PD-L1 Amplification Research Articles

Prognostic relevance of integrated genetic profiling in adult T-cell leukemia/lymphoma

AbstractAdult T-cell leukemia/lymphoma (ATL) is a heterogeneous group of peripheral T-cell malignancies characterized by human T-cell leukemia virus type-1 infection, whose genetic profile has recently been fully investigated. However, it is still poorly understood how these alterations affect clinical features and prognosis. We investigated the effects of genetic alterations commonly found in ATL on disease phenotypes and clinical outcomes, based on genotyping data obtained from 414 and 463 ATL patients using targeted-capture sequencing and single nucleotide polymorphism array karyotyping, respectively. Aggressive (acute/lymphoma) subtypes were associated with an increased burden of genetic and epigenetic alterations, higher frequencies of TP53 and IRF4 mutations, and many copy number alterations (CNAs), including PD-L1 amplifications and CDKN2A deletions, compared with indolent (chronic/smoldering) subtypes. By contrast, STAT3 mutations were more characteristic of indolent ATL. Higher numbers of somatic mutations and CNAs significantly correlated with worse survival. In a multivariate analysis incorporating both clinical factors and genetic alterations, the Japan Clinical Oncology Group prognostic index high-risk, older age, PRKCB mutations, and PD-L1 amplifications were independent poor prognostic factors in aggressive ATL. In indolent ATL, IRF4 mutations, PD-L1 amplifications, and CDKN2A deletions were significantly associated with shorter survival, although the chronic subtype with unfavorable clinical factors was only marginally significant. Thus, somatic alterations characterizing aggressive diseases predict worse prognosis in indolent ATL, among which PD-L1 amplifications are a strong genetic predictor in both aggressive and indolent ATL. ATL subtypes are further classified into molecularly distinct subsets with different prognosis. Genetic profiling might contribute to improved prognostication and management of ATL patients.

Clinical significance of PD-1/PD-Ls gene amplification and overexpression in patients with hepatocellular carcinoma.

Background: The remarkable clinical activity of PD-1 antibody in advanced hepatocellular carcinoma (HCC) highlights the importance of PD-1/PD-L1-mediated immune escape as therapeutic target in HCC. However, the frequency and prognostic significance of PD-Ls genetic alterations in HCC remain unknown.Methods: Fluorescence in situ hybridization were used to determine PD-Ls genetic alterations, and qPCR data coupled with immunofluorescence were used to measure the mRNA and protein levels of PD-Ls. Clinical relevance and prognostic value of 9p24.1 genetic alterations were investigated on tissue microarray containing three independent cohorts of 578 HCC patients. The results were further validated in an independent cohort of 442 HCC patients from The Cancer Genome Atlas (TCGA) database.Results: In total, 7.1%-15.0% for amplification and 15.8%-31.3% for polysomy of 9p24.1 were revealed in three cohorts of HCC patients, similar to the objective response rate of PD-1 antibody in HCC. Patients with 9p24.1 genetic alterations significantly and independently correlated with unfavorable outcomes than those without. FISH and qPCR data coupled with immunofluorescence revealed that genetic alterations of 9p24.1 robustly contributed to PD-L1 and PD-L2 upregulation. In addition, increased expression of PD-L1 instead of PD-L2 also predicted poor survival by multivariate analyses. Meanwhile, high infiltration of PD-1+ immune cells also indicated dismal survival in HCC.Conclusions: Amplification or higher expression of PD-L1 significantly and independently correlated with unfavorable survival in HCC patients, authenticating the PD-1/PD-L1 axis as rational immunotherapeutic targets for HCC.

Development and validation of a novel clinical fluorescence in situ hybridization assay to detect JAK2 and PD-L1 amplification: a fluorescence in situ hybridization assay for JAK2 and PD-L1 amplification

The amplification of chromosome 9p24.1 encoding PD-L1, PD-L2, and JAK2 has been reported in multiple types of cancer and is associated with poor outcome, upregulation of PD-L1, and activation of the JAK/STAT pathway. We have developed a novel fluorescence in situ hybridization assay which combines 3 probes mapping to 9p24.1 with a commercial chromosome 9 centromere (CEN9) probe for detection of the JAK2/9p24.1 amplification. JAK2 fluorescence in situ hybridization was compared with array-based comparative genomic hybridization in 34 samples of triple negative breast cancer tumor. By array-based comparative genomic hybridization, 15 had 9p24.1 copy-number gain (log2ratio>0.3) and 19 were classified as non-gain (log2ratio≤0.3). Copy-number gain was defined as JAK2/CEN9 ratio ≥1.1 or average JAK2 signals≥3.0. Twelve of 15 samples with copy-number gain by array-based comparative genomic hybridization were also detected by fluorescence in situ hybridization. Eighteen of 19 samples classified as copy-number non-gain by array-based comparative genomic hybridization were concordant by array-based comparative genomic hybridization. The sensitivity and specificity of the fluorescence in situ hybridization assay was 80% and 95%, respectively (P=0.02). The sample with the highest level of amplification by array-based comparative genomic hybridization (log2ratio=3.6) also scored highest by fluorescence in situ hybridization (ratio=8.2). There was a correlation between the expression of JAK2 and amplification status (Mean 633 vs 393, P=0.02), and there was a trend of association with PD-L1 RNA expression (Mean 46 vs 22, P=0.11). No significant association was observed between PD-L1 immunohistochemistry expression and copy-number gain status. In summary, the novel array-based comparative genomic hybridization assay for detection of chromosome 9p24.1 strongly correlates with the detection of copy-number gain by array-based comparative genomic hybridization. In triple negative breast cancer, this biomarker may identify a relevant subset of patients for targeted molecular therapies.

