Search

Tumor marker levels and cfDNA mutant allele fractions across 19 time intervals for eight patients that were serially monitored via blood draws.

Complete patient-level and mutation-level concordance information, mutation identities, and monitoring status for each patient in study.

Supplementary Data from Origins and Timing of Emerging Lesions in Advanced Renal Cell Carcinoma

Comparisons of CA 19-9 concentration and cfDNA mutant allele fractions in plasma. (A-E) cfDNA mutant allele fractions ("cfDNA percentage") versus CA 19-9 units per milliliter (U/mL) were determined at similar times for five patients with three or more serial blood draws. cfDNA percentage represents the mutant allele fraction of the most abundant cfDNA mutation, as determined at time zero, for each patient. (F) cfDNA mutant allele fractions and CA 19-9 marker measurements for patient 56 over the course of four time points while on therapy (gemcitabine nab-paclitaxel). Note that draw number 1 occurred 17 days after diagnosis.

List of genes and exons captured in the Guardant360 cfDNA sequencing test.

Supplementary Figure from Origins and Timing of Emerging Lesions in Advanced Renal Cell Carcinoma

Renal cell carcinoma (RCC) with venous tumor thrombus (VTT) arising from the primary tumor occurs in approximately 10% of cases and is thought to represent more advanced disease. The intravascular nature of VTT suggests that it may serve as a source for hematogenous metastases. RCC with VTT and distant metastasis provides unique opportunities to examine the origins and emergence timing of these distinct tumor lesions, and to identify molecular correlates with disease state. We performed multi-region exome and RNA-sequencing analysis of 16 patients with RCC with VTT, with eight patients also having sequenced metastasis, to identify genomic alterations, biological pathways, and evolutionary processes contributing to VTT and metastasis, and to ask whether metastasis arises directly from or independent of VTT. No specific genomic alterations were associated with VTT. Hallmark copy-number alterations (deletions of 14q, 8p, and 4q) were associated with metastasis and disease recurrence, and secondary driver alterations tended to accumulate in metastatic lineages. Mismatch repair mutational signatures co-occurred across most tumors, suggesting a role for intracellular DNA damage in RCC. Robust phylogenetic timing analysis indicated that metastasis typically emerged before VTT, rather than deriving from it, with the earliest metastases predicted to emerge years before diagnosis. As a result, VTT in metastatic cases frequently derived from a metastatic lineage. Relative to the primary tumor, VTT upregulated immediate-early genes and transcriptional targets of the TNFα/NF-κB pathway, whereas metastases upregulated MTOR and transcriptional targets downstream of mTORC1 activation.Implications:These results suggest that VTT and metastasis formation occur independently, VTT presence alone does not necessarily imply more advanced disease with inevitably poor prognosis.

Characteristics of the 26 patients in study.

Empirical determination of the limit of detection. Defined oncogenic driver and background SNVs (a), oncogene and tumor suppressor indels (b), and oncogenic fusions (c) were titrated to specified VAFs, and multiple replicates were measured to determine the 95% LOD. (d) Cell lines comprising defined gene amplifications were titrated to bracket the expected LOD, and the 95% LOD was determined from statistical detection models as per the CLSI Classical Approach.

Genes on the Guardant360 panel.

Supplemental Methods, Supplemental References, Supplemental Figure Legends, Figure S1-13, Table S1-5 and S14

Digital Sequencing analytical performance metrics summary table.

cfDNA variants associated with resistance to on-label targeted therapies across six cancer types

Purpose: To analytically and clinically validate a circulating cell-free tumor DNA sequencing test for comprehensive tumor genotyping and demonstrate its clinical feasibility.Experimental Design: Analytic validation was conducted according to established principles and guidelines. Blood-to-blood clinical validation comprised blinded external comparison with clinical droplet digital PCR across 222 consecutive biomarker-positive clinical samples. Blood-to-tissue clinical validation comprised comparison of digital sequencing calls to those documented in the medical record of 543 consecutive lung cancer patients. Clinical experience was reported from 10,593 consecutive clinical samples.Results: Digital sequencing technology enabled variant detection down to 0.02% to 0.04% allelic fraction/2.12 copies with ≤0.3%/2.24–2.76 copies 95% limits of detection while maintaining high specificity [prevalence-adjusted positive predictive values (PPV) >98%]. Clinical validation using orthogonal plasma- and tissue-based clinical genotyping across >750 patients demonstrated high accuracy and specificity [positive percent agreement (PPAs) and negative percent agreement (NPAs) >99% and PPVs 92%–100%]. Clinical use in 10,593 advanced adult solid tumor patients demonstrated high feasibility (>99.6% technical success rate) and clinical sensitivity (85.9%), with high potential actionability (16.7% with FDA-approved on-label treatment options; 72.0% with treatment or trial recommendations), particularly in non–small cell lung cancer, where 34.5% of patient samples comprised a directly targetable standard-of-care biomarker.Conclusions: High concordance with orthogonal clinical plasma- and tissue-based genotyping methods supports the clinical accuracy of digital sequencing across all four types of targetable genomic alterations. Digital sequencing's clinical applicability is further supported by high rates of technical success and biomarker target discovery. Clin Cancer Res; 24(15); 3539–49. ©2018 AACR.

Driver alteration prevalence in the cfDNA cohort versus in TCGA

Quantitative precision of Digital Sequencing results. (a) SNV VAFs in 375 consecutive SNV control runs. (b) Indel and fusion VAFs (left y-axis) and CNA copy number (right y-axis) in consecutive CFI control runs. (c) Variation distribution of DS results in 375 consecutive analyses of a single SNV control lot (blue) and multiple consecutive lots of indel (red), fusion (grey), and CNA (orange) controls.

Detailed results of Digital Sequencing and clinical ddPCR comparison. Results shown in individual 2x2 tables for each biomarker tested. *Note, all discordances were below DFCI ddPCR's reportable range. Detailed clinical information for each is presented in Supplemental Table 3.

Discordant results between Digital Sequencing and clinical ddPCR. Details of the cases discordant between DS and Dana-Farber Clinical Laboratory's ddPCR results with orthogonal follow-up data. *Note, while approved specifically for T790M+ NSCLC, osimertinib has been reported to have clinical activity in T790-wild type patients as well; however, osimertinib efficacy in T790M-wild type patients resistant to 1st generation EGFR TKIs is unclear.

Purpose: Cell-free DNA (cfDNA) sequencing provides a noninvasive method for obtaining actionable genomic information to guide personalized cancer treatment, but the presence of multiple alterations in circulation related to treatment and tumor heterogeneity complicate the interpretation of the observed variants.Experimental Design: We describe the somatic mutation landscape of 70 cancer genes from cfDNA deep-sequencing analysis of 21,807 patients with treated, late-stage cancers across >50 cancer types. To facilitate interpretation of the genomic complexity of circulating tumor DNA in advanced, treated cancer patients, we developed methods to identify cfDNA copy-number driver alterations and cfDNA clonality.Results: Patterns and prevalence of cfDNA alterations in major driver genes for non–small cell lung, breast, and colorectal cancer largely recapitulated those from tumor tissue sequencing compendia (The Cancer Genome Atlas and COSMIC; r = 0.90–0.99), with the principal differences in alteration prevalence being due to patient treatment. This highly sensitive cfDNA sequencing assay revealed numerous subclonal tumor-derived alterations, expected as a result of clonal evolution, but leading to an apparent departure from mutual exclusivity in treatment-naïve tumors. Upon applying novel cfDNA clonality and copy-number driver identification methods, robust mutual exclusivity was observed among predicted truncal driver cfDNA alterations (FDR = 5 × 10−7 for EGFR and ERBB2), in effect distinguishing tumor-initiating alterations from secondary alterations. Treatment-associated resistance, including both novel alterations and parallel evolution, was common in the cfDNA cohort and was enriched in patients with targetable driver alterations (>18.6% patients).Conclusions: Together, these retrospective analyses of a large cfDNA sequencing data set reveal subclonal structures and emerging resistance in advanced solid tumors. Clin Cancer Res; 24(15); 3528–38. ©2018 AACR.

