Abstract 4059: Identification of pre-existing neoantigen-specific T cells in patients receiving checkpoint inhibitor therapy using a deep learning antigen prediction model

Immune Checkpoint Blockade (ICB) has revolutionized the treatment of patients with cancer but provides durable clinical benefit to only a minority of patients. Neoantigens, antigens derived from tumor mutations, are emerging as key targets of T cells in patients with clinical responses to ICB whose tumors exhibit high mutational burden. T cells specific to some neoantigens appear to be naturally generated in these patients and can be further expanded upon ICB. Accurate neoantigen identification may fuel therapeutic advances by driving potent anti-neoantigen immune responses through therapeutic immunization or adoptive cell therapy. However, identification of neoantigens and neoantigen-reactive T cells remains a significant challenge. Current methods are time consuming and labor intensive and require non-routine clinical specimen collections. Here, we apply deep learning to a large HLA peptide and genomic dataset from NSCLC patients to create a computational model of minimal epitope presentation for neoantigen prediction. We demonstrate that our model allows rapid identification of neoantigens and neoantigen reactive T cells using routine clinical biopsies and blood specimens. This methodology does not require HLA-specific reagents and enables identification of T cell responses for every patient. To identify neoantigen-specific T cells, selected peptides, prioritized by our prediction model, were synthesized as minimal epitopes and added to short-term in vitro stimulation (IVS) cultures to expand neoantigen-reactive T cells. The presence of antigen-specific T cells was assessed two weeks later using IFN-gamma (IFN-g) ELISpot against pools of the prioritized neoantigens. Patient-specific peptide pools and IFN-g ELISpot enabled identification of neoantigen-reactive T cells in 5/9 (56%) patients on ICB. Deconvolution yielded an average of 2 recognized neoantigens per patient, with increased responses over time observed in one patient where multiple time points were available. Neoantigen responses were confirmed in repeat IVS cultures where additional samples were available. Importantly, expansion of HLA-matched healthy controls in the presence of patient peptide pools did not result in any responses by IFN-g ELISpot. In addition, antigen-reactive T cells were sorted for single cell sequencing on the 10x platform using CD137 as an HLA-independent T-cell activation marker. In summary, pre-existing neoantigen specific T cells and TCR clones can be rapidly identified utilizing routine clinical specimens, through (i) a powerful neoantigen prediction model, (ii) short-term in vitro T-cell expansion followed by IFN-gamma ELISpot, and (iii) HLA-independent selection and single cell sequencing. This methodology may allow rapid design of neoantigen-targeting vaccines or selection of TCRs for cell therapeutic approaches.

Citation Format: Christine D. Palmer, Brendan Bulik, Aaron Yang, Meghan Hart, Matthew Davis, Joshua Francis, Fujiko Duke, Rita Zhou, Jennifer Busby, Tyler Murphy, Andrew Clark, Lauren Young, Mike Zhong, Kamilah Caldwell, Jesse Abhyankar, Alison Cook, Tommy Boucher, Raphaël Rousseau, Cynthia Voong, Naiyer Rizvi, Mark Frattini, Roman Yelensky, Karin Jooss, Mojca Skoberne. Identification of pre-existing neoantigen-specific T cells in patients receiving checkpoint inhibitor therapy using a deep learning antigen prediction model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4059.

Abstract 4445: MHC class II antigen identification for cancer immunotherapy by deep learning on tumor HLA peptides

A key goal in immuno-oncology is the identification of tumor antigens recognized by T cells. Significant progress has been made in predicting MHC class I presentation of tumor-specific antigens (peptides) recognized by CD8 reactive T cells. However, predicting antigens recognized by CD4 T cells that are presented by the MHC class II has proven to be more challenging. Studies have shown binding affinity to be less predictive for MHC class II presentation than for class I. Class II antigen-directed therapeutics require accurate antigen identification from patient samples, which remains elusive today.

Methods: We focused initially on antigen presentation by the HLA-DR and generated a dataset of human tumor transcriptomes and HLA-DR immunopeptidomes from resections of B cell lymphomas (N=39). Transcriptomes were obtained by NGS of exome-captured cDNA and immunopeptidomes by immunoprecipitation using the HLA-DR specific Ab L243 and MS/MS. Each sample was typed for HLA-DRB1,3,4,5 using standard methods. Additionally, we obtained published class II mass spectrometry data for two B cell lines, each of which expressed a single common HLA class II allele (HLA-DRB1*15:01 and HLA-DRB5*01:01). RNA sequencing data was not available for either cell line; therefore, we substituted RNA-sequencing data from a different B cell line, B721.221. We combined these data to train a deep learning model of Class II HLA peptide presentation. Our model addressed two key challenges: (1) learning HLA-allele-specific models from tumor and normal data where each sample expressed up to 4 unique HLA-DR alleles and (2) reflecting information about all aspects of HLA presentation, including gene expression, antigen processing and stable binding of peptides to HLA. We evaluated the performance of the model on two independent test datasets. First, we tested the model on HLA peptides from a held-out sample from the lymphoma training dataset. Second, we tested the model on a separate public cell line expressing both HLA-DRB1*15:01 and HLA-DRB5*01:01.

Results: An average of 567 training and 203 testing peptides at q<0.01 and >50M unique transcriptome reads were obtained from the lymphoma samples. From cell-lines, 433 peptides for training and 223 peptides for testing were used. The model demonstrated a significant improvement in prediction accuracy, achieving >10%-point gain in area under the ROC curve (ROC AUC) vs a standard binding affinity-based predictor. On the held-out lymphoma test data, it achieved a ROC AUC of .95 vs the binding affinity model ROC AUC .80. On the cell-line test data, it achieved a ROC AUC of .90 vs the binding affinity model ROC AUC of .79.

Conclusion: We used a large dataset of transcriptomes and HLA peptidomes to train a deep learning model for HLA class II antigen presentation. The new model significantly outperforms standard methods and advances in silico HLA class II antigen selection for personalized cancer immunotherapy.

Citation Format: Tommy Boucher, Matthew Davis, Christine Palmer, Tyler Murphy, Andrew Clark, Fujiko Duke, Aaron Yang, Lauren Young, Karin Jooss, Mojca Skoberne, Josh Francis, Roman Yelensky, James Sun, Jennifer Busby. MHC class II antigen identification for cancer immunotherapy by deep learning on tumor HLA peptides [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4445.

Deep learning using tumor HLA peptide mass spectrometry datasets improves neoantigen identification

Neoantigens, which are expressed on tumor cells, are one of the main targets of an effective antitumor T-cell response. Cancer immunotherapies to target neoantigens are of growing interest and are in early human trials, but methods to identify neoantigens either require invasive or difficult-to-obtain clinical specimens, require the screening of hundreds to thousands of synthetic peptides or tandem minigenes, or are only relevant to specific human leukocyte antigen (HLA) alleles. We apply deep learning to a large (N = 74 patients) HLA peptide and genomic dataset from various human tumors to create a computational model of antigen presentation for neoantigen prediction. We show that our model, named EDGE, increases the positive predictive value of HLA antigen prediction by up to ninefold. We apply EDGE to enable identification of neoantigens and neoantigen-reactive T cells using routine clinical specimens and small numbers of synthetic peptides for most common HLA alleles. EDGE could enable an improved ability to develop neoantigen-targeted immunotherapies for cancer patients.

Genomic profiling of ER+ breast cancers after short-term estrogen suppression reveals alterations associated with endocrine resistance

Inhibition of proliferation in estrogen receptor-positive (ER⁺) breast cancers after short-term antiestrogen therapy correlates with long-term patient outcome. We profiled 155 ER⁺/human epidermal growth factor receptor 2-negative (HER2^-) early breast cancers from 143 patients treated with the aromatase inhibitor letrozole for 10 to 21 days before surgery. Twenty-one percent of tumors remained highly proliferative, suggesting that these tumors harbor alterations associated with intrinsic endocrine therapy resistance. Whole-exome sequencing revealed a correlation between 8p11-12 and 11q13 gene amplifications, including FGFR1 and CCND1, respectively, and high Ki67. We corroborated these findings in a separate cohort of serial pretreatment, postneoadjuvant chemotherapy, and recurrent ER⁺ tumors. Combined inhibition of FGFR1 and CDK4/6 reversed antiestrogen resistance in ER⁺FGFR1/CCND1 coamplified CAMA1 breast cancer cells. RNA sequencing of letrozole-treated tumors revealed the existence of intrachromosomal ESR1 fusion transcripts and increased expression of gene signatures indicative of enhanced E2F-mediated transcription and cell cycle processes in cancers with high Ki67. These data suggest that short-term preoperative estrogen deprivation followed by genomic profiling can be used to identify druggable alterations that may cause intrinsic endocrine therapy resistance.

Biological and clinical evidence for somatic mutations in BRCA1 and BRCA2 as predictive markers for olaparib response in high-grade serous ovarian cancers in the maintenance setting

To gain a better understanding of the role of somatic mutations in olaparib response, next-generation sequencing (NGS) of BRCA1 and BRCA2 was performed as part of a planned retrospective analysis of tumors from a randomized, double-blind, Phase II trial (Study 19; D0810C00019; NCT00753545) in 265 patients with platinum-sensitive high-grade serous ovarian cancer. BRCA1/2 loss-of-function mutations were found in 55% (114/209) of tumors, were mutually exclusive, and demonstrated high concordance with Sanger-sequenced germline mutations in matched blood samples, confirming the accuracy (97%) of tumor BRCA1/2 NGS testing. Additionally, NGS identified somatic mutations absent from germline testing in 10% (20/209) of the patients. Somatic mutations had >80% biallelic inactivation frequency and were predominantly clonal, suggesting that BRCA1/2 loss occurs early in the development of these cancers. Clinical outcomes between placebo- and olaparib-treated patients with somatic BRCA1/2 mutations were similar to those with germline BRCA1/2 mutations, indicating that patients with somatic BRCA1/2 mutations benefit from treatment with olaparib.