Immune Escape in Breast Cancer During In Situ to Invasive Carcinoma Transition.

To investigate immune escape during breast tumor progression, we analyzed the composition of leukocytes in normal breast tissues, ductal carcinoma in situ (DCIS), and invasive ductal carcinomas (IDC). We found significant tissue and tumor subtype-specific differences in multiple cell types including T cells and neutrophils. Gene expression profiling of CD45+CD3+ T cells demonstrated a decrease in CD8+ signatures in IDCs. Immunofluorescence analysis showed fewer activated GZMB+CD8+ T cells in IDC than in DCIS, including in matched DCIS and recurrent IDC. T-cell receptor clonotype diversity was significantly higher in DCIS than in IDCs. Immune checkpoint protein TIGIT-expressing T cells were more frequent in DCIS, whereas high PD-L1 expression and amplification of CD274 (encoding PD-L1) was only detected in triple-negative IDCs. Coamplification of a 17q12 chemokine cluster with ERBB2 subdivided HER2+ breast tumors into immunologically and clinically distinct subtypes. Our results show coevolution of cancer cells and the immune microenvironment during tumor progression.Significance: The design of effective cancer immunotherapies requires the understanding of mechanisms underlying immune escape during tumor progression. Here we demonstrate a switch to a less active tumor immune environment during the in situ to invasive breast carcinoma transition, and identify immune regulators and genomic alterations that shape tumor evolution. Cancer Discov; 7(10); 1098-115. ©2017 AACR.See related commentary by Speiser and Verdeil, p. 1062This article is highlighted in the In This Issue feature, p. 1047.

Checkpoint inhibitor is active against large cell neuroendocrine carcinoma with high tumor mutation burden

BackgroundLarge cell neuroendocrine tumor (LCNEC) of the lung is a rare and aggressive tumor similar to small cell lung cancer (SCLC). Thus, it is often treated similarly to SCLC in the front-line setting with a platinum doublet. However, treatment for patients beyond the first line remains undefined.Case presentationWe report the case of a patient with stage IB LCNEC (PD-L1 negative but positive for PD-L1 amplification and tumor mutation burden high) who progressed after adjuvant chemotherapy after surgery and subsequent therapy with an antibody drug conjugate targeting a neuroendocrine-specific cell surface marker but achieved a significant and durable response with pembrolizumab, a humanized IgG4 monoclonal anti-PD-1 antibody.ConclusionsImmunotherapy with checkpoint inhibitors is an effective treatment option for patients with metastatic LCNEC, even if PD-L1 expression is negative.

Abstract 3698: Associations of the immune microenvironment with PD-L1 copy number alterations and PD-L1 expression in resected non-small cell lung cancer

Abstract Background: The relationships between tumor PD-L1 expression and the tumor immune microenvironment (TIM) have gained attention in a variety of cancers including non-small cell lung cancer (NSCLC). However, the relevance of TIM according to the tumor PD-L1 copy number alterations (CNAs) whose significance have been shown in several malignancies remains to be uncovered. Methods: We evaluated PD-1+, CD8+, and Foxp3+ lymphocytes in tumor nest and surrounding stroma separately using microarrayed 636 resected NSCLC. Their associations with PD-L1 CNAs and PD-L1 expression were analyzed. Tumor-infiltrating lymphocytes (TILs) were counted under high-power fields (×200). PD-L1 CNAs were assessed with fluorescent in situ hybridization, and were classified into three categories (amplification, polysomy, and disomy). PD-L1 expression was evaluated using immunohistochemistry (clone E1L3N), and positivity of ≥5% was used as the cut-off. Results: In the entire cohort, the median age was 68 years (23-88 years). The cohort comprised 444 (68.0%) men and 197 (30.2%) never-smokers. The median follow-up duration was 3.6 years. The numbers of TILs in stroma were much greater than those in tumor nest. Smoking history and squamous histology were significantly associated with immune cell infiltration in both stroma and tumor nest, except for CD8+ TILs in stroma. There were 20 tumors with PD-L1 amplification (1.7%) and 84 tumors with polysomy (13.2%). PD-L1 expression was positive in 201 tumors (30.8%). In tumor nest, the numbers of individual TILs significantly increased according to the increase of PD-L1 copy number (Kruskal-Wallis, p < 0.001 for PD-1+, p < 0.001 for CD8+, and p = 0.0087 for Foxp3+). In stroma, only PD1+ TILs were significantly associated with PD-L1 CNAs (p = 0.0032), and CD8+ TILs tended to correlate with PD-L1 CNAs (p = 0.079). The numbers of CD8+, PD-1+, and FOXP3+ TILs were significantly greater in PD-L1-positive tumors than in PD-L1-negative tumors, in both tumor nest and stroma (p < 0.001). Neither kind of TILs, PD-L1 expression, nor PD-L1 CNAs alone was independent prognostic factor for overall survival in a multivariate analysis with Cox hazards models. However, when the PD-L1 copy numbers were combined with Foxp3+ TILs, cases with tumor high PD-L1 copy numbers (≥5 copies per nucleus) and high Foxp3+ TILs had independently worse prognoses in both tumor nest and stroma infiltration (hazard ratio (HR), 4.38; 95% confidence interval, 2.26-8.49, for tumor nest model. HR, 1.96; 95% CI, 1.04-3.68, for stroma model). Conclusion: CD8+, Foxp3+, and PD-L1+ lymphocytes in tumor nest were significantly associated with PD-L1 CNAs and PD-L1 expression. Cases with both increased PD-L1 copy numbers and Foxp3+ TILs have independently worse survival outcomes. Additional data of other TIL subgroups related to PD-L1 CNAs and expression will be discussed. Citation Format: Katsuhiro Yoshimura, Yusuke Inoue, Tomoaki Kahyo, Akikazu Kawase, Masayuki Tanahashi, Hiroshi Ogawa, Naoki Inui, Kazuhito Funai, Kazuya Shinmura, Hiroshi Niwa, Takafumi Suda, Haruhiko Sugimura. Associations of the immune microenvironment with PD-L1 copy number alterations and PD-L1 expression in resected non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3698. doi:10.1158/1538-7445.AM2017-3698

