treated metastatic RCC (n=105, previously treated with immune checkpoint inhibitor [ICI]; n=19, previously treated ICI naive); 145 had clear cell RCC and 2 had non-clear cell RCC. In this exploratory analysis, we evaluated the association between clinical outcomes and gene expression signatures and DNA variants for individual RCC-specific driver genes of interest based on published reports. Methods: Patients (pts) with metastatic RCC were treated with lenva 20 mg orally once daily + pembro 200 mg intravenously once every 3 weeks. The analysis population included pts with treatment-naive (n=10) and ICI pretreated (n=70) disease with evaluable RNA-sequencing data for the 18-gene T-cell–inflamed gene expression profile (TcellinfGEP) and for 11 other signatures (angiogenesis; glycolysis; gMDSC; hypoxia; mMDSC; MVD; MYC; proliferation; RAS; stroma/EMT/TGFβ; WNT) and whole exome sequencing (WES) data for DNA variants for individual genes ( VHL, PBRM1, BAP1, and SETD2). Specimens were collected prior to the start of treatment. The associations between each signature score and ORR and PFS per immune-related RECIST were evaluated using logistic regression and Cox proportional hazards, respectively. One-sided P values for TcellinfGEP (hypothesized positive association) and two-sided P values for all other signatures (no hypothesized association) were adjusted for multiplicity using the Hochberg step-up procedure; significance was prespecified at α=0.05. The association between DNA variants for individual genes and ORR was evaluated descriptively. Clinical data cutoff was August 18, 2020. Results: Of 147 treated pts, RNA sequencing and WES data were available for 80 (54%) and 60 (41%), respectively. TcellinfGEP was not associated with ORR ( P=0.827) or PFS ( P=0.741), nor were the other 11 signatures before or after adjustment for TcellinfGEP. ORR for DNA variants reported in the table. Conclusions: In this exploratory analysis of pts with metastatic RCC enrolled in Study 111/KEYNOTE-146 treated with lenva + pembro, responses were observed regardless of biomarker status. There were no statistically significant associations between gene signatures and clinical outcomes. Clinical benefit was observed regardless of VHL, PBRM1, BAP1, or SETD2 mutation status. Analyses in larger randomized datasets will provide additional information on the role of biomarkers in RCC. Clinical trial information: NCT02501096. [Table: see text]

Transcriptomic Determinants of Response to Pembrolizumab Monotherapy across Solid Tumor Types.

To explore relationships between biological gene expression signatures and pembrolizumab response. RNA-sequencing data on baseline tumor tissue from 1,188 patients across seven tumor types treated with pembrolizumab monotherapy in nine clinical trials were used. A total of 11 prespecified gene expression signatures [18-gene T-cell-inflamed gene expression profile (TcellinfGEP), angiogenesis, hypoxia, glycolysis, proliferation, MYC, RAS, granulocytic myeloid-derived suppressor cell (gMDSC), monocytic myeloid-derived suppressor cell (mMDSC), stroma/epithelial-to-mesenchymal transition (EMT)/TGFβ, and WNT] were evaluated for their relationship to objective response rate (per RECIST, version 1.1). Logistic regression analysis of response for consensus signatures was adjusted for tumor type, Eastern Cooperative Oncology Group performance status, and TcellinfGEP, an approach equivalent to evaluating the association between response and the residuals of consensus signatures after detrending them for their relationship with the TcellinfGEP (previously identified as a determinant of pembrolizumab response) and tumor type. Testing of the 10 prespecified non-TcellinfGEP consensus signatures for negative association [except proliferation (hypothesized positive association)] with response was adjusted for multiplicity. Covariance patterns of the 11 signatures (including TcellinfGEP) identified in Merck-Moffitt and The Cancer Genome Atlas datasets showed highly concordant coexpression patterns in the RNA-sequencing data from pembrolizumab trials. TcellinfGEP was positively associated with response; signatures for angiogenesis, mMDSC, and stroma/EMT/TGFβ were negatively associated with response to pembrolizumab monotherapy. These findings suggest that features beyond IFNγ-related T-cell inflammation may be relevant to anti-programmed death 1 monotherapy response and may define other axes of tumor biology as candidates for pembrolizumab combinations. See related commentary by Cho et al., p. 1479.

FDA-Approved and Emerging Next Generation Predictive Biomarkers for Immune Checkpoint Inhibitors in Cancer Patients.