A computational approach to distinguish somatic vs. germline origin of genomic alterations from deep sequencing of cancer specimens without a matched normal

A key constraint in genomic testing in oncology is that matched normal specimens are not commonly obtained in clinical practice. Thus, while well-characterized genomic alterations do not require normal tissue for interpretation, a significant number of alterations will be unknown in whether they are germline or somatic, in the absence of a matched normal control. We introduce SGZ (somatic-germline-zygosity), a computational method for predicting somatic vs. germline origin and homozygous vs. heterozygous or sub-clonal state of variants identified from deep massively parallel sequencing (MPS) of cancer specimens. The method does not require a patient matched normal control, enabling broad application in clinical research. SGZ predicts the somatic vs. germline status of each alteration identified by modeling the alteration’s allele frequency (AF), taking into account the tumor content, tumor ploidy, and the local copy number. Accuracy of the prediction depends on the depth of sequencing and copy number model fit, which are achieved in our clinical assay by sequencing to high depth (>500x) using MPS, covering 394 cancer-related genes and over 3,500 genome-wide single nucleotide polymorphisms (SNPs). Calls are made using a statistic based on read depth and local variability of SNP AF. To validate the method, we first evaluated performance on samples from 30 lung and colon cancer patients, where we sequenced tumors and matched normal tissue. We examined predictions for 17 somatic hotspot mutations and 20 common germline SNPs in 20,182 clinical cancer specimens. To assess the impact of stromal admixture, we examined three cell lines, which were titrated with their matched normal to six levels (10–75%). Overall, predictions were made in 85% of cases, with 95–99% of variants predicted correctly, a significantly superior performance compared to a basic approach based on AF alone. We then applied the SGZ method to the COSMIC database of known somatic variants in cancer and found >50 that are in fact more likely to be germline.

We introduce SGZ, a computational method for predicting somatic vs. germline origin and homozygous vs. heterozygous or sub-clonal state of variants identified from deep massively parallel sequencing of clinical formalin-fixed, paraffin embedded (FFPE) cancer specimens. The method does not require fresh tissue or a patient matched normal control, enabling broad application in clinical research. It supports functional prioritization and interpretation of alterations discovered on routine testing and may inform clinical decision making and ultimately expand treatment choices for cancer patients.

Abstract 629: An integrated genomic and proteomic analysis of human tumors reveals key factors in neoantigen identification and enables epitope prediction for cancer immunotherapy

Background: Immunotherapy has become essential in the treatment of advanced cancer, with the advent of drug approvals in multiple indications and functional cures for a minority of patients. However, overall survival gains from immunotherapy have been limited and substantial progress remains. The discovery of tumor-specific neoantigens as principal targets of T cells unleashed by immune checkpoint blockade was a key breakthrough in immunotherapy research. Although example neoantigens have been reported, the impact of genomic features on HLA presentation and neoantigen identification has not been comprehensively assessed. Moreover, in silico methods for predicting which mutations lead to neoantigens are currently inadequate for most therapeutic applications, with a false positive rate of >95%.

Methods: We performed deep tumor/normal exome sequencing (Illumina-based, average unique depth 557x/149x), tumor transcriptome sequencing (avg. 111M unique reads), and mass spec HLA class I peptide sequencing (Thermo Lumos, avg. 3,482 peptides) on fresh-frozen primary tumor specimens from 23 patients (22 NSCLC, 1 OC). Somatic mutations were called with a custom GATK-based pipeline, gene expression assessed with RSEM, and peptides matched to mass spectra using COMET/Percolator. We examined the effect of NGS parameters on neoantigen identification, analyzed HLA presentation as a function of genomic features, and created a novel deep learning model for epitope prediction.

Results: We identified 6,252 nonsynonymous somatic mutations across the cohort (avg. 272 per patient) and found 3,874 (62%) transcribed in mRNA. Down-sampling tumor DNA NGS data to a more typical average unique coverage of ~200x and RNA to ~60M reads revealed 3,275 transcribed mutations, an overall 15% loss of candidate neoantigens and >25% loss in low tumor content samples, demonstrating the importance of sequencing depth. Gene expression measurement at the RNA level predicted HLA presentation more strongly than expected, with peptides from genes detected in RNA found presented >20X more often than peptides from non-detected genes. With this and public datasets we trained a model to predict HLA presentation and compared it to standard binding affinity prediction (e.g., NetMHC or MHCflurry) at the strong-binder 50nM threshold for well-characterized alleles. Whereas binding affinity-based prediction gave a positive predictive value (PPV) of only 3%, our integrated deep learning approach showed a PPV of >25% at equivalent recall, a near ten-fold gain.

Conclusions: A novel genomic and proteomic analysis of human tumors establishes NGS requirements for neoantigen identification, characterizes key correlates of HLA presentation, and enables prediction for personalized cancer immunotherapy research.

Citation Format: Brendan Bulik-Sullivan, Jennifer Busby, Matthew Davis, Andrew Clark, Tyler Murphy, Michele Busby, Fujiko Duke, Jessica Smith, Jason Hudak, Adnan Derti, Josh Francis, Roman Yelensky. An integrated genomic and proteomic analysis of human tumors reveals key factors in neoantigen identification and enables epitope prediction for cancer immunotherapy [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 629. doi:10.1158/1538-7445.AM2017-629

Rucaparib in relapsed, platinum-sensitive high-grade ovarian carcinoma (ARIEL2 Part 1): an international, multicentre, open-label, phase 2 trial

BACKGROUND

Poly(ADP-ribose) polymerase (PARP) inhibitors have activity in ovarian carcinomas with homologous recombination deficiency. Along with BRCA1 and BRCA2 (BRCA) mutations genomic loss of heterozygosity (LOH) might also represent homologous recombination deficiency. In ARIEL2, we assessed the ability of tumour genomic LOH, quantified with a next-generation sequencing assay, to predict response to rucaparib, an oral PARP inhibitor.

METHODS

ARIEL2 is an international, multicentre, two-part, phase 2, open-label study done at 49 hospitals and cancer centres in Australia, Canada, France, Spain, the UK, and the USA. In ARIEL2 Part 1, patients with recurrent, platinum-sensitive, high-grade ovarian carcinoma were classified into one of three predefined homologous recombination deficiency subgroups on the basis of tumour mutational analysis: BRCA mutant (deleterious germline or somatic), BRCA wild-type and LOH high (LOH high group), or BRCA wild-type and LOH low (LOH low group). We prespecified a cutoff of 14% or more genomic LOH for LOH high. Patients began treatment with oral rucaparib at 600 mg twice per day for continuous 28 day cycles until disease progression or any other reason for discontinuation. The primary endpoint was progression-free survival. All patients treated with at least one dose of rucaparib were included in the safety analyses and all treated patients who were classified were included in the primary endpoint analysis. This trial is registered with ClinicalTrials.gov, number NCT01891344. Enrolment into ARIEL2 Part 1 is complete, although an extension (Part 2) is ongoing.

FINDINGS

256 patients were screened and 206 were enrolled between Oct 30, 2013, and Dec 19, 2014. At the data cutoff date (Jan 18, 2016), 204 patients had received rucaparib, with 28 patients remaining in the study. 192 patients could be classified into one of the three predefined homologous recombination deficiency subgroups: BRCA mutant (n=40), LOH high (n=82), or LOH low (n=70). Tumours from 12 patients were established as BRCA wild-type, but could not be classified for LOH, because of insufficient neoplastic nuclei in the sample. The median duration of treatment for the 204 patients was 5·7 months (IQR 2·8-10·1). 24 patients in the BRCA mutant subgroup, 56 patients in the LOH high subgroup, and 59 patients in the LOH low subgroup had disease progression or died. Median progression-free survival after rucaparib treatment was 12·8 months (95% CI 9·0-14·7) in the BRCA mutant subgroup, 5·7 months (5·3-7·6) in the LOH high subgroup, and 5·2 months (3·6-5·5) in the LOH low subgroup. Progression-free survival was significantly longer in the BRCA mutant (hazard ratio 0·27, 95% CI 0·16-0·44, p<0·0001) and LOH high (0·62, 0·42-0·90, p=0·011) subgroups compared with the LOH low subgroup. The most common grade 3 or worse treatment-emergent adverse events were anaemia or decreased haemoglobin (45 [22%] patients), and elevations in alanine aminotransferase or aspartate aminotransferase (25 [12%]). Common serious adverse events included small intestinal obstruction (10 [5%] of 204 patients), malignant neoplasm progression (10 [5%]), and anaemia (nine [4%]). Three patients died during the study (two because of disease progression and one because of sepsis and disease progression). No treatment-related deaths occurred.

INTERPRETATION

In patients with BRCA mutant or BRCA wild-type and LOH high platinum-sensitive ovarian carcinomas treated with rucaparib, progression-free survival was longer than in patients with BRCA wild-type LOH low carcinomas. Our results suggest that assessment of tumour LOH can be used to identify patients with BRCA wild-type platinum-sensitive ovarian cancers who might benefit from rucaparib. These results extend the potential usefulness of PARP inhibitors in the treatment setting beyond BRCA mutant tumours.

FUNDING

Clovis Oncology, US Department of Defense Ovarian Cancer Research Program, Stand Up To Cancer-Ovarian Cancer Research Fund Alliance-National Ovarian Cancer Coalition Dream Team Translational Research Grant, and V Foundation Translational Award.

Triple-negative breast cancers with amplification of JAK2 at the 9p24 locus demonstrate JAK2-specific dependence

Amplifications at 9p24 have been identified in breast cancer and other malignancies, but the genes within this locus causally associated with oncogenicity or tumor progression remain unclear. Targeted next-generation sequencing of postchemotherapy triple-negative breast cancers (TNBCs) identified a group of 9p24-amplified tumors, which contained focal amplification of the Janus kinase 2 (JAK2) gene. These patients had markedly inferior recurrence-free and overall survival compared to patients with TNBC without JAK2 amplification. Detection of JAK2/9p24 amplifications was more common in chemotherapy-treated TNBCs than in untreated TNBCs or basal-like cancers, or in other breast cancer subtypes. Similar rates of JAK2 amplification were confirmed in patient-derived TNBC xenografts. In patients for whom longitudinal specimens were available, JAK2 amplification was selected for during neoadjuvant chemotherapy and eventual metastatic spread, suggesting a role in tumorigenicity and chemoresistance, phenotypes often attributed to a cancer stem cell-like cell population. In TNBC cell lines with JAK2 copy gains or amplification, specific inhibition of JAK2 signaling reduced mammosphere formation and cooperated with chemotherapy in reducing tumor growth in vivo. In these cells, inhibition of JAK1-signal transducer and activator of transcription 3 (STAT3) signaling had little effect or, in some cases, counteracted JAK2-specific inhibition. Collectively, these results suggest that JAK2-specific inhibitors are more efficacious than dual JAK1/2 inhibitors against JAK2-amplified TNBCs. Furthermore, JAK2 amplification is a potential biomarker for JAK2 dependence, which, in turn, can be used to select patients for clinical trials with JAK2 inhibitors.