PD-L1 protein expression in tumour cells and immune cells in mismatch repair protein-deficient and -proficient colorectal cancer: the foundation study using the SP142 antibody and whole section immunohistochemistry

AimsRoutine application of PD-L1 immunohistochemistry (IHC) in colorectal cancer (CRC) is limited due to lack of standardized scoring criteria, antibody clones, and intratumoral staining heterogeneity. We assessed PD-L1 protein expression...

The conundrum of the Epstein-Barr virus-associated gastric carcinoma in the Americas.

Epstein-Barr virus-associated gastric carcinoma shows a higher prevalence in the Americas than Asia. We summarize all studies of Epstein Barr virus-associated gastric carcinoma in the Americas, focusing on host characteristics, environmental associations and phylogeographic diversity of Epstein-Barr virus strains. In the Americas, the prevalence of Epstein Barr virus-associated gastric carcinoma is 11.4%, more frequent in males and portray predominantly diffuse-type histology. EBERs, EBNAs, BARTs and LMP are the highest expressed genes; their variations in healthy individuals may explain the phylogeographic diversity of Epstein-Barr virus across the region. Gastric cancer cases harbor exclusively the western genotype (subtype D and kept Xho I site), suggesting a disrupted co-evolution between the pathogen and its host. Epstein-Barr virus-associated gastric carcinoma molecular subtype cases from The Cancer Genome Atlas display PIK3CA gene mutations, amplification of JAK2, PD-L1 and PD-L2 and CpG island methylator phenotype, leading to more extensive methylation of host and viral genomes than any other subtypes from the study. Environmental conditions include negative- and positive- associations with being firstborn child and smoking, respectively. A marginal association with H. pylori has also been reported. Lymphoepithelioma-like carcinoma is associated with Epstein Barr virus in 80%–86% of cases, most of which have been included as part of Epstein Barr virus-associated gastric carcinoma series (prevalence 1.1%–7.6%). Whether these cases represent a variant of Epstein-Barr virus-associated gastric carcinoma is discussed. We propose novel research strategies to solve the conundrum of the high prevalence of Epstein-Barr virus-associated gastric carcinoma in the Americas.

Mutation patterns in genes encoding interferon signaling and antigen presentation: A pan-cancer survey with implications for the use of immune checkpoint inhibitors.

Blockade of immune checkpoints has become a powerful tool in cancer medicine, which is effective across various solid cancer types and hematologic malignancies. While immunohistochemical detection of PD-L1 expression in tumor cells, immune cells, or both has been introduced as predictive biomarker in several clinical trials, shortcomings and limitations of this approach were quickly recognized. As a single biomarker is unlikely to adequately reflect the complex interplay between immune cells and cancer, various genetic determinants of therapy success, including microsatellite instability, mutational burden, and PD-L1 amplification, are being investigated. Very recent work indicates that mutations in B2M, JAK1, and JAK2 render melanoma resistant to immune checkpoint blockade, thus serving as negative response predictors. Using the TCGA dataset, we performed a pan-cancer analysis of potentially damaging mutations in key genes implicated in antigen presentation and interferon-gamma signaling and investigated associations with transcript levels of immune checkpoint genes, cytolytic activity, and mutational burden. For B2M, JAK1, and JAK2, we observed overall mutation frequencies of 1.8%, 2%, and 2.6%, respectively, and found significant associations with mutational burden. On pathway level, melanoma as well as bladder, gastric, and lung cancer were most frequently affected by putative resistance mutations with mutation rates of 27%-50% in the antigen presentation pathway and of 16%-21% in the interferon signaling pathway. Our analysis suggests that a significant number of tumors harbor mutations that may negatively interfere with immune checkpoint inhibition, or confer a higher likelihood of resistance for which a second hit is ultimately required. Since these mutations are prevalent in treatment-naïve tumors, genetic screening prior to therapy might complement current approaches at predicting response to immune checkpoint blockade.

