Integrative common and rare variant analyses provide insights into the genetic architecture of liver cirrhosis

We report a multi-ancestry genome-wide association study on liver cirrhosis and its associated endophenotypes, alanine aminotransferase (ALT) and γ-glutamyl transferase. Using data from 12 cohorts, including 18,265 cases with cirrhosis, 1,782,047 controls, up to 1 million individuals with liver function tests and a validation cohort of 21,689 cases and 617,729 controls, we identify and validate 14 risk associations for cirrhosis. Many variants are located near genes involved in hepatic lipid metabolism. One of these, PNPLA3 p.Ile148Met, interacts with alcohol intake, obesity and diabetes on the risk of cirrhosis and hepatocellular carcinoma (HCC). We develop a polygenic risk score that associates with the progression from cirrhosis to HCC. By focusing on prioritized genes from common variant analyses, we find that rare coding variants in GPAM associate with lower ALT, supporting GPAM as a potential target for therapeutic inhibition. In conclusion, this study provides insights into the genetic underpinnings of cirrhosis.

Health State Utilities Associated with False-Positive Cancer Screening Results

INTRODUCTION

Early cancer detection can significantly improve patient outcomes and reduce mortality rates. Novel cancer screening approaches, including multi-cancer early detection tests, have been developed. Cost-utility analyses will be needed to examine their value, and these models require health state utilities. The purpose of this study was to estimate the disutility (i.e., decrease in health state utility) associated with false-positive cancer screening results.

METHODS

In composite time trade-off interviews using a 1-year time horizon, UK general population participants valued 10 health state vignettes describing cancer screening with true-negative or false-positive results. Each false-positive vignette described a common diagnostic pathway following a false-positive result suggesting lung, colorectal, breast, or pancreatic cancer. Every pathway ended with a negative result (no cancer detected). The disutility of each false positive was calculated as the difference between the true-negative and each false-positive health state, and because of the 1-year time horizon, each disutility can be interpreted as a quality-adjusted life-year decrement associated with each type of false-positive experience.

RESULTS

A total of 203 participants completed interviews (49.8% male; mean age = 42.0 years). The mean (SD) utility for the health state describing a true-negative result was 0.958 (0.065). Utilities for false-positive health states ranged from 0.847 (0.145) to 0.932 (0.059). Disutilities for false positives ranged from - 0.031 to - 0.111 (- 0.041 to - 0.111 for lung cancer; - 0.079 for colorectal cancer; - 0.031 to - 0.067 for breast cancer; - 0.048 to - 0.088 for pancreatic cancer).

CONCLUSION

All false-positive results were associated with a disutility. Greater disutility was associated with more invasive follow-up diagnostic procedures, longer duration of uncertainty regarding the eventual diagnosis, and perceived severity of the suspected cancer type. Utility values estimated in this study would be useful for economic modeling examining the value of cancer screening procedures.

Multidimensional biomarker predicts disease control in response to immunotherapy in recurrent or metastatic head and neck squamous-cell carcinoma

PURPOSE

Anti-PD-1 therapy provides clinical benefit in 40-50% of patients with relapsed and/or metastatic head and neck squamous cell carcinoma (RM-HNSCC). Selection of anti- PD-1 therapy is typically based on patient PD-L1 immunohistochemistry (IHC) which has low specificity for predicting disease control. Therefore, there is a critical need for a clinical biomarker that will predict clinical benefit to anti-PD-1 treatment with high specificity.

METHODS

Clinical treatment and outcomes data for 103 RM-HNSCC patients were paired with RNA-sequencing data from formalin-fixed patient samples. Using logistic regression methods, we developed a novel biomarker classifier based on expression patterns in the tumor immune microenvironment to predict disease control with monotherapy PD-1 inhibitors (pembrolizumab and nivolumab). The performance of the biomarker was internally validated using out-of-bag methods.

RESULTS

The biomarker significantly predicted disease control (65% in predicted non-progressors vs. 17% in predicted progressors, p < 0.001) and was significantly correlated with overall survival (OS; p = 0.004). In addition, the biomarker outperformed PD-L1 IHC across numerous metrics including sensitivity (0.79 vs 0.64, respectively; p = 0.005) and specificity (0.70 vs 0.61, respectively; p = 0.009).

CONCLUSION

This novel assay uses tumor immune microenvironment expression data to predict disease control and OS with high sensitivity and specificity in patients with RM-HNSCC treated with anti-PD-1 monotherapy.

Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study

BACKGROUND

Multicancer early detection (MCED) blood tests can detect a cancer signal from circulating cell-free DNA (cfDNA). PATHFINDER was a prospective cohort study investigating the feasibility of MCED testing for cancer screening.

METHODS

In this prospective cohort study done in oncology and primary care outpatient clinics at seven US health networks, a convenience sample of adults aged 50 years or older without signs or symptoms of cancer consented to MCED testing. We collected blood, analysed cfDNA, and returned results to participants' doctors. If a methylation signature indicative of cancer was detected, predicted cancer signal origin(s) informed diagnostic assessment. The primary outcome was time to, and extent of, diagnostic testing required to confirm the presence or absence of cancer. This trial is registered at ClinicalTrials.gov, NCT04241796, and is completed.

FINDINGS

Between Dec 12, 2019, and Dec 4, 2020, we recruited 6662 participants. 4204 (63·5%) of 6621 participants with analysable results were women, 2417 (36·5%) were men, and 6071 (91·7%) were White. A cancer signal was detected in 92 (1·4%) of 6621 participants with analysable results. 35 (38%) participants were diagnosed with cancer (true positives) and 57 (62%) had no cancer diagnosis (false positives). Excluding two participants whose diagnostic assessments began before MCED test results were reported, median time to diagnostic resolution was 79 days (IQR 37-219): 57 days (33-143) in true-positive and 162 days (44-248) in false-positive participants. Most participants had both laboratory tests (26 [79%] of 33 with true-positive results and 50 [88%] of 57 with false-positive results) and imaging (30 [91%] of 33 with true-positive results and 53 [93%] of 57 with false-positive results). Fewer procedures were done in participants with false-positive results (17 [30%] of 57) than true-positive results (27 [82%] of 33) and few had surgery (one with a false-positive result and three with a true-positive result).

INTERPRETATION

This study supports the feasibility of MCED screening for cancer and underscores the need for further research investigating the test's clinical utility.

FUNDING

GRAIL.

Complex effects of sequence variants on lipid levels and coronary artery disease

Many sequence variants have additive effects on blood lipid levels and, through that, on the risk of coronary artery disease (CAD). We show that variants also have non-additive effects and interact to affect lipid levels as well as affecting variance and correlations. Variance and correlation effects are often signatures of epistasis or gene-environmental interactions. These complex effects can translate into CAD risk. For example, Trp154Ter in FUT2 protects against CAD among subjects with the A1 blood group, whereas it associates with greater risk of CAD in others. His48Arg in ADH1B interacts with alcohol consumption to affect lipid levels and CAD. The effect of variants in TM6SF2 on blood lipids is greatest among those who never eat oily fish but absent from those who often do. This work demonstrates that variants that affect variance of quantitative traits can allow for the discovery of epistasis and interactions of variants with the environment.

Genome-wide meta-analysis identifies 93 risk loci and enables risk prediction equivalent to monogenic forms of venous thromboembolism

We report a genome-wide association study of venous thromboembolism (VTE) incorporating 81,190 cases and 1,419,671 controls sampled from six cohorts. We identify 93 risk loci, of which 62 are previously unreported. Many of the identified risk loci are at genes encoding proteins with functions converging on the coagulation cascade or platelet function. A VTE polygenic risk score (PRS) enabled effective identification of both high- and low-risk individuals. Individuals within the top 0.1% of PRS distribution had a VTE risk similar to homozygous or compound heterozygous carriers of the variants G20210A (c.*97 G > A) in F2 and p.R534Q in F5. We also document that F2 and F5 mutation carriers in the bottom 10% of the PRS distribution had a risk similar to that of the general population. We further show that PRS improved individual risk prediction beyond that of genetic and clinical risk factors. We investigated the extent to which venous and arterial thrombosis share clinical risk factors using Mendelian randomization, finding that some risk factors for arterial thrombosis were directionally concordant with VTE risk (for example, body mass index and smoking) whereas others were discordant (for example, systolic blood pressure and triglyceride levels).

Cholesterol not particle concentration mediates the atherogenic risk conferred by apolipoprotein B particles: a Mendelian randomization analysis

BACKGROUND AND AIMS

The causal contribution of apolipoprotein B (apoB) particles to coronary artery disease (CAD) is established. We examined whether this atherogenic contribution is better reflected by non-high-density lipoprotein cholesterol (non-HDL-C) or apoB particle concentration.

