Genomic Correlates Of Response Research Articles

Immune checkpoint inhibition for the treatment of metastatic breast cancer with high tumor mutational burden: Real world clinical and genomic correlates of response.

1078 Background: Pembrolizumab has received tumor-agnostic approval for the treatment of unresectable/metastatic solid tumors carrying a tumor mutational burden (TMB) >10 mut/Mb. The registrational KEYNOTE-158 study, however, did not include any patients with metastatic breast cancer (MBC). Here we present a retrospective study of patients with TMB-high MBC treated with immune checkpoint inhibition (ICI) and report real-world efficacy endpoints, including progression free survival as well as clinical and genomic characteristics that may be associated with duration of response. Methods: All patients with MBC treated with an ICI at MSK and with TMB-high tumors (>10 mut/Mb) assessed by tumor-normal targeted sequencing (MSK-IMPACT) were included. Clinicopathological and demographic data were obtained via the MSK Breast Cancer Translational Program. Mutational signatures were determined using the Signature Multivariate Analysis (SigMA) algorithm. Real world progression free survival (PFS) was determined clinically and analyzed using Kaplan-Meier estimates. Univariate and multivariate analyses were performed using Cox-proportional hazards models. Results: Of 688 patients with TMB-high breast cancer at MSKCC, 87 had received ICI-based therapy for estrogen receptor positive (ER+), triple negative breast cancer (TNBC), and HER2+ MBC. Of those, 45 patients were treated with ICI monotherapy (29 ER+, 11 TNBC, 5 HER2), 31 were treated with ICI + chemotherapy (12 ER+, 19 TNBC), and 11 were treated with combinations of ICI and targeted therapy (8 ER+, 1 TNBC, 2 HER2). Median age at time of ICI treatment across all patients was 60. ICI was the median 5th line therapy for all patients (range 1-16). Median PFS of ICI for all patients was 3.17 months (95%CI 2.27-5.5). Multivariable analysis did not show statistically significant differences in PFS based upon treatment regimen (monotherapy vs. chemoIO HR 1.03, 95%CI 0.54-1.96, p= 0.92, monotherapy vs combination IO HR 1.66, 95%CI 0.69-4.02, p=0.26), ER status (ER- vs ER+, HR 1.20, 95%CI 0.63-2.28, p=0.57), or HER2 status (HER2- vs HER2+, HR 1.63, 95%CI 0.51-5.20, p=0.41. However, longer PFS was significantly associated with TMB>30 (HR 0.35, 95%CI 0.13-0.92, p=0.03) and dominant APOBEC signature score (APOBEC- vs APOBEC+, HR 0.49, 95%CI 0.25-0.98, p=0.04). Shorter PFS was significantly associated with later treatment line at the time of ICI exposure (continuous, HR 1.22, 95%CI 1.11-1.36, p=0.0001). Conclusions: In this heavily pretreated real-world cohort, ICI in TMB high MBC shows modest clinical activity. Our analysis suggests that ICI may be more effective in MBC patients that are less heavily pre-treated, have TMB>30, and have a predominantly APOBEC mutational signature. Further translational work is necessary to explore the mechanisms of these findings.

Genomic correlates of response and resistance to immune checkpoint inhibitors (ICI) in patients with metastatic urothelial carcinoma (mUC) in a real-world setting.

Genomic correlates of response and resistance to immune checkpoint inhibitors (ICI) in patients with metastatic urothelial carcinoma (mUC) in a real-world setting.

Radiation in a new era of Multiple Myeloma management: Patterns of utilization, clinical, radiologic, and biochemical outcomes, and possible genomic correlates of response

Radiation in a new era of Multiple Myeloma management: Patterns of utilization, clinical, radiologic, and biochemical outcomes, and possible genomic correlates of response

Abstract 1126: Disease monitoring with comprehensive genomics provides evidence of mechanism of action and immune evasion in patients receiving an individualized neoantigen cancer vaccine