A patient’s response to immune checkpoint inhibitors (ICIs) is a complex quantitative trait, and determined by multiple intrinsic and extrinsic factors. Three currently FDA-approved predictive biomarkers (progra1mmed cell death ligand-1 (PD-L1); microsatellite instability (MSI); tumor mutational burden (TMB)) are routinely used for patient selection for ICI response in clinical practice. Although clinical utility of these biomarkers has been demonstrated in ample clinical trials, many variables involved in using these biomarkers have poised serious challenges in daily practice. Furthermore, the predicted responders by these three biomarkers only have a small percentage of overlap, suggesting that each biomarker captures different contributing factors to ICI response. Optimized use of currently FDA-approved biomarkers and development of a new generation of predictive biomarkers are urgently needed. In this review, we will first discuss three widely used FDA-approved predictive biomarkers and their optimal use. Secondly, we will review four novel gene signature biomarkers: T-cell inflamed gene expression profile (GEP), T-cell dysfunction and exclusion gene signature (TIDE), melanocytic plasticity signature (MPS) and B-cell focused gene signature. The GEP and TIDE have shown better predictive performance than PD-L1, and PD-L1 or TMB, respectively. The MPS is superior to PD-L1, TMB, and TIDE. The B-cell focused gene signature represents a previously unexplored predictive biomarker to ICI response. Thirdly, we will highlight two combined predictive biomarkers: TMB+GEP and MPS+TIDE. These integrated biomarkers showed improved predictive outcomes compared to a single predictor. Finally, we will present a potential nucleic acid biomarker signature, allowing DNA and RNA biomarkers to be analyzed in one assay. This comprehensive signature could represent a future direction of developing robust predictive biomarkers, particularly for the cold tumors, for ICI response.

YAP1 transcription factor expression to define subsets of cancers with T-cell inflamed phenotype in a pan-tumor analysis across 33 tumor types.

2610 Background: Immune checkpoint blockade (ICB) has become an established treatment option for the majority of solid tumors. PD-L1 expression, tumor mutation burden (TMB) and mismatch repair (MMR) deficiency are established but suboptimal predictive markers to select patients for ICB therapy. The identification of better predictive biomarkers to complement existing biomarkers is an area of need. We previously identified Yes Associated Protein 1 (YAP1) gene expression as a marker of inflamed tumor phenotype in small cell lung cancer. We sought to elucidate the role of YAP1 as a tumor agnostic biomarker of inflamed tumor phenotype. Methods: We obtained the publicly available TCGA normalized RSEM expression data (version dated September 29, 2019) from the GDC legacy archive (https://portal.gdc.cancer.gov/legacy-archive) for this analysis. We used the shinySISPA method ( https://bbisr.shinyapps.winship.emory.edu/shinySISPA/ ) to classify the primary tumor samples into YAP1 high or low on the basis of the normalized expression profile for YAP1 gene. The T-cell–inflamed gene expression profile score for each sample was calculated as weighted sum of the normalized expression values of the 18 genes ( PSMB10, HLA-DQA1, HLA-DRB1, CMKLR1, HLA-E, NKG7, CD8A, CCL5, CXCL9, CD27, CXCR6, IDO1, STAT1, TIGIT, LAG3, CD274, PDCD1LG2, CD276) as originally published in the Patent filed under “WO201609437” and validated as a predictor of clinical efficacy in patients treated with anti PD1 immunotherapy drug, pembrolizumab. Results: A total of 11283 samples from 33 different histologic tumor types contained in the TCGA database were included in this analysis. A small but meaningful subset of cases 271 (2%) were classified as YAP1 high with a wide range in proportion of YAP1 high tumors of 0.67% to 12.09% across the 33 histologic tumor types. Adrenocortical, cholangiocarcinoma, chromophobe RCC, DLBCL, and mesothelioma had a high rate of YAP1 high tumors (> 10% of cases). Overall trend showed a higher median interferon gamma GEP score in YAP1 high versus YAP1 low tumors with a GEP score of 11.2 vs. 10.7 respectively. A minority of histologic tumor subtypes (HNSCC, clear cell RCC, sarcoma, uterine carcinosarcoma, testicular GCT, melanoma, mesothelioma and ovarian cancer) showed a reverse trend of lower GEP score in association with YAP1 high status. Ongoing validation of these findings in an independent cohort of clinical samples and correlation of YAP1 status with other predictive biomarkers (TMB, PD-L1 and MSS status) will be presented at the meeting. Conclusions: YAP1 is highly expressed in a small but meaningful subsets of cancers and is associated with the inflamed phenotype in the majority of cancers. YAP1 expression is most common in rarer tumor types and in histologic types where currently available biomarkers have not been shown to predict the benefit of ICB.