Noninvasive detection of clinically relevant copy number alterations in diffuse large B-cell lymphoma.

7507 Background: Somatic copy number alterations (SCNAs) are common and clinically important genomic events in lymphomas. For example, MYC and BCL2 amplifications are associated with adverse outcomes (Quesada, ASH 2016), while PD-L1 ( CD274) amplifications are associated with improved response to checkpoint inhibitors (Ansell, NEJM 2015). However, noninvasive detection of these events from circulating tumor DNA (ctDNA) remains difficult. Using CAPP-Seq, a targeted high-throughput sequencing platform, we developed a method to profile both focal and broad SCNAs from plasma. Methods: We profiled plasmas from a cohort of 75 pretreatment diffuse large B-cell lymphoma patients and 48 healthy controls. Focal SCNAs were evaluated at ultra-high depths (~10,000x), allowing for detection of lesions at ~1% ctDNA fraction. Thresholds were tuned to allow a false positive rate of 1%, which was empirically validated in an independent healthy cohort (n = 15), yielding a panel-wide false discovery rate of ~2.3% (0% in our genes of interest). Sequencing reads outside the targeted regions were separately pooled and analyzed to evaluate arm and chromosome level SCNAs. Results: We detected SCNAs in clinically relevant genes at the frequencies reported in literature, including amplifications in MYC (8.0%), BCL2 (24.0%), and BCL6 (14.7%) and deletions in TP53 (13.3%) and CDKN2A (9.3%). Remarkably, 26.7% of the cohort demonstrated amplification of both PD-L1 and PD-L2 ( PDCD1LG2). Furthermore, we discovered amplifications in PD-L2, but not PD-L1, in 13.3% of our patients. Interestingly, PD-L1 amplifications were more common in patients with relapsed lymphoma than in those with treatment-naïve disease (43.5% vs 19.2%, p = 0.02). Most PD-L1 amplifications were focal (65%) while the remainder typically involved > 80% of Chr9p. Corresponding tissue profiling data is in progress and will also be presented. Conclusions: Noninvasive sampling of lymphoma ctDNA enables detection of both focal and broad SCNAs, including amplifications of MYC, BCL2, and PD-L1. The ability to noninvasively profile copy number altered regions allows for biopsy-free discovery of clinically significant structural alterations in lymphoma patients.

Genomic profiling of squamous malignancies across anatomic sites.

11512 Background: Primary squamous cell carcinomas (SCCs) have diverse etiologies, but can share genomic features. We reviewed the genomic profiles of a series of SCC cases of differing anatomic origin. Methods: Hybrid-capture based genomic profiling of 182 or 236 or 315 genes was performed on 4783 squamous malignancies in the course of clinical care, with baits for HPV6, 11, 16, and 18, and assessment of tumor mutation burden (TMB; mutations/Mb) and microsatellite instability. Results: Sites of origin were head and neck (HNSCC, n = 1300), cervical (cSCC; n = 318), anal (aSCC, n = 248), esophageal (n = 242), lung (lSCC, n = 2386), and cutaneous (sSCC, n = 289) SCC cases. For HNSCC, cSCC, and aSCC (collectively termed HCA SCC), 395 (30%), 215 (68%), and 211 (83%) were HPV positive, respectively. For HCA SCC, the most common GA were in TP53 (45%), CDKN2A (29%), PIK3CA (24%), TERT (21%), and FAT1 (14%). The most frequent GA differentially associated with HPV status were in PIK3CA (34.9% versus 16.0%), CYLD (11.4% versus 1.4%) and PTEN (14.8% versus 6.1%) for HPV+ cases, and TP53 (3.8% versus 76.5%), CDKN2A ()1.4% versus 49.8%), and TERT (4.3% versus 33.0%) for HPV- cases. Mean TMB for HPV+ and HPV- cases were 6.6 (STDEV 7.3) and 13.7 (STDEV 29.7), respectively. TMB of all SCC cases was significantly different (p < 10-12) when stratified by HPV status. For lSCC and eSCC, the most common GA were found in TP53 (86%) CDKN2A (40%), and PIK3CA (26%) and mean TMB was 11.6 with HPV found in 3.1% of cases. In sSCC, the most common GA were in TP53 (85.5%), CDKN2A (54.3%), and TERT(44.0%), and mean TMB was 59.5 with HPV in 3.1% of cases. Subsets of SCC cases had defining and targetable GA including bi-allelic deletion of SMARCB1 ( < 0.3%), amplification of PD-L1 (~2% ), and various kinase fusions. Cases demonstrating radiologic response to immunotherapy and matched targeted therapies, as well as subsequent development of multiple mechanisms of acquired resistance, will be presented. Conclusions: HPV driven SCC have similar genomic profiles regardless of of site origin, and have a significantly lower median TMB than HPV negative SCC. Early consistency of responses of SCC to matched therapies may strengthen the case for site independent genomic predictors of therapy response.

Comprehensive analysis of potential immunotherapy genomic biomarkers by profiling paired tumor/normal exome of 1,000 Chinese cancer patients.