Abstract Therapeutic vaccines hold promise to broaden the potency of immune checkpoint blockade (ICB) therapy in tumors lacking immune reactivity. A heterologous chimpanzee adenovirus (ChAd68) and self-amplifying mRNA (samRNA)-based individualized neoantigen vaccine regimen in combination with nivolumab 480 mg IV and ipilimumab 30mg SC (NCT03639714) has previously demonstrated safety, durable immunogenicity, and clinical benefit in patients with previously treated metastatic disease. Genomic correlates of response were studied over time in 29 patients (13 MSS-CRC, 13 GEA, 3 NSCLC) to understand novel mechanisms of action. Exome and transcriptome sequencing from archival biopsies was used for neoantigen selection. Monthly circulating tumor DNA (ctDNA) samples were collected for monitoring using a comprehensive tumor-naïve and tumor-informed hybrid-capture based ctDNA assay. Paired pre- and post-vaccine tumor transcriptomes were analyzed for 10 patients with 6 having accompanying DNA T cell receptor Β CDR3 repertoire sequencing (TCRseq) in biopsies and longitudinal PBMCs. Prior to vaccination patient tumors were not enriched for immune infiltration or tumor mutation burden (TMB), median 4.3 mut/Mb (range: 2-17 mut/Mb). Minimal neoantigen and mutation drift was observed with a median of 92.5% of neoantigens (range: 45-100%) and a median of 84% (range: 24-99%) of individual mutations detected in biopsies and ctDNA. Notably, paired pre- and post-vaccine biopsy gene expression analyses show upregulation in gene signatures associated with immune infiltration aligning with evidence of T cell expansion measured by significantly expanding CDR3 clonotypes (p &lt0.01). Longitudinal TCRseq in PBMCs demonstrate vaccine induced TCR repertoire dynamics and expanding and contracting clones observed in tumor biopsies could be monitored throughout treatment. In 4 patients the most drastic TCR repertoire changes were observed at time points measured after a 2nd dose of ChAd68. Lastly, we observe evidence of acquired immune evasion through ctDNA monitoring in two patients each following a year of study treatment. One GEA patient acquired HLA-LOH after remaining stable on treatment and one MSS-CRC with a molecular response (MR) for &gt7 months acquired novel biallelic loss-of-function mutations in TAP1 following 1 year of study treatment. We demonstrate that our neoantigen-directed immunotherapy regimen drives durable immune pressure on the tumor in patients with advanced disease where CPI alone has provided minimal benefit. Further, the evidence of acquired resistance supports the induction of immune pressure on tumors following individualized neoantigen vaccination. Comprehensive ctDNA longitudinal monitoring enables real-time assessment of clinical response and acquired resistance. Citation Format: Desiree Schenk, Rita Zhou, Olivia Petrillo, Alexis Mantilla, Italo Faria do Valle, Steven Maron, Brian S. Henick, Chih-Yi Liao, Daniel V.T. Catenacci, Sameek Roychowdhury, Benjamin Solomon, Alexander I. Spira, Ankur Dhanik, Andrew R. Fergusson, Karin Jooss, Matthew Davis. Disease monitoring with comprehensive genomics provides evidence of mechanism of action and immune evasion in patients receiving an individualized neoantigen cancer vaccine [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1126.

Abstract 1238: Comprehensive ctDNA monitoring provides early signal of clinical benefit with a novel personalized neoantigen directed immunotherapy for late-stage cancer patients

Abstract Neoantigen directed immunotherapy holds promise to increase the likelihood of patients with solid tumor devoid of immune infiltration benefiting from immune checkpoint immunotherapy (CPI). A heterologous prime-boost vaccination approach consisting of Chimpanzee Adenovirus (ChAd) prime and multiple self-amplifying mRNA (SAM) boosts, delivering 20 neoantigens, has been evaluated in a Phase 1/2 clinical trial in late-stage solid tumor patients in combination with nivolumab and ipilimumab (NCT03639714). Neoantigen dynamics, tumor burden and genomic correlates of response were studied over time in 20 patients (8 GEA, 2 NSCLC, 10 MSS-CRC). Exome sequences from archival (sample used for neoantigen selection), baseline (start of immunization) and on-treatment biopsies were analyzed for 20, 16 and 10 patients respectively. Paired pre- and post-vaccine tumor transcriptomes were analyzed for 6 patients. Personalized capture baits were designed for all non-synonymous mutations detected in archival biopsies (mean 146; range: 67-402) for ctDNA monitoring. Longitudinal ctDNA samples were collected monthly on treatment (mean 7; range: 1-18). ctDNA duplex UMI libraries were captured and sequenced to a target mean raw depth >80,000x and reduced to 3x per strand consensus duplex reads. The majority of vaccine neoantigens were detected in ctDNA (87%; range 45%-100%) and mean neoantigen variant allele frequency (VAF) strongly correlated with all monitored mutations VAF (R2 = 0.90, p < 0.0001) through treatment. The percentage of vaccine neoantigens detected was higher than that for all monitored mutations in the same samples with a median of 80% (21%-98%) in ctDNA and 70% (44%-100%) in biopsies. Five of 9 MSS-CRC patients with measurable baseline ctDNA achieved molecular responses (MR, >50% reduction in ctDNA from baseline) that correlated with OS and PFS, and in some patients, was accompanied by radiologic tumor shrinkage. One MSS-CRC patient with MR for >7 months acquired novel biallelic loss-of-function mutations in TAP1 following 1 year of study treatment. Differential gene expression analysis from paired pre- and post-vaccine biopsies (including 2 MSS-CRC pairs with MR) demonstrated significant upregulation in gene signatures associated with immune-inflamed tumor microenvironments including interferon alpha and gamma responses. We demonstrate that tumor-informed neoantigen selection and vaccine manufacturing while patients receive chemotherapy is feasible, since the majority of neoantigens are retained in the tumor post-chemotherapy. Further, neoantigen-directed immunotherapy appears to drive clinical benefit in patients with advanced MSS-CRC tumors, where CPI alone has provided minimal benefit. Comprehensive ctDNA longitudinal monitoring enables real time assessment of clinical response and acquired resistance. Citation Format: Desiree Schenck, Rita Zhou, Alexis Mantilla, Oliver Spiro, Taylor Patch, Adrienne Johnson, Daniel Navarro Gomez, Brian S. Henick, Chih-Yi Liao, Sameek Roychowdhury, Steve Maron, Benjamin Solomon, Alexander I. Spira, Daniel V. Catenacci, Andrew R. Fergusson, Raphael F. Rousseau, Karin Jooss, Matthew J. Davis. Comprehensive ctDNA monitoring provides early signal of clinical benefit with a novel personalized neoantigen directed immunotherapy for late-stage cancer patients [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 1238.