METHOD AND RESULTS

We performed Mendelian randomization (MR) analysis using 235 variants as genetic instruments; testing the relationship between their effects on the exposures, non-HDL-C and apoB, and on the outcome CAD using weighted regression. Variant effect estimates on the exposures came from the UK Biobank (N = 376 336) and on the outcome from a meta-analysis of five CAD datasets (187 451 cases and 793 315 controls). Subsequently, we carried out sensitivity and replication analyses.In univariate MR analysis, both exposures associated with CAD (βnon-HDL-C = 0.40, P = 2.8 × 10-48 and βapoB = 0.38, P = 1.3 × 10-44). Adding effects on non-HDL-C into a model that already included those on apoB significantly improved the genetically predicted CAD effects (P = 3.9 × 10-5), while adding apoB into the model including non-HDL-C did not (P = 0.69). Thirty-five per cent (82/235) of the variants used as genetic instruments had discordant effects on the exposures, associating with non-HDL-C/apoB ratio at P < 2.1 × 10-4 (0.05/235). Fifty-one variants associated at genome-wide significance.

CONCLUSION

Many sequence variants have discordant effects on non-HDL-C and apoB. These variants allowed us to show that the causal mechanism underlying the relationship between apolipoprotein B particles and CAD is more associated with non-HDL-C than apoB particle concentration.

Multiomics study of nonalcoholic fatty liver disease

Nonalcoholic fatty liver (NAFL) and its sequelae are growing health problems. We performed a genome-wide association study of NAFL, cirrhosis and hepatocellular carcinoma, and integrated the findings with expression and proteomic data. For NAFL, we utilized 9,491 clinical cases and proton density fat fraction extracted from 36,116 liver magnetic resonance images. We identified 18 sequence variants associated with NAFL and 4 with cirrhosis, and found rare, protective, predicted loss-of-function variants in MTARC1 and GPAM, underscoring them as potential drug targets. We leveraged messenger RNA expression, splicing and predicted coding effects to identify 16 putative causal genes, of which many are implicated in lipid metabolism. We analyzed levels of 4,907 plasma proteins in 35,559 Icelanders and 1,459 proteins in 47,151 UK Biobank participants, identifying multiple proteins involved in disease pathogenesis. We show that proteomics can discriminate between NAFL and cirrhosis. The present study provides insights into the development of noninvasive evaluation of NAFL and new therapeutic options.

Interim results of PATHFINDER, a clinical use study using a methylation-based multi-cancer early detection test

3010

Background: PATHFINDER (NCT04241796) is an interventional, prospective study evaluating implementation of a blood-based multi-cancer early detection (MCED) test that uses targeted methylation-based cfDNA analysis to detect multiple cancer types and simultaneously predict cancer signal origin (CSO). We present a prespecified interim analysis of PATHFINDER evaluating an MCED test in a clinical setting. Methods: Participants (pts; ≥50y) were enrolled into 2 risk cohorts: non-elevated and elevated (smoking history, prior cancer [ > 3y post treatment], or genetic predisposition). MCED test results (cancer signal detected/not detected) were returned to investigators; pts with a signal detected also received a CSO prediction and underwent further diagnostic testing by their medical team. The primary objective was to assess the extent of diagnostic testing needed to achieve diagnostic resolution (eg, time to resolution, number/type of tests). Secondary endpoints included positive predictive value (PPV) and a measure of test satisfaction (following diagnostic resolution [signal detected] and post test [signal not detected]). Results: PATHFINDER consented 6796 pts before closing accrual on 12/4/20; as of October 6, 2020, 4086 consented, 4047 enrolled, and 4033 analyzable pts were included in the interim analysis (62.4% female, 92.1% white). Two study-related adverse events (anxiety of mild severity) were reported. Cancer signal was detected in 1.5% (62/4033) of pts; 40/62 reached diagnostic resolution to date. Kaplan-Meier estimate of median time to resolution was 78 (95% CI, 54-151) days. Among 40 pts that reached diagnostic resolution, ≥1 imaging test was performed in 93% (37/40); ≥1 invasive procedure was performed in 72% (13/18) versus 18% (4/22) of pts with diagnostic resolution of cancer versus no cancer, respectively. Based on results to date, PPV was 45% (95% CI, 30.7-60.2%; 18/40). Of 18 cancer diagnoses, 11 were solid tumors (3 stage IV, 6 stages I-III, 1 metastatic recurrence, 1 missing stage), and 7 were hematologic malignancies (1 stage IV, 4 stages I-III, 2 without AJCC stage). Accuracy of the top CSO prediction in true positives was 82.4% (95% CI, 59.0-93.8%; 14/17). Most pts were satisfied with the test (43.7% extremely satisfied, 30.7% very satisfied, 14.6% satisfied). Signal detection rate and test satisfaction were similar in the 2 risk cohorts; PPV tended to be higher in the elevated risk cohort, as expected. Conclusions: An interim analysis of this return of results study demonstrated promising MCED test results. Of 40 pts achieving diagnostic resolution, nearly half had a diagnostic workup confirming cancer; CSO was predicted with high accuracy for detected cancers. Taken together with the rarity of adverse events and high test satisfaction, these results support the feasibility of clinical implementation. Full enrollment cohort data will be available at the meeting. Clinical trial information: NCT04241796.

A prespecified interim analysis of the PATHFINDER study: Performance of a multicancer early detection test in support of clinical implementation

3070

Background: A multi-cancer early detection (MCED) test that uses targeted methylation-based cfDNA technology to detect cancer and predict cancer signal origin (CSO) has potential to efficiently identify malignancies for which effective screening modalities do not exist. A previous version of a blood-based MCED test demonstrated favorable classification and test characteristics. Samples from the ongoing PATHFINDER study were reanalyzed in a prespecified interim analysis to evaluate performance of a more recent version of the test with an updated classifier (eg, updated CSO localization, hematological signal threshold) that is planned for clinical implementation as a general multi-cancer screening tool. Methods: PATHFINDER (NCT04241796) is an interventional, prospective study in which results (cancer signal detected/not detected and predicted CSO) using a previous version of the MCED test are returned to investigators, and those with a signal detected undergo further diagnostic testing. In this prespecified interim analysis, samples from those enrolled as of October 6, 2020 were reanalyzed with the more recent version of the MCED test (these results were not returned to investigators). The positive predictive value (PPV) for cancer detection, overall CSO accuracy, and concordance between the two test versions were assessed. Results: A total of 4011/4047 (99%) participants (pts) were analyzable (mean [SD] age 63.9 [8.7] years, 62% female, 92% white, 24% with prior cancer history, 39% ever smoker [4% current], 6% with genetic cancer predisposition). Cancer signal was detected in 0.95% (38/4011). A total of 27/38 also had signal detected by the previous version of the MCED test, including 19 who reached diagnostic resolution (13 with cancer diagnosis and 6 without); 11/38 were discordant positives. Nine different cancer types were detected in the 13 pts (2 stage I, 3 stage II, 2 stage III, and 3 stage IV); 1 had no AJCC stage expected, 1 metastatic recurrence and 1 stage evaluation underway. A conservative minimal PPV assuming all discordant positives are false positives, was 43.3% (13/30, 95% CI 27.4-60.8%) based on 19 pts with diagnostic resolution and 11 discordant positives. High negative percent agreement (PA) 99.7% (99.5-99.8%) between the two test versions was observed. Positive PA of 43.5% (95% CI, 31.9-55.9%) was consistent with the more stringent threshold for hematologic signal in the recent MCED version, as most discrepant cases had hematologic CSO with the previous MCED test. Among 13 detected cancers, accuracy of the top CSO prediction was 92.3% (12/13, 95% CI 66.7-99.6%). Conclusions: In this prespecified interim analysis, the more recent version of the MCED test detected cancers with high PPV and high accuracy of CSO prediction, supporting readiness for use in clinical practice. Full enrollment cohort data will be available at the meeting. Clinical trial information: NCT04241796.

Abstract CT291: The PATHFINDER Study: Assessment of the implementation of an investigational multi-cancer early detection test into clinical practice

Background: Early detection decreases patient mortality for many cancers, but prevailing paradigms focus on distinct screening modalities for each cancer type. A minimally invasive test that detects multiple tumor types could substantially reduce cancer mortality and improve screening efficiency. In preliminary studies, we showed that a targeted methylation cell-free DNA (cfDNA) test can detect cancer and determine the tissue of origin (TOO) of abnormally methylated cfDNA with high sensitivity and specificity; however, blood was collected at the time of diagnosis and results were not disclosed to inform care. The current study includes disclosure of results to the study investigator to understand the care pathways prompted by a “signal detected” test result, and the benefits, harms, and burdens associated with a multi-cancer detection test.