Impact of PD-L1 and T-cell inflamed gene expression profile on survival in advanced ovarian cancer

ObjectiveProgrammed death ligand 1 (PD-L1) expression affects tumor evasion of immune surveillance. The prognostic value and relationship of PD-L1 expression to T-cell–inflamed immune signatures in ovarian cancer are unclear. The...

YAP1 positive small-cell lung cancer subtype is associated with the T-cell inflamed gene expression profile and confers good prognosis and long term survival.

9019 Background: The dominant expression of transcription factors ASCL1, NeuroD1, YAP1 or POU2F3 characteristically defines four small cell lung cancer (SCLC) subtypes (SCLC-A, SCLC-N, SCLC-Y and SCLC-P). The clinical validation and biological relevance of these emerging SCLC subtypes is currently lacking. Methods: Using the Illumina TruSeq RNA Exome Kit, we generated RNA-Seq data from 61 cases of SCLC and pulmonary carcinoid to interrogate gene expression differences in SCLC subtypes as well as in survival outliers (top and bottom decile) matched for clinically relevant prognostic factors and treatment. We also assessed YAP1 protein expression in a blinded fashion by immunohistochemistry in 130 SCLC cases. Results: We successfully classified 68% of SCLC into one of the four SCLC subtypes whereas 81.5% of carcinoids did not fit into any of these categories. GSEA for differentially expressed genes between outlier subgroups showed significant upregulation of interferon gamma and interferon alpha response genes in late survivors. Moreover, a previously validated 18-gene T-cell inflamed gene expression profile was upregulated in late survivors and in the SCLC-Y subtype. Furthermore, the SCLC-Y subtype and late survivors showed higher expression of HLA gene family and reduced expression of cancer testis antigens. The median (95%CI) OS was 14 (4.3, 28.8), 16.7 (0.9, NA), 8.1 (2, 9.7) and 20.1 (0.6, 39.5) months respectively, for SCLC-A, N, P and Y subtypes. YAP-1 protein expression was positive in 17 of 130 (13%) SCLC cases. The majority of cases with positive YAP1 expression by immunohistochemistry, 12 of 17 cases (70.6%), were limited stage SCLC at the time of original diagnosis. Conclusions: SCLC subtypes have clinical implication as predictive and prognostic biomarker. SCLC-Y subtype is enriched for T-cell inflamed phenotype and long term survival, and may predict for clinical benefit of immunotherapy.

Pembrolizumab (pembro) plus olaparib in patients (pts) with docetaxel-pretreated metastatic castration-resistant prostate cancer (mCRPC): KEYNOTE-365 cohort A efficacy, safety, and biomarker results.

5544 Background: Pembro + olaparib has shown antitumor activity and acceptable safety in docetaxel-pretreated pts with mCRPC enrolled in cohort A of the phase I/II KEYNOTE-365 study (NCT02861573). Updated results with new biomarker data are reported. Methods: Pts with docetaxel-pretreated mCRPC who progressed within 6 mo of screening received pembro 200 mg IV Q3W + olaparib 400-mg capsule or 300-mg tablet PO BID. Pts might have received 1 other chemotherapy and ≤2 second-generation androgen-receptor targeted therapies. Primary end points: PSA response rate (decrease ≥50% from baseline, confirmed by a second value ≥3 wks later), ORR per RECIST v1.1, and safety. Key secondary end points: DCR, DOR, rPFS, and OS. Biospecimens (eg, blood, tissue) were collected for biomarker analysis (tissue PD-L1 expression, androgen receptor variant 7 [AR-v7] expression in circulating tumor cells [CTCs], and a T-cell–inflamed gene expression profile [GEP]). ctDNA was analyzed by Guardant Health 360 (GH360) and Omni (GH Omni) assays. FFPE tissue was analyzed by FoundationOne CDx (F1CDx) assay. Results: 84 of 87 enrolled pts were treated; 48/84 (57.1%) had measurable disease. Median (range) time from enrollment to data cutoff was 3.6 mo (0.0-29.2) for all pts and 26.7 mo (21.2-29.2) for 41 pts with ≥27 wks’ follow-up. Confirmed PSA response rate was 9% (95% CI, 3.5-16.8) in 82 pts with a baseline PSA assessment. Median time to PSA progression: 3.8 mo (95% CI, 2.9-4.4). In 24 pts with measurable disease and ≥27 wks’ follow-up, ORR was 8.3% (95% CI, 1.0-27.0; 2 PRs) and DCR ≥6 mo was 20.8% (95% CI, 7.1-42.2). Median (range) DOR was NR (12.0+ to 21.4+ mo); 2 pts had DOR ≥12 mo. In all pts, median rPFS was 4.3 mo (95% CI, 3.4-7.7) and median OS was 14.4 mo (95% CI, 8.1-18.5). Grade ≥3 TRAEs occurred in 29 pts (35%); 2 pts died of TRAEs (1 myocardial infarction, 1 unknown). Overall, 26% had PD-L1+ tumors (combined positive score ≥1). Of 31 pts with CTC data, 12.9% were AR-v7+. No BRCA1/2 mutation was detected by GH360 (n=42). Of 57 pts analyzed by GH Omni, 2 had BRCA2 mutations, 1 had a BRCA1 mutation, 4 had ATM mutations, 1 had a CHEK1 mutation, and 6 had CDK12 mutations. Of 49 pts analyzed by F1CDx, 4 had BRCA mutations; 1 pt had a copy number loss mutation not detected by ctDNA analysis. GEP was not associated with ORR or PSA response. Conclusions: Pembro + olaparib continued to show activity and acceptable safety in pts with docetaxel-pretreated mCRPC. A phase III study of this combination is ongoing (KEYLYNK-010, NCT03834519). Clinical trial information: NCT02861573 .

