Abstract 764: Fully automated sample-to-report NGS workflow for comprehensive genomic profiling for myeloid neoplasms

Introduction: Myeloid malignancies are associated with a broad and diverse set of genomic alterations, including SNVs, insertions, deletions and gene fusions. Comprehensive characterization of genetic mutations in hematological disorders currently requires a variety of diagnostic tests and takes multiple days to complete. We developed a fully automated NGS Myeloid Assay that offers an easy to use sample-to-report workflow and the capability for processing up to 8 samples per day.

Methods: The Genexus System is comprised of two software linked instruments, the Genexus Purification System and the Genexus Integrated Sequencer. The Genexus Purification System was used to isolate the DNA and RNA from blood or bone marrow samples from precharacterized myeloid samples representative of Acute Myeloid Leukemia (AML) and Myelodysplastic Syndrome (MDS) and from the blood of healthy donors. The Genexus Integrated Sequencer was used to dilute the nucleic acids to optimal concentration and to sequence the samples in replicates with Oncomine Myeloid Genexus v2 Assay. Six DNA and RNA samples were sequenced per run per day along with commercially available analytical controls and a No Template Control. The report was generated by the Genexus Software analysis pipeline optimized to detect different variant types with high sensitivity and specificity.

Results: The purification workflows were tested with blood input of 50-400uL for DNA and 50-150uL for RNA. Genexus Purification System extracted and quantified nucleic acids showed input dependent yields. DNA and RNA yields obtained with 50uL sample inputs consistently met the 27.5ng DNA and 15ng RNA minimum requirements for Genexus Integrated Sequencer workflows. DNA libraries had >97% Uniformity of Amplicon Coverage and >95% Target Base Coverage at 350x. The percentage ratio of Mapped Reads for DNA and RNA libraries was approximately 80:20. Detected genetic variations included key hotspots in CEBPA, FLT3, IDH1/2, NPM1, NRAS, RUNX1, and U2AF1 genes that are prevalent in AML and MDS. Genexus Variant Calling results showed high reproducibility and high concordance to the Ion GeneStudio S5 sequencing platform (>95%). The analytical controls, AcroMetrix Oncology Hotspot Control, Seraseq Myeloid Mutation DNA Mix and Seraseq Myeloid Fusion RNA Mix, were sequenced with Sensitivity and PPV >95%.

Conclusion: The Genexus System offers an automated sample-to-report workflow with minimal hands-on-time and run results in 30 hours which allows an easy to use solution for next day turnaround time. When used with the Oncomine Myeloid GX v2 Assay, it provides accurate and comprehensive information on diverse mutations including fusions that are relevant to the study of myeloid cancers.

For research use only. Not for use in diagnostic procedures.

Citation Format: Marina Sedova, Alexy Ongpin, Jennifer Burke, Collyn Seeger, Sarah Brozio, Janice Au-Young, Jiajie Huang, Thilanka Jayaweera, Iris Casuga, Milton Huynh, Fiona Hyland. Fully automated sample-to-report NGS workflow for comprehensive genomic profiling for myeloid neoplasms [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 764.

Abstract 2079: Prediction of DDR and other mutation signatures using targeted panels for FFPE samples

Mutational processes can cause driver mutations and are considered the primary cause of tumorigenesis. Cosmic signatures classify environmental or intrinsic mutational processes. Identification of signatures including DNA Damage Response (DDR) signatures enables research into the origin of cancer mutagenesis and into treatment optimization. Typically, Mutation signatures are identified using Whole Genome Sequencing or Whole Exome Sequencing. We demonstrate signature prediction of mutation signatures with two amplification-based targeted panels, which have a high sequencing success rate for FFPE samples. ~2000 FFPE Samples from a pan-solid tumor cohort were sequenced with either of two AmpliSeq panels (OCAplus and OTMLA), sizes 1.4 Mb and 1.7 Mb, to optimize a method to identify mutation signatures. First, we identified somatic SNVs by filtering out likely population germline mutations, and used these to construct the single base change substitution (SBS) matrix. The COSMIC cancer signatures use properties of the whole genome: we normalized to extend signature detection to targeted panel data. We adjusted the mutation frequencies observed using the ratio of trinucleotide counts in the genome and the ratio in the panel. Next, we measured the cosine similarity between the normalized sample and the COSMIC signatures. We selected the signatures with a strong match (>0.7) to the normalized sample. We further impute the signatures using a reduced candidate set, to assess the signature fit in the sample and reduce false positives. These steps provide the optimized signatures. We detected optimized signatures in 33% of the samples. A single signature was detected in 9% of the samples, 2 signatures in 6%, and >2 signatures in 17% of the samples. ~3% of samples showed at least one MMR signatures (SBS6, SBS14, SBS15, SBS20, SBS21, SBS26 and SBS44). In all 10 OCAPlus samples with MMR signatures, we also detected mutation(s) in an MMR gene. APOBEC signatures, SBS2 and SBS13, were observed in 3% of samples sequenced with OCAPlus panel and TML panel. We found 3 samples showing HRR signature SBS3. We successfully identified the putative causal DDR mutation in a number of samples. For example, of 62 samples with NTHL1 mutations, 60 had NTHL1 related signature SBS30. A sample with MUTYH mutations was assigned the SBS36 signature. We sequenced matched Tumor/Normal pairs; in most cases, the tumor sample showed a stronger mutational signature, with their corresponding normal sample showing either no signature or a weaker matching signature. We assessed reproducibility of the mutational signature. Looking at duplicate and triplicate replicates, we found that replicates consistently displayed the same mutational signatures. We show identification of mutation signatures using targeted panels designed for FFPE samples.