Methods: PATHFINDER (NCT04241796) is a prospective, longitudinal, multi-center clinical study. Approximately 6,200 participants with varying levels of cancer risk will be enrolled at 5-10 clinical sites in the United States. Participants must be ≥50 years of age. If participants have at least 1 of the following–a smoking history of ≥100 cigarettes, a genetic cancer predisposition, or a history of invasive or hematologic malignancy with definitive treatment completed &gt;3 years prior to enrollment–they are eligible for the Elevated Risk Cohort (n = 4340); otherwise, they are eligible for the Non-Elevated Risk Cohort (n = 1860). Individuals being evaluated for cancer or who have a history of invasive or hematologic malignancy within 3 years of enrollment will be excluded. Blood drawn at study sites will be analyzed at GRAIL, Inc. (Menlo Park, CA). Participants and study investigators will be informed of the test results. Participants with “signal not detected” results will be advised to continue routine clinical care, including age- and risk-based cancer screening. Those with “signal detected” test results will be informed of the predicted origin and undergo diagnostic evaluation. The primary study objective is to determine the extent of diagnostic testing required to achieve diagnostic resolution following a “signal detected” test result, defined as the date when the study investigator determines to end diagnostic evaluation triggered by a “signal detected” test result. Health resource utilization, the number and types of tests and time required to reach this point, is the primary endpoint. Secondary objectives include evaluation of test performance (specificity, positive predictive value, and TOO accuracy), and assessment of participants' perceptions about the test and changes in anxiety and health-related quality of life. Exploratory aims include assessment of the number and types of invasive cancer diagnoses among participants with “signal not detected” test results and changes in attitude toward adherence to guideline-recommended screening.

Citation Format: Lincoln Nadauld, Charles H. McDonnell III, Minetta C. Liu, Eric Klein, Tomasz M. Beer, Deborah Schrag, Andrew G. Hudnut, Richard Whittington, Bruce Taylor, Joanna Tierney, Catherine Marinac, Jafi Lipson, Margarita Lopatin, Karen C. Chung, Eric Fung, Anne-Renee Hartman, Geoffrey R. Oxnard. The PATHFINDER Study: Assessment of the implementation of an investigational multi-cancer early detection test into clinical practice [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 CT291.

Tissue/Site-Agnostic Study of Ribociclib for Tumors With Cyclin D-CDK4/6 Pathway Genomic Alterations: A Phase II, Open-Label, Single-Arm Basket Study

PURPOSE

As part of the Novartis Signature Program, this study evaluated the efficacy of ribociclib (selective cyclin-dependent kinase 4/6 [CDK4/6] inhibitor) in patients with cyclin D-CDK4/6 pathway-aberrant tumors.

METHODS

This was a phase II, single-arm, signal-seeking study in patients with advanced malignancies that had progressed on or after standard treatment. Prior identification of tumor CDK4/6 mutation or amplification, CCND1/3 amplification, or CDKN2A mutation or loss was required. Clinical benefit (defined as the proportion of patients with response or stable disease at ≥ 16 weeks) was the primary end point.

RESULTS

From 61 centers in the United States, 106 patients (median age, 62.5 years) were enrolled across multiple malignancies. The patient population was heavily pretreated (median number of prior therapies, three; range, 0 to 19). Median progression-free survival was 1.8 months (95% CI, 1.8 to 1.9). In patients with solid tumors, the clinical benefit rate was 18.1% (n = 19 of 105) and the overall response rate was 2.9% (n = 3 of 105); three partial responses occurred in patients with adenocarcinoma (unknown primary), soft tissue sarcoma, and urothelial carcinoma. No tumor cohort met the prespecified criteria for success. The most common adverse events suspected to be related to treatment were neutropenia (30.2%; decreased neutrophils, 15.1%), fatigue (31.1%), and nausea (29.2%). Fatigue and nausea were typically mild. Only one incident of febrile neutropenia was experienced (grade 3).

CONCLUSION

No new or unexpected safety signals were observed in this heavily pretreated patient population. Although responses were seen in tumors with CCND1-CDK4/6 amplifications, the primary end point was not met, suggesting additional evaluation of ribociclib, possibly as combination therapy, is needed.

Abstract 2278: A high throughput method for the optimization of digital PCR assays for personalized circulating tumor DNA detection

Single color digital PCR (sc-dPCR) is a robust approach for the quantitation of low allelic fraction mutations in clinical oncology samples. More recently this technology has been employed to identify mutations from circulating tumor DNA (ctDNA) that has been extracted from the blood samples of cancer patients. The use of digital PCR has great potential for non-invasive longitudinal monitoring via liquid biopsies. However, this application requires low input DNA volumes and relies on a single nucleotide variant (SNV) to distinguish between normal and ctDNA, necessitating that sc-dPCR primer binding is both highly efficient and specific. These stringent requirements make assay optimization a tedious process that greatly limits the rate at which personalized detection panels can be generated. We have developed a high throughput method to optimize sc-dPCR assays utilizing Next Generation Sequencing (NGS) technology to assess amplification more quickly and with more flexibility than traditional gel based analysis.

Using our assay optimization approach, a segment of each gene containing a tumor specific SNV was incorporated into the genome of Saccharomyces cerevisiae. These renewable positive control colonies were cultured in a 96 well plate format and pooled to mimic the low allelic frequency conditions of ctDNA. The presence of each tumor specific SNV was confirmed by preparing and sequencing a library containing the unique barcode region of each colony. Using bulk PCR, up to 96 primer sets were tested at one annealing temperature in a singleplex format. Alternatively, we multiplexed up to 11 primers in each well, greatly increasing the number of assays that can be developed per plate. Using this multiplexed format, we introduced a thermal gradient across the plate to identify the optimal annealing temperature of each primer set in a single run. A parallel experiment with identical PCR conditions was run using NA18507 human DNA to act as a negative control for primer specificity.

All amplicons in each PCR condition were uniquely indexed and sequenced using an NGS platform. Using a ratio of the number of reads associated with on target and non-mutation specific amplicon sequences for each primer set, the success of each assay was determined. This method was also used to identify specific mismatches incorporated in the primer sequence that increased binding specificity. Using a sequencing based analysis method, we have observed that sc-dPCR assays can be optimized rapidly across multiple mutations, making them more accessible for personalized monitoring.

Citation Format: Maya M. Arce, Christina Wood-Bouwens, Derrick Haslem, Billy T. Lau, John Bell, Alison Almeda, Matt Kubit, Bryce Moulton, Robin Romero, Robert P. St. Onge, Lincoln Nadauld, Hanlee P. Ji. A high throughput method for the optimization of digital PCR assays for personalized circulating tumor DNA detection [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 2278.

Utilizing data from patient-derived xenograft mouse models of human tumors to inform clinical decision making in Molecular Tumor Boards (MTB) deliberations

e14660

Background: Molecular Tumor Boards (MTB) are often the critical decision-making step in identifying genome-guided treatments for patients with difficult-to-treat cancers, e.g. BRAF mutated metastatic colon cancers. A common challenge for MTBs is prioritizing between two or more actionable variants in a tumor. A potential solution to this challenge is to incorporate drug response in Patient-derived xenografts (PDX) models into MTB deliberations. The goal of this study was to evaluate the feasibility of using PDX models to elucidate drug-effectiveness in BRAF-mutated cancer as an example of a common clinical scenario. Methods: We selected BRAF-mutated PDX models from the JAX PDX resource based on the presence of an activating BRAF mutation and a second actionable variant from the JAX Cancer Treatment Profile (CTP). Somatic mutation data from PDX tumors were presented to members of the Intermountain MTB. PDX models were then treated with drugs recommended by the MTB; outcomes based on tumor growth inhibition (TGI) were shared with the MTB. Results: Gene/variant targets, associated drugs for the 2nd mutation and responses are described in the table. The MTB members determined that PDX data presented in TGI format is helpful in MTB deliberations. Activity of BRAF-targeted therapy was expected while the low activity of olaparib in the BRCA1-mutated colon cancer model was unexpected. The MTB then discussed molecular mechanisms that contributed to these outcomes. Conclusions: The pilot study demonstrated that utilizing PDX drug response data as an additional molecular annotation for MTB deliberations is feasible. Future studies will further optimize this process. [Table: see text]

Characterization of the tumor mutation burden in hepatobiliary tumors

295

Background: Hepatobiliary tumors are aggressive tumors with emerging evidence for increasing sensitivity to immune checkpoint inhibitors (ICI). Tumor mutation burden (TMB) was found to be a quantitative biomarker associated with production of neoantigens within the tumor and predict the sensitivity to immune therapy. Herein, we explore the TMB as a potential biomarker of response to immune therapy in hepatobiliary tumors. Methods: We retrospectively assessed all patients with hepatobiliary malignancies who have undergone next generation sequencing (NGS) between January 2013 and September 2018. We then analyzed the tumor mutation burden of these tumors and also identified frequency of patients with no clinically actionable mutations. Results: Of the 65 total patients with hepatobiliary tumors, 49 patients (75%) had at least one clinically actionable mutation while 16 patients (25%) had no clinically actionable mutations. Among 65 patients, 44 patients had hepatocellular carcinoma, 15 patients had cholangiocarcinoma and 6 patients had gallbladder carcinoma. The TMB data is available for 15 patients. The mean TMB reported was 2.7 (1.16 – 4.25), which suggests low mutation burden in general in all our HB tumors. Among the patients with available TMB, the underlying risk factor was noted as hepatitis C in 3, NASH in 1, others in 6, unknown in 5 patients. Conclusions: Our data suggests the TMB in hepatobiliary tumors is low in general irrespective of their underlying risk factors. Future larger studies are needed to evaluate TMB as a potential biomarker in hepatobiliary tumors to help select patients that will benefit from immune therapy.