Clinical Actionability of Comprehensive Genomic Profiling for Management of Rare or Refractory Cancers

BACKGROUND

The frequency with which targeted tumor sequencing results will lead to implemented change in care is unclear. Prospective assessment of the feasibility and limitations of using genomic sequencing is critically important.

METHODS

A prospective clinical study was conducted on 100 patients with diverse-histology, rare, or poor-prognosis cancers to evaluate the clinical actionability of a Clinical Laboratory Improvement Amendments (CLIA)-certified, comprehensive genomic profiling assay (FoundationOne), using formalin-fixed, paraffin-embedded tumors. The primary objectives were to assess utility, feasibility, and limitations of genomic sequencing for genomically guided therapy or other clinical purpose in the setting of a multidisciplinary molecular tumor board.

RESULTS

Of the tumors from the 92 patients with sufficient tissue, 88 (96%) had at least one genomic alteration (average 3.6, range 0-10). Commonly altered pathways included p53 (46%), RAS/RAF/MAPK (rat sarcoma; rapidly accelerated fibrosarcoma; mitogen-activated protein kinase) (45%), receptor tyrosine kinases/ligand (44%), PI3K/AKT/mTOR (phosphatidylinositol-4,5-bisphosphate 3-kinase; protein kinase B; mammalian target of rapamycin) (35%), transcription factors/regulators (31%), and cell cycle regulators (30%). Many low frequency but potentially actionable alterations were identified in diverse histologies. Use of comprehensive profiling led to implementable clinical action in 35% of tumors with genomic alterations, including genomically guided therapy, diagnostic modification, and trigger for germline genetic testing.

CONCLUSION

Use of targeted next-generation sequencing in the setting of an institutional molecular tumor board led to implementable clinical action in more than one third of patients with rare and poor-prognosis cancers. Major barriers to implementation of genomically guided therapy were clinical status of the patient and drug access. Early and serial sequencing in the clinical course and expanded access to genomically guided early-phase clinical trials and targeted agents may increase actionability.

IMPLICATIONS FOR PRACTICE

Identification of key factors that facilitate use of genomic tumor testing results and implementation of genomically guided therapy may lead to enhanced benefit for patients with rare or difficult to treat cancers. Clinical use of a targeted next-generation sequencing assay in the setting of an institutional molecular tumor board led to implementable clinical action in over one third of patients with rare and poor prognosis cancers. The major barriers to implementation of genomically guided therapy were clinical status of the patient and drug access both on trial and off label. Approaches to increase actionability include early and serial sequencing in the clinical course and expanded access to genomically guided early phase clinical trials and targeted agents.

TP53 Alterations Correlate with Response to VEGF/VEGFR Inhibitors: Implications for Targeted Therapeutics

TP53 tumor-suppressor gene mutations are among the most frequent abnormalities in cancer, affecting approximately 40% of patients. Yet, there is no accepted way to target these alterations in the clinic. At the same time, antagonists of VEGFR or its ligand are best-selling oncology drugs, with multiple, expensive compounds approved. Although only a subset of patients benefit from these antiangiogenesis agents, no relevant biomarker has been identified. Interestingly, TP53 mutations upregulate VEGF-A and VEGFR2. We prospectively enrolled 500 patients, to be interrogated by comprehensive genomic profiling (CGP) (next-generation sequencing, 236 genes), and to be matched, whenever possible, with targeted agents. Herein, we analyze outcomes based on VEGF/VEGFR inhibitor treatment and presence of TP53 mutations. Of the 500 patients, 188 (37.6%; with ≥1 alteration) were treated; 106 (56% of 188) had tumors that harbored TP53 mutations. VEGF/VEGFR inhibitor therapy was independently associated with improvement in all outcome parameters [rate of stable disease (SD) ≥6 months/partial and complete remission (PR/CR); (31% versus 7%; TP53-mutant patients (who received no other molecular-matched agents) treated with versus without VEGF/VEGFR inhibitors), time-to-treatment failure, and overall survival (multivariate analysis: all P ≤ 0.01)] for the patients harboring TP53-mutant cancers, but improvement was not seen in any of these parameters for patients with TP53 wild-type neoplasms. We conclude that TP53 mutations predict sensitivity to VEGF/VEGFR inhibitors in the clinic. TP53 alterations may therefore be a ready biomarker for treatment with antiangiogenesis agents, a finding of seminal importance across the cancer field. Mol Cancer Ther; 15(10); 2475-85. ©2016 AACR.

Abstract 3576: Broad analysis of recurrent somatic mutations in cancer reveals a common novel non-coding mutation in the promoter of PMS2 associated with greatly increased tumor mutation load

Purpose

Tumor mutation load is an emerging prognostic and diagnostic marker for many cancers. Sensitivity to immunotherapeutic agents, which stimulate an antitumor immune response by selectively inhibiting immunosuppressive cell surface ligands, is known to correlate with high mutation load in the tumor. Numerous somatic and germline defects can cause genomic instability, including alterations affecting the mismatch repair pathway, DNA polymerases, and cell cycle regulators. We describe here the discovery of previously unreported mutations in the promoter of the PMS2 gene that are associated with significantly increased tumor mutation load.

Methods

Comprehensive genomic profiling by hybridization capture of exonic regions from either 236 or 315 cancer-related genes and select introns from 19 genes commonly rearranged in cancer was used to characterize more than 60,000 clinical FFPE cancer specimens. At least 50 ng of extracted DNA was analyzed per sample and the constructed libraries were sequenced to high, uniform median coverage (&gt;500x). Samples were assessed for base substitutions, short insertions and deletions, copy number alterations and gene fusions/rearrangements. Mutation load is assessed as the number of somatic coding point mutations per megabase of targeted territory.

Results

Mutation load from targeted cancer gene analysis recapitulates previous results evaluating whole exome and whole genome mutation load in tumors and cell lines. A novel mutation hotspot was identified in the promoter of PMS2, which codes for the PMS2 protein, a dimerization partner of MLH1 and integral to the DNA mismatch repair complex. Promoter mutations were found in 7.5% of melanoma specimens (n = 101/1348) and 17% of skin squamous cell carcinomas (n = 30/175). In both diseases, PMS2 promoter mutations are the most significant genomic correlate of high mutation load. In melanoma, PMS2 promoter mutant specimens have 4x the median mutation load of the general melanoma population. Skin squamous cell carcinomas show a 2.5-fold increase. Functional characterization is underway to support the hypothesis that these mutations lead to modified PMS2 transcriptional activity, which is known to cause hypermutation.

Conclusions

The growing corpus of cancer genome data continues to enable novel discoveries in cancer biology. Non-coding and regulatory mutations have not been the target of focused study, and our findings extend the small set of regulatory mutations thought to affect tumor development. Our discovery highlights the power of large-scale genomic analysis to uncover additional disease mechanisms, and furthers our understanding of how cells maintain genome integrity. With the documented association of mutation load and immunotherapy sensitivity, investigation into the response of PMS2 promoter mutant tumors is warranted.

Citation Format: Zachary R. Chalmers, Franklin W. Huang, Laurie M. Gay, Siraj M. Ali, Roman Yelensky, Juliann Chmielecki, Jeffery S. Ross, Vincent A. Miller, Philip J. Stephens, Levi A. Garraway, Garrett M. Frampton. Broad analysis of recurrent somatic mutations in cancer reveals a common novel non-coding mutation in the promoter of PMS2 associated with greatly increased tumor mutation load. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3576.

Patient-derived xenotransplants can recapitulate the genetic driver landscape of acute leukemias

Genomic studies have identified recurrent somatic mutations in acute leukemias. However, current murine models do not sufficiently encompass the genomic complexity of human leukemias. To develop pre-clinical models, we transplanted 160 samples from patients with acute leukemia (AML, MLL, B-ALL and T-ALL) into immunodeficient mice. Of these, 119 engrafted with expected immunophenotype. Targeted sequencing of 374 genes and 265 frequently rearranged RNAs detected recurrent and novel genetic lesions in 48 paired primary tumor (PT) and patient-derived xenotransplant (PDX) samples. Overall, the frequencies of 274 somatic variant alleles correlated between PT and PDX samples, although the data were highly variable for variant alleles present at 0-10%. 17% of variant alleles were detected in either PT or PDX samples only. Based on variant allele frequency changes, 24 PT-PDX pairs were classified as concordant while the other 24 pairs showed various degree of clonal discordance. There was no correlation of clonal concordance with clinical parameters of diseases. Significantly more bone marrow samples than peripheral blood samples engrafted discordantly. These data demonstrate the utility of developing PDX banks for modeling human leukemia, and emphasize the importance of genomic profiling of PDX and patient samples to ensure concordance before performing mechanistic or therapeutic studies.

Nonamplification ERBB2 genomic alterations in 5605 cases of recurrent and metastatic breast cancer: An emerging opportunity for anti-HER2 targeted therapies

BACKGROUND

Activating, nonamplification ERBB2 mutations (ERBB2mut) are not detected by immunohistochemistry (IHC) or fluorescence in situ hybridization (FISH), but are detected by DNA sequencing and may predict clinical responses to human epidermal growth factor receptor (HER2)-targeted therapy. The authors queried 5605 advanced/metastatic breast cancers (mBC) to uncover the frequency of ERBB2mut genomic alterations. Clinical responses to anti-HER2 therapeutics were identified.

METHODS

DNA was extracted from 40 µm of formalin-fixed paraffin-embedded (FFPE) sections. Comprehensive genomic profiling (CGP) was used to evaluate up to 315 genes (592× mean coverage depth). Results were analyzed for base substitutions, short indels, copy number changes, and selected rearrangements.