11605 Background: With the broadening landscape of immunotherapy use, it is important to identify patients who are likely to benefit from the therapy. We reported the comprehensive analyses of potential predictive biomarkers for PD1/PD-L1 inhibitors based on Chinese patients’ exome profiling data. Methods: Over 1,000 cancer patients from 70 hospitals across 20 provinces in China were recruited and the whole exome of tumor/normal samples of each patient were sequenced. Four potential genomic biomarkers: tumor mutation burden (TMB), mismatch repair deficiency (MMR), microsatellite instability (MSI), and PD-L1 (CD274) amplification (PD-L1 AMP) were analyzed in this Chinese cohort and compared with the mainly Caucasian cohort in TCGA database. Results: At least one of the four preselected genomic biomarkers was identified in 40.8% of this Chinese cancer patient cohort by clinical whole exome sequencing (CWES) analysis. Similar to TCGA cohort, the top 3 high TMB tumor types are lung, esophagus and colorectal cancer. Chinese hepatocellular carcinoma (HCC) patients showed higher TMB than the TCGA cohort (Median: 106 vs. 65 non-synonymous mutations per tumor NMT), which might be due to different etiologies. Five late stage cancer patients (3 lung, 1 melanoma and 1 gallbladder) using PD1/PD-L1 blockade with high TMB showed durable clinical benefit (SD≥6months/PR/CR). Comparing with TCGA data, Chinese colorectal cancer cohort (252 patients) had relative lower MSI-high or MMR related mutations (8.33% vs. 12.5%, and 3.97% vs. 10.0% respectively). PD-L1 AMP most frequently occurred in lung squamous (14.3% VS 9.8% TCGA), HER2-positive breast cancer (8.8% VS 6.8% TCGA with unknown HER2 status) and sarcoma (6.0% VS 9.4%TCGA). One Chinese renal cell carcinoma patient with PD-L1 AMP received anti-PD-1 treatment and on-going PFS is 7 months by far. Conclusions: Our CWES analysis and limited clinical follow-up observations suggested that about 40% Chinese cancer patients had at least one of the 4 potential immunotherapy predictive biomarker mutations. Long-term follow-up is needed to verify the validity of those markers.

Heterogeneity of Tumor and Immune Cell PD-L1 Expression and Lymphocyte Counts in Surgical NSCLC Samples

BackgroundImmune-checkpoint inhibitors against programmed cell death protein 1 (PD-1)/programmed death-ligand 1 (PD-L1) have shown remarkable therapeutic activity in non–small-cell lung cancer (NSCLC). However, biomarker-based patient selection remains a challenge. Our aim was to assess the heterogeneity of various immune markers between different tumor areas of surgically resected NSCLC specimens. Materials and MethodsWe included 94 adenocarcinoma (ADC) and 50 squamous cell carcinoma (SCC) specimens. Two distinct tumor areas of each tumor sample were selected and incorporated into tissue microarrays. PD-L1 expression in tumor cells (TCs) and immune cells (ICs) was assessed using clone SP142 (Ventana). PDL1 gene amplification was assessed using fluorescence in situ hybridization. CD3 and CD8 densities were quantified using digital image-based analysis. Heterogeneity was assessed using kappa agreement index (KI) and intraclass correlation coefficient. ResultsPrevalence of PD-L1 expression was 16.8% in TCs and 27.8% in ICs. Eleven tumors (7.6%) showed PDL1 amplification. In ADC, KI of PD-L1 TC and IC expression between cores was 0.465 and 0.260, compared with 0.274 and 0.124 in SCC, respectively. Higher concordance was observed for PDL1 amplification (KI, 0.647 in ADC and KI, 1 in SCC). Eleven (61.1%) of 18 amplified cores showed PD-L1 staining in < 5% of TCs. Intraclass correlation coefficients for CD3 and CD8 were 0.293 and 0.186 in ADC and 0.489 and 0.610 in SCC samples, respectively. ConclusionsWe found significant heterogeneity of PD-L1 expression in both ADC and SCC samples, especially in the IC compartment. Heterogeneous expression of PD-L1 could misclassify patients for PD-1/PD-L1-directed therapies.

Overexpression and gene amplification of PD-L1 in cancer cells and PD-L1+ immune cells in Epstein–Barr virus-associated gastric cancer: the prognostic implications

Cancer cells use PD-L1 to evade antitumor immunity through interaction with programmed cell death protein 1 (PD-1) on T cells. Recent whole-genome sequence studies revealed frequent gene amplification of PD-L1 in Epstein–Barr virus-associated gastric cancer (EBVaGC). To investigate the significance of PD-L1 in cancer cells and their microenvironment in EBVaGC, we studied PD-L1 expression by analysis of the public database and immunohistochemistry with fluorescent in situ hybridization (FISH) of the PD-L1 gene. Analysis of the database from The Cancer Genome Atlas also disclosed high expression of PD-L1 in EBVaGC compared with other molecular subtypes of GC. Expression of PD-L1 was frequently detected in cancer cells of EBVaGC (33/96; 34%), with infiltration of PD-L1+ immune cells in its stroma (43/96; 45%). Both expression of PD-L1 in cancer cells and PD-L1+ immune cell infiltration in EBVaGC were significantly correlated with diffuse histology according to Lauren's classification and tumor invasion (pT1b or more). As a prognostic indicator, PD-L1 expression in cancer cells correlated with poor outcomes in both overall survival and disease-specific survival (P=0.0498, 0.007). PD-L1-positive cancers had dense infiltration of PD-L1+ immune cells as well as CD8+ and PD-1+ cells in EBVaGC. FISH analysis of representative samples of the tumor demonstrated gene amplification of PD-L1 in 11% of cases. PD-L1-amplified cells corresponded to PD-L1-positive cells showing high-intensity immunohistochemical staining among cancer cells showing weak or moderate intensities. Taken together, PD-L1 expression in cancer cells and their microenvironment may contribute to the progression of EBVaGC, and gene amplification occurs as clonal evolution during progression. This specific subtype of GC infected with EBV is potentially a good candidate for immunotherapy targeting of the PD-L1/PD-1 axis.