Abstract 219: Development of TEAPOT algorithm to reconstruct individual ovarian tumors' evolutionary history based upon bulk and single cell whole exome sequencing data

Background. Mutation detection through genetic testing is playing an increasingly important role in personalized precision medicine in cancer. However, current tests identifying driver mutations as therapeutic targets are based on detection of common mutations in cancer genes. These tests are not patient specific and do not address intra-tumor heterogeneity. Ubiquitous intra-tumor genetic heterogeneity is a mechanism of drug resistance and cancer recurrence.

Methods. Approximately 16-24 microsamples are acquired to represent the entire cancer cell population for every ovarian tumor. Each microsample consists of a few cells within a clone and is selected to substitute for a single cell and overcome the large allele dropout rate commonly seen in single genome amplification and sequencing. TEAPOT (Tumor Evolution Assay for Personalized Oncology Therapy) algorithm has been developed to reconstruct a tumor's evolutionary history through integration of whole exome sequencing data from the bulk primary tumor and 16-24 microsamples taken from the bulk tumor. The evolutionary history for an individual tumor is expressed as a rooted and binary tumor developmental tree representing the mitotic process starting from an ancestral cancer cell. Individual mutations are assigned to the cells where they originally occur. The offspring size carrying a mutation was estimated based on tumor purity, variant allele frequency and the variant's copy number.

Results. TEAPOT algorithm builds a tumor's evolutionary history with the following features: 1) a tumor's evolutionary history is unique for each ovarian cancer patient; 2) the size of a tree is proportional to the number of microsamples selected; 3) 16-24 microsamples builds a tree with 5 or more generations; 4) TEAPOT detects a driver mutation's occurrence at a specific developmental stage such as 1-cell, 2-cell, 4-cell, etc; 5) The size of offspring carrying a mutation thus the intra-tumor prevalence of the mutation can be estimated; 6) multiple driver mutations can be located separately in different clones. Therefore, TEAPOT provides a quantitative description of intra-tumor genetic heterogeneity and identifies sub-clonal driver mutations in a tumor.

Conclusion. TEAPOT reconstructs a tumor's developmental process thus providing a patient-specific evolutionary history. Quantitation of intra-tumor prevalence of driver mutations may inform selection of an effective targeted agent and may provide rationale for cocktail treatment targeting multiple driver mutations simultaneously. TEAPOT can be also used for other solid and liquid cancers. A driver mutation's role in a patient may be functionally defined and quantitated based upon the growth advantage (fitness) it confers on its host cells in the reconstructed tumor evolutionary history.

Citation Format: Jianshu Zhang, Helaman Escobar, Harshmi Shah, Mickey Miller, Yang Wei, Kristen Schneider, Michelle Knirr, Kenny Day, Christopher Johnson, Baoli Yang, Eric Devor, Kristina Thiel, Lincoln Nadauld, Kimberly Leslie, Donghai Dai. Development of TEAPOT algorithm to reconstruct individual ovarian tumors' evolutionary history based upon bulk and single cell whole exome sequencing data [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 219.

Abstract 438: High-quality CNV segments from low-coverage whole genome sequencing from FFPE cancer biopsies based on an evaluation of multiple CNV tools

Changes in DNA copy number, i.e., somatic CNVs, are common genetic aberrations in cancers. The effects of CNV include alteration in gene dosage across large segments of the cancer genome affecting the expression of cancer driver genes by amplifications, or cancer suppressor genes by deletions. In addition, CNVs are markers of underlying rearrangements within or between chromosomes and there is increasing evidence supporting a greater role for CNVs in developing and maintaining neoplastic cell population diversity.

Copy number aberrations can be estimated from next generation sequencing data, with high sensitivity and genomic resolution by sequencing the whole genome (WGS). For this study, we demonstrated that high quality CNV calls can be extracted in a fast and cost-effective way from low-coverage whole genome sequencing. Novaseq S2 flowcells (Illumina Inc) enables to obtain an average coverage of 3-4x per sample after pooling up to 96 samples per flowcell.

We examined three different copy number detection tools (CNVkit, BicSeq, and seqCBS) from paired tumor and normal WGS using microarray data as a reference. Pearson correlations were computed between the reference and CNVs from the WGS in two fashion; i) segment based and ii) gene based. The segment based comparison used sliding window of 100 K bp while gene based comparison used segments at the gene level.

We found high correlations between microarray and WGS segments. The highest correlations were obtained by CNVkit, ranging from 0.964 to 0.985 (SD: 0.973 - 0.007) and BicSeq, ranging from 0.963 to 0.986 (SD: 0.975 - 0.008). These results open the prospect of assessing large cancer cohorts of hundreds of samples at a reasonable cost. We are planning to apply this method to a large cohort of Stage III colon cancer patients and determine the clinical relevance of CNVs for survival.

Citation Format: HoJoon Lee, Li Charlie Xia, Stephanie Greer, John Bell, Sue M. Grimes, Christina Wood Bouwens, Giwon Shin, Billy TC Lau, Lucas Johnson, Noemi Andor, Kenneth Day, Mickey Miller, Helaman Escobar, Lincoln Nadauld, Hanlee P. Ji, Paul Van Hummelen. High-quality CNV segments from low-coverage whole genome sequencing from FFPE cancer biopsies based on an evaluation of multiple CNV tools [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 438.

Abstract 4331: High-throughput whole-genome sequencing of formalin fixed, paraffin-embedded tissues from colorectal cancer patients

Formalin-fixed, paraffin-embedded (FFPE) storage is a universally-adopted, cost-effective and long-term solution for tissue storage. Given the abundance of FFPE clinical samples, they provide a good source of tissue for genome sequencing studies. However, the extraction of nucleic acids from FFPE tissues yield highly fragmented DNA of variable quality. The poor quality and low yield make it difficult to generate adequate sequencing libraries for genomic studies.

For a population cancer genome study, we optimized a procedure for whole-genome sequencing on FFPE tumor and normal tissue of colorectal cancer patients. This pilot study aimed to first identify changes in DNA copy number using low-pass sequencing, with the additional goal of later adding structural variant and gene fusion detection and single nucleotide variant analysis with deeper sequencing. We processed, sequenced and analyzed 193 patients. These FFPE samples had varying storage characteristics. We optimized the initial extraction quality thresholds, DNA extraction and library preparation to generate high quality sequencing data.

We identified an initial sample quality and library threshold to hold the best correlation for sequencing quality. Optimal yield and DNA size from extracted FFPE was achieved by using Promega's LEV DNA FFPE Extraction Kit. Library prep optimization involved using mechanical shearing of DNA to increase overall library size for structural variant detection as well as using unique dual index adapters to identify index swapping artifacts during sequencing. An additional cleanup was also performed post-fragmentation to increase overall library size. Overall, we successfully developed a robust protocol for conducting whole genome sequencing on FFPE tumor samples.

Citation Format: Matthew Kubit, Christina Wood-Bouwens, Sue Grimes, John Bell, GiWon Shin, Billy Lau, Mickey Miller, Kenneth Day, Helaman Escobar, Hanlee Ji, Lincoln Nadauld, Paul Van Hummelen. High-throughput whole-genome sequencing of formalin fixed, paraffin-embedded tissues from colorectal cancer patients [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 4331.