RESULTS

Of 5605 cases, 698 (12.5%) featured ERBB2 alterations, including 596 (10.6%) ERBB2 amplifications (ERBB2amp) and 138 (2.4%) ERBB2mut; 38 cases (0.7%) had co-occurring ERBB2amp and ERBB2mut. ERBB2mut predominantly affected the kinase (124 cases; 90%) or extracellular (15 cases; 11%) domains. Both primary BC (52 cases; 38%) and metastatic site biopsies (86 cases; 62%) were found to harbor ERBB2mut, which were distributed across carcinoma not otherwise specified (NOS) (69 cases; 50%), invasive ductal carcinoma (IDC) (40 cases; 29%), invasive lobular carcinoma (ILC) (27 cases; 20%), and mucinous mBC (2 cases; 1%). Genes commonly coaltered with ERBB2 were tumor protein 53 (TP53) (49%); phosphatidylinositol 3-kinase catalytic subunit alpha (PIK3CA) (42%); cadherin 1, type 1 (CDH1) (37%); MYC (17%); and cyclin D1 protein (CCND1) (16%). CDH1 mutations were enriched in ERBB2mut mBC (P<0.0006) and associated with recurrent mBC. Selected patients with ERBB2mut, without ERBB2amp, who responded to anti-HER2 targeted therapies are presented herein.

CONCLUSIONS

Within this large series, 1.8% of cases harbored ERBB2mut, which are undetectable by standard-of-care IHC or FISH tests. Metastatic BC driven by ERBB2mut respond to anti-HER2 targeted therapies, and expanding clinical trials designed to detect ERBB2mut by CGP and optimize targeted treatments are warranted. Cancer 2016. © 2016 American Cancer Society. Cancer 2016;122:2654-2662. © 2016 American Cancer Society.

Comprehensive Genomic Profiling Identifies a Subset of Crizotinib-Responsive ALK-Rearranged Non-Small Cell Lung Cancer Not Detected by Fluorescence In Situ Hybridization

INTRODUCTION

For patients with non-small cell lung cancer (NSCLC) to benefit from ALK inhibitors, sensitive and specific detection of ALK genomic rearrangements is needed. ALK break-apart fluorescence in situ hybridization (FISH) is the U.S. Food and Drug Administration approved and standard-of-care diagnostic assay, but identification of ALK rearrangements by other methods reported in NSCLC cases that tested negative for ALK rearrangements by FISH suggests a significant false-negative rate. We report here a large series of NSCLC cases assayed by hybrid-capture-based comprehensive genomic profiling (CGP) in the course of clinical care.

MATERIALS AND METHODS

Hybrid-capture-based CGP using next-generation sequencing was performed in the course of clinical care of 1,070 patients with advanced lung cancer. Each tumor sample was evaluated for all classes of genomic alterations, including base-pair substitutions, insertions/deletions, copy number alterations and rearrangements, as well as fusions/rearrangements.

RESULTS

A total of 47 patients (4.4%) were found to harbor ALK rearrangements, of whom 41 had an EML4-ALK fusion, and 6 had other fusion partners, including 3 previously unreported rearrangement events: EIF2AK-ALK, PPM1B-ALK, and PRKAR1A-ALK. Of 41 patients harboring ALK rearrangements, 31 had prior FISH testing results available. Of these, 20 were ALK FISH positive, and 11 (35%) were ALK FISH negative. Of the latter 11 patients, 9 received crizotinib based on the CGP results, and 7 achieved a response with median duration of 17 months.

CONCLUSION

Comprehensive genomic profiling detected canonical ALK rearrangements and ALK rearrangements with noncanonical fusion partners in a subset of patients with NSCLC with previously negative ALK FISH results. In this series, such patients had durable responses to ALK inhibitors, comparable to historical response rates for ALK FISH-positive cases.

IMPLICATIONS FOR PRACTICE

Comprehensive genomic profiling (CGP) that includes hybrid capture and specific baiting of intron 19 of ALK is a highly sensitive, alternative method for identification of drug-sensitive ALK fusions in patients with non-small cell lung cancer (NSCLC) who had previously tested negative using standard ALK fluorescence in situ hybridization (FISH) diagnostic assays. Given the proven benefit of treatment with crizotinib and second-generation ALK inhibitors in patients with ALK fusions, CGP should be considered in patients with NSCLC, including those who have tested negative for other alterations, including negative results using ALK FISH testing.

Comprehensive Genomic Profiling of Clinically Advanced Medullary Thyroid Carcinoma

OBJECTIVE

The aim of this study was to determine the genomic alterations of cancer-related genes in advanced medullary thyroid carcinoma during the course of clinical care.

METHODS

Hybrid-capture-based comprehensive genomic profiling was performed on 34 consecutive medullary thyroid carcinoma cases to identify all four classes of genomic alterations, and outcome for an index patient was collected.

RESULTS

RET was mutated in 88% (30/34) of cases, with RET M918T being responsible for 70% (21/30) of the RET alterations. The other RET alterations were RET E632_L633del, C634R, C620R, C618G/R/S, V804M, and RET amplification. Two of the four RET wild-type patients harbored mutations in KRAS or HRAS (1/34 each). The next most frequent genomic alterations were amplifications of CCND1, FGF3, and FGF19 and alterations in CDKN2A (3/34 each). One case with a RET M918T mutation developed acquired resistance to progressively dose-escalated vandetanib. When the mTOR inhibitor everolimus was added to continued vandetanib treatment, the patient achieved a second 25% reduction of tumor volume (RECIST 1.1) for 8 months.

CONCLUSIONS

Comprehensive genomic profiling identified the full breadth of RET alterations in metastatic medullary thyroid carcinoma and possible cooperating oncogenic driver alterations. This approach may refine the use of targeted therapy for these patients.

Comprehensive genomic profiling of anal squamous cell carcinoma reveals distinct genomically defined classes

BACKGROUND

Squamous cell cancers of the anal canal (ASCC) are increasing in frequency and lack effective therapies for advanced disease. Although an association with human papillomavirus (HPV) has been established, little is known about the molecular characterization of ASCC. A comprehensive genomic analysis of ASCC was undertaken to identify novel genomic alterations (GAs) that will inform therapeutic choices for patients with advanced disease.

PATIENTS AND METHODS

Hybrid-capture-based next-generation sequencing of exons from 236 cancer-related genes and intronic regions from 19 genes commonly rearranged in cancer was performed on 70 patients with ASCC. HPV status was assessed by aligning tumor sequencing reads to HPV viral genomes. GAs were identified using an established algorithm and correlated with HPV status.

RESULTS

Sixty-one samples (87%) were HPV-positive. A mean of 3.5 GAs per sample was identified. Recurrent alterations in phosphoinositol-3-kinase pathway (PI3K/AKT/mTOR) genes including amplifications and homozygous deletions were present in 63% of cases. Clinically relevant GAs in genes involved in DNA repair, chromatin remodeling, or receptor tyrosine kinase signaling were observed in 30% of cases. Loss-of-function mutations in TP53 and CDKN2A were significantly enhanced in HPV-negative cases (P < 0.0001).

CONCLUSIONS

This is the first comprehensive genomic analysis of ASCC, and the results suggest new therapeutic approaches. Differing genomic profiles between HPV-associated and HPV-negative ASCC warrants further investigation and may require novel therapeutic and preventive strategies.

Comprehensive Genomic Profiling of Carcinoma of Unknown Primary Site: New Routes to Targeted Therapies

IMPORTANCE

For carcinoma of unknown primary site (CUP), determining the primary tumor site may be uninformative and often does not improve outcome.

OBJECTIVE

To discover opportunities for targeted therapies in patients with CUP not currently searched for in routine practice.

DESIGN, SETTING, AND PARTICIPANTS

Comprehensive genomic profiling on 200 CUP formalin-fixed paraffin-embedded specimens (mean, 756× coverage) using the hybrid-capture-based FoundationOne assay at academic and community oncology clinics.

MAIN OUTCOMES AND MEASURES

Presence of targetable genomic alterations (GAs) in CUP and responses to targeted therapies.

RESULTS

There were 125 adenocarcinomas of unknown primary site (ACUPs) and 75 carcinomas of unknown primary site without features of adenocarcinoma (non-ACUPs). At least 1 GA was found in 192 (96%) of CUP specimens, with a mean (SD) of 4.2 (2.8) GAs per tumor. The most frequent GAs were in TP53 (110 [55%]), KRAS (40 [20%]), CDKN2A (37 [19%]), MYC (23 [12%]), ARID1A (21 [11%]), MCL1 (19 [10%]), PIK3CA (17 [9%]), ERBB2 (16 [8%]), PTEN (14 [7%]), EGFR (12 [6%]), SMAD4 (13 [7%]), STK11 (13 [7%]), SMARCA4 (12 [6%]), RB1 (12 [6%]), RICTOR (12 [6%]), MLL2 (12 [6%]), BRAF (11 [6%]), and BRCA2 (11 [6%]). One or more potentially targetable GAs were identified in 169 of 200 (85%) CUP specimens. Mutations or amplifications of ERBB2 were more frequent in ACUPs (13 [10%]) than in non-ACUPs (3 [4%]). Alterations of EGFR (10 [8%] vs 2 [3%]) and BRAF (8 [6%] vs 3 [4%]) were more common in ACUPs than in non-ACUPs. Strikingly, clinically relevant alterations in the receptor tyrosine kinase (RTK)/Ras signaling pathway including alterations in ALK, ARAF, BRAF, EGFR, FGFR1, FGFR2, KIT, KRAS, MAP2K1, MET, NF1, NF2, NRAS, RAF1, RET, and ROS1 were found in 90 (72%) ACUPs but in only 29 (39%) non-ACUPs (P < .001).

CONCLUSIONS AND RELEVANCE

Almost all CUP samples harbored at least 1 clinically relevant GA with potential to influence and personalize therapy. The ACUP tumors were more frequently driven by GAs in the highly druggable RTK/Ras/mitogen-activated protein kinase (MAPK) signaling pathway than the non-ACUP tumors. Comprehensive genomic profiling can identify novel treatment paradigms to address the limited options and poor prognoses of patients with CUP.

RAS/MAPK Activation Is Associated with Reduced Tumor-Infiltrating Lymphocytes in Triple-Negative Breast Cancer: Therapeutic Cooperation Between MEK and PD-1/PD-L1 Immune Checkpoint Inhibitors

PURPOSE

Tumor-infiltrating lymphocytes (TIL) in the residual disease (RD) of triple-negative breast cancers (TNBC) after neoadjuvant chemotherapy (NAC) are associated with improved survival, but insight into tumor cell-autonomous molecular pathways affecting these features are lacking.

EXPERIMENTAL DESIGN

We analyzed TILs in the RD of clinically and molecularly characterized TNBCs after NAC and explored therapeutic strategies targeting combinations of MEK inhibitors with PD-1/PD-L1-targeted immunotherapy in mouse models of breast cancer.