Abstract P2-04-18: JAK2 and PD-L1 amplification enhance the dynamic expression of PD-L1 in triple negative breast cancer

Abstract Background: Triple negative breast cancer (TNBC) is a heterogeneous disease. Amplification of chromosome 9p24.1 encoding JAK2 and PD-L1 has been reported in up to 25% of TNBC and is associated with poor clinical outcome. In lymphoma, JAK2 is a transcriptional activator of both PD-1 ligands, and chromosome 9p copy number gain has been associated with therapeutic activity of nivolumab. We evaluated the interaction of JAK2 and PD-L1 expression in TNBC. Methods: 9p24.1 amplification in 4 TNBC cell lines (MDA-MB-231, MDA-MB-436, HCC1937, and HCC70) was measured using array comparative genomic hybridization (aCGH). Amplification was defined as aCGH log2ratio&amp;gt;2.0. Cell surface expression of PD-L1 was detected by flow cytometry and compared with the median fluorescence intensity (MFI) of isotype control Ig. To selectively inhibit JAK2, lentiviral vectors encoding two different shRNA were generated. JAK2, pSTAT1 and pSTAT3 expression were measured by immunoblot. The effects of the anti-JAK1/2 inhibitor ruxolitinib and interferon-gamma (IFN-γ) induction of PD-L1 was measured. Results: 9p24.1 copy number loss was measured in MDA-MB-231 (log2ratio =-1), neutral in HCC1937 ((log2ratio =0), gained in MDA-MB-436 ((log2ratio =+1) and amplified in HCC70 ((log2ratio =+2). No correlation was observed between PD-L1 expression and the 9p24.1 amplification, with the MFI ratio range from 5 to 16.5, mean 8.3 by flow cytometry. TNBC cell lines had higher baseline expression of PD-L1 compared to the ER+ cell lines T47D and MCF-7 (ratio, 0 p=0.1). Low dose IFN-γ (1.0-10.0 ng/ml) rapidly induced expression of PD-L1 in MDA-MB-231 (1.5 fold increase) and HCC70 (3.3 fold increase) with significant activation of the JAK2/STAT1 pathway. The induction of pSTAT1 and PD-L1 expression by IFN-γ was blocked with low dose (1mM) JAK1/2 inhibitor ruxolitinib. Knockdown of JAK2 with shRNA (&amp;gt;80%) did not impact PD-L1 baseline expression in MDA-MB-231 and HCC70 but abrogated IFN-γ –mediated induction of PD-L1 and the phosphorylation of STAT1. Conclusion: These data suggest that TNBC cell lines have baseline PD-L1 expression, but the cells with 9p24.1 amplification are highly sensitive to PD-L1 induction with IFN-γ which can be abrogated with inhibition with a JAK1/2 inhibitor or shRNA. Synergistic inhibition of PD1/PD-L1 and JAK2 may have therapeutic efficacy in the subset of TNBC with 9p24.1 amplification, and the dynamic effects of tumor PD-L1 expression in response to local inflammation should be considered in the evaluation of PD-1/PD-L1 checkpoint blockade. Citation Format: Chen M, Pockaj B, Andreozzi M, Barrett MT, Ocal IT, McCullough AE, Krishna S, Anderson KS. JAK2 and PD-L1 amplification enhance the dynamic expression of PD-L1 in triple negative breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-04-18.

MA14.01 Updated Dataset Assessing Tumor Mutation Burden (TMB) as a Biomarker for Response to PD-1/PD-L1 Targeted Therapies in Lung Cancer (LC)