Association of the neutrophil-to-lymphocyte ratio prior to checkpoint blockade immunotherapy (CBI) or radiation plus CBI with overall survival in melanoma patients

200

Background: Studies in certain cancer types have indicated that radiation therapy prior to the use of CBI provides a survival benefit. This benefit has not been clearly defined for patients with metastatic melanoma. Additionally, the neutrophil-to-lymphocyte ratio (NLR) may be a potential biomarker. Methods: This retrospective study was performed in patients diagnosed with melanoma between January 2007 and August 2016 who received CBI with or without previous radiation treatment at Intermountain Healthcare (Utah, USA). Cases were identified from electronic medical records and data was manually extracted through August of 2017. The neutrophil-to-lymphocyte ratio (NLR) was calculated from the absolute neutrophil and lymphocyte counts of a complete blood cell count with differentials performed as a routine standard of care procedure in melanoma patients prior to therapy initiation. Overall survival was defined as the length of time (d) from start of CBI to death as of August, 2017. Results: Forty-six melanoma patients were initially identified. Of these, thirteen patients were excluded due to lack of follow-up data (n = 9), radiation performed after CBI (n = 3), or concurrent radiation and CBI (n = 1). The final analysis consisted of 33 subjects separated below (NLR < 3.12, n = 16) and above (NLR ≥ 3.12, n = 17) the NLR median. Age, height, body mass, and body mass index were not significantly different between groups (p-range: 0.11-0.60). Results from the Kaplan-Meier curve indicate that a NLR above the median associates with lower overall survival (Mantel, p = 0.04) in melanoma patients receiving CBI with or without previous radiation treatment. In a separate analysis of this cohort, overall survival was not significantly influenced by radiation therapy prior to CBI. Conclusions: Although prior radiation therapy offered no survival advantage for patients receiving CBI, NLR less than 3.12 was associated with an increase in overall survival. Further studies are need to explore NLR as a biomarker.

DNA damage repair (DDR) pathway defects in gastrointestinal (GI) malignancies

647

Background: The clinical significance of the genomic alterations associated with DDR pathway in GI tumors (besides MMR defects) is largely unknown. These patients can potentially derive benefit from targeted therapy with poly ADP ribose polymerase (PARP) inhibitors, which have already shown promising activity in ovarian, breast and prostate cancers. In this study, we investigated the frequency and clinical significance of DDR repair defects (other than MMR defects) in GI tumors. Methods: We performed a retrospective analysis of all patients who had tumor next generation sequencing performed between January 2013 and August 2017 on GI cancers harboring DDR pathway defects. Data including demographics, clinical history, and treatment were extracted from patients' records. Results: Of 299 patients with GI tumors sequenced, 35 cases (12%) were noted to have DDR defects. The most commonly mutated genes – 6 (17%) BRCA2, 5 (14%) PALB2, 4 (11%) ATM, 3 (8.6%) BRCA1, 2 (5.7%) each of NBN, MUTYH, ERCC3, PARP1 amplification and 1 (2.8%) each of ERCC2, CDK12, and PARP2 amplification. Two patients had both ATM and BRCA2 mutations. Combination of ATM and MRE11, ATM and BRCA1, BRCA1 and ERCC6, BRCA2 and CDK12 were noted in 1 patient each. Of the 23 patients with available clinical data, the median age at diagnosis was 65 (range 30–85) years with male and female prevalence rates of 60.8% and 39.2%, respectively. Stage at diagnosis was I (n = 3), II (n = 3), III (n = 8), and IV (n = 9). The primary site of tumor was found in 8 (34.8%) colon, 4 (17.4%) liver, 4 (17.4%) pancreas, 2 (8.7%) esophagus, 2 (8.7%) anus, 2 (8.7%) appendix, 1 (4.3%) rectum. Seventeen patients received platinum‐based therapy, 7 were treated with PARP inhibitors and 5 patients received both platinum and PARP inhibitor. Median overall survival from diagnosis for patients with stage I/II was 65.3 months, stage III was 23.1 months, and stage IV was 22.4 months. The median survival of patients treated with olaparib was 24.5 months. Conclusions: DDR pathway defects in GI tumors are uncommon. However, they can potentially be targeted with PARP inhibitors with durable survival. Future clinical trials are warranted to explore the role to PARP inhibitors in these unique subset of patients.

Mutational landscape of metastatic colorectal cancer: Aggregate insights from a molecular tumor board

837

Background: The mutational landscape of metastatic colorectal cancer (CRC) is being elucidated by next-generation sequencing of both primary tumors and metastatic foci. In this study, we examined the results of all stage IV CRC cases submitted to our molecular tumor board (MTB). Methods: We performed a retrospective analysis of all patients who had next-generation sequencing performed between January 2015 and August 2017 on either their primary tumor and/or metastasis. Cases were presented at a MTB convened twice monthly. For the purposes of this study only pathogenic mutations were notated, not variants of unknown significance (VUS). Results: Eighty-seven unique patients had 97 specimens sequenced (28 primary tumors and 69 metastases). The primaries averaged 3 mutations per specimen (range: 1-6) whereas the metastases averaged 4 (range: 1-14, p=.25). The most common anatomic sites of submitted metastatic tissue were the liver (n=35, average mutations=4), followed by the lungs (n=10, average mutations=3) and the omentum/peritoneum (n=8, average mutations=3). Two patients had both a primary tumor and a metastasis sequenced, with a 33% rate of concordance in inter-specimen mutations. Five patients had multiple metastases sequenced, with a 53% rate of concordance in inter-specimen mutations; in every case of longitudinal sequencing, mutational burden increased in metastases over time. The most common mutation was apc (21% of all mutations), followed by p53 (16%) and then kras (13%). Candidacy for EGFR-directed therapy was found in 8 cases, and mismatch repair defects were detected in 5 cases. Conclusions: In stage IV CRC cases sent to our MTB, tissue from metastases was more commonly submitted for analysis than from the primary tumors. There is not necessarily concordance between mutations in primary tumors and metastases, nor among multiple metastases in the same patient. Sequencing at multiple time points in the disease course may allow observation of clonal evolution and dynamic adaptation of therapeutic targets.

Utility of next-generation sequencing in clinical decision making in hepatocellular carcinoma (HCC)

295

Background: HCC is a heterogeneous disease with diverse genomic alterations. The prior genomic studies have identified common alterations in TERT, P53, WNT pathways. However, most of these alterations are not targetable with current FDA approved targeted agents. In this study, we determine the clinical impact of targeted next generation sequencing in patients with advanced HCC. Methods: We retrospectively assessed all patients with gastrointestinal (GI) malignancies who have undergone next generation sequencing (NGS) between January 2013 and August 2017. The primary endpoint is to determine the frequency of clinically actionable mutations in HCC. Secondary endpoint is to identify number of patients eligible for current FDA approved targeted agents. Results: Of the 299 consecutive GI tumors sequenced, 29 cases were identified as HCC. Clinically actionable mutations were noted in 22 of 29 patients (pts) (76%). Most pts (52%) were found to harbor more than one potentially actionable genetic alterations (n = 15). The common pathways involved: P53 in 9 pts (22%), cell-cycle regulation in 7 pts (17.1%) and mitogen-activated protein kinase in 6 pts (14.6%). Other pathways involved were DNA repair in 4 pts (9.8%), WNT in 3 pts (7.3%), MYC in 3 pts (7.3%), NOTCH in 2 pts (4.9%), HNPCC in 2 pts (4.9%), APC in 2 pts (4.9%), phosphatidylinositol 3-kinase-AKT-mTOR in 1 patient (pt) (2.4%), BRINP in 1 pt (2.4%) and angiogenesis in 1 pt (2.4%). We noted 10 pts (34.5%) harbor alterations that could potentially be targeted with FDA approved treatments such as palbociclib (for CCND1/2 amplification), PARP inhibitors (for DNA repair defects), and immunotherapy (for MMR defects) on precision medicine clinical trials such as TAPUR or MATCH. Conclusions: Mutational profiling using a targeted NGS panel identified clinically actionable alterations in nearly 75% of advanced HCC patients. Almost one third of these patients were potential candidates for current FDA approved treatments. NGS and enrollment in clinical trials should be considered in all fit patients with HCC who had progression on current standard of care treatments.