RESULTS

Presence of TILs in the RD was significantly associated with improved prognosis. Genetic or transcriptomic alterations in Ras-MAPK signaling were significantly correlated with lower TILs. MEK inhibition upregulated cell surface MHC expression and PD-L1 in TNBC cells both in vivo and in vitro. Moreover, combined MEK and PD-L1/PD-1 inhibition enhanced antitumor immune responses in mouse models of breast cancer.

CONCLUSIONS

These data suggest the possibility that Ras-MAPK pathway activation promotes immune-evasion in TNBC, and support clinical trials combining MEK- and PD-L1-targeted therapies. Furthermore, Ras/MAPK activation and MHC expression may be predictive biomarkers of response to immune checkpoint inhibitors.

Comprehensive Genomic Profiling Identifies Frequent Drug-Sensitive EGFR Exon 19 Deletions in NSCLC not Identified by Prior Molecular Testing

PURPOSE

Reliable detection of drug-sensitive activating EGFR mutations is critical in the care of advanced non-small cell lung cancer (NSCLC), but such testing is commonly performed using a wide variety of platforms, many of which lack rigorous analytic validation.

EXPERIMENTAL DESIGN

A large pool of NSCLC cases was assayed with well-validated, hybrid capture-based comprehensive genomic profiling (CGP) at the request of the individual treating physicians in the course of clinical care for the purpose of making therapy decisions. From these, 400 cases harboring EGFR exon 19 deletions (Δex19) were identified, and available clinical history was reviewed.

RESULTS

Pathology reports were available for 250 consecutive cases with classical EGFR Δex19 (amino acids 743-754) and were reviewed to assess previous non-hybrid capture-based EGFR testing. Twelve of 71 (17%) cases with EGFR testing results available were negative by previous testing, including 8 of 46 (17%) cases for which the same biopsy was analyzed. Independently, five of six (83%) cases harboring C-helical EGFR Δex19 were previously negative. In a subset of these patients with available clinical outcome information, robust benefit from treatment with EGFR inhibitors was observed.

CONCLUSIONS

CGP identifies drug-sensitive EGFR Δex19 in NSCLC cases that have undergone prior EGFR testing and returned negative results. Given the proven benefit in progression-free survival conferred by EGFR tyrosine kinase inhibitors in patients with these alterations, CGP should be considered in the initial presentation of advanced NSCLC and when previous testing for EGFR mutations or other driver alterations is negative. Clin Cancer Res; 22(13); 3281-5. ©2016 AACR.

Abstract P6-03-12: Comprehensive genomic profiling of clinically advanced mucinous carcinoma of the breast

Background: Mucinous carcinoma of the breast (mucBC) is generally associated with a favorable prognosis, but on occasion, may have an aggressive clinical course in which it is commonly refractory to cytotoxic chemotherapy. The low incidence of mucBC (∼2% of breast cancers) precludes the development of consensus based guidelines for management of these relapsed/refractory cases. We performed hybrid-capture based comprehensive genomic profiling (CGP) to identify potential therapy targets not routinely searched for in clinical management of metastatic mucBC.

Methods: DNA was extracted from 40 microns of FFPE sections from 22 cases of stage IV mucBC. Comprehensive genomic profiling (CGP) was performed using a hybrid-capture, adaptor ligation based next generation sequencing assay to a mean coverage depth of &gt;550X. The results were analyzed for all classes of genomic alterations (GA) including base substitutions, insertions and deletions, select rearrangements, and copy number changes. Clinically relevant genomic alterations (CRGA) were defined as those identifying anti-cancer drugs on the market or in registered clinical trials.

Results: The median age of the 22 mucBC patients was 57 years (range 32 to 79 years). Samples were from breast (11), lymph nodes (3), chest wall (2), liver (2), soft tissue (2), bone (1) and pleura (1). Three mucBC were grade 1, 17 were grade 2 and 2 were grade 3. Twenty-one (95%) mucBC were ER+, 19 (86%) were PR+ and 4 (18%) were HER2+ by IHC and/or FISH. There were 129 GA identified on the 22 mucBC (5.9 per tumor) including 51 CRGA with a mean of 2.3 per tumor. Amplifications of FGFR1 and ZNF703 were found in 8 out of 22 cases (36%) on the same amplicon. Other most frequently altered genes were TP53 (32%), CCND1 and FGF3/4/19 often co-amplified together (27%). ERBB2/HER2 alterations were found on 5 cases (23%) including amplifications on all 4 HER2+ cases by IHC and/or FISH, and ERBB2 substitution D769Y on one additional mucBC. CRGA were found on some other 20 genes included PIK3CA (5), BRCA1 (1), TSC2 (1), STK11 (1), AKT3 (1), and ESR1 (1).

Conclusions: The subset of relapsed/refractory mucBC presents a management challenge, but comprehensive genomic profiling offers avenues for benefit from targeted therapy. MucBC relative to breast cancer is predominantly ER+, enriched for FGFR1 amplification, 36% vs 11% from TCGA ER+ breast cancer (N=601) with Fisher's test p-value &lt;0.005. Moreover, metastatic mucBC appears more often to have ERBB2/HER2 alterations (23%) than typical mucBC cured by local treatments. Comprehensive genomic profiling uncovers a variety of genomic targets in metastatic mucBC that could facilitate the introduction of targeted therapies for patients with this challenging disease.

Citation Format: Wang K, Ali SM, Khaira D, Elvin JA, Vergilio J-A, Suh J, Yelensky R, Lipson D, Chmielecki J, Miller VA, Stephens PJ, Ross JS. Comprehensive genomic profiling of clinically advanced mucinous carcinoma of the breast. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P6-03-12.

Abstract P3-07-05: Non-amplification ERBB2 genomic alterations in 5,605 cases of refractory and metastatic breast cancer: An emerging opportunity for anti-HER2 targeted therapies

Background: Non-amplification ERBB2 alterations (ERBB2 mut) in advanced/metastatic breast cancer (mBC) are not detected by IHC or FISH, but when detected by DNA sequencing assays can lead to clinical responses to anti-HER2 targeted therapy. We queried a database of more than 43,000 clinical cases to uncover the frequency, type and associated genomic alterations (GA) in mBC driven by ERBB2 mut and highlight clinical responses to small molecule drug and antibody-based anti-HER2 therapeutics.

Methods: DNA was extracted from 40 microns of FFPE sections from 5,605 mBC. Comprehensive genomic profiling (CGP) was performed using a hybrid-capture, adaptor ligation based next generation sequencing assay of up to 315 genes to a mean coverage depth of &gt;600X. The results were analyzed for base substitutions, short insertions and deletions, selected rearrangements, and copy number changes.

Results: 698 (12.5%) of 5,605 mBC featured ERBB2 alterations. 596 (10.6%) featured ERBB2 amplifications and 137 (2.4%) featured ERBB2mut. 35 (0.6%) of total mBC had both ERBB2amp and ERBB2mut, which accounted for 5.0% of all ERBB2 altered mBC. The 137 ERBB2mut mBC cases had a median age of 61 years (range 29 to 93 years) and were sequenced to a mean depth of 600X. Samples utilized for CGP included 52 (38%) from the patient's primary BC and 85 (62%) from metastatic sites including bone/soft tissue/skin (12%), liver (20%), LN (14%), serous cavities (6%), lung (4%) and miscellaneous sites (6%). 71 (52%) mBC were submitted as carcinoma NOS, 44 (32%) as IDC, 22 (16%) as ILC and 1 (1%) as mucinous mBC. Of the 137 ERBB2mut cases, 8 featured more than 1 ERBB2 mut. There were 124 (85%) ERBB2 kinase domain mutations and 15 (10%) extra-cellular domain ERBB2mut. The most common genes co-altered in ERBB2mut mBC were TP53 (49%), PIK3CA (42%), CDH1 (37%), MYC (17%), and CCND1 (16%). The enrichment of ERBB2mut in CDH1 mut mBR was significant (p=0.0006) and associated with relapsed lobular mBC. Multiple case examples of kinase domain and extra-cellular domain ERBB2mut mBC responding to a variety of anti-HER2 targeted therapies will be presented.

Conclusions: In this large series of 5,605 mBC, 20% of the total ERBB2 alterations were non-amplification ERBB2mut not detectable by standard of care IHC and FISH slide-based HER2 tests. Given the demonstration of ERBB2mut driven mBC responsive to anti-HER2 targeted therapies in this study, expansion of clinical trials designed to detect these ERBB2mut cases with CGP and optimize the targeted therapies for these patients is strongly recommended.

Citation Format: Ross JS, Wang K, Ali SM, Chumsri S, Elvin JA, Vergilio J-A, Suh J, Yelensky R, Lipson D, Chmielecki J, Miller VA, Stephens PJ. Non-amplification ERBB2 genomic alterations in 5,605 cases of refractory and metastatic breast cancer: An emerging opportunity for anti-HER2 targeted therapies. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P3-07-05.

Comprehensive genomic profiling (CGP) to assess mutational load in gastric and esophageal adenocarcinomas: Implications for immunotherapies

66

Background: Gastroesophageal cancer (GEC) present a management challenge, particularly when ERBB2 is not amplified and cytotoxic therapy has failed. Tumor mutational load is linked to predicted benefit from immune checkpoint inhibitors in advanced cancers. Using CGP we assessed the relationship between mutational burden and clinically relevant genomic alterations in GEC samples in the course of routine clinical care. Methods: DNA was extracted from 40 microns of FFPE sections from patients with GEC. CGP was performed on hybridization-captured, adaptor ligation based libraries to a mean coverage depth of greater than 500x for 236 or 315 cancer-related genes. Mutational load was characterized as the number of somatic base substitutions and short indels per megabase. Samples were stratified by histologic subtype, and presence or absence of therapeutically relevant receptor tyrosine kinase (RTK) alterations. The majority of samples were from patients with advanced disease. Results: The genomic profiles from a total of 736 gastric and 862 esophageal carcinomas were assessed. Median patient age was 57 for gastric and 60 for esophageal cases. There was no significant difference in mutational load between gastric and esophageal samples. 10th, 25th, median, 75th, and 90th percentiles of mutation load were 1.3, 2.7, 6.3, 11.3, and 20.0 for gastric cancers and 1.8, 3.6, 6.4, 11.3, and 17.5 for esophageal cancers. 27 (3.7%) gastric and 20 (2.3%) esophageal cases had alterations in mismatch repair genes MLH1, MSH2, MSH6 which was associated with ~2.5 fold increased mutational load. Mutational load was not significantly associated with the presence of RTK alterations, which occurred in 155 (21%) gastric and 355 (41%) esophageal cases. Conclusions: CGP in the course of clinical care can be used to assess mutational load in GEC. Alterations in RTKs were not exclusive of the high mutation subgroup, contrary to some previous reports. Mutations in DNA mismatch repair genes were associated with higher mutation burden as expected. Incorporation of CGP into ongoing prospective immunotherapy trials and clinical practice is needed to refine these relationships.