Association of gene expression with clinical outcomes in patients with renal cell carcinoma treated with pembrolizumab in KEYNOTE-427.

5024 Background: We assessed the association of baseline RNA-sequencing–based gene expression signatures and DNA alterations with response or resistance to pembrolizumab in patients with advanced renal cell carcinoma in cohorts A (clear cell; n = 110) and B (non-clear cell; n = 165) of the phase 2 KEYNOTE-427 study (NCT02853344). Methods: Using RNA-sequencing, we analyzed the association of gene expression signatures (18-gene T-cell–inflamed gene expression profile [GEP]; 10 non–T-cell–inflamed GEP canonical signatures [angiogenesis, gMDSC, glycolysis, hypoxia, mMDSC, MYC, proliferation, RAS, stromal/EMT/TGFβ, WNT]) quantifying tumor microenvironment elements (TME) with objective response rate (ORR) and progression-free survival (PFS). Canonical signatures were derived from 2 databases (TCGA, Moffit) using an algorithm that included genes based on their correlation to reference signatures in the literature. Signature definitions were finalized before linking to the clinical data, and significance was prespecified at 0.10 given the potential for limited power. Canonical signatures were analyzed through regression testing of response for association with consensus signatures after adjusting for T-cell–inflamed GEP and International Metastatic RCC Database Consortium scores in the model. P values were adjusted for multiplicity. Using whole exome sequencing, we also summarized the association of renal cell carcinoma driver gene mutations with ORR. Clinical data cutoff: Jan 30, 2019. Results: Patient characteristics for this analysis were comparable to the overall population. In cohort A, T-cell–inflamed GEP (n = 78) was statistically significantly associated with a better ORR ( P = 0.021; AUROC = 0.65) but not PFS ( P = 0.116). No other TME canonical signatures showed a correlation with ORR or PFS. ORR was estimated for mutations (Table). Conclusions: RNA-sequencing–based, T-cell–inflamed GEP was associated with ORR in patients with clear cell renal cell carcinoma receiving first-line pembrolizumab. Precision was limited by sample size for estimating ORR by specific gene mutation status. Evaluation of tissue-based biomarkers in larger studies are planned. Biomarker analyses from patients in cohort B will also be presented. Clinical trial information: NCT02853344 . [Table: see text]

T-cell–inflamed gene expression profile (GEP) and PD-L1 expression in patients (pts) with esophageal cancer (EC).