Abstract 5762: Genomic correlates of response to immune checkpoint blockade in Epstein-Barr virus-associated pulmonary lymphoepithelioma-like carcinoma

Abstract Introduction: Pulmonary lymphoepithelioma-like carcinoma (PLELC) is a rare and distinct subtype of non-small cell lung cancer (NSCLC). The use of immune checkpoint inhibitors (ICIs) for advanced PLELC and tumor genomic features underlying varied treatment responses remains poorly investigated. Methods: The study involved 33 PLELC patients with clinically annotated outcomes to ICIs. Targeted sequencing covering 520 cancer related genes were performed in 22 patients to identify single nucleotide variations (SNVs), copy number variations (CNVs), tumor mutation burden (TMB) and microsatellite instability. Mutation profiles of PLELC were compared with other NSCLC subtypes including lung adenocarcinoma (LUAD) and squamous cell lung carcinoma (LUSC) as well as EBV (Epstein-Barr virus) positive nasopharyngeal carcinoma (NPC) in public databases. Associations between PLELC mutation profiles and treatment responses to ICIs were analyzed. An external dataset of ICI-treated melanoma was used in further analysis of B2M. Results: The objective response rate (ORR), disease control rate (DCR), durable clinical benefit (DCB) rate, and median PFS was 18.2%, 60.6%, 45.5%, and 4.5 months respectively. Genetic alterations were frequently identified in B2M (9 of 22, 41%), CDKN1B (7 of 22, 32%), TP53 (6 of 22, 27%), SGK1 (5 of 22, 23%), NFKBIA (4 of 22, 18%) and TNFAIP3 (4 of 22, 18%). CNVs were comprehensively presented in PLELC while SVNs were less common. In our study, 8 of 9 tumors with B2M mutations and all tumors with CDKN1B, SGK1, TNFAIP3 mutations harbored copy number amplifications. Mutation profiles of PLELC were largely different from those of LUAD and LUSC but were similar to NPC. The median TMB was 2.99 mutations/Mb. All tumors were microsatellite stable (MSS). Patients with higher TMB showed improved DCR (p=0.029), PFS (p=0.056) and overall survival (p=0.038) than those with lower TMB. Patients with lower baseline plasma EBV DNA level also had longer PFS (p<0.001). The presence of B2M amplifications was significant associated with higher TMB level (p=0.022). The PFS of 7 patients with high TMB and amplified B2M were significantly longer than that of other patients (p=0.046). In the external dataset, higher expression of B2M was associated with better ICI responses for melanoma (p<0.05). Pathways involving anti-tumor immune responses such as NK cell-mediated cytotoxicity, antigen processing and presentation were correlated with high B2M expression while Wnt, TGF-beta and MAPK pathways that define poor response to ICI were correlated with low B2M expression. Conclusion: Our study observed comprehensive CNVs in PLELC patients which is a unique molecular feature compared with other NSCLC subtypes and EBV-positive NPC. Higher TMB level and amplifications of B2M in PLELC patients may be used as potential biomarkers to indicate better treatment responses to ICIs. Citation Format: Li-na He, Xuanye Zhang, Xinze Lv, Yuxi Zhou, Min Li, Ting Hou, Wenjie Sun, Li Zhang, ShaoDong Hong. Genomic correlates of response to immune checkpoint blockade in Epstein-Barr virus-associated pulmonary lymphoepithelioma-like carcinoma [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 5762.

Genomic correlates of response to capecitabine and temozolomide (CAPTEM) in pancreatic neuroendocrine tumors.

4124 Background: Despite the frequent use of capecitabine and temozolomide (CAPTEM) to treat metastatic, well-differentiated pancreatic neuroendocrine tumors (PNETs), no reliable genomic predictors of response currently exist. PNETs commonly harbor mutations in MEN1, ATRX, DAXX, and the PI3K/AKT/mTOR pathway. We sought to determine whether the mutational status of these genes correlates with response to CAPTEM. Methods: A retrospective cohort of PNET cases seen at Cedars-Sinai Medical Center or from Perthera’s Real-World Evidence Database included 23 patients who were treated with CAPTEM in 1st or 2nd line and had targeted next-generation sequencing (NGS) of their tumors available. Genomic alterations were correlated with progression-free survival (PFS) using multivariate Cox regression analysis. Results: We analyzed 23 PNET patients, 4 (17.4%) of whom had documented functional tumors. We identified MEN1 mutations as positively associated with CAPTEM response, but this effect was less pronounced for the subset with co-occurring DAXX mutations, which are commonly found alongside MEN1 alterations. With and without accounting for line of therapy, we found that PFS on CAPTEM was significantly longer in MEN1-mutated, DAXX-wildtype tumors compared to other mutation profiles ( P < 0.01, see Table). ATRX (67%) and PTEN (33%) alterations were also enriched in the MEN1-mutated/ DAXX-wildtype subset; however, other PI3K/AKT/mTOR alterations were common across all MEN1-mutated cases. Conclusions: We describe a novel genomic signature ( MEN1 mut/ DAXX wt) that correlates with PNET response to CAPTEM therapy and is exploratory in nature. Prospective validation of these associations is warranted while taking into account other therapies, histopathologic factors, and other genomic correlates. [Table: see text]