Immune checkpoint inhibitors (ICPIs) nivolumab and pembrolizumab have been FDA-approved in non-small cell LC (NSCLC). Current IHC based diagnostics are challenged by assay and slide scoring issues and modest predictive value, and more robust and comprehensive biomarkers of ICPI efficacy are needed. A discovery set of 64 NSCLCs treated with ICPIs suggested that high TMB (≥15 mutations/Mb) significantly correlated with longer time on drug (Spigel et al., ASCO 2016, Abstract:9017). Comprehensive genomic profiling (CGP) was performed during the course of clinical care. TMB was assessed as the number of somatic, coding, base substitution and indels per Mb of genome. Microsatellite instability-high (MSI-H) or stable (MSS) status was determined using a proprietary algorithm. 15,529 LCs: 66% adenocarcinoma, 1% sarcomatoid, 14% NSCLC NOS, 11% squamous, 5% small cell, and 2% large cell were assessed. TMB was similar across all lung histologies (median: 6.3, 8.1, 9.0, 9.9, 9.9, and 10.8); the median was 7.6 for all LC cases (TMB ≥15 in 24% of cases), compared to 4.5 for 80,000+ samples of diverse tumor types in the database. Of LCs assessed 0.3% were MSI-H, of which 30/31 were TMB-high; however, 24% of MSS-stable cases were also TMB-high. PD-L1 amplification and DNA repair pathway mutation (MLH1, MSH2, POLE) were found in 1.0% and 1.1% of LC cases analyzed, respectively. Tumors harboring known drivers (ALK, ROS1, EGFR, BRAF V600E, MET splice) had low TMB (median: 2.5, 3.6, 3.8, 3.8, 4.5), whereas tumors with KRAS mutation, non-V600E BRAF mutation, PD-L1 amplification, or DNA repair alterations were more likely to be TMB-high (median: 9.0, 10.8, 14.4, 21.6). High TMB may be a predictive biomarker of response to ICPIs. Several factors including lack of a known driver, MSI-H status, PD-L1 amplification, and DNA repair mutation correlated with high TMB (P<0.0001 for all cases). However, 95% of TMB-high cases assessed were MSS and lacked both PD-L1 amplification and DNA repair mutation, and thus would likely not be selected for immunotherapy by assessment of individual genomic alterations or MSI status alone. A validation cohort of NSCLC patients treated with anti-PD-1/PD-L1 therapies including analysis of clinical outcome, TMB, genomic profile, and available clinicopathologic characteristics will be presented. CGP of LC to simultaneously determine TMB, MSI status, PD-L1 amplification, and the presence of driver alterations may provide clinically useful predictors of response to ICPI and other targeted therapies using a single platform, but prospective clinical trials are needed to confirm these observations.

P2.01-063 PDL1, JAK2 and PTEN Copy Number Alterations Synergistically Upregulate PD-L1 Expression in NSCLC

Predictive biomarkers research in anti-PD-1/PD-L1 immunotherapy treatment is still at an early stage of development. Recently, amplification of PDL1 gene (9p24.1) has been described in NSCLC in correlation with PD-L1 protein expression. In addition, other tumor-constitutive alterations such as JAK2 amplifications (322kb upstream of PDL1) or PTEN deletions are also known to modulate PD-L1 expression. We aimed to determine PDL1, JAK2 and PTEN copy number alterations (CNAs) and subsequent PD-L1 protein expression in NSCLC. A total of 171 NSCLC patients (121 ADC and 50 SCC) were included. Clinical, histopathological and molecular data were collected. In resected early-stage diseases, two distinct histologic areas from FFPE tumoral tissue were included for each patient in 8 tissue microarrays (TMAs). PD-L1 expression analysis was assessed by IHC using PD-L1 #SP142 clone (Ventana) and positive cut off was defined at >1%. Moreover, H-score semi-quantitative approach was used to generate a score from 0 to 300. PDL1, JAK2 and PTEN CNAs were studied by FISH using commercial and non-commercial probes hybridized with respective centromere enumeration probes. Amplifications were defined as mean gene by mean centromere ratio ≥2.0, deletions as mean gene by mean centromere ratio ≤0.8, gains as mean gene ≥2.5, and high gains as mean gene ≥4.0. PD-L1 expression was positive in 40 out of 171 cases (23.3%), with an average H-score of 177 and significantly associated with ADC solid histological pattern (p=0.012), KRAS mutations (p=0.001), the presence of TILs (p=0.001), and active smoking status (p=0.031). PDL1 gene CNAs were seen in 68/159 assessable cases (42.8%). We found 14 tumors with PDL1 amplification (8.8%), 21 PDL1 high gains (13%) and 33 PDL1 gains (20.8%). Twelve out of 14 FISH amplified cases had PD-L1 positive expression. Thus, FISH predicted positive PD-L1 IHC result with a low sensitivity (31.6%) but a high specificity (98.6%). Among PD-L1 expressing tumors (n=40), seven cases had JAK2 amplifications (6 of them with PDL1 gene coamplification) and eight showed PTEN deletions (3 of them were PDL1 amplified). Differences in H-score intensity between these groups were observed: JAK2-PDL1 coamplified cases had near 2-fold increase in PD-L1 expression than PDL1 alone (average H-score: 282 vs. 148). PDL1 gene amplification is synergistically regulating PD-L1 protein expression. In addition, JAK2 amplification upregulates PD-L1 expression, following the concept of cooperative oncogenic effects of genes within the PDJ amplicon. PDL1, JAK2 and PTEN CNAs analysis may be relevant for anti-PD-1/PD-L1 patient selection.

Classical Hodgkin Lymphoma with Reduced β2M/MHC Class I Expression Is Associated with Inferior Outcome Independent of 9p24.1 Status.