Mutual exclusivity statistics for cfDNA alterations in lung adenocarcinoma, breast cancer, and colorectal cancer

Per-sample mutation burden distribution by tumor type. Per-sample adjusted mutation count relative to other samples within clinical indication. Green bars indicate the 95th percentile.

Quantitative accuracy of Digital Sequencing. Quantitative correlation between expected and observed VAFs for SNVs (a) and CNAs (b) in the analytical validation studies. (c) DS copy number calls compared to those reported in the Cancer Cell Line Encyclopedia (CCLE).

A growing body of evidence suggests that T-cell responses against neoantigens are critical regulators of response to immune checkpoint blockade. We previously showed that circulating neoantigen-specific CD8 T cells in patients with lung cancer responding to anti-Programmed death-ligand 1 (PD-L1) (atezolizumab) exhibit a unique phenotype with high expression of CD57, CD244, and KLRG1. Here, we extended our analysis on neoantigen-specific CD8 T cells to patients with metastatic urothelial cancer (mUC) and further profiled total CD8 T cells to identify blood-based predictive biomarkers of response to atezolizumab.We identified tumor neoantigens from 20 patients with mUC and profiled their peripheral CD8 T cells using highly multiplexed combinatorial tetramer staining. Another set of patients with mUC treated with atezolizumab (n=30) or chemotherapy (n=40) were selected to profile peripheral CD8 T cells by mass cytometry. Using single-cell transcriptional analysis (single-cell RNA sequencing (scRNA-seq)), together with CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) and paired T-cell receptor (TCR) sequencing, we further characterized peripheral CD8 T cells in a subset of patients (n=16).High frequency of CD57 was observed in neoantigen-specific CD8 T cells in patients with mUC responding to atezolizumab. Extending these findings to bulk CD8 T cells, we found higher frequency of CD57 expressing CD8 T cells before treatment in patients responding to atezolizumab (n=20, p0.01) but not to chemotherapy. These findings were corroborated in a validation cohort (n=30, p0.01) and notably were independent of known biomarkers of response. scRNA-seq analysis identified a clonally expanded cluster enriched within CD57Collectively, we show high frequencies of CD57 among neoantigen-specific and bulk CD8 T cells in patients responding to atezolizumab. The TCR repertoire overlap between peripheral CD57

Most colorectal (CRC) tumors are dependent on EGFR/KRAS/BRAF/MAPK signaling activation. ARID1A is an epigenetic regulator mutated in approximately 5% of non-hypermutated CRC tumors. Here we show that anti-EGFR but not anti-VEGF treatment enriches for emerging ARID1A mutations in CRC patients. In addition, we find that patients with ARID1A mutations, at baseline, are associated with worse outcome when treated with cetuximab- but not bevacizumab-containing therapies; thus, this suggests that ARID1A mutations may provide both an acquired and intrinsic mechanism of resistance to anti-EGFR therapies. We find that, ARID1A and EGFR-pathway genetic alterations are mutually exclusive across lung and colorectal cancers, further supporting a functional connection between these pathways. Our results not only suggest that ARID1A could be potentially used as a predictive biomarker for cetuximab treatment decisions but also provide a rationale for exploring therapeutic MAPK inhibition in an unexpected but genetically defined segment of CRC patients.

BackgroundA growing body of evidence suggests that T-cell responses against neoantigens are critical regulators of response to immune checkpoint blockade. We previously showed that circulating neoantigen-specific CD8 T cells in patients with lung cancer responding to anti-Programmed death-ligand 1 (PD-L1) (atezolizumab) exhibit a unique phenotype with high expression of CD57, CD244, and KLRG1. Here, we extended our analysis on neoantigen-specific CD8 T cells to patients with metastatic urothelial cancer (mUC) and further profiled total CD8 T cells to identify blood-based predictive biomarkers of response to atezolizumab.MethodsWe identified tumor neoantigens from 20 patients with mUC and profiled their peripheral CD8 T cells using highly multiplexed combinatorial tetramer staining. Another set of patients with mUC treated with atezolizumab (n=30) or chemotherapy (n=40) were selected to profile peripheral CD8 T cells by mass cytometry. Using single-cell transcriptional analysis (single-cell RNA sequencing (scRNA-seq)), together with CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) and paired T-cell receptor (TCR) sequencing, we further characterized peripheral CD8 T cells in a subset of patients (n=16).ResultsHigh frequency of CD57 was observed in neoantigen-specific CD8 T cells in patients with mUC responding to atezolizumab. Extending these findings to bulk CD8 T cells, we found higher frequency of CD57 expressing CD8 T cells before treatment in patients responding to atezolizumab (n=20, p+ CD8 T cells in responding patients characterized by higher expression of genes associated with activation, cytotoxicity, and tissue-resident memory markers. Furthermore, compared with CD57− CD8 T cells, TCRs of CD57+ CD8 T cells showed increased overlap with the TCR repertoire of tumor-infiltrating T cells.ConclusionsCollectively, we show high frequencies of CD57 among neoantigen-specific and bulk CD8 T cells in patients responding to atezolizumab. The TCR repertoire overlap between peripheral CD57+ CD8 T cells and tumor-infiltrating lymphocytes suggest that accumulation of peripheral CD57+ CD8 T cells is reflective of an ongoing antitumor T-cell response. Our findings provide evidence and rationale for using circulating CD8 T cells expressing CD57 as a readily accessible blood-based biomarker for selecting patients with mUC for atezolizumab therapy.