Long-term Follow-up and Patterns of Response, Progression, and Hyperprogression in Patients after PD-1 Blockade in Advanced Sarcoma.

Programmed cell death protein 1 (PD-1) blockade can mediate objective responses in advanced sarcomas, but their durability has not been established and it is unclear if hyperprogressive disease (HPD) occurs in sarcomas treated with PD-1 inhibitors. We pooled patients who were treated prospectively with nivolumab or pembrolizumab as monotherapy or with bempegaldesleukin, epacadostat, ipilimumab, or talimogene laherparepvec. We did a new independent assessment for HPD and analyzed clinical, pathologic, and genomic data from baseline tumor biopsies. Our primary endpoint was the incidence of HPD; secondary endpoints were clinical or genomic correlates of response or HPD. We treated 134 patients with advanced sarcoma from 2015 to 2019. Twenty-one patients (16%) had a complete or partial response (CR/PR), and 30% of responses were durable for over 2 years. Forty-eight (36%) patients had stable disease (SD), 45 (34%) had progressive disease without HPD (PD), and 15 (11%) had HPD. Five patients (4%) were not evaluable for HPD. The sarcoma subtypes, sites of metastasis, clinical course, and genomic alterations in patients with PD and HPD were similar, except HPD tumors were smaller at baseline. In patients with advanced sarcoma, PD-1 blockade can mediate durable responses. HPD occurs in sarcoma at an incidence that is similar to what has been reported in other solid tumors, but patients with HPD were clinically and biologically similar to those who had PD. Further research is required to establish whether HPD is a biologically distinct phenomenon and whether a theoretical risk of HPD should influence patient management.

DNMT3A mutation to identify a subset of non-small cell lung cancers with increased sensitivity to PD-(L)1 blockade.

9113 Background: Despite significant improvements in overall survival with PD-(L)1 inhibition, the majority of patients with metastatic NSCLC do not respond to immune checkpoint inhibition (ICI). Growing evidence suggests the importance of genomic alterations in modulating anti-cancer immune response and predicting ICI efficacy in solid tumors. However, the genomic correlates of response and resistance to ICI in NSCLC are still largely unknown. Methods: Patients with advanced NSCLC treated with PD-(L)1 blockade whose tumors underwent comprehensive genomic profiling were included. Mutation enrichment analysis was performed to identify genomic alterations enriched in responders versus (vs) non-responders to ICI. Loss-of function mutations annotated as oncogenic by OncoKB, and missense mutations predicted to be deleterious by SIFT/Polyphen-2 were considered. Cox proportional hazards models was used to estimate hazard ratios (HRs) in univariable and multivariable models. Results: : Among 600 NSCLCs, we identified deleterious mutations in the DNA methyltransferase 3A ( DNMT3A) gene as the most significant alteration enriched in responders versus non-responders to PD-(L)1 blockade (q-Value < 0.05). DNMT3AMUT (7.3%, N = 44) and DNMT3A wild-type ( DNMT3AWT) cases (92.7%, N = 556) were well balanced in terms of baseline clinicopathologic features, including PD-L1 expression, sex, performance status, age, concurrent genomic alterations, and smoking history. DNMT3AMUT tumors had a significantly higher median TMB compared to DNMT3AWT cases (12.1 vs 9.8 mutations/megabase, P = 0.03). DNMT3A loss was associated with significantly higher objective response rate (ORR, 50% vs 20.5%, P < 0.001), longer median progression-free (mPFS, 9.2 vs 2.9 months, HR 0.60, P < 0.01) and overall survival (mOS, 23.1 s 12.1 months, HR 0.59, P = 0.01) among DNMT3AMUT compared to DNMT3AWT NSCLCs. Loss-of function mutation in DNMT3A was confirmed be an independent predictor of improved PFS (HR 0.61, P = 0.01) and OS (HR 0.62, P = 0.04) at multivariable analysis. DNMT3A mutation had no impact on OS among patients with advanced NSCLC who did not receive ICI (HR 1.18, P = 0.22), nor among those with early-stage resected NSCLC (HR 1.17, P = 0.48), suggesting that DNMT3A mutation is predictive, rather than prognostic, of ICI efficacy. Although a subset of DNMT3A mutations could have potentially arisen from tumor-associated hematopoietic cells, the DNMT3A allele fraction-to-tumor purity ratio was ≥0.5 in more than 50% of cases, suggesting that a proportion of these mutations were derived from lung cancer cells. Conclusions: Loss-of-function mutation in DNMT3A may identify a new genomically defined subset of NSCLC with increased sensitivity to PD-(L)1 blockade. Additional studies are ongoing to determine the exact source of DNMT3A mutation (clonal hematopoiesis vs tumor) and their relative contribution to ICI efficacy.