26 Background: GEP and PD-L1 expression have been associated with anti–PD-1/PD-L1 therapy. In this retrospective observational study we explored the prognostic value of GEP and PD-L1 expression in pts with EC receiving standard-of-care therapy (SOC). Methods: Tumor tissue samples collected from 2005 to 2017 were procured from Yonsei Cancer Center (South Korea), Memorial Sloan Kettering Cancer Center (USA) and Aarhus University Hospital (Denmark). GEP score was derived from an 18-gene signature using extracted tumor RNA analyzed by NanoString nCounter; GEP high/intermediate (GEP-H/I) and low were defined by a cutoff of –1.540, consistent with pembrolizumab clinical trials. PD-L1 expression was assessed by PD-L1 IHC 22C3 pharmDx assay (Agilent); positive was defined as combined positive score (CPS) ≥ 10, where CPS is the the number of PD-L1–positive cells (tumor cells, lymphocytes and macrophages) divided by the total number of viable tumor cells, multiplied by 100. Associations of GEP score and PD-L1 expression with clinicopathologic variables were analyzed by chi-square test and multiple logistic regression models. Overall survival (OS) from diagnosis date to death date/last follow-up was analyzed using Cox proportional hazards models adjusting for age, sex, stage, region and ECOG performance status (PS). Results: 294 samples with both PD-L1 and GEP data were analyzed. Median age was 65 y (range 33-88); 85% were from men, 58% were stage IV, 63% were esophageal adenocarcinoma (EAC) and 37% were esophageal squamous cell carcinoma (ESCC). Overall 36% of tumors were GEP-H/I: 46% in EAC vs 18% in ESCC. GEP was not associated with OS overall (adjusted hazard ratio [aHR] –0.90; 95% CI 0.68-1.18) or in pts with EAC (aHR 0.93; 95% CI 0.68-1.27) or ESCC (aHR 0.76; 95% CI 0.40-1.44). 21% of tumors were PD-L1-CPS ≥ 10: 18% in EAC and 26% in ESCC. PD-L1 expression was associated with ECOG PS (adjusted odds ratio 0.520; 95% CI 0.309-0.875; P = 0.014) but was not associated with OS overall (aHR 0.89; 95% CI 0.64-1.24) or in pts with EAC (aHR 0.97; 95% CI 0.63-1.49) or ESCC (aHR 1.31; 95% CI 0.73-2.34). Conclusions: Our results suggest that T-cell–inflamed GEP and PD-L1 expression may not be prognostic in pts with EC who received SOC.

T-Cell–Inflamed Gene-Expression Profile, Programmed Death Ligand 1 Expression, and Tumor Mutational Burden Predict Efficacy in Patients Treated With Pembrolizumab Across 20 Cancers: KEYNOTE-028

Biomarkers that can predict response to anti-programmed cell death 1 (PD-1) therapy across multiple tumor types include a T-cell-inflamed gene-expression profile (GEP), programmed death ligand 1 (PD-L1) expression, and tumor mutational burden (TMB). Associations between these biomarkers and the clinical efficacy of pembrolizumab were evaluated in a clinical trial that encompassed 20 cohorts of patients with advanced solid tumors. KEYNOTE-028 ( ClinicalTrials.gov identifier: NCT02054806) is a nonrandomized, phase Ib trial that enrolled 475 patients with PD-L1-positive advanced solid tumors who were treated with pembrolizumab 10 mg/kg every 2 weeks for 2 years or until confirmed disease progression or unacceptable toxicity occurred. The primary end point was objective response rate (ORR; by RECIST v1.1, investigator review). Secondary end points included safety, progression-free survival (PFS), and overall survival (OS). Relationships between T-cell-inflamed GEP, PD-L1 expression, and TMB and antitumor activity were exploratory end points. ORRs (with 95% CIs) ranged from 0% (0.0% to 14.2%) in pancreatic cancer to 33% (15.6% to 55.3%) in small-cell lung cancer. Across cohorts, median (95% CI) PFS ranged from 1.7 months (1.5 to 2.9 months) to 6.8 months (1.9 to 14.1 months) in pancreatic and thyroid cancers, respectively, and median OS from 3.9 months (2.8 to 5.5 months) to 21.1 months (9.1 to 22.4 months) in vulvar and carcinoid tumors, respectively. Higher response rates and longer PFS were demonstrated in tumors with higher T-cell-inflamed GEP, PD-L1 expression, and/or TMB. Correlations of TMB with GEP and PD-L1 were low. Response patterns indicate that patients with tumors that had high levels of both TMB and inflammatory markers (GEP or PD-L1) represent a population with the highest likelihood of response. Safety was similar and consistent with prior pembrolizumab reports. A T-cell--inflamed GEP, PD-L1 expression, and TMB predicted response to pembrolizumab in multiple tumor types. These biomarkers (alone/in combination) may help identify patients who have a higher likelihood of response to anti-PD-1 therapies across a broad spectrum of cancers.