PurposeRenal cell carcinoma (RCC) with venous tumor thrombus (VTT) arising from the primary tumor occurs in 4-10% of cases and is associated with advanced disease. RCC with VTT and distant metastasis represents a unique clinical entity, and provides opportunities to examine the origins and relative timing of tumor lesion emergence and to identify molecular correlates with disease state.Experimental DesignWe performed genomic and evolutionary analyses on 16 RCC patients with VTT, with eight also having metastases, using multi-region exome and RNA sequencing.ResultsNo genomic alterations were specifically associated with the VTT or metastasis lesions; each tumor had multiple hallmark driver alterations, consistent with advanced disease state. We found that 21% (3/14) of clear-cell RCC cases could be assigned a previously defined “evolutionary subtype”. Somatic mutation signatures were largely consistent with previously established RCC signatures, and showed low heterogeneity across regions of each tumor. Mismatch repair and homologous recombination (“BRCA-ness”) deficiency signatures consistently co-occurred across most tumors, suggesting a pervasive role for intracellular DNA damage in RCC and the potential for related treatment strategies. Phylogenetic timing analysis of metastatic cases suggested that in most tumors, metastases branched from the primary tumor prior to formation of VTT and in some cases before diversification of the primary tumor. Both VTT and the earliest metastases were predicted to emerge many years prior to diagnosis. Transcriptional landscape analysis identified key differences distinguishing each lesion type from primary tumor: VTT upregulated TNFα signaling and associated inflammatory pathways, whereas metastases upregulated MTOR signaling.ConclusionsOur results provide a map of how RCC tumors can evolve, with metastatic clones typically emerging early in RCC development and taking hold via MTOR signaling, and later formation of VTT via local inflammatory processes.Statement of Translational RelevanceRenal cell carcinoma (RCC) is a deadly and relatively common malignancy, which often presents as or progresses to metastatic disease. We used multi-region sequencing of RCC patients with venous tumor thrombus (VTT) and metastasis to ask how and when new lesions arise from the primary tumor, and what genomic factors contribute to their spread. Phylogenetic analysis of patients with VTT and co-presenting metastases suggested that in most cases, the VTT and metastases derive from distinct tumor clones. Moreover, metastatic clones often appear many years prior to diagnosis. We found that local TNFα inflammation may contribute to VTT formation, whereas MTOR signaling is associated with metastases. Our study sheds light on the relationship of VTT and metastases, suggests therapeutic and biomarker strategies for RCC, and points to the need for early detection studies in RCC to better understand when metastases emerge and to identify at-risk patients.

Tumor enrichment in low tumor content tissues, those below 20% tumor content depending on the method, is required to generate quality data reproducibly with many downstream assays such as next generation sequencing. Automated tissue dissection is a new methodology that automates and improves tumor enrichment in these common, low tumor content tissues by decreasing the user-dependent imprecision of traditional macro-dissection and time, cost, and expertise limitations of laser capture microdissection by using digital image annotation overlay onto unstained slides. Here, digital hematoxylin and eosin (H&E) annotations are used to target small tumor areas using a blade that is 250 µm2 in diameter in unstained formalin fixed paraffin embedded (FFPE) or fresh frozen sections up to 20 µm in thickness for automated tumor enrichment prior to nucleic acid extraction and whole exome sequencing (WES). Automated dissection can harvest annotated regions in low tumor content tissues from single or multiple sections for nucleic acid extraction. It also allows for capture of extensive pre- and post-harvest collection metrics while improving accuracy, reproducibility, and increasing throughput with utilization of fewer slides. The described protocol enables digital annotation with automated dissection on animal and/or human FFPE or fresh frozen tissues with low tumor content and could also be used for any region of interest enrichment to boost adequacy for downstream sequencing applications in clinical or research workflows.

Immune checkpoint inhibitors targeting the PD-1/PD-L1 axis lead to durable clinical responses in subsets of cancer patients across multiple indications, including non-small cell lung cancer (NSCLC), urothelial carcinoma (UC) and renal cell carcinoma (RCC). Herein, we complement PD-L1 immunohistochemistry (IHC) and tumor mutation burden (TMB) with RNA-seq in 366 patients to identify unifying and indication-specific molecular profiles that can predict response to checkpoint blockade across these tumor types. Multiple machine learning approaches failed to identify a baseline transcriptional signature highly predictive of response across these indications. Signatures described previously for immune checkpoint inhibitors also failed to validate. At the pathway level, significant heterogeneity is observed between indications, in particular within the PD-L1+ tumors. mUC and NSCLC are molecularly aligned, with cell cycle and DNA damage repair genes associated with response in PD-L1- tumors. At the gene level, the CDK4/6 inhibitor CDKN2A is identified as a significant transcriptional correlate of response, highlighting the association of non-immune pathways to the outcome of checkpoint blockade. This cross-indication analysis reveals molecular heterogeneity between mUC, NSCLC and RCC tumors, suggesting that indication-specific molecular approaches should be prioritized to formulate treatment strategies., PD-L1 immune checkpoint inhibition has been used for several tumour types. Here, the authors use immunohistochemistry, tumour mutation burden and RNA-seq data from 366 patients with different indications to identify molecular signatures of response to atezolizumab and reveal pathway heterogeneity and the involvement of non-immune pathways.

Personalized cancer immunotherapies (pCIT) rely on targeting of somatic cancer mutations, which in their translated peptide form are known as neoantigens. Each tumor has a unique set of mutations, and thus, the vast majority of cancer neoantigens targeted by these personalized therapies are private to an individual's tumor. We sought to understand whether a single strategy for targeting private neoantigens—prioritizing clonal neoantigens over subclonal ones—could apply equally well across primary and metastatic lesions in patients with advanced metastatic solid tumors, and across different cancer indications. Previous studies of largely primary tumors in non-metastatic settings have suggested that indications such as melanoma and NSCLC consistently have low mutational heterogeneity and a preponderance of clonal neoantigens, whereas other indications such as CRC, RCC, and breast cancer often have a higher degree of mutational heterogeneity and fewer clonal neoantigens. It remains unclear whether standard clonality metrics (e.g., cancer cell fraction or "CCF") can accurately predict global mutation clonality across tumor lesions and whether CCF could offer any predictive benefit over simple variant allele frequencies. By multi-region sequencing analysis of five metastatic solid tumors across four indications (CRC, NSCLC, RCC, UBC), with all lesions sampled at the same time point, we characterized neoantigen heterogeneity and whether the clonality and expression level of mutations would influence the likely immunogenicity of neoantigen candidates selected for pCIT. We show that the ability to target cancer neoantigens effectively depends on the type of lesion sampled and on indication. In NSCLC and UBC, where most mutations were shared across tumor lesions, effective targeting could be accomplished by sampling either the primary or certain metastatic lesions. However, CRC and RCC tumors had more complex, branching mutational phylogenies, leading to non-overlapping sets of mutations across different tumor lesions. On average, across the five metastatic cases, a given tumor sample’s predictive value for global mutation clonality varied from 20% (RCC) to 70% (NSCLC). Enrichment of clonal neoantigens could be accomplished by prioritizing mutations with higher variant allele frequency (VAF), as expected, but using CCF for mutation prioritization led to poorer enrichment of clonal neoantigens. In one case (UBC), we observed truncal neoantigen reduction via down-regulation of mutant allele expression, suggesting that early immune recognition of tumors is an important factor in pCIT targeting. Our data suggest that indication-specific neoantigen targeting strategies, which consider mutation presence and expression across multiple tumor lesions, may be necessary for pCIT to be broadly effective. Citation Format: Amy Lo, Andrew Wallace, Daniel Oreper, Nicolas Lounsbury, Charles Havnar, Carmina Espiritu, Ximo Pechuan-Jorge, Richard Bourgon, Ryan Jones, Katrina Krogh, Guang-Yu Yang, Oliver Zill. Multi-region sequencing analysis of metastatic solid tumors to inform targeting of personalized cancer immunotherapies [abstract]. In: Abstracts: AACR Virtual Special Conference: Tumor Immunology and Immunotherapy; 2020 Oct 19-20. Philadelphia (PA): AACR; Cancer Immunol Res 2021;9(2 Suppl):Abstract nr PO092.