Genetic ancestry and clinical outcomes to immune checkpoint inhibitors among seven common cancers.

10536 Background: Prior studies and clinical trials report associations between self-reported race and clinical outcomes to Immune Checkpoint Inhibitors (ICIs). However, comprehensive studies of ancestry-associated differences in clinical outcomes have not been performed. We derived genetic ancestry scores and assessed clinical outcomes in 1341 patients with cancer treated with ICIs. Methods: Patients at the Dana-Farber Cancer Institute treated with ICIs only and with relevant cancer types and targeted exome sequencing data (Oncopanel) were included. Relevant cancer types included colorectal adenocarcinoma (CRC), esophagogastric adenocarcinoma (EGC), head and neck squamous cell carcinoma (HNSCC), melanoma, non-small cell lung cancer (NSCLC), renal cell carcinoma (RCC), and urothelial carcinoma (UC). We developed a bioinformatics pipeline to infer fine-scale genetic ancestry for each patient (n=1341) directly from tumor sequencing data by leveraging off and on-target sequenced reads and external ancestry reference panels. Three ancestry scores were determined (African, East Asian, European). Overall survival (OS) and time-to-treatment failure (TTF) were compared by Cox logistic regression between ancestral populations. Hazard ratio (HR) was derived using multivariable analysis, adjusted for single versus combination therapy, prior lines of therapy, and tumor mutational burden (TMB, as percentiles). Results: Median follow-up was 37.8 months (m; interquartile range: 35.7-39.5m). Common cancer types included CRC (n=52), EGC (n=114), HNSCC (n=88), melanoma (n=274), NSCLC (n=571), RCC (n=99), and UC (n=143). A higher East Asian ancestry (EAS) was significantly associated with worse OS ( p=0.03) and TTF ( p=0.002) in patients with RCC, independent of the histologic subtype (Table). There was no significant association between any of the three ancestral populations and clinical outcomes in the other 6 cancer types. Conclusions: We described clinical outcomes to ICIs across three global populations in 7 cancers. As the medical field re-evaluates the use of self-reported race in clinical decision-making, we utilize a novel ancestry pipeline that can be readily applied to tumor-only sequencing panels and better characterize non-white populations. We find no ancestry differences in clinical outcomes except in patients with RCC treated with ICIs which will require future validation. We plan to analyze genomic correlates of response by ancestry in each of the cancer types to better understand these diverge clinical behaviors.[Table: see text]

Abstract 5859: Genomic correlates of response to immune checkpoint inhibitors in advanced head and neck squamous cell carcinoma

Abstract Background: Objective response rates for patients with advanced head and neck squamous cell carcinoma (HNSCC) receiving PD-1/L1 immune checkpoint inhibitors (ICI) without chemotherapy remain under 25%. Some studies suggest that HNSCC patients with human papillomavirus (HPV)-associated disease and high tumor mutational burden (TMB) derive greater benefit, although these findings do not yet guide patient selection. We hypothesized that uniform genomic analysis of a larger cohort may help place the predictive value of TMB in a broader biological context and identify more robust predictors of response. Methods: 274 patients with biopsy-proven HNSCC treated with ICI at our institution from June 2014 to September 2018 were identified. Following targeted massively parallel sequencing, single nucleotide variant calling, and quality control measures, 127 samples were available for analysis. Clinical records were reviewed to determine radiographic response and overall survival. Responders included complete or partial responders by RECIST v1.1. Nonresponders included those with progressive or stable disease. We evaluated associations between genomic features and response using standard statistical methods. Results: TMB for the entire cohort was significantly higher among responders (p = 0.01, Fisher's); however, it had poor predictive power (AUROC = 0.68). Among the entire cohort, somatic mutations in NOTCH1 were enriched in responders (12/26 R, 17/101 NR, p = 0.003). After correcting for TMB, NOTCH1 remained associated with complete or partial response (p < 0.05, logistic regression). After adjusting for mutations in NOTCH1, no significant difference in TMB was observed between responders and nonresponders (p > 0.05, logistic regression). Among HPV-negative responders, somatic mutations in CDK12, a transcription-associated kinase whose loss is associated with genomic instability, neoantigen burden, and T cell infiltration, were enriched (3/26 R, 0/101 NR, p = 0.008, Fisher's). Conclusion: Through uniform genomic analysis of the largest ICI-treated HNSCC cohort to-date, we validated prior findings regarding the significance of TMB and NOTCH1 as correlates of response for ICI in HNSCC. We also show that TMB alone is insufficient as a predictor of response. We identify CDK12 as a variant with a mechanistic relationship to tumor immunology as a possible correlate of response in HPV-negative patients, which warrants further investigation. Citation Format: Nisarg A. Patel, Natalie I. Vokes, Haitham Elmarakeby, Glenn J. Hanna, Eliezer M. Van Allen. Genomic correlates of response to immune checkpoint inhibitors in advanced head and neck squamous cell carcinoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5859.