Abstract A32: MCL1 gene amplification in breast cancer is associated with TNBC status and can respond to a sorafenib/vorinostat regimen

Background: MCL1 encodes the induced myeloid leukemia cell differentiation protein Mcl-1, a member of the BCL-2 family which functions to inhibit apoptosis. Mcl-1 over-expression has been associated with high tumor grade and adverse prognosis in triple negative breast cancer (TNBC) but therapies specifically leading to inhibition of MCL-1 have not been identified.

Methods: Comprehensive genomic profiling (CGP) using hybridization capture of 3,769 exons from 315 cancer-related genes and 47 introns of 19 genes commonly rearranged in cancer was applied to ≥50ng of DNA extracted from 2,824 consecutive BC and sequenced to high, uniform median coverage (&gt;600X). The original primary BC was assayed in 44% of cases and a sample from a metastatic focus was assayed in 56% of cases.

Results: Of 2824 consecutive BC cases, 200 (7.1%) cases harbored MCL1 amplification. Of these MCL1-amplified cases, 146 (73%) were TNBC and 54 were non-TNBC (p&lt;0.0001). Twelve of the latter cases (22%) were ERBB2 (HER2) amplified and slide-based HER2 status concordance with CGP was 99%. MCL1 amplification was also observed in the TCGA dataset for 32/123 (26%) of TNBC (p=0.008). Of the MCL1 amplified TNBC cases, 88% were high grade and 98% were stage IV at the time of CGP. Genes co-altered within MCL1 amplified TNBC included TP53 (86%), MYC (41%), MYST3 (21%), LYN (20%), CCNE1 (19%), PIK3CA (18%), and AKT3 (15%). Two MCL1 amplified TNBC patients were treated with a multi-drug regimen based on sorafenib and vorinostat and experienced significant clinical benefit.

Conclusions: MCL1 amplification is a frequent feature in advanced stage and high grade TNBC, and correspondingly such MCL1 amplified tumors very seldom harbor co-amplifications of ERBB2. Clinical observation suggests that treatment with sorafenib and vorinostat in heavily pre-treated MCL1 amplified patients may be correlated with clinical benefit, consistent with historic preclinical investigation. These preliminary findings suggest that MCL1 amplified TNBC may be able to benefit from combination targeted therapy, and warrant further systematic investigation.

Citation Format: JS Ross, K Wang, A Johnson, J Watson, C Hatzis, L Pusztai, J Chmielecki, R Yelensky, D Lipson, JA Elvin, J Vergilio, J Suh, VA Miller, K Dicke, PJ Stephens, SM Ali. MCL1 gene amplification in breast cancer is associated with TNBC status and can respond to a sorafenib/vorinostat regimen. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Breast Cancer Research; Oct 17-20, 2015; Bellevue, WA. Philadelphia (PA): AACR; Mol Cancer Res 2016;14(2_Suppl):Abstract nr A32.

Evaluation of microsatellite instability (MSI) status in gastrointestinal (GI) tumor samples tested with comprehensive genomic profiling (CGP)

528

Background: PD-1 inhibitor therapy (pembrolizumab) has demonstrated promising activity in MSI-H (MSI-High) colorectal cancer (CRC) and MSI-H non-CRC compared to MSS tumors, supporting MSI as predictive of anti-PD-1/L1 therapy independent of tumor histology. We developed and validated a novel computational method to assess MSI status using next generation sequencing (NGS) data generated from CGP of clinical cases. We then sought to determine the frequency of MSI in a diverse sample of clinical cases of GI tumors previously assayed with CGP (FoundationOne). Methods: To determine MSI status, 114 intronic homopolymer repeat loci (10-20bp long in the human reference genome) with adequate coverage on the CGP panel were analyzed for length variability and compiled into an overall MSI score via principal components analysis. Translation of the MSI score to MSI-H or MSS (MSI-Stable) was established using a training data set. Results: Our methodology was validated in 69 samples with MSI status (20 MSI-H, 49 MSS) previously determined by standard clinical testing, either PCR or immunohistochemistry staining. Our concordance was 97% (65/67). We then applied our method to 2545 GI cancers assayed with CGP in the course of clinical care, with small bowel carcinoma having the highest rate of MSI-H (Table). The 3.5% frequency of MSI-H in CRC is lower than expected, possibly attributed to sampling bias against MSI-H CRC in the cohort. In CRC, we discovered significant differences in mutational profile: histone methyltransferase MLL2/3 mutations were enriched in MSI-H CRC (P<0.0001, Fisher’s Exact test), while TP53mutations were enriched in MSS CRC (P<0.0001, Fisher’s). Conclusions: CGP can be used to detect MSI with high accuracy. In addition, application of this approach to the 2545 GI cancers demonstrates the use of CGP to identify differences in mutational profiles between MSI-H and MSS cancers. Broadened application of these methods to more tumor types may provide target enrichment for emerging immunotherapies. [Table: see text]

Abstract A11: NGS-based tumor genomic profiling to identify ovarian cancer patients who benefit from the PARP inhibitor rucaparib

Background: PARP inhibitors (PARPi) are synthetically lethal to tumor cells with homologous recombination deficiency (HRD). HRD can result from deleterious BRCA1/2 mutations (BRCAmut) or other mechanisms that have not been fully elucidated. Regardless of mechanism, HRD leads to a common phenotype of genome-wide loss of heterozygosity (LOH). It has been hypothesized that this genomic phenotype can be used to identify BRCA wild-type (BRCAwt) HRD tumors likely sensitive to PARPi. Using comprehensive next generation sequencing (NGS)-based tumor genomic profiling, we developed an HRD assay for potential use as a companion diagnostic for rucaparib in high-grade ovarian cancer (HGOC) by combining tumor BRCA1/2 status and quantification of genomic LOH.

Methods: In the phase 2 study ARIEL2 Part 1 (NCT01891344), pre-treatment screening biopsies and archival formalin-fixed paraffin embedded tumor specimens were profiled using Foundation Medicine's NGS-based HRD assay, which detects all classes of genomic alterations, including base substitutions, insertions/deletions, and homozygous deletions in BRCA1/2. Genomic LOH was assessed by sequencing &gt;3,500 evenly-distributed single nucleotide polymorphisms across the genome and quantifying the extent of genomic LOH. A pre-specified genomic LOH cutoff was determined using publicly available SNP array data of ovarian tumors to predict platinum sensitivity as a surrogate marker for PARPi sensitivity. Response was assessed by RECIST v1.1 and GCIG CA-125 response criteria.

Results: As of July 1 2015, 195 archival tumor and 152 screening biopsy samples (142 matched pairs) from 206 HGOC patients enrolled (204 patients treated) in ARIEL2 Part 1 were successfully profiled using the NGS-based HRD assay. Some screening biopsies were not suitable for successful NGS-based HRD assessment primarily because of insufficient tumor nuclei or inadequate tumor volume. Most matched pairs of archival and pre-trial screening samples exhibited similar genomic LOH profiles (r=0.86); however, 14% of screening samples had higher genomic LOH compared with archival samples collected more than one year earlier. All BRCA1/2 germline and somatic mutated tumors had high genomic LOH in the screening samples. Receiver operating characteristic analysis of genomic LOH showed utility in identifying RECIST/CA-125 responders to rucaparib (AUC=0.72, p&lt;1e-4), with slightly better predictive utility using screening samples compared to archival samples (AUC=0.72 vs 0.69). Using the pre-specified genomic LOH cutoff, high genomic LOH tumors were detected in 54% of evaluable BRCAwt patients; significantly different overall response rates were found in patients with high vs low genomic LOH tumors (48% vs 26%; chi-square p=0.0074).

Conclusions: We developed an NGS-based HRD assay that assesses tumor BRCA1/2 and genomic LOH to prospectively identify HGOC patients who may benefit from rucaparib treatment. The optimized NGS-based HRD assay will be prospectively tested in the ongoing portion of the phase 2 study (ARIEL2 Part 2, NCT01891344) and a phase 3 maintenance study (ARIEL3, NCT01968213) that will investigate rucaparib in HGOC.

Citation Format: Iain A. McNeish, Kevin K. Lin, James X. Sun, Sandra Goble, Amit Oza, Robert L. Coleman, Clare L. Scott, Gottfried Konecny, Anna V. Tinker, David M. O'Malley, Rebecca Kristeleit, Ling Ma, James D. Brenton, Katherine Bell-McGuinn, Ana Oaknin, Alexandra Leary, Elaina Mann, Heidi Giordano, Roman Yelensky, Mitch Raponi, Elizabeth Swisher. NGS-based tumor genomic profiling to identify ovarian cancer patients who benefit from the PARP inhibitor rucaparib. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: Exploiting Vulnerabilities; Oct 17-20, 2015; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(2 Suppl):Abstract nr A11.

Characterization of mutational load in patients with advanced urothelial cancer

460

Background: The efficacy of novel immunotherapeutic agents (e.g., PD-1/PD-L1 and CTLA4 inhibitors) in advanced cancer is linked to tumor mutational load. We assessed mutational load via comprehensive genomic profiling (CGP) performed in the course of clinical care for patients with advanced urothelial cancer (UC). Methods: DNA was extracted from 40 microns of FFPE sections from 760 consecutive patients with relapsed/metastatic UC. CGP was performed on hybridization-captured, adaptor ligation based libraries to a mean coverage depth of 646X for at least 3,230 exons of 182 cancer-related genes plus 37 introns from 14 genes frequently rearranged in cancer. Mutational load was characterized as the number of such SUBs or INDELs per megabase (Mb) after filtering to remove known somatic and deleterious mutations, given that these are selected for with hybrid capture. Results: The median age of the patients was 66 years old with a 3:1 male:female ratio. Mean mutations per megabase were assessed as a range of 0 to 79, and the 25th, median, and 75th quartile thresholds were 3.2, 5.5, and 8.8. MSH6 alterations in 0.8% cases had a median of 15.2 mutations per megabase. However, the ten most frequently altered genes in this series- TERT (64%), TP53 (54%), CDKN2A (35%), CDKN2B (28%), KDM6A (24%), ARID1A (23%), MLL2 (22%), PIK3CA (21%), RB1 (21%) and FGFR3 (19%) - were not associated with differences in mutational load. Conclusions: A highly variable mutational load was seen in patients with advanced UC. A subpopulation of MSH6-altered UC had very high mutational loads relative to other UCs. Further correlation to clinical outcomes will be investigated to assess the correlation between mutational load and response to immunotherapy.