Sequencing of the mutant allele fraction of circulating cell-free DNA (cfDNA) derived from tumors is increasingly utilized to detect actionable genomic alterations in cancer. We conducted a prospective blinded study of a comprehensive cfDNA sequencing panel with 54 cancer genes. To evaluate the concordance between cfDNA and tumor DNA (tDNA), sequencing results were compared between cfDNA from plasma and genomic tumor DNA (tDNA). Utilizing next generation digital sequencing technology (DST), we profiled approximately 78,000 bases encoding 512 complete exons in the targeted genes in cfDNA from plasma. Seventy-five patients were prospectively enrolled between February 2013 and March 2014, including 61 metastatic cancer patients and 14 clinical stage II CRC patients with matched plasma and tissue samples. Using the 54-gene panel, we detected at least one somatic mutation in 44 of 61 tDNA (72.1%) and 29 of 44 (65.9%) cfDNA. The overall concordance rate of cfDNA to tDNA was 85.9%, when all detected mutations were considered. We collected serial cfDNAs during cetuximab-based treatment in 2 metastatic KRAS wild-type CRC patients, one with acquired resistance and one with primary resistance. We demonstrate newly emerged KRAS mutation in cfDNA 1.5 months before radiologic progression. Another patient had a newly emerged PIK3CA H1047R mutation on cfDNA analysis at progression during cetuximab/irinotecan chemotherapy with gradual increase in allele frequency from 0.8 to 2.1%. This blinded, prospective study of a cfDNA sequencing showed high concordance to tDNA suggesting that the DST approach may be used as a non-invasive biopsy-free alternative to conventional sequencing using tumor biopsy.

Extensive biological and clinical evidence supports the notion many CRC tumors are addicted to EGFR/KRAS/MAPK signaling. For example, mutations in KRAS, BRAF and other MAPK pathway components confer intrinsic or acquired resistance to anti-EGFR therapy. However, a significant proportion of acquired resistance cases have remained unexplained. ARID1A mutations are found in approximately 11% of CRC tumors, but a connection between ARID1A mutations and EGFR/MAPK signaling activation in CRC has not been previously suspected. By leveraging the genomic biomarker data collected from 333 patients at baseline and at progression who participated in the CALGB/SWOG 80405 trial, we found that mutations in ARID1A, a key component of SWI/SNF complex, were enriched in the cell free DNA of 6/16 of the patients whose tumors had become resistant to anti-EGFR therapy with ARID1A mutations detected at either timepoint (adj. p = 0.03, OR = 0.09, two-tailed fisher's exact test). In contrast, there was no evidence of enriched ARID1A mutations in patients treated with bevacizumab. To investigate the potential role of a broader ARID1A deficiency phenotype in resistance to cetuximab therapy, we developed an ARID1A mutant-like signature from TCGA that captures the transcriptional profile characteristic of ARID1A mutants. Using this signature to stratify CALGB patients based on gene expression data collected on tissue at diagnosis, we also found patients with an ARID1A mutant-like gene expression profile had worst outcome in patients treated with cetuximab than bevacizumab with adjustment for established clinical variables for both OS [p = 0.0002, hazard ratio (HR) 6.2, 95% confidence interval (CI) 2.4-16], and PFS [p = 0.0008, HR 4.7 (1.9 -12)], suggesting that ARID1A mutations may be also a mechanism of intrinsic resistance to anti-EGFR therapies. Gene expression analysis demonstrated that ARID1A deficiency leads to re-activation of an EGFR-like signature, suggesting reactivation of this pathway as a possible mechanism of resistance. Furthermore, CRISPR knockout of ARID1A in the cetuximab-sensitive CRC cell line NCI-H508 conferred elevated MAPK signaling relative to parental line and resistance to cetuximab in culture. Consistent with these findings, we also observed strong mutual exclusivity between ARID1A mutations and mutations in the EGFR/MAPK signaling pathway in more than 40,000 lung and colorectal cancer patients profiled in the FoundationCore® database. Strikingly, in lung cancers where EGFR mutations structural variants (SV) are more prevalent, EGFR SV was the top mutually exclusive alteration with ARID1A mutations in lung cancers. Of 5980 lung cancer patients with ARID1A and/or EGFR SVs, only 100 patients (1.7%) had both SVs. Taken together, our data suggest that ARID1A loss-of-function mutations may promote resistance to cetuximab by driving an EGFR-like transcription program in the absence of ligand-dependent activation of EGFR. Our results suggest that ARID1A defects could be potentially used as an exclusion biomarker for cetuximab treatment decisions, and they provide a rationale for exploring therapeutic MAPK inhibition in ARID1A mutant CRC patients. Citation Format: Radia Marie Johnson, Xueping Qu, Joshua Thomas, Yvonne Kschonsak, Ling-Yuh Huw, Fang-Shu Ou, Ethan Sokol, Nnamdi Ihuegbu, Oliver Zill, Omar Kabbarah, Anneleen Daemen, Richard Bourgon, Alan Venook, Federico Innocenti, Heinz-Josef Lenz, Felipe de Sousa e Melo, Carlos Bais. ARID1A mutations induce an EGFR-like gene expression signature and confer intrinsic and acquired resistance to cetuximab treatment in first line metastatic CRC [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 LB-063.

Purpose: Cell-free DNA (cfDNA) sequencing provides a noninvasive method for obtaining actionable genomic information to guide personalized cancer treatment, but the presence of multiple alterations in circulation related to treatment and tumor heterogeneity complicate the interpretation of the observed variants.Experimental Design: We describe the somatic mutation landscape of 70 cancer genes from cfDNA deep-sequencing analysis of 21,807 patients with treated, late-stage cancers across >50 cancer types. To facilitate interpretation of the genomic complexity of circulating tumor DNA in advanced, treated cancer patients, we developed methods to identify cfDNA copy-number driver alterations and cfDNA clonality.Results: Patterns and prevalence of cfDNA alterations in major driver genes for non–small cell lung, breast, and colorectal cancer largely recapitulated those from tumor tissue sequencing compendia (The Cancer Genome Atlas and COSMIC; r = 0.90–0.99), with the principal differences in alteration prevalence being due to patient treatment. This highly sensitive cfDNA sequencing assay revealed numerous subclonal tumor-derived alterations, expected as a result of clonal evolution, but leading to an apparent departure from mutual exclusivity in treatment-naïve tumors. Upon applying novel cfDNA clonality and copy-number driver identification methods, robust mutual exclusivity was observed among predicted truncal driver cfDNA alterations (FDR = 5 × 10−7 for EGFR and ERBB2), in effect distinguishing tumor-initiating alterations from secondary alterations. Treatment-associated resistance, including both novel alterations and parallel evolution, was common in the cfDNA cohort and was enriched in patients with targetable driver alterations (>18.6% patients).Conclusions: Together, these retrospective analyses of a large cfDNA sequencing data set reveal subclonal structures and emerging resistance in advanced solid tumors. Clin Cancer Res; 24(15); 3528–38. ©2018 AACR.