Genomic correlates of response to PD-1 axis inhibitors (IO) in metastatic urothelial cancer (mUC) using explainable artificial intelligence (AI).

570 Background: AI is increasingly being used in clinical cancer genomics research. Probabilistic Graphical Models (PGMs) are AI algorithms that capture multivariate, mutli-level dependencies in complex patterns in large datasets while retaining human interpretability. We hypothesize that PGMs can identify clinical and genomic features that correlate with IO response in patients (pts) with mUC. Methods: In this retrospective study eligibility criteria were: diagnosis of mUC, receipt of IO for mUC, comprehensive genomic profiling data available from CLIA certified labs. The Bayesian Network (BN, PGM based AI) was used to discover clinical characteristics and selected genomic alterations relevant to IO response by RECIST 1.1 (investigator assessed). Results: Overall, 95 pts (73 men) with mUC were evaluated. 45 (47%) were ever smokers.The presented BN correctly captured the clinical landscape of mUC explaining significant relationship between included variables (p<0.0001). Ever smokers and pts with de novo metastasis had higher TMB and better response to IO. Inactivating MLL2 alterations were more prevalent in non-smokers, and negatively correlated with response to IO. FGFR3 alterations did not predict response to IO. Significant associations are presented in Table. Conclusions: These hypothesis-generating data (by a novel approach, i.e. PGM based AI) showed that smoking and high-TMB were associated with improved response to IO; in contrast, inactivating MLL2 alternations and visceral metastasis predicted inferior response. FGFR3 alterations did not correlate with response. This model validated previous findings and found new hypothesis-generating relationship, such as altered MLL2 gene; external validation is needed.[Table: see text]

1574P - Survival prolongation by rationale innovative genomics (SPRING): An international WIN consortium phase I study exploring safety and efficacy of avelumab, palbociclib, and axitinib in advanced non-small cell lung cancer (NSCLC) with integrated genomic and transcriptomic correlates

1574P - Survival prolongation by rationale innovative genomics (SPRING): An international WIN consortium phase I study exploring safety and efficacy of avelumab, palbociclib, and axitinib in advanced non-small cell lung cancer (NSCLC) with integrated genomic and transcriptomic correlates

Role for immune checkpoint blockade in BRCA2-mutant prostate cancer.

59 Background: Except for rare cases with microsatellite instability, prostate cancer has low mutation burden and low response rate to immune checkpoint blockade (ICB). Genomic correlates of response to ICB in microsatellite stable (MSS) prostate cancer are currently unknown. Here we describe an exceptional response to ICB in BRCA2-mutant prostate cancer and explore immunologic features of BRCA-mutant tumors. Methods: A patient with BRCA2-mutant MSS prostate cancer who progressed on PARP inhibitor therapy was treated with pembrolizumab. Another 8 cases of BRCA2-mutant prostate cancer were evaluated for lymphocyte infiltration and PD-L1 expression by a pathologist. The Cancer Genome Atlas (TCGA) dataset was used to evaluate BRCA-mutant tumors from multiple histology. Presence of putative pathogenic mutations in BRCA1/2 were tested for association with mRNA levels of CD8+ T cell marker ( CD8A), immune checkpoint genes (in PD-1, CTLA-4 pathways), and expression of endogenous retroviruses (ERV). Markers of homologous recombination (HR) defects, namely number of Signature 3 mutations and large ( > 3bp) indels, were also analyzed. Results: Initial patient with BRCA2-mutant MSS prostate cancer had a rapid and ongoing response to pembrolizumab. Analysis of local cohort will be presented. In the TCGA prostate cancer dataset, BRCA2-mutant tumors showed significant overexpression of CD8A, PD-1 and CTLA-4 pathway genes, and ERV3-2. Similar results were observed in BRCA1-mutant (but not BRCA2-mutant) ER+ HER2− breast cancer, but not in BRCA-mutant triple-negative breast and ovarian cancer. Analyses of HR defect markers corroborated these observations. Conclusions: BRCA-mutation accompanied by ERV3-2 overexpression, observed in BRCA2-mutant prostate and BRCA1-mutant ER+ HER2− breast cancer, is associated with CD8+ T cell infiltration and checkpoint pathway upregulation. The effect of BRCA1/2 mutation on immunogenicity may be very tissue and/or cell-of-origin specific, with different cancer types having different levels of immune activation. The association between ERV expression and BRCA2-mutation in prostate cancer requires further investigation. BRCA2-mutant prostate cancers are a class of MSS tumors that may be sensitive to ICB.