Cell cycle checkpoint defects in gastrointestinal malignancies

680

Background: Cyclin Dependent Kinases (CDKs) play a significant role in cell cycle regulation. Aberrations involving the cell cycle pathway genes can lead to uncontrolled cell proliferation and genomic instability. These could potentially be targeted with CDK4/6 inhibitors. The frequency and type of these alterations in GI tumors is largely unknown. Methods: We analyzed the frequency of abnormalities in cell cycle genes in patients with diverse GI malignancies (colorectal, liver, pancreas, gastroesophageal, anal, appendix) that underwent next generation sequencing from January 2013 to August 2017. Results: Aberrations in the cell cycle pathway were identified in 33 of 299 (11%) of cancers. The frequency of aberrations was as follows: CDKN2A/B in 10 (30.3%), CCND1 in 7 patients (pts) (21.2%), CCND2 in 2 pts (6%), CEBPA in 2 pts (6%), CDK6 in 2 pts (6%), CDK8 in 2 pts (6%) and CDK2 in 1 (3%). Alteration involving multiple genes of cell cycle noted in 7 patients (21.2%) with combination of CCND1 and CDKN2A being most common combination. The cell cycle checkpoint defects were most frequently seen in 9 pts with colon (27%), 8 pts with hepatobiliary (27%), 8 pts with pancreatic (24%), 7 pts with esophageal (21%), and less commonly in small bowel (6%) and GIST (6%). Conclusions: The alterations in the cell cycle pathway are most common in certain GI tumors mainly colon, pancreatic, hepatobiliary and esophageal tumors. Future clinical trials exploring the potential role of targeted agents such as CDK4/6 inhibitors alone or in combination with other targeted agents such as MEK inhibitors requires further exploration in these tumors.

Abstract 998: OPeN: the oncology precision network data sharing consortium

Background: The advent of Next-Generation Sequencing (NGS), and other molecular diagnostic technologies, has enabled the use of genomic information to guide targeted treatment in cancer patients. While this precision oncology approach can yield exciting clinical outcomes, the innumerable genomic variants identified in individual tumors effectively establishes each case as a unique N=1 clinical presentation. This scenario is contrary to a basic dogma of medical practice where historical cases and treatment outcomes guide future management and therapeutic decisions. Aggregation of large data sets, on a multi-institutional basis, has the potential to overcome the N=1 paradox and yield management insights in the implementation of precision oncology.

Methods: We have formed the Oncology Precision Network (OPeN), an oncology data sharing consortium, to aggregate big data sets consisting of clinical, genomic, pharmacological, and treatment response data from diverse patient cases. Data from Intermountain Healthcare, Stanford University, and Swedish Cancer Institute-Providence St. Joseph Health, as well as other institutions, comprises the database and is derived from 79 hospitals, over 800 physician clinics and more than 50,000 annual cases.

Results: The OPeN database can be interrogated by variant type, specific therapeutics, clinical outcomes, and by grouped variables, in a structured data format. The overarching IT platform is a cloud based, open source, triple store precision oncology solution, Syapse. These data are yielding valuable insights, including tumor mutational burden (TMB) scores and their correlation to immunotherapy response, clinical response in various drug-gene combinations, and therapy-specific adverse events.

Conclusions: We anticipate this resource will be used by the Molecular Tumor Boards of contributing institutions for clinical interpretation, and by treating providers to overcome the N=1 challenge associated with precision oncology.

Citation Format: Lincoln Nadauld, Derrick Haslem, Paul D. Tittel, Mariko Tameishi, Thomas Brown, James Ford. OPeN: the oncology precision network data sharing consortium [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 998. doi:10.1158/1538-7445.AM2017-998

Abstract 5359: An ultrasensitive and highly reproducible hybridization capture-based Next-Generation Sequencing clinical assay to enable precision oncology in patients with solid tumors

Sequencing whole cancer exome and whole genome provides useful information to study cancer evolution. However, cost and amount of data may become overwhelming. Enrichment capture-based methods to design custom targeted gene panel have rapidly evolved recently in cancer genomics arena. Custom targeted gene sequencing entails several advantages (i.e., better quality, ethical, affordable, technically suitable, personalized and reimbursable) for cancer patients over whole exome and whole genome sequencing. With this in mind, and to enable precision oncology in patients with solid tumors, we developed ICG100 2.0 panel, a hybridization capture-based next-generation sequencing assay for targeted sequencing of all exons and flanking introns of 162 commonly mutated cancer genes in formalin-fixed, paraffin-embedded (FFPE) tumors. To establish ICG100 2.0 panel, we compared three commercially available capture based technologies and evaluated reproducibility, sensitivity, specificity and the detection limit for low-frequency variants using internally developed bioinformatics pipeline. Cell lines, reference standards/synthetic DNA and solid tumor samples with the known genetic information was utilized in this analysis. Results were derived from MiSeq and NextSeq platforms and cross-compared on other platforms, including MassArray and ddPCR for establishing concordance and uniformity. Intra and interrun replicates were utilized to assess the quality, precision and reproducibility of variant calling. Mean depth of coverage was observed at &gt;300X with &gt;99% sensitivity and specificity. These findings and observations will guide other clinical laboratories to establish new assay that require less DNA input, enzyme based fragmentation and reduced preparation time. While we show that the three capture based methods (after comparison) had an overall accuracy in SNP and CNV detection similar to each other with minor differences, we describe an approach to assess and establish the best assay from the clinical standpoint to guide treatment decisions and match cancer patients to the most appropriate clinical trials.

Citation Format: Moises Hernandez, Sara King, Sharanya Raghunath, Christopher Johnson, David Loughmiller, Lincoln Nadauld, Pravin J. Mishra. An ultrasensitive and highly reproducible hybridization capture-based Next-Generation Sequencing clinical assay to enable precision oncology in patients with solid tumors [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 5359. doi:10.1158/1538-7445.AM2017-5359

Abstract 1436: Identification of novel tumor suppressor candidates in familial cholangiocarcinoma using sequencing-based Megabase-scale haplotypes from germline and cancer genomes

Cholangiocarcinoma (bile duct cancer) is a rare epithelial malignancy with an extremely poor prognosis characterized by a 5-year survival of less than 10%. While this tumor type can be amenable to specific combination chemotherapy regimens, it is typically diagnosed at an advanced stage and inevitably progresses towards distant metastasis. We identified a family with a predisposition to cholangiocarcinoma, having multiple affected family members, and sought to identify the underlying germline mutation(s).

We employed linked-read sequencing, a microfluidic technology where high-molecular weight DNA molecules (&gt;20 kb) are partitioned into droplets and labeled with unique barcodes, enabling computational reconstruction of the original large molecules following traditional short-read sequencing, to identify entire megabase-scale regions of the genome that segregate with the disease. We applied this technology to an entire generation of a family predisposed to cholangiocarcinoma, including a mix of 8 affected and unaffected siblings. We additionally utilized each affected individual’s cancer genome to search for a common second hit that would serve as further confirmation. Our analysis yielded a small set of four candidate pathogenic variants. Functional validation of the candidate genes in a zebrafish model, using morpholino-based gene silencing, identified one candidate whose silencing led to a reproducible phenotype consistent with known aspects of the candidate gene’s function. The candidate variant occurs in a TGF family-interacting domain, and has not been extensively characterized in other cancers. Cross-validation with TCGA datasets showed the gene was altered in 5% of cholangiocarcinomas, all of which displayed homozygous deletions. This identification of a germline allele associated with familial cholangiocarcinoma represents a unique discovery with clinical implications for the described family and warrants further investigation in other patients who display features of an inherited cholagniocarcinoma. This finding also raises the possibility of TGF pathway targeting as a therapeutic strategy in select cholangiocarcinomas, and highlights the power of whole-genome phasing to discover novel inherited tumor suppressor variants.

Citation Format: Stephanie Greer, Lincoln Nadauld, Billy Lau, Laura Miotke, Erik Hopmans, Christina M. Wood, John M. Bell, David A. Jones, Hanlee P. Ji. Identification of novel tumor suppressor candidates in familial cholangiocarcinoma using sequencing-based Megabase-scale haplotypes from germline and cancer genomes [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 1436. doi:10.1158/1538-7445.AM2017-1436

Abstract 2712: Joint somatic mutation and germline variant identification and scoring from tumor molecular profiling and ct-DNA monitoring of cancer patients by high-throughput sequencing

Cancer tumor profiling by targeted resequencing of actionable cancer genes is rapidly becoming the standard approach for selecting targeted therapies and clinical trials in refractory cancer patients. In this clinical scenario, a tumor sample is obtained from an FFPE block and sequenced by targeted next-generation sequencing (NGS) to uncover actionable somatic mutations in relevant cancer genes. Some of the challenges that arise in analyzing tumor-derived NGS data include distinguishing between somatic and germline variants in the absence of normal tissue data, recognizing pathogenic germline variants, and identifying sequencing errors (which occur at about 0.5% rate). Additional challenges arise when considering other clinical applications of NGS such as sequencing cell-free tumor DNA (cf-DNA) from plasma samples to monitor disease response or disease recurrence. Here we present a principled approach to identify both single-nucleotide and small insertion/deletion somatic mutations and germline variants from NGS data of tumor tissue that leverages the allelic fraction patterns in tumors and prior information from external databases through the use of a Bayesian Network algorithm. Our approach allows us to score each putative mutation or variant with respect to its probability of belonging to each variant class, versus classification as a sequencing error. The method enables the joint calling of related samples form the same patient, such as cases where a cf-DNA sample and primary tumor sample are both profiled improving sensitivity and specificity. We validated our method by analyzing data obtained with the TOMA OS-Seq targeted sequencing RUO assay for 98 cancer genes from a mixture of well-known genomes, patient case triads (where normal, tumor and cf-DNA are available), and a retrospective analysis of tumor patient data that underwent clinical tumor profiling for therapy selection.