Purpose: To analytically and clinically validate a circulating cell-free tumor DNA sequencing test for comprehensive tumor genotyping and demonstrate its clinical feasibility. Experimental Design: Analytic validation was conducted according to established principles and guidelines. Blood-to-blood clinical validation comprised blinded external comparison with clinical droplet digital PCR across 222 consecutive biomarker-positive clinical samples. Blood-to-tissue clinical validation comprised comparison of digital sequencing calls to those documented in the medical record of 543 consecutive lung cancer patients. Clinical experience was reported from 10,593 consecutive clinical samples. Results: Digital sequencing technology enabled variant detection down to 0.02% to 0.04% allelic fraction/2.12 copies with ≤0.3%/2.24–2.76 copies 95% limits of detection while maintaining high specificity [prevalence-adjusted positive predictive values (PPV) >98%]. Clinical validation using orthogonal plasma- and tissue-based clinical genotyping across >750 patients demonstrated high accuracy and specificity [positive percent agreement (PPAs) and negative percent agreement (NPAs) >99% and PPVs 92%–100%]. Clinical use in 10,593 advanced adult solid tumor patients demonstrated high feasibility (>99.6% technical success rate) and clinical sensitivity (85.9%), with high potential actionability (16.7% with FDA-approved on-label treatment options; 72.0% with treatment or trial recommendations), particularly in non–small cell lung cancer, where 34.5% of patient samples comprised a directly targetable standard-of-care biomarker. Conclusions: High concordance with orthogonal clinical plasma- and tissue-based genotyping methods supports the clinical accuracy of digital sequencing across all four types of targetable genomic alterations. Digital sequencing's clinical applicability is further supported by high rates of technical success and biomarker target discovery. Clin Cancer Res; 24(15); 3539–49. ©2018 AACR.

Background About 65% of breast cancers express the estrogen receptor α and the mainstay of treatment are therapies that result in estrogen receptor modulation (selective estrogen receptor modulators, SERMs) or estrogen deprivation (aromatase inhibitors, AIs). Even though endocrine therapy has resulted in reduced recurrence and mortality, a significant portion of patients relapse with metastatic disease and subsequently progress while on therapy for advanced disease (endocrine resistance). Recent evidence showed that activating hot spot mutations in the ligand binding domain of the ERα (ESR1) are acquired on treatment (frequency of 20%) and can drive resistance to endocrine therapy, especially AIs. ESR1 mutations can be detected by evaluation of circulating tumor DNA (ctDNA), a method where circulating DNA fragments with tumor-specific sequence alterations are identified in the blood of patients. Methods This is a retrospective evaluation of 9 patients with hormone receptor positive (HR+) metastatic breast cancer (MBC) who had progressed on multiple lines of endocrine therapy (ET) and were found to have ESR1 mutations in ctDNA. Patients had blood drawn for ctDNA analysis either at progression to serve as a baseline before starting a new regimen or to monitor response to ongoing treatment. Guardant360™(Guardant Health) involves ctDNA isolation from plasma using a Qiagen circulating nucleic acid kit, then a panel of 68 gene mutations associated with solid tumors as reported in the COSMIC database sequenced using single-molecule digital sequencing technology. Results All of the patients had MBC and were luminal subtype except for one HER2+, and most had invasive ductal carcinoma although 2 patients were invasive lobular carcinoma (22%). Most patients had both bone and visceral involvement (78%), only two patients had bone only metastasis. The patients were generally heavily pretreated with an average of 3 lines of ETs and 6 lines of therapy altogether (chemotherapy + ET). Duration on endocrine therapy ranged from 23 months to 7 years (mean 4.3 years). All patients were found to have ESR1 mutations on ctDNA, the range of percentage of mutant allele was 0.28-23.76%. Three patients had tissue sent for NGS and none of the tissue samples had an ESR1 mutation detected although they were biopsied at various time points in treatment. One of those patients had two ESR1 mutations in ctDNA, which were not detected on tissue sent for NGS one year prior, and had not been on ET for several years. One patient with abdominal carcinomatosis from lobular carcinoma who had been on ETs therapies for 6 years was found to have 4 distinct ESR1 mutations in a single blood draw, suggesting sub-clonal evolution of resistance. One patient also had 5 circulating tumor cells, all of which had ESR1 mutations detected when circulating tumor cells were individually sequenced. Conclusions ctDNA is a sensitive test for detection of ESR1 in HR+ MBC patients, with the advantage of being a blood based assay which lends itself to serial analysis. In this patient population ctDNA can be a helpful tool to predict response to ET and predict treatment failure. Citation Format: Austin L, Rodriguez A, Jaslow R, Fortina P, Nagy R, Zill O, Talasaz A, Cristofanilli M. Detection of activating estrogen receptor 1 (ESR1) in circulating tumor DNA (ctDNA) in hormone-receptor positive metastatic breast cancer (MBC). [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 P3-05-02.

The genetic bases for species-specific traits are widely sought, but reliable experimental methods with which to identify functionally divergent genes are lacking. In the Saccharomyces genus, interspecies complementation tests can be used to evaluate functional conservation and divergence of biological pathways or networks. Silent information regulator (SIR) proteins in S. bayanus provide an ideal test case for this approach because they show remarkable divergence in sequence and paralog number from those found in the closely related S. cerevisiae. We identified genes required for silencing in S. bayanus using a genetic screen for silencing-defective mutants. Complementation tests in interspecies hybrids identified an evolutionarily conserved Sir-protein-based silencing machinery, as defined by two interspecies complementation groups (SIR2 and SIR3). However, recessive mutations in S. bayanus SIR4 isolated from this screen could not be complemented by S. cerevisiae SIR4, revealing species-specific functional divergence in the Sir4 protein despite conservation of the overall function of the Sir2/3/4 complex. A cladistic complementation series localized the occurrence of functional changes in SIR4 to the S. cerevisiae and S. paradoxus branches of the Saccharomyces phylogeny. Most of this functional divergence mapped to sequence changes in the Sir4 PAD. Finally, a hemizygosity modifier screen in the interspecies hybrids identified additional genes involved in S. bayanus silencing. Thus, interspecies complementation tests can be used to identify (1) mutations in genetically underexplored organisms, (2) loci that have functionally diverged between species, and (3) evolutionary events of functional consequence within a genus.

High-quality, well-annotated genome sequences and standardized laboratory strains fuel experimental and evolutionary research. We present improved genome sequences of three species of Saccharomyces sensu stricto yeasts: S. bayanus var. uvarum (CBS 7001), S. kudriavzevii (IFO 1802(T) and ZP 591), and S. mikatae (IFO 1815(T)), and describe their comparison to the genomes of S. cerevisiae and S. paradoxus. The new sequences, derived by assembling millions of short DNA sequence reads together with previously published Sanger shotgun reads, have vastly greater long-range continuity and far fewer gaps than the previously available genome sequences. New gene predictions defined a set of 5261 protein-coding orthologs across the five most commonly studied Saccharomyces yeasts, enabling a re-examination of the tempo and mode of yeast gene evolution and improved inferences of species-specific gains and losses. To facilitate experimental investigations, we generated genetically marked, stable haploid strains for all three of these Saccharomyces species. These nearly complete genome sequences and the collection of genetically marked strains provide a valuable toolset for comparative studies of gene function, metabolism, and evolution, and render Saccharomyces sensu stricto the most experimentally tractable model genus. These resources are freely available and accessible through www.SaccharomycesSensuStricto.org.