Comprehensive Genomic Profiling of Advanced Penile Carcinoma Suggests a High Frequency of Clinically Relevant Genomic Alterations

BACKGROUND

Advanced penile squamous cell carcinoma (PSCC) is associated with poor survival due to the aggressiveness of the disease and lack of effective systemic therapies. Comprehensive genomic profiling (CGP) was performed to identify clinically relevant genomic alterations (CRGAs).

MATERIALS AND METHODS

DNA was extracted from 40 μm of formalin-fixed, paraffin-embedded sections in patients with advanced PSCC. CGP was performed on hybridization-captured, adaptor ligation-based libraries to a mean coverage depth of 692× for 3,769 exons of 236 cancer-related genes plus 47 introns from 19 genes frequently rearranged in cancer. CRGAs were defined as genomic alterations (GAs) linked to targeted therapies on the market or under evaluation in mechanism-driven clinical trials.

RESULTS

Twenty male patients with a median age of 60 years (range, 46-87 years) were assessed. Seventeen (85%) cases were stage IV and three cases (15%) were stage III. CGP revealed 109 GAs (5.45 per tumor), 44 of which were CRGAs (2.2 per tumor). At least one CRGA was detected in 19 (95%) cases, and the most common CRGAs were CDKN2A point mutations and homozygous deletion (40%), NOTCH1 point mutations and rearrangements (25%), PIK3CA point mutations and amplification (25%), EGFR amplification (20%), CCND1 amplification (20%), BRCA2 insertions/deletions (10%), RICTOR amplifications (10%), and FBXW7 point mutations (10%).

CONCLUSION

CGP identified CRGAs in patients with advanced PSCC, including EGFR amplification and PIK3CA alterations, which can lead to the rational administration of targeted therapy and subsequent benefit for these patients.

IMPLICATIONS FOR PRACTICE

Few treatment options exist for patients with advanced penile squamous cell carcinoma (PSCC). Outcomes are dismal with platinum-based chemotherapy, with median survival estimated at 1 year or less across multiple series. Biological studies of patients with PSCC to date have principally focused on human papillomavirus status, but few studies have elucidated molecular drivers of the disease. To this end, comprehensive genomic profiling was performed in a cohort of 20 patients with advanced PSCC. Findings of frequent mutations in CDKN2A, NOTCH1, PIK3CA, and EGFR (all in excess of 20%) point to potential therapeutic avenues. Trials of targeted therapies directed toward these mutations should be explored.

Comprehensive genomic profiling of 295 cases of clinically advanced urothelial carcinoma of the urinary bladder reveals a high frequency of clinically relevant genomic alterations

BACKGROUND

In the current study, the authors present a comprehensive genomic profile (CGP)-based study of advanced urothelial carcinoma (UC) designed to detect clinically relevant genomic alterations (CRGAs).

METHODS

DNA was extracted from 40 µm of formalin-fixed, paraffin-embedded sections from 295 consecutive cases of recurrent/metastatic UC. CGP was performed on hybridization-captured, adaptor ligation-based libraries to a mean coverage depth of 688X for all coding exons of 236 cancer-related genes plus 47 introns from 19 genes frequently rearranged in cancer, using process-matched normal control samples as a reference. CRGAs were defined as GAs linked to drugs on the market or currently under evaluation in mechanism-driven clinical trials.

RESULTS

All 295 patients assessed were classified with high-grade (International Society of Urological Pathology classification) and advanced stage (stage III/IV American Joint Committee on Cancer) disease, and 294 of 295 patients (99.7%) had at least 1 GA on CGP with a mean of 6.4 GAs per UC (61% substitutions/insertions/deletions, 37% copy number alterations, and 2% fusions). Furthermore, 275 patients (93%) had at least 1 CRGA involving 75 individual genes with a mean of 2.6 CRGAs per UC. The most common CRGAs involved cyclin-dependent kinase inhibitor 2A (CDKN2A) (34%), fibroblast growth factor receptor 3 (FGFR3) (21%), phosphatidylinositol 3-kinase catalytic subunit alpha (PIK3CA) (20%), and ERBB2 (17%). FGFR3 GAs were diverse types and included 10% fusions. ERBB2 GAs were equally divided between amplifications and substitutions. ERBB2 substitutions were predominantly within the extracellular domain and were highly enriched in patients with micropapillary UC (38% of 32 cases vs 5% of 263 nonmicropapillary UC cases; P<.0001).

CONCLUSIONS

Using a CGP assay capable of detecting all classes of GA simultaneously, an extraordinarily high frequency of CRGA was identified in a large series of patients with advanced UC. Cancer 2016;122:702-711. © 2015 American Cancer Society.

A metastatic colon adenocarcinoma harboring BRAF V600E has a durable major response to dabrafenib/trametinib and chemotherapy

The subset of metastatic colorectal adenocarcinomas that harbor BRAF V600E mutations are aggressive tumors with significantly shortened survival and limited treatment options. Here we present a colorectal cancer patient whose disease progressed through standard chemotherapy and who developed liver metastasis. Comprehensive genomic profiling (FoundationOne^®) identified a BRAF V600E mutation in the liver lesion, as well as other genomic alterations consistent with colorectal cancers. Combination therapy of dabrafenib and trametinib with standard cytotoxic chemotherapy resulted in a durable major ongoing response for the patient. This report illustrates the utility of comprehensive genomic profiling with personalized targeted therapy for aggressive metastatic colorectal adenocarcinomas.

RAS/MAPK Activation Is Associated with Reduced Tumor-Infiltrating Lymphocytes in Triple-Negative Breast Cancer: Therapeutic Cooperation Between MEK and PD-1/PD-L1 Immune Checkpoint Inhibitors

PURPOSE

Tumor-infiltrating lymphocytes (TIL) in the residual disease (RD) of triple-negative breast cancers (TNBC) after neoadjuvant chemotherapy (NAC) are associated with improved survival, but insight into tumor cell-autonomous molecular pathways affecting these features are lacking.

EXPERIMENTAL DESIGN

We analyzed TILs in the RD of clinically and molecularly characterized TNBCs after NAC and explored therapeutic strategies targeting combinations of MEK inhibitors with PD-1/PD-L1-targeted immunotherapy in mouse models of breast cancer.

RESULTS

Presence of TILs in the RD was significantly associated with improved prognosis. Genetic or transcriptomic alterations in Ras-MAPK signaling were significantly correlated with lower TILs. MEK inhibition upregulated cell surface MHC expression and PD-L1 in TNBC cells both in vivo and in vitro. Moreover, combined MEK and PD-L1/PD-1 inhibition enhanced antitumor immune responses in mouse models of breast cancer.

CONCLUSIONS

These data suggest the possibility that Ras-MAPK pathway activation promotes immune-evasion in TNBC, and support clinical trials combining MEK- and PD-L1-targeted therapies. Furthermore, Ras/MAPK activation and MHC expression may be predictive biomarkers of response to immune checkpoint inhibitors.

Kinase fusions are frequent in Spitz tumours and spitzoid melanomas

Spitzoid neoplasms are a group of melanocytic tumours with distinctive histopathological features. They include benign tumours (Spitz naevi), malignant tumours (spitzoid melanomas) and tumours with borderline histopathological features and uncertain clinical outcome (atypical Spitz tumours). Their genetic underpinnings are poorly understood, and alterations in common melanoma-associated oncogenes are typically absent. Here we show that spitzoid neoplasms harbour kinase fusions of ROS1 (17%), NTRK1 (16%), ALK (10%), BRAF (5%) and RET (3%) in a mutually exclusive pattern. The chimeric proteins are constitutively active, stimulate oncogenic signalling pathways, are tumourigenic and are found in the entire biologic spectrum of spitzoid neoplasms, including 55% of Spitz naevi, 56% of atypical Spitz tumours and 39% of spitzoid melanomas. Kinase inhibitors suppress the oncogenic signalling of the fusion proteins in vitro. In summary, kinase fusions account for the majority of oncogenic aberrations in spitzoid neoplasms and may serve as therapeutic targets for metastatic spitzoid melanomas.

Comprehensive genomic profiling of extrahepatic cholangiocarcinoma reveals a long tail of therapeutic targets

AIM

We queried whether extrahepatic cholangiocarcinoma featured clinically relevant genomic alterations that could lead to targeted therapy.

METHODS

Comprehensive genomic profiling by hybridisation capture of up to 315 genes was performed on 99 clinically advanced extrahepatic cholangiocarcinoma.

RESULTS

There were 60 male and 39 female patients with a median age of 60.5 years. A total of 400 alterations were identified (mean 4.0; range 0-13) in 84 genes. Eighty-two (83%) of extrahepatic cholangiocarcinoma patients featured at least one clinically relevant genomic alterations including KRAS (43%); ERBB2 (9%), PTEN (7%); ATM and NF1 (6%) and CCND1, FBXW7, GNAS, MDM2 and NRAS (all at 5%). BRAF, BRCA2, CDK4, CDK6, FGFR1, FGFR3, PTCH1, RAF1 and STK11 were each altered in a single patient. No IDH1/2 mutations or FGFR2 gene fusions were identified.

CONCLUSIONS

Comprehensive genomic profiling of extrahepatic cholangiocarcinoma differs significantly from intrahepatic cholangiocarcinoma and pancreatic adenocarcinoma, and reveals diverse opportunities for the use of targeted therapies.

Evaluation of 122 advanced-stage cutaneous squamous cell carcinomas by comprehensive genomic profiling opens the door for new routes to targeted therapies

BACKGROUND

The authors hypothesized that comprehensive genomic profiling of advanced-stage cutaneous squamous cell carcinoma (cSCC) could identify genomic-derived drug targets of therapy for patients with conventional therapy-resistant disease.