Abstract 706: In vivo tumor-associated mutation screen identifies PI3K activation as a mechanism of resistance to PD-1 blockade

Abstract Known genomic correlates of response to immunotherapy do not perfectly predict clinical outcome, supporting the existence of unknown mechanisms of resistance to tumor immunity. We hypothesize that somatic, cancer-associated mutations account for heterogeneity in spontaneous response to tumors and response to immunotherapy. We have undertaken a systematic in vivo screen to identify mechanisms of resistance to tumor immunity in order to define a comprehensive set of therapeutic targets and provide biomarkers of sensitivity to immunotherapy. Mouse tumor cell lines (MC38 colon carcinoma or B16 melanoma) were engineered to express a library of barcoded open reading frames (ORFs) mutagenized to encode known cancer-associated somatic mutations. Tumor-bearing animals were treated with anti-PD-1. A mutation in Phospho-Inositol 3 Kinase (PI3K), PIK3CA c.3140A>G, increased in representation in anti-PD-1 treated tumors but not in immunodeficient animals, suggesting that activity of the mutant allele conferred selective growth advantage in the setting of tumor immunity. This mutation encodes a constitutively active mutant catalytic domain, PIK3CA H1047R. MC38 tumors homogeneously expressing H1047R and implanted into wild-type mice failed to respond to anti-PD-1 therapy, while tumors expressing a control gene regressed. Pharmacologic PI3K inhibition resensitized tumors to treatment with anti-PD-1. PD-1-treated PIK3CA H1047R tumors had fewer infiltrating CD8+ T cells as measured by immunohistochemistry and flow cytometry. Single-cell RNA-seq of tumor-infiltrating immune cells revealed a population of myeloid cells expressing known immune inhibitory proteins that differentially enriched in PIK3CA H1047R-expressing tumors. Our data suggest that PI3K has, in addition to its well-described oncogenic role, a role in tumor immune evasion mediated by establishment of an inhibitory myeloid microenvironment. As such, activating mutations in PI3K may be useful as a biomarker of poor response to immunotherapy, and these studies provide a rationale for therapeutic combination trials of PI3K inhibition with checkpoint blockade and other myeloid-targeting immunotherapies. Citation Format: Natalie B. Collins, Brian C. Miller, Kevin Bi, Rose Al Abosy, Kathleen Yates, Yashaswi Shrestha, John Doench, Jesse Boehm, W Nicholas Haining. In vivo tumor-associated mutation screen identifies PI3K activation as a mechanism of resistance to PD-1 blockade [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 706.

Abstract IA26: Genomic correlates of response and resistance to immune checkpoint blockade

Abstract The rise of multiple immunotherapies, such as immune checkpoint blockade, has significantly improved outcomes for multiple patient populations, including melanoma, lung adenocarcinoma, and urothelial cancers, among others. In parallel, the expansion of precision cancer medicine through robust interrogation of cancer genomes at the point of care has enhanced patient stratification for multiple treatment selection strategies and cancer types. The intersection of clinical cancer genomics and immuno-oncology has created an opportunity to learn how the tumor and germline genomes, along with transcriptomic measurements of the immune microenvironment, impact clinical immune checkpoint blockade response. Here, we describe the current landscape of tumor and microenvironmental genomic correlates of immune checkpoint blockade across cancer subtypes, and examine how emerging genomic studies of immunotherapy responders and non-responders may inform patient selection strategies for clinical trial design in these diverse clinical settings. We explore how these approaches, paired with emerging computational biology methodology, may also enable discovery of new mechanisms of selective treatment response. Finally, we examine emerging genomic data on clinical resistance to immunotherapies, from both signaling pathway and neoantigen evolution perspectives, and discuss the implications for combination treatment strategies. Overall, the convergence of comprehensive clinical genomics and immuno-oncology treatment strategies may inform precision medicine for immunotherapies across cancer types. Citation Format: Eliezer M. Van Allen. Genomic correlates of response and resistance to immune checkpoint blockade. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2016 Oct 20-23; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2017;5(3 Suppl):Abstract nr IA26.