Background: Cell-free circulating tumor DNA analysis provides a non-invasive method for obtaining actionable genomic information to guide personalized cancer treatment. Deep sequencing of cell-free DNA (cfDNA) can potentially provide insights into tumor heterogeneity across multiple tumor sites in a patient, including emerging treatment-resistant subclones. However, the increased informational complexity of polyclonal cfDNA in circulation poses analysis challenges, particularly in tumors with abundant copy number alterations. To facilitate interpretation of this added complexity, we developed methods to identify cfDNA copy-number driver alterations and cfDNA clonality, Methods: We analyzed a large clinical sequencing database of somatic point mutations and copy number alterations from targeted cfDNA sequencing of 21,807 consecutive patients across >50 cancer types (Guardant Health, CA). We evaluated a minimal cfDNA clonality model that relies on the relationship between variant allele frequency (VAF) in cfDNA and the level of tumor DNA in circulation (ctDNA level), while accounting for copy number alterations. Results: We found that the initial simple model of cfDNA clonality performed well on >90% of samples, given a relatively small targeted genomic region (70 genes, 150 kb). However, normalizing VAF by copy number is subject to error in some samples due to the effect of ctDNA level on variant detection, variable unique molecule coverage across samples, and non-linearity of VAF at high copy number. Therefore, we developed an improved cfDNA clonality model that incorporated these analytical and biological features, which was then trained on a portion of the large cfDNA data set. Our cfDNA clonality model accurately distinguished subclonal resistance from driver alterations in a test set of over 5,000 lung, colorectal, and breast cancer patients. Although numerous subclonal tumor-derived alterations were apparent in the initial test data set, leading to an apparent departure from mutual exclusivity in treatment-naïve tumors, robust mutual exclusivity was observed among cfDNA clonal driver alterations when our cfDNA clonality analysis method was applied. These results suggest our analytical approach can be used to identify treatment-associated emerging resistance alterations in patients from a single blood draw, including parallel evolution of distinct subclonal alterations. Conclusion: Managing cancer will likely depend on identifying emerging treatment-resistant subclones at or in anticipation of progression. Highly accurate deep sequencing of cfDNA, along with comprehensive models of cfDNA clonality, can elucidate subclonal structure of the tumor and identify emerging treatment resistance. Citation Format: Stephen Fairclough, Oliver Zill, Catalin Barbacioru, Justin Odegaard, Richard B. Lanman, AmirAli Talasaz, Darya Chudova. A method for differentiating clonal driver mutations from subclonal emerging resistance mutations in circulating cell-free DNA [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 2190.

Patients with pancreatic and biliary carcinomas lack personalized treatment options, in part because biopsies are often inadequate for molecular characterization. Cell-free DNA (cfDNA) sequencing may enable a precision oncology approach in this setting. We attempted to prospectively analyze 54 genes in tumor and cfDNA for 26 patients. Tumor sequencing failed in 9 patients (35%). In the remaining 17, 90.3% (95% confidence interval, 73.1%–97.5%) of mutations detected in tumor biopsies were also detected in cfDNA. The diagnostic accuracy of cfDNA sequencing was 97.7%, with 92.3% average sensitivity and 100% specificity across five informative genes. Changes in cfDNA correlated well with tumor marker dynamics in serial sampling (r = 0.93). We demonstrate that cfDNA sequencing is feasible, accurate, and sensitive in identifying tumor-derived mutations without prior knowledge of tumor genotype or the abundance of circulating tumor DNA. cfDNA sequencing should be considered in pancreatobiliary cancer trials where tissue sampling is unsafe, infeasible, or otherwise unsuccessful. Significance: Precision medicine efforts in biliary and pancreatic cancers have been frustrated by difficulties in obtaining adequate tumor tissue for next-generation sequencing. cfDNA sequencing reliably and accurately detects tumor-derived mutations, paving the way for precision oncology approaches in these deadly diseases. Cancer Discov; 5(10); 1040–8. ©2015 AACR. This article is highlighted in the In This Issue feature, p. 1005

Analysis of cell-free circulating tumor DNA (ctDNA) enables non-invasive and serially repeatable genomic profiling of advanced cancer patients, providing options when tissue biopsy is contra-indicated or of insufficient quantity (20-25% of solid tumor patients). Liquid biopsy studies to-date have been limited to modest-sized cohorts and case studies. Here we present genomic profiling results from liquid biopsies of >10,000 advanced cancer patients in whom Guardant360TM (G360) was ordered for clinical care. G360 provides deep-coverage (15,000x average) and highly accurate sequencing of ctDNA across a 70-gene target-capture panel. It detects single nucleotide variants, indels, gene amplifications, and fusions across all NCCN-recommended solid tumor genomic alterations with analytic specificity >99.9999% and analytic sensitivity Of the >10,000 patients (>50 solid cancer types), 80% of patients had at least one somatic alteration detected in ctDNA (median = 3; mean = 4.3 alterations). The most common cancers were lung (32%), gastrointestinal (23%), and breast (14%). Mutational spectra across the 70 analyzed genes were similar to TCGA results with the notable exception of increased prevalence of resistance mutations in the liquid biopsy cohort, likely owing to ongoing/prior therapy. Canonical drivers in NSCLC patients were generally mutually exclusive, although activating EGFR or RAS mutations were observed in 12.5% of patients with fusions (5/40). Colorectal cancers showed both mutual exclusivity among KRAS, NRAS, and BRAF drivers, and convergent evolution toward downstream pathway activation under anti-EGFR therapy, with some samples showing multiple resistance mechanisms. Mutation patterns for the most frequently mutated genes were generally well correlated between ctDNA and published tissue data, including for commonly altered tumor suppressor genes (for TP53: Pearson r = 0.94, Spearman ρ = 0.80). G360 is ordered prior to initial treatment when biopsy tissue is exhausted, unobtainable or under-genotyped, but more often is ordered upon progression to identify evolving resistance mechanisms that may be targetable. Thus, relative frequencies of genomic alterations were expected to differ from published frequencies established in unselected, early-stage, treatment-naïve cohorts. Nonetheless, the specific patterns of alterations in ctDNA from this unprecedented liquid biopsy cohort largely recapitulated patterns observed in published tissue sequencing studies. Moreover, ctDNA sequencing clearly and robustly identified convergent evolution of drug resistance occurring under therapy. Citation Format: Oliver A. Zill, Kimberly C. Banks, Coyt Jackson, Stefanie Mortimer, Arthur Baca, Becky Nagy, Richard B. Lanman, Helmy Eltoukhy, AmirAli Talasaz. Comparison of over 10,000 clinical NGS circulating tumor DNA profiles to tissue-derived genomic compendia. [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 4343.