In classical Hodgkin lymphoma (cHL), malignant Hodgkin Reed-Sternberg (HRS) cells evade antitumor immunity by multiple mechanisms, including perturbed antigen presentation and enhanced PD-1 signaling. HRS cell expression of the PD-1 ligands is attributable, in part, to copy number alterations of 9p24.1/CD274(PD-L1)/PDCD1LG2(PD-L2) Amplification of PD-L1/PD-L2 is associated with advanced clinical stage and inferior progression-free survival (PFS) following first-line (induction) therapy. The relationships between altered expression of β2-microglobulin (β2M), MHC class I, and MHC class II by HRS cells, PD-L1/PD-L2 amplification, and clinical outcome in cHL are poorly defined. We assessed these variables in diagnostic biopsy specimens from 108 patients with cHL who received uniform treatment and had long-term follow-up and found decreased/absent expression of β2M/MHC class I in 79% (85/108) and decreased/absent expression of MHC class II in 67% (72/108) of cases. Patients with decreased/absent β2M/MHC class I had shorter PFS, independent of PD-L1/PD-L2 amplification and advanced stage. Decreased or absent MHC class II was unrelated to outcome. These results suggest that MHC class I-mediated antigen presentation by HRS cells is an important component of the biological response to standard chemo/radiotherapy. The paucity of β2M/MHC class I expression on HRS cells also prompts speculation regarding alternative mechanisms of action of PD-1 blockade in cHL. Cancer Immunol Res; 4(11); 910-6. ©2016 AACR.

117P - Evaluation of tumor- and stromal immune marker heterogeneity in non-small cell lung cancer

Immunotherapy is the current standard in second-line treatment for non-small cell lung cancer (NSCLC) patients. However, accurate identification of patients who benefit based on immune marker (IM) expression remains a challenge. Here, we aimed to estimate the potential impact of heterogeneity on the assessment of IM expression in both lung adenocarcinoma (ADC) and squamous-cell carcinoma (SCC). A total of 144 surgically treated NSCLC patients were included, 94 ADCs and 50 SCCs. Two tumor cores per patient were incorporated into tissue microarrays (TMA). CD3 and CD8 were assessed by immunohistochemistry (IHC) and PD-L1 by IHC and FISH. Expression of IM was analyzed both in tumor cells (TC) and tumor stroma (TS). A PD-L1 positivity threshold of ≥1% was established for IHC and PDL1 gene amplification was defined as a PDL1/CEP9 ratio of ≥2. In each core, CD3 and CD8 positive cells were recorded quantitatively and the CD8/CD3 ratio was calculated. Heterogeneity of IM between corresponding tumor cores was analyzed using kappa agreement index. Finally, IM expression was correlated with clinical, pathological and molecular variables. Patients' median age was 65, 80% were male and 46% were current smokers. PD-L1 was expressed in TC and TS in 22% and 62% of cases, respectively. Amplification of PDL1 was found in 12 of cases, of which 8 (67%) were IHC PD-L1 positive in TC. Heterogeneity of TC PD-L1 positivity was 8% in ADC and 6% in SCC. Regarding TS PD-L1 expression, discordance between corresponding cores was found in 19% and 34% of ADC and SCC cases, respectively. PDL1 amplification was discordant between cores in 5 of 12 (42%) cases. Per core, median CD3 and CD8 positive cells were 436 and 159, respectively. Median CD8/CD3 ratio was 0.23 in ADC and 0.53 in SCC. Taking these values as cut-off points, CD8/CD3 ratio was discordant in 22% and 10% of ADC and SCC cases, respectively. Finally, TC PD-L1 expression was correlated with CD8 infiltrate. (p < 0.05 for ADC and SCC). PD-L1 IHC expression appears to be more heterogeneous when assessed in the TS compared to the TC compartment, especially in SCC histology. Intra-tumor heterogeneity has to be taken into account when selecting patients for immunotherapy based on PD-L1 positivity or lymphocyte presence.

Integration of genomics and histology revises diagnosis and enables effective therapy of refractory cancer of unknown primary with PDL1 amplification.

Identification of the tissue of origin in cancer of unknown primary (CUP) poses a diagnostic challenge and is critical for directing site-specific therapy. Currently, clinical decision-making in patients with CUP primarily relies on histopathology and clinical features. Comprehensive molecular profiling has the potential to contribute to diagnostic categorization and, most importantly, guide CUP therapy through identification of actionable lesions. We here report the case of an advanced-stage malignancy initially mimicking poorly differentiated soft-tissue sarcoma that did not respond to multiagent chemotherapy. Molecular profiling within a clinical whole-exome and transcriptome sequencing program revealed a heterozygous, highly amplified KRAS G12S mutation, compound-heterozygous TP53 mutation/deletion, high mutational load, and focal high-level amplification of Chromosomes 9p (including PDL1 [CD274] and JAK2) and 10p (including GATA3). Integrated analysis of molecular data and histopathology provided a rationale for immune checkpoint inhibitor (ICI) therapy with pembrolizumab, which resulted in rapid clinical improvement and a lasting partial remission. Histopathological analyses ruled out sarcoma and established the diagnosis of a poorly differentiated adenocarcinoma. Although neither histopathology nor molecular data were able to pinpoint the tissue of origin, our analyses established several differential diagnoses including triple-negative breast cancer (TNBC). We analyzed 157 TNBC samples from The Cancer Genome Atlas, revealing PDL1 copy number gains coinciding with excessive PDL1 mRNA expression in 24% of cases. Collectively, these results illustrate the impact of multidimensional tumor profiling in cases with nondescript histology and immunophenotype, show the predictive potential of PDL1 amplification for immune checkpoint inhibitors (ICIs), and suggest a targeted therapeutic strategy in Chromosome 9p24.1/PDL1-amplified cancers.