Abstract A16: Defining molecular mechanisms of resistance to tumor immunity

Abstract Recent success of immune checkpoint blockade solidifies the importance of the immune system in the defense against cancer. The clinical impact of the immune response is, however, very heterogeneous, with some patients achieving dramatic responses while others fail to respond. Known genomic correlates of response to immunotherapy are not perfectly predictive of clinical outcome, supporting the existence of unknown mechanisms of resistance to tumor immunity. I hypothesize somatic acquired mutations of individual tumors may account for heterogeneity in the spontaneous response to tumors and response to immunotherapy. I have undertaken a systematic in vivo screen to identify mechanisms of resistance to tumor immunity in order to discover new mechanisms of immune resistance, define a comprehensive set of therapeutic targets and provide biomarkers of sensitivity to immunotherapeutic strategies. Mouse tumor cell lines (MC38 colon carcinoma or B16 melanoma) were engineered to express a library of barcoded open reading frames (ORFs) mutagenized to encode known cancer-associated somatic mutations from the Pan Cancer analysis within The Cancer Genome Atlas (TCGA). These cell lines form tumors when implanted subcutaneously in immunocompetent animals. Tumor-bearing animals were then subjected to immunotherapy with either therapeutic vaccination or checkpoint blockade with anti-PD-1. Barcode relative representation was measured by next generation sequencing at the time of tumor implantation and at the time of tumor harvest post-immunotherapy. Barcoded mutant ORFs that confer immune resistance increased in representation under immune pressure in comparison to untreated or immunodeficient animals. A mutation in Phospho-Inositol 3 Kinase (PI3K), PIK3CA c.3140A>G, consistently increased in representation in both B16 and MC38 immunotherapy-treated tumors. This mutation encodes a constitutively active mutant catalytic domain of PI3K, PIK3CA H1047R. MC38 tumors homogenously expressing PIK3CA H1047R and implanted into wild type mice failed to respond to anti-PD-1 therapy, while tumors expressing a control gene or non-scoring mutation in PI3K regressed after treatment with anti-PD-1. Pharmacologic PI3K inhibition resensitized tumors to treatment with anti-PD-1. PD-1-treated PIK3CA H1047R tumors had fewer infiltrating CD8+ T cells as measured by immunohistochemistry and flow cytometry of tumor infiltrating lymphocytes. I conclude that PI3K has, in addition to its well-described oncogenic role, a role in tumor immune evasion. As such, activation of PI3K may be useful as a predictor of resistance to immunotherapy. Importantly, these findings also provide a rationale for therapeutic combination trials of immune checkpoint blockade and PI3K inhibition. Citation Format: Natalie B. Collins, Robert Manguso, Hans Pope, W. Nicholas Haining. Defining molecular mechanisms of resistance to tumor immunity. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2016 Oct 20-23; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2017;5(3 Suppl):Abstract nr A16.

Abstract OT2-01-10: Pilot study of carboplatin, nab-paclitaxel and pembrolizumab for metastatic triple-negative breast cancer (ongoing clinical trial)

Abstract BACKGROUND. Triple-negative breast cancer (TNBC) is associated with an aggressive phenotype and decreased survival. TNBC is characterized by tumor-infiltrating lymphocytes (TIL) which predict for a better prognosis and likely reflect immune recognition of tumor-associated antigens by TIL. However, potent immune suppressive signals exist in the tumor microenvironment such as those mediated by PD-1 with its ligand, PD-L1. Therefore, to test the validity of decreasing PD-1/PD-L1-mediated immune suppression, a Phase Ib study of single-agent pembrolizumab in 32 patients with advanced TNBC showed a partial response of 16.1% and stable disease of 9.7%, thereby attesting to the effectiveness of single-agent pembrolizumab in these patients. Other studies have demonstrated that cytotoxic chemotherapy favorably modulates immunity against cancer and there is therefore a strong rationale to combine chemotherapy with an immune modulator such as pembrolizumab for the treatment of mTNBC. TRIAL DESIGN. This is an investigator-initiated, industry-sponsored (Merck) pilot study of carboplatin (C), nab-paclitaxel (N) and pembrolizumab (P) in 30 patients with metastatic (m) TNBC. Eligible patients will receive 3 cycles of CNP, with each cycle consisting of C (AUC 6 on days 1 of a 21-day cycle), N (100 mg/m2 IV on days 1, 8 and 15 of a 21-day cycle), and P (200 mg IV on day 15 of each cycle). After completion of 3 cycles CNP, patients with responding or stable disease by RECIST 1.1 criteria will be eligible for additional cycle(s) of CNP. ELIGIBILITY CRITERIA. Patients must have radiologically measurable mTNBC, an ECOG performance status of 0-1, must not have received more than 2 prior therapies for this disease, and must be willing to undergo a preliminary biopsy of a metastatic focus for research purposes. A second post-treatment biopsy will be encouraged but will not be mandated. SPECIFIC AIMS. The primary objective is to determine overall response rate (ORR) in patients treated with CNP. The secondary objectives are to determine progression-free survival (PFS) and safety/tolerability of CNP. Correlative objectives include the identification of pathologic and genomic correlates of response to CNP. STATISTICAL METHODS. Clinical response will be scored using RECIST 1.1 criteria. Under the proposed treatment, the expected clinical response is about 35%. With the precision of the 2-sided 95% confidence interval for the response rate set to 0.17 (the distance to the expected response rate of 35%), the sample size required for the study is 30 patients. The true response rate of therapy will be estimated based on the number of responses using a binomial distribution and its confidence intervals will be estimated using Wilson's method. The Kaplan-Meier method will be used to estimate PFS. Factors including pathologic and genomic correlates that predict survival outcomes will be identified by Cox model or extensions of the Cox model. TARGET ACCRUAL. We plan to enroll 30 patients over 2 years, with the first patient expected to be enrolled in September 2016. CONTACT INFORMATION. Joseph Baar, MD, PhD. Seidman Cancer Center of University Hospitals Case Medical Center. E-mail: joseph.baar@uhhospitals.org. Citation Format: Baar J, Abraham J, Silverman P, Budd GT, Vinayak S, Varadan V, Moore H, Montero A, Fu P. Pilot study of carboplatin, nab-paclitaxel and pembrolizumab for metastatic triple-negative breast cancer (ongoing clinical trial) [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 OT2-01-10.