METHODS

Comprehensive genomic profiling of 315 cancer genes was applied to 50 ng of DNA from 122 cSCC cases for the evaluation of all classes of genomic alterations (GAs). Clinically relevant genomic alterations (CRGAs) were defined as those identifying anticancer drugs on the market or in registered clinical trials.

RESULTS

There were 21 women (17%) and 101 men (83%) with a median age of 64.9 years (range, 21-87 years). Eleven cSCC cases (9%) were histologic AJCC grade 1, 69 (57%) were grade 2, and 42 (34%) were grade 3. The primary cSCC was used for sequencing in 77 cases (63%). Metastatic lesions were sequenced in 37% of cases. There were 1120 total GAs identified (average of 9.2 GAs per tumor), with 100% of cases harboring at least 1 alteration. Of the 122 cSCCs, 107 (88%) harbored at least 1 CRGA (2.5 CRGAs per cSCC) includingNOTCH1 (43%); patched 1 (PTCH1) (11%); BRCA2 (10%); HRAS (8%); ataxia telangiectasia mutated (ATM) (7%); erb-B2 receptor tyrosine kinase 4 (ERBB4) (7%); neurofibromatosis type 1 (NF1) (7%); erb-B2 receptor tyrosine kinase 2 (ERBB2) (6%); phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) (6%); cyclin D1 (CCND1) (6%); epidermal growth factor receptor (EGFR) (5%); and F-box and WD repeat domain containing 7, E3 ubiquitin protein ligase (FBXW7) (5%).

CONCLUSIONS

In the current study, approximately 88% of patients with cSCC were found to harbor clinically relevant GAs that have the potential to guide the treatment of patients with advanced-stage tumors with targeted therapeutic agents. Cancer 2016;122:249-257. © 2015 American Cancer Society.

RICTOR Amplification Defines a Novel Subset of Patients with Lung Cancer Who May Benefit from Treatment with mTORC1/2 Inhibitors

We identified amplification of RICTOR, a key component of the mTOR complex 2 (mTORC2), as the sole actionable genomic alteration in an 18-year-old never-smoker with lung adenocarcinoma. Amplification of RICTOR occurs in 13% of lung cancers (1,016 cases) in The Cancer Genome Atlas and at a similar frequency in an independent cohort of 1,070 patients identified by genomic profiling. In the latter series, 11% of cases harbored RICTOR amplification as the only relevant genomic alteration. Its oncogenic roles were suggested by decreased lung cancer cell growth both in vitro and in vivo with RICTOR ablation, and the transforming capacity of RICTOR in a Ba/F3-cell system. The mTORC1/2 inhibitors were significantly more active against RICTOR-amplified lung cancer cells as compared with other agents targeting the PI3K-AKT-mTOR pathway. Moreover, an association between RICTOR amplification and sensitivities to mTORC1/2 inhibitors was observed. The index patient has been treated with mTORC1/2 inhibitors that led to tumor stabilization for more than 18 months.

SIGNIFICANCE

RICTOR amplification may define a novel and unique molecular subset of patients with lung cancer who may benefit from treatment with mTORC1/2 inhibitors.

The distribution of BRAF gene fusions in solid tumors and response to targeted therapy

Although the BRAF V600E base substitution is an approved target for the BRAF inhibitors in melanoma, BRAF gene fusions have not been investigated as anticancer drug targets. In our study, a wide variety of tumors underwent comprehensive genomic profiling for hundreds of known cancer genes using the FoundationOne™ or FoundationOne Heme™ comprehensive genomic profiling assays. BRAF fusions involving the intact in-frame BRAF kinase domain were observed in 55 (0.3%) of 20,573 tumors, across 12 distinct tumor types, including 20 novel BRAF fusions. These comprised 29 unique 5' fusion partners, of which 31% (9) were known and 69% (20) were novel. BRAF fusions included 3% (14/531) of melanomas; 2% (15/701) of gliomas; 1.0% (3/294) of thyroid cancers; 0.3% (3/1,062) pancreatic carcinomas; 0.2% (8/4,013) nonsmall-cell lung cancers and 0.2% (4/2,154) of colorectal cancers, and were enriched in pilocytic (30%) vs. nonpilocytic gliomas (1%; p < 0.0001), Spitzoid (75%) vs. nonSpitzoid melanomas (1%; p = 0.0001), acinar (67%) vs. nonacinar pancreatic cancers (<1%; p < 0.0001) and papillary (3%) vs. nonpapillary thyroid cancers (0%; p < 0.03). Clinical responses to trametinib and sorafenib are presented. In conclusion, BRAF fusions are rare driver alterations in a wide variety of malignant neoplasms, but enriched in Spitzoid melanoma, pilocytic astrocytomas, pancreatic acinar and papillary thyroid cancers.

Comprehensive Genomic Profiling of Advanced Esophageal Squamous Cell Carcinomas and Esophageal Adenocarcinomas Reveals Similarities and Differences

BACKGROUND

Esophageal squamous cell carcinomas (ESCCs) and esophageal adenocarcinomas (EACs) account for >95% of esophageal malignancies and represent a major global health burden. ESCC is the dominant histology globally but represents a minority of U.S. cases, with EAC accounting for the majority of U.S.

CASES

The patient outcomes for advanced ESCC and EAC are poor, and new therapeutic options are needed. Using a sensitive sequencing assay, we compared the genomic profiles of ESCC and EAC with attention to identification of therapeutically relevant genomic alterations.

METHODS

Next-generation sequencing-based comprehensive genomic profiling was performed on hybridization-captured, adaptor ligation-based libraries to a median coverage depth of >650× for all coding exons of 315 cancer-related genes plus selected introns from 28 genes frequently rearranged in cancer. Results from a single sample were evaluated for all classes of genomic alterations (GAs) including point mutations, short insertions and deletions, gene amplifications, homozygous deletions, and fusions/rearrangements. Clinically relevant genomic alterations (CRGAs) were defined as alterations linked to approved drugs and those under evaluation in mechanism-driven clinical trials.

RESULTS

There were no significant differences by sex for either tumor type, and the median age for all patients was 63 years. All ESCCs and EACs were at an advanced stage at the time of sequencing. All 71 ESCCs and 231 EACs featured GAs on profiling, with 522 GAs in ESCC (7.4 per sample) and 1,303 GAs in EAC (5.6 per sample). The frequency of clinically relevant GAs in ESCC was 94% (2.6 per sample) and 93% in EAC (2.7 per sample). CRGAs occurring more frequently in EAC included KRAS (23% EAC vs. 6% ESCC) and ERBB2 (23% EAC vs. 3% ESCC). ESCC samples were enriched for CRGA in PIK3CA (24% ESCC vs. 10% EAC), PTEN (11% ESCC vs. 4% EAC), and NOTCH1 (17% ESCC vs. 3% EAC). Other GAs that differed significantly between histologic tumor types included SMAD4 (14% EAC vs. 1% ESCC), RB1 (14% ESCC vs. 2% EAC), SOX2 (18% ESCC vs. 1% EAC), and NFE2L2 (24% ESCC vs. 1% EAC).

CONCLUSION

ESCC and EAC share similarly high frequencies of overall and clinically relevant genomic alterations; however, the profiles of genomic alterations in the two diseases differ widely, with KRAS and ERBB2 far more frequently altered in EAC compared with ESCC and with mammalian target of rapamycin (MTOR) pathway genes (PIK3CA and PTEN) and NOTCH1 more frequently altered in ESCC compared with EAC. Comprehensive genomic profiling highlights the promise of identifying clinically relevant genomic alterations in both ESCC and EAC and suggests new avenues for molecularly directed therapies in esophageal cancer.

Abstract B06: Distinct genomic profiles and targetable alterations revealed by FoundationOne® Heme in hematolymphoid malignancies in adolescents and young adults

Background: Hematolymphoid malignancies (HLM) in adolescent and young adult (AYA, age 15-39) patients exhibit distinct clinical and routine pathologic characteristics. Comparative analysis of genomic alterations (GA) of HLM in AYA versus younger and older patient cohorts has not been reported previously.

Methods: DNA was extracted and captured using custom baits targeting all exons of 405 cancer-related genes and select introns of 31 genes commonly rearranged in cancer (n=24), or both DNA and RNA were extracted and captured using custom baits for 405 genes and 265 frequently rearranged genes, respectively (n=49) (FoundationOne® Heme). All captured libraries were sequenced to high uniform depth in a CLIA-certified, CAP-accredited laboratory with an average depth of &gt;500x for DNA and &gt;10M unique pairs for RNA (Illumina HiSeq). Sequence data were assessed for base substitutions, insertions/deletions (indels), copy number alterations, and select gene rearrangements.

Results: The assays identified an average of 3.1 cancer-related GAs per HLM (range 0-22), including 135 base substitutions, 45 indels, 36 gene fusions or rearrangements, 13 gene deletions, and 12 gene amplifications. The most frequently altered genes in this AYA HLM cohort were TP53 (22%), NRAS (12%), KRAS (11%), PTPN11 (10%), MYC (8%), B2M (8%), SOCS1 (7%), FLT3 (7%), and CDKN2A (7%). Clinically relevant GAs (targetable by therapies approved or in clinical trials) were found in 64% of cases, including 60% of lymphoma, 71% of acute myeloid leukemia (AML), and 79% of acute lymphoblastic leukemia (ALL). Clinically relevant GAs identified in this AYA cohort included NRAS and KRAS (23%), targetable by MEK inhibitors, FLT3 (7%), targetable by multiple agents, as well as partial tandem duplication of MLL and rearrangement of IGH/MYC. AML AYA samples contained a high frequency of PTPN11 mutations (41%) which can activate the RAS/MAP/ERK, PI3K, and Src pathways. The enrichment of PTPN11 mutations in 7/17 (41%) AYA AML vs. 0/10 (0%) pediatric AML (age&lt;15) and 1/ 80 (1%) of older adult AML was significant (p=0.0001). A higher frequency of KRAS mutations in AYA AML (29%) vs. older adults (6%) was additionally observed (p=0.01). 2 of 9 AYA diffuse large B-cell lymphoma (DLBCL) cases were found to harbor ALK fusions, a potential enrichment relative to older patients which will be further explored.

Conclusions: FoundationOne Heme can detect all classes of genomic alterations in AYA hematolymphoid malignancies. Significant molecular distinctions identified in this AYA population implicate age-related mechanisms in the development and progression of HLM, and provide novel potential therapeutic approaches to improve clinical outcomes in this distinct patient cohort.