Background: Metastatic breast cancer (MBC) is an incurable disease with complex molecular features including somatic mutations that evolve in relation to genomic instability and selective treatment pressure. Patients with treatment-refractory MBC may benefit from tumor genomic evaluation using next generation sequencing (NGS). Furthermore, analysis of circulating tumor DNA (ctDNA) in patients with advanced disease offers the possibility of non-invasive molecular monitoring. Methods: A patient with MBC was tested at each progression with a ctDNA NGS panel (Guardant360™) that includes all NCCN-recommended somatic genomic variants for solid tumors and sequences complete exons of >50 genes to report single nucleotide variants (SNVs), fusions, amplifications, and indels with high sensitivity and ultra-high specificity (>99.9999%). The patient was diagnosed with invasive breast cancer at age 44 and treated with surgery and hormonal therapy. At age 61, she had axillary adenopathy and liver metastases. Treatment details are in Table 1. Results: ctDNA analysis was performed at the time of metastatic diagnosis and at 5 additional time points over the course of treatment. All samples revealed an ERBB2 exon 19 indel (p.Leu755_Glu757delinsSer), and multiple SNVs and gene amplifications. ERBB2 amplification was seen in 4 of 6 samples. Mutant allele fractions (Table 1) correlated with clinical response to treatment and progression. Conclusions: Analysis of ctDNA in this patient identified an ERBB2 exon 19 indel, which are present in 2-4% of non-small cell lung cancers but 1-2% in breast cancer. Treatment with anti-HER2 monoclonal antibody or dual anti-EGFR/ERBB2 tyrosine kinase inhibitor therapies may show clinical benefit. ctDNA analysis can detect emergence of actionable resistance mutations with the advantage of serial evaluation, allowing capture of inter- and intra-tumor heterogeneity and illustration of molecular progression and response. Table 1.Blood DrawDisease status at time of blood drawMutant allele fraction of ERBB2 indel (* = ERBB2 amplification)Treatment1Initial diagnosis9%trastuzumab, pertuzumab and docetaxel (TPD), cycle 12Stable disease1%*TPD cycle 23Progressing10%*TPD cycle 34Progressing62%*trastuzumab, emtansine (TDM1) 3 cycles5Progressing70%*vinorelbine, trastuzumab, everolimus6Partial response1%flourouracil, epirubicin, cyclophoasphamide Citation Format: Laura Austin, Rebecca Nagy, Oliver Zill, Richard B. Lanman, AmirAli Talasaz, Massimo Cristofanilli. Managing metastatic breast cancer via serial monitoring with circulating cell-free tumor DNA next generation sequencing testing. [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 172.

e20643Background: Somatic mutations (SNVs) in ALK are a common mechanism of resistance to crizotinib, and may respond to next-generation ALK inhibitors. Historically, identification of these ALK SN...

e23021Background: In non-small cell lung cancer (NSCLC), targeted treatment with first line tyrosine kinase inhibitors (TKIs) selects for the evolution of resistance at EGFR T790M in 50% of cases. ...

LBA11501 Background: Next-generation sequencing (NGS) of circulating tumor DNA (ctDNA) enables non-invasive profiling of solid tumor cancers. Liquid biopsy studies to date have been limited to modest-size cohorts and case studies. Methods: Somatic genomic profiles of over 15,000 patients with advanced-stage clinical cancer were determined by a highly accurate, deep-coverage (15,000x) ctDNA NGS test targeting 70 genes (Guardant360). Frequencies of somatic ctDNA alterations per gene were compared to those previously described in tissue sequencing projects (e.g., TCGA). Accuracy of ctDNA sequencing (PPV) was assessed by comparing with matched tissue tests for 386 patients. Results: The cohort consisted of lung (37%), breast (14%), colorectal (10%) and other cancers (38%), with ctDNA clinical sensitivity of 86%, 83%, 85%, and 78%, respectively. Cancer-type-specific frequencies and mutual exclusivity patterns among major driver alterations largely recapitulated those seen in tissue sequencing studies. Mutation frequencies per codon correlated well between ctDNA and published tissue data, both for commonly altered tumor suppressors and for oncogenes (Pearson correlations: TP53, r = 0.94; KRAS, r = 0.99; PIK3CA, r = 0.99). The overall accuracy of ctDNA sequencing in comparison with matched tissue tests was 87% (336/386). The accuracy increased to 98% when blood and tumor were collected less than six months apart. Four distinct classes of clinical outcome benefits have been observed by liquid biopsy: 1) actionable mutations in cases with tissue QNS ( ALK fusion, or EGFR or BRAF activating mutations in lung; ERBB2 amp in gastric), 2) actionable resistance mutations at time of progression ( MET amp or EGFRT790M in lung), 3) evolution of sensitivity upon progression ( ERBB2-amplified metastatic breast cancer with triple negative primary), 4) under-genotyped tumors ( BRAFV600E or ERBB2indel in lung). Conclusions: Somatic alteration patterns in ctDNA samples largely agree with tissue alteration patterns, with the exception of resistance mutations. Clinical outcome benefits have been observed for patients treated based on ctDNA findings.

Background Metastatic breast cancer (BC) is an incurable condition and treated with palliative intent. Standard diagnostic imaging and serum markers have limited sensitivity and are not recommended in clinical practice. Micrometastatic disease in the bone marrow (DTC) and peripheral blood (CTCs) is a recognized prognostic marker but with limited clinical utility. The detection of asymptomatic disease using a sensitive, reproducible and robust blood-based molecular test, or molecular residual disease (MoRD) could potentially represents a tool with the capability to design early therapeutic interventions and improve outcome. Circulating tumor DNA (ctDNA) has the potential to reflect residual tumor burden with higher diagnostic accuracy. We performed a pilot study in patients with high-risk primary BC. Methods This is a prospective study of 30 patients with either locally advanced BC who had completed primary therapy (21 patients) and had no evidence of disease (NED), or were metastatic but treated with curative intent and currently NED or stable (9 patients). Plasma was analyzed for ctDNA either after completing neo-adjuvant therapy (NAT), for recurrence monitoring after surgery, or when there was a clinical suspicion of recurrence. Guardant360™(Guardant Health) is a ctDNA next generation sequencing panel which produces a quantitative measurement of the mutant allele fraction for single nucleotide variants in 54 genes and copy number variants in 3 genes (panel was expanded to 68 genes in Feb '15) using digital sequencing technology. Results Baseline ctDNA analysis was done for 30 patients and 25 (83%) had serial draws for a total of 76 samples. All patients were stage 3-4 except for two stage 2 patients. ctDNA or MoRD was detected in 17 (57%) of patients and in 39 (51%) of the samples. Of the 18 patients treated with NAT, 11 achieved pCR or had minimal residual disease. Of these 11, six had no ctDNA detected after surgery, 3 had mixed results of no ctDNA and low volume ctDNA alterations on different draws, and 2 had persistent mutations on 2 draws. Ten of these 11 patients remain NED with median follow up of 24 months, while the one patient who recurred had persistent low volume missense ctDNA alterations on serial draws, first detected 6 months before clinically evident recurrence. Of the 7 patients with significant residual disease (less than PR) after NAT, 6 had post-surgical ctDNA detected and 5 have recurred at a median of 13 months after surgery. Four of those patients had ctDNA tested prior to recurrence and all had alterations detected in the blood prior to clinical recurrence. Lastly, one HER2+ metastatic patient treated with curative intent with a subsequent negative PET scan and no ctDNA detected after HER2-targeted therapy progressed on CT 2 months later and repeat ctDNA revealed EGFR mutant allele fraction of 51% and ERBB2 amplification. Conclusions The evaluation of ctDNA in high-risk BC patients can identify MoRD and predict for clinical recurrence. Patients with no or low volume ctDNA after primary treatment remained NED longer than those with multiple or high volume alterations. Future studies will validate these early observations and aid in selecting patients for additional systemic therapy with the hope of improving outcome. Citation Format: Austin L, Jaslow R, Fortina P, Nagy R, Zill O, Talasaz A, Cristofanilli M. Circulating tumor DNA (ctDNA) for detection of molecular residual disease (MoRD) in breast cancer. [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 P6-03-06.