Citation Format: Francisco M. De La Vega, Ryan T. Koehler, Yannick Pouliot, Yosr Bouhlal, Austin So, Federico Goodsaid, Sean Irvine, Len Trigg, Lincoln Nadauld. Joint somatic mutation and germline variant identification and scoring from tumor molecular profiling and ct-DNA monitoring of cancer patients by high-throughput sequencing. [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 2712.

Abstract 3650: Multi-pronged approach to establish control standards for somatic mutations in next generation sequencing (NGS) oncology test

In recent years, somatic mutation testing via Next Generation Sequencing (NGS) has emerged as a powerful tool to detect targetable mutations in diseases such as cancer. These include mutations such as BRAF V600E, which can lead to targeted drug therapy; therefore, improving the standard of care in cancer diagnostics. To generate clinically meaningful results, we need to define standards that can be applied to laboratories performing somatic testing on tumor samples. With ongoing debate about the use of positive or negative controls in somatic testing assays, we designed an experiment to test the concordance of clinically actionable mutations commonly seen in tumor testing. We use cell line controls and commercially engineered DNA to validate concordance of mutations at specific allelic ratios. Our results depict an experiment design to determine if a genomic test can detect actionable mutations with high levels of accuracy and precision.

In this study, we designed a multi-pronged approach to evaluate the need for control standards by the ICG100 clinical NGS panel. To establish a baseline negative control, we performed concordance testing on the NA12878 cell line. We compared SNPs, insertions and deletions found using our methodology to the variants reports by NIST. We found our results to be concordant with NIST at a sensitivity of 92% and a specificity of 93%. To establish a baseline for positive controls, we utilized commercially engineered DNA which contains variants spiked-in at known allelic frequencies. This allowed us to determine if the ICG100 panel was able to retrieve specific somatic mutations. We were able to detect high confidence somatic mutations, such as BRAFV600E, as well as BRAF V600G which was spiked in at a lower expect allelic frequency of 4%. Moreover, we are able to detect with high confidence mutations such as ALK F1174L, EGFR G719S, PIK3CA H1047R, and MET Y1247D, and we were able to reproduce the results across multiple cell lines. Overall, our results show the need to utilize controls as standard protocol which can help assess the proficiency of a clinical assay. With rapid advances is NGS testing methodologies, there is a need for establishing standardized controls that asses the performance of a somatic tumor testing.

Citation Format: Christopher Johnson, Sharanya Raghunath, Jackie Wayne, Aimee Shamo, Patrick Bradley, Moises Hernandez, David Loughmiller, Jason Gillman, Derrick Haslem, Gary Stone, Lincoln Nadauld, Pravin J. Mishra. Multi-pronged approach to establish control standards for somatic mutations in next generation sequencing (NGS) oncology test. [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 3650.

Abstract A2-41: The Amplinome™ Test – A droplet digital PCR (ddPCR) based pan-cancer test to assess gene amplification status in solid tumors

Gene amplifications constitute &gt;30% of all actionable mutation types present in cancer. However, despite their prevalence and the availability of targeted drug therapies, the development of molecular assays to detect gene amplifications has been limited by technical challenges such as limited DNA quantity and tumor heterogeneity. To address this, we describe the Amplinome™ test, a test that leverages the high precision and accuracy of droplet digital PCR (ddPCR) to determine the amplification status of 12 commonly amplified genes targeted by FDA approved drugs.

Twelve taqman assays were designed to target AURKA, BRAF, CDK4, CDK6, EGFR, ERBB2/HER2, ERBB3/HER3, FGFR1, JAK2, MET, SRC, and VEGFA, and duplexed with the reference gene RNaseP (Epoch, Bothell, WA). Each duplexed assay was then assembled with 2 μL of purified DNA and analyzed via ddPCR (Bio-rad QX-200 ddPCR Platform, Hercules CA). Genes were identified as “amplified” if the observed target:reference ratio was &gt;1.48 (p &lt; 0.005), which adjusts the ASCO-CAP threshold of 2.2 recommended for HER2 amplification for a minimum tumor content of 40%.

The Amplinome test identified 29 of 49 patients having actionable amplifications in at least one of the 12 assayed genes. In contrast, only 6 of 49 patients were identified as having amplifications in these genes through a commercially available sequencing test. An overall concordance rate of 90% (532/588 calls) between the two tests was observed. Discordant calls (56/588) arose from amplifications identified by the Amplinome test but not sequencing (55/56). One discordant call (HER3) was identified as amplified via sequencing, but confirmed to be unamplified via FISH.

The observed 10-fold increase in sensitivity versus a sequencing-based clinical test suggests that the prevalence of actionable amplifications may be significantly underestimated in cancer. This sensitivity, combined with its ability to provide results within an actionable timeframe (&lt;1 week), the underlines the potential value of Amplinome™ test in providing treatment options for oncologists and their patients.

Citation Format: Amy Wong, Jennifer Pecson, Girish Putcha, Lincoln Nadauld, Austin P. So. The Amplinome™ Test – A droplet digital PCR (ddPCR) based pan-cancer test to assess gene amplification status in solid tumors. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A2-41.

Abstract 1114: Genomics in cancer patient care: Bench to bedside and beyond

Burgeoning sequencing technologies and the slow pace to use genomics data clinically has been largely hampered by lack of unestablished clinical utilities. There is an immediate need to store, analyze and retrieve meaningful clinical information from high throughput genomics data. At Intermountain Cancer Genomics (ICG) we created a complete translational pipeline that integrates these major core aspects: a clinical laboratory that generates data targeted within our gene panel, a bioinformatics pipeline that accounts for the quality of the data and further generates a set of variants that can be interpreted and targeted clinically to improve the quality of a patient's life.

At the ICG clinical laboratory we have developed a comprehensive ICG100 targeted sequencing panel under CLIA-CAP guidelines. DNA extracted from patient's tumor specimen gets sequenced for the entire coding region of 96 cancer-related genes which are often altered in cancer. These targeted regions are sequenced on Illumina's MiSeq sequencing platform using an in-solution, oligo-capture sequencing method. This test offers high coverage (&gt;100X) and detects all classes of genomic alterations, including indels, translocations, copy number alterations (CNAs), and point mutations across the exons of 96 genes. This approach is viable and well-suited for all sample types including FFPE, fresh tissue and plasma.

We have developed a comprehensive bio-analytics pipeline that accommodates the diverse variants generated by the unique sequencing chemistry. Comparison of ICG100 to a commercially available CLIA-certified sequencing test that also detects copy number alterations, reveals high concordance across the spectrum of variant types. In addition, the ICG100 test detected 10 additional CNAs across 6 separate samples that were not identified in the commercially available test, suggesting an increased sensitivity with ICG100.

This integrated service utilizes a collaborative molecular tumor board that consist of subject expert scientists and physicians. Interdisciplinary tumor board suggests effective treatment options based on genomics data and clinical relevance. ‘Actionable’ genomic alterations are categorized as such if linked to an approved therapy in the solid tumor examined or another solid tumor. This test can be ordered by oncologist through a simple web-based interface where genomic results and molecular tumor board interpretation can be viewed. Additionally, oncologists can make a treatment selection and order drug at their convenience.

The ICG100 test therefore not only is cost-effective but offers higher sensitivity, coverage, superior ability for clinical management with identification of actionable CNAs, genomics-driven personalized treatment and precision cancer care.

Citation Format: Sharanya Raghunath, David Loughmiller, Aimee Shamo, Jackie Wayne, Patrick Bradley, Jason Gillman, Gary Stone, Derrick Haslem, Lincoln Nadauld, Pravin J. Mishra. Genomics in cancer patient care: Bench to bedside and beyond. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1114. doi:10.1158/1538-7445.AM2015-1114

Abstract 625: The frequency of gene amplifications in cancer revealed by a droplet digital PCR (ddPCR) based pan-cancer gene panel test

The development of molecular assays designed to detect gene amplifications has largely been hampered by technical challenges such as limited DNA quantity and tumor heterogeneity, which demand methods of very high precision and sensitivity. The recent introduction of affordable digital PCR platforms, such as droplet digital PCR (ddPCR), that are capable of providing single molecule resolution of target abundances thus provides a unique opportunity to address this gap in molecular diagnostics. A ddPCR-based test was therefore developed under CLIA-CAP guidelines to determine the amplification status of 12 commonly amplified genes targeted by FDA approved drugs. Termed the Amplinome Test, this test was applied to 49 clinical samples received over a period of 6 months and compared to the results obtained from a commercially available clinical sequencing test applied that also reports copy number alterations (CNAs). An overall concordance rate of 90% (532/588 calls) was observed between the two tests across the entire gene set, 5 of which were identified as amplified. There were 56 discordant copy number amplifications between the two testing modalities. The vast majority (55/56) of discordant calls arose from gene amplifications identified by the Amplinome test, but not detected by sequencing, indicating an 11-fold increase in sensitivity in detecting amplifications. One discordant call (HER3) was identified as amplified via sequencing, but confirmed to be unamplified via FISH. At the patient level, the Amplinome test identified 5-fold more patients as having actionable amplifications in at least one of the 12 assayed genes versus clinical sequencing (29 vs. 6). Clinical management was altered in 14 of the 29 patients with an actionable CNA identified on the Amplinome test; those patients received targeted therapy directed against the amplified gene. The ddPCR-based Amplinome test thus provides a highly sensitive method for measuring gene amplifications in cancer that alters patient management, and suggests that the prevalence of actionable amplifications may be significantly underestimated by standard clinical next-generation sequencing tests.

Citation Format: Austin P. So, Amy Wong, Jennifer Pecson, Girish Putcha, Gregory Jensen, Michael Lucero, Gary Stone, Jason Gillman, Pravin Mishra, David Loughmiller, Derrick S. Haslem, Lincoln Nadauld. The frequency of gene amplifications in cancer revealed by a droplet digital PCR (ddPCR) based pan-cancer gene panel test. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 625. doi:10.1158/1538-7445.AM2015-625

A quality outcomes analysis following treatment with personalized genomic cancer medicine

12

Background: Personalized genomic cancer medicine is an approach that has long shown efficacy in molecularly-defined subsets of breast and lung cancers, amongst others, but has not been employed more broadly, in part, due to limitations in testing technologies. Recent advances in genomic technologies have increasingly alleviated these constraints, thereby enabling precision cancer medicine. Data regarding the quality outcomes of patients treated with precision cancer medicine is an important next step along the path to widespread employment of this approach. Methods: We performed an IRB-approved retrospective analysis evaluating the use of targeted therapies matched to patients’ molecular aberrations as determined by genomic testing and recommended by a Molecular Tumor Board (MTB). This study assessed whether a detailed molecular profile of advanced solid malignancies including, but not limited to, the lung, gastrointestinal tract, bladder, prostate, ovary, uterus and skin, that were subsequently treated with matched targeted therapies resulted in improvements in three quality outcome measures: 1) Progression Free Survival (PFS), 2) treatment related morbidity, and 3) cost of treatment. Results: Preliminary results of the cohort analysis suggest that the costs associated with genomic targeted therapy are comparable to standard therapies; however, the costs of treatment related morbidities is significantly lower for patients receiving genomic cancer medicine compared to standard chemotherapy approaches. In line with these findings, overall treatment related morbidities are significantly reduced in the genomic cancer medicine cohort compared to a control cohort. Data regarding the Progression Free Survival are pending at the time of this report. Conclusions: These retrospective data suggest that personalized genomic cancer medicine approaches result in decreased morbidities and cost savings compared to standard chemotherapeutic approaches. In patients with advanced cancer, genomic-based treatments appear to be cost-effective, safe and viable option for treating advanced cancer patients. Additional data regarding survival outcomes are required to determine efficacy of treatment.

Abstract 3488: Surveying colorectal cancer genome for clinically actionable genomic amplifications with droplet digital PCR

New cancer therapies are increasingly geared towards exploiting critical genetic and genomic features specific to the tumor. These mutations and genomic aberrations are the basis for precision cancer medicine. Thus, rapid molecular characterization of clinical cancer samples has become increasingly important for cancer targeted therapy development. Our study addresses this need by identifying ‘druggable’ gene amplifications in colorectal cancer (CRC). Along these lines, we developed an accurate and rapid droplet-digital PCR (ddPCR) assay to analyze cancer-specific copy number variations (CNV) in 13 genes with oncogenic function that are the targets of specific cancer therapies. These genes were selected based on the availability of approved and early-phase targeted therapeutics that inhibit their oncogenic function. Using TCGA data, we vetted our targets by identifying frequent CRC amplifications and a correlation with gene expression.

The ddPCR method involves emulsifying matched-normal cancer sample DNA that provides specific advantages for highly sensitive and specific detection of cancer CNVs. Post-amplification, emulsion droplets are streamed single-file into a capillary that leads past a two-color detector, where the positive droplets for the target and reference genes are “counted” for quantitation. This technology requires nanogram amounts of genomic DNA, thus facilitating the study of clinical cancer samples from small biopsies. It works well with genomic DNA extracted from formalin fixed paraffin embedded tumor samples. Furthermore, we demonstrated high sensitivity for detecting copy number amplifications even in samples containing a prominent fraction of normal tissue.

We are extending our analysis to the full 13 “druggable” gene targets and evaluating patterns of mutual exclusivity and co-occurrence among a cohort of over 200 CRC tumors with information on clinical outcome. For example, we have demonstrated FGFR1 amplifications in 5.2% of our CRC tumor population. For ERBB2 variations in the same cohort, we detected amplifications in 3.6% of our population. ERBB2 and FGFR1 amplification events displayed a mutually exclusive pattern of segregation (p value &lt;0.0001 per Chi-squared test with Yates correction). This efficient and inexpensive assay offers a significant potential to extend personalized therapeutic options available to CRC patients.

Citation Format: Rowza T. Rumma, Laura Miotke, Lincoln Nadauld, Georges Natsoulis, Michael DiMaio, Moe Jalali, Hanlee Ji. Surveying colorectal cancer genome for clinically actionable genomic amplifications with droplet digital PCR. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3488. doi:10.1158/1538-7445.AM2013-3488

Quantitative and Sensitive Detection of Cancer Genome Amplifications from Formalin Fixed Paraffin Embedded Tumors with Droplet Digital PCR

For the analysis of cancer, there is great interest in rapid and accurate detection of cancer genome amplifications containing oncogenes that are potential therapeutic targets. The vast majority of cancer tissue samples are formalin fixed and paraffin embedded (FFPE) which enables histopathological examination and long term archiving. However, FFPE cancer genomic DNA is oftentimes degraded and generally a poor substrate for many molecular biology assays. To overcome the issues of poor DNA quality from FFPE samples and detect oncogenic copy number amplifications with high accuracy and sensitivity, we developed a novel approach. Our assay requires nanogram amounts of genomic DNA, thus facilitating study of small amounts of clinical samples. Using droplet digital PCR (ddPCR), we can determine the relative copy number of specific genomic loci even in the presence of intermingled normal tissue. We used a control dilution series to determine the limits of detection for the ddPCR assay and report its improved sensitivity on minimal amounts of DNA compared to standard real-time PCR. To develop this approach, we designed an assay for the fibroblast growth factor receptor 2 gene (FGFR2) that is amplified in a gastric and breast cancers as well as others. We successfully utilized ddPCR to ascertain FGFR2 amplifications from FFPE-preserved gastrointestinal adenocarcinomas.

APRIL/TRDL-1, a tumor necrosis factor-like ligand, stimulates cell death

We have examined the activity of a new member of the tumor necrosis factor (TNF) family identified through Expressed Sequence Tag database searching using TNFalpha protein as the search query. We have termed this protein TNF-related death ligand-la (TRDL-1alpha). Traditional cDNA library screening identified two additional splice variants designated TRDL-1beta and TRDL-1gamma that differed from TRDL-1alpha by the deletion of two small regions within the protein coding region. TRDL-1alpha is identical in sequence to the recently described molecule, APRIL, that may induce cell proliferation. We found, however, that purified, FLAG-tagged TRDL-1alpha caused Jurkat cell death with kinetics that paralleled FasL. In vitro binding experiments demonstrated that TRDL-1alpha coprecipitated Fas and HVEM and suggested TRDL-1alpha as an alternate ligand for these receptors. TRDL-1 localized to chromosome 17p13.3 and its expression was widespread in normal tissues. Examination of 48 tumor samples revealed high levels of TRDL-1 expression in several tumors, including those from the gastrointestinal tract. Expression of TRDL-1 in COS-1 cells confirmed membrane association of TRDL-1, typical of TNF family members.