Background: Analysis of cell-free circulating tumor DNA (ctDNA) by next-generation sequencing (NGS) enables non-invasive real-time testing. The Guardant360® (G360) ctDNA panel includes all NCCN-recommended solid tumor genomic targets and sequences complete exons of 68 genes for single nucleotide variants (SNVs), and fusions, amplifications and indels for a subset. We conducted updated validity studies for this larger panel. Methods: Analytical sensitivity was assessed via serial dilution studies of cell-free DNA (cfDNA) with known alterations spiked into healthy cfDNA background. SNVs detected with G360 were compared to whole-exome sequencing at an outside lab for analytic specificity. Tissue NGS for fusions, SNVs, indels, and/or amplifications was compared to G360 results. Consecutive G360 tests over 12 months were analyzed for assay success rate, yield, mutant allele fractions (MAF), and actionability. Results: Serial dilution studies confirmed the previously reported 0.1% lower limit of detection. The accuracy and analytic specificity studies were 98.6% concordant, with 3 putative false positives. Sanger sequencing by a 3rd lab for these discordant calls confirmed the G360 calls as true positives, verifying the previously reported >99.9999% specificity. Retrospective review of outside tumor test reports of 56 patients with varied genomic alterations found 90% concordance with G360 (77/86 variants: 4/4 fusions, 9/9 indels, 13/17 amps, 51/56 SNVs). Of 5,491 tests, 99.5% were successfully reported with somatic alterations found in 76% [median MAF 0.42% (range 0.1-92.8%)]. Lung, breast and colon cancers were most common (see Table). Overall, 10% of patients had >1 on-label matched therapy identified, 54% off-label, and 77% had clinical trial options. On-label matched therapy rates rose for lung adenocarcinoma (15%), breast (16%) and colon (16%) cancers (Table 1). Conclusion: The updated G360 test achieves single-molecule sensitivity to 0.1% MAF and analytical specificity of >99.9999%, as reported previously. Clinical concordance of plasma to tissue-based NGS across all four types of genomic alterations was high, and actionability rose modestly, especially in lung, breast and colon cancer. Clinical utility included biopsy avoidance, rescue of tissue samples with insufficient DNA and tissue-false negatives due to tumor evolution and heterogeneity. Table 1Cancer Type% of 4,888 unique patients% with ctDNA alterations identified% with on-label therapy options identified% with off-label therapy options identified% with clinical trial options identifiedMost commonly involved gene2nd most commonly involved gene3rd most commonly involved geneLung - ALL33%82%10%52%78%TP53 (55%)KRAS (18%)EGFR (18%)Lung adenocarcinoma11%82%15%56%78%TP53 (51%)EGFR (23%)KRAS (19%)Breast15%74%16%58%79%TP53 (43%)PIK3CA (28%)ERBB2/ESR1 (10%) / (9%)CRC12%82%16%72%88%TP53 (61%)APC (38%)KRAS (38%)PDAC7%70%3%57%72%TP53 (47%)KRAS (42%)PIK3CA (6%)Ovarian3%81%1%42%81%TP53 (69%)PIK3CA (15%)KRAS (11%)CUP3%85%0%64%83%TP53 (40%)KRAS (24%)PIK3CA (14%)All100%76%10%54%77%TP53 (50%)KRAS (17%)PIK3CA (14%) Citation Format: Kimberly C. Banks, Stefanie A. W. Mortimer, Oliver A. Zill, Richard B. Lanman, Helmy Eltoukhy, AmirAli Talasaz. Genomic profiling of over 5,000 consecutive cancer patients with a CLIA-certified cell-free DNA NGS test: Analytic and clinical validity and utility. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr B140.

Introduction: EGFR T790M has emerged as an important biomarker to predict response to 3rd-generation tyrosine kinase inhibitors (TKI). Therapies such as AZD 9291 and Rociletinib demonstrated impressive results in the clinical setting in patients with non-small cell lung cancer (NSCLC). However just as this resistance biomarker (T790M) has developed as a result of treatment with prior EGFR TKIs, EGFR C797S mutation is an acquired resistance mechanism to EGFR T790M inhibitors. This mutation was recently reported for the first time in clinical practice utilizing tissue-biopsy based next generation sequencing (NGS) and in the clinical trial setting in cell-free circulating DNA (cfDNA). Here we report the first case of a patient with NSCLC and emergence of C797S mutation utilizing a cfDNA biopsy-free NGS panel in serial assessments in the clinical setting. Methods: Patients with NSCLC were seen at Levine Cancer Institute and had serial blood draws to assess molecular aberrations in cfDNA over time for comparison with radiographic reviews. Blood samples were sent to Guardant360 for assessment by NGS with a targeted cfDNA NGS panel for 68 genes with complete exon sequencing for all 28 exons in EGFR, and other genes in the panel. Mutant allele fractions (MAF) are reported as% of mutant DNA molecules divided by total molecules (mostly leukocyte DNA-derived) overlapping the same mutated nucleotide base position. Results: A 50 year old African-American female never-smoker with lung adenocarcinoma developed a tissue-biopsy confirmed EGFR T790M mutation after 12 months on erlotinib for EGFR exon 19 deletion. She had significant clinical and radiographic responses to a 3rd generation TKI started May 2014. cfDNA NGS tests in March and April 2015 showed no genomic alterations detectable. In late May 2015, she developed clinical symptoms of cough, dyspnea and fatigue but no definitive radiographic evidence of progression on scans June 2015. She received steroids and antibiotics and continued treatment with the 3rd generation TKI. However, corresponding cfDNA NGS testing revealed emergence of multiple low level mutations and the original deletion in EGFR (EGFR L747_P753 DelInsS, T790M, C797S, and A755G variant of uncertain significance) and TP53 (all Discussion: Complete exon sequencing of cfDNA upon clinical progression uncovered the simultaneous re-emergence of the original driver EGFR exon 19 del and resistance driver mutation EGFR T790M as well as the emergence of the next resistance driver mutation EGFR C797S. The latter mutation interferes with covalent binding of a 3rd generation TKI and under treatment pressure, the tumor appears to have simultaneously evolved two different EGFR C797S clones at low MAFs. cfDNA can be a dynamic measure of tumor response, easily obtained from a patient on any visit with potential for detection before macroscopic/radiographic evidence of progression, as demonstrated in this case with tumor evolution of the EGFR T790M and C797S alterations. Citation Format: Kathryn F. Mileham, Qing Zhang, Carol J. Farhangfar, Daniel E. Haggstrom, Stephen Fairclough, Oliver A. Zill, Daniel R. Carrizosa, Richard B. Lanman, Edward S. Kim. Development of EGFR C797S mutation in serial liquid biopsy assessments in the clinical practice setting. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr C80.

107 Background: Pancreatobiliary carcinomas (PC) carry a poor prognosis but have not yet benefitted from the revolution in precision oncology, in part because biopsy tissue is often inadequate for ...