Citation Format: Fiona Hyland, Ajithavalli Chellappan, Chintan Vora, Shilpa Nair, Jagannath Patro, Ritika Raj, Rushikesh Kanap. Prediction of DDR and other mutation signatures using targeted panels for FFPE samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2079.

Challenges in bioinformatics approaches to tumor mutation burden analysis

Several immune checkpoint inhibitors (ICIs) have already been introduced into clinical practice or are in advanced phases of clinical experimentation. Extensive efforts are being made to identify robust biomarkers to select patients who may benefit from treatment with ICIs. Tumor mutation burden (TMB) may be a relevant biomarker of response to ICIs in different tumor types; however, its clinical use is challenged by the analytical methods required for its evaluation. The possibility of using targeted next-generation sequencing panels has been investigated as an alternative to the standard whole exome sequencing approach. However, no standardization exists in terms of genes covered, types of mutations included in the estimation of TMB, bioinformatics pipelines for data analysis, and cut-offs used to discriminate samples with high, intermediate or low TMB. Bioinformatics serve a relevant role in the analysis of targeted sequencing data and its standardization is essential to deliver a reliable test in clinical practice. In the present study, cultured and formalin-fixed, paraffin-embedded cell lines were analyzed using a commercial panel for TMB testing; the results were compared with data from the literature and public databases, demonstrating a good correlation. Additionally, the correlation between high tumor mutation burden and microsatellite instability was confirmed. The bioinformatics analyses were conducted using two different pipelines to highlight the challenges associated with the development of an appropriate analytical workflow.

Validation of a Targeted Next-Generation Sequencing Panel for Tumor Mutation Burden Analysis: Results from the Onconetwork Immuno-Oncology Consortium

Tumor mutation burden (TMB) is evaluated as a biomarker of response to immunotherapy. We present the efforts of the Onconetwork Immuno-Oncology Consortium to validate a commercial targeted sequencing test for TMB calculation. A three-phase study was designed to validate the Oncomine Tumor Mutational Load (OTML) assay at nine European laboratories. Phase 1 evaluated reproducibility and accuracy on seven control samples. In phase 2, six formalin-fixed, paraffin-embedded samples tested with FoundationOne were reanalyzed with the OTML panel to evaluate concordance and reproducibility. Phase 3 involved analysis of 90 colorectal cancer samples with known microsatellite instability (MSI) status to evaluate TMB and MSI association. High reproducibility of TMB was demonstrated among the sites in the first and second phases. Strong correlation was also detected between mean and expected TMB in phase 1 (r² = 0.998) and phase 2 (r² = 0.96). Detection of actionable mutations was also confirmed. In colorectal cancer samples, the expected pattern of MSI-high/high-TMB and microsatellite stability/low-TMB was present, and gene signatures produced by the panel suggested the presence of a POLE mutation in two samples. The OTML panel demonstrated robustness and reproducibility for TMB evaluation. Results also suggest the possibility of using the panel for mutational signatures and variant detection. Collaborative efforts between academia and companies are crucial to accelerate the translation of new biomarkers into clinical research.

Abstract 6472: Clonal lineage and somatic hypermutation analysis of chronic lymphocytic leukemia by long-amplicon IGH chain sequencing

Background: Current next-generation sequencing (NGS) approaches for analyzing SHM commonly rely on multiplex primers targeting the framework 1 (FR1) or leader region of the IGH variable gene in combination with joining gene primers to amplify rearranged IGH chains from gDNA template. Limitations include the potential for joining gene mutations to interfere with primer binding and an inability to evaluate isotype. Here we present a method for translational research investigations of IGH chain SHM employing multiplex FR1 and isotype (constant gene) specific primers (Oncomine IGH-LR assay primers) to amplify IGH chains from RNA template. We evaluated performance by comparing SHM values obtained from NGS of RNA from 54 CLL samples amplified using Oncomine IGH-LR primers to values obtained by Sanger sequencing or NGS of RNA amplified using FR1 or leader region/J gene primers.

Methods: IGH chains from 54 CLL samples derived from two separate sequencing sites (Site 1: 24 samples, Site 2: 30 samples) were amplified from peripheral blood using Oncomine IGH-LR assay followed by sequencing via the Ion Gene Studio S5. Clonotyping, clonal lineage identification and somatic hypermutation analysis was performed by Ion Reporter via comparison to the IMGT reference database. Oncomine IGH-LR assay SHM values were compared to those obtained via NGS-based sequencing utilizing FR1/J gene primers (Site 1) or Sanger sequencing utilizing IGH-leader or FR1 and joining gene primers (Site 2).

Results: IGHV SHM values were highly concordant between NGS approaches (Spearman cor =.957, Site 1) and between Sanger sequencing and NGS approaches (Spearman cor = .849, Site 2). Sequence data obtained using Oncomine IGH-LR assay enables more in-depth clonal lineage analysis, including evaluation of isotype representation and subclonal evolution.

Conclusions: These results support the robustness and reliability of multiplex FR1 and constant gene based IGH chain amplification for the translational research characterization of somatic hypermutation in CLL and other B cell neoplasms.

Citation Format: Jayde Chang, Zadie Davis, Graeme Quest, Harriet Feilloter, Michelle Toro, Geoffrey Lowman, Loni Pickle, Fiona Hyland, Timothy Looney. Clonal lineage and somatic hypermutation analysis of chronic lymphocytic leukemia by long-amplicon IGH chain sequencing [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 6472.

Abstract 158: Custom primer design pipeline and analysis workflow for targeted methylation sequencing using NGS Ion AmpliSeq technology

Changes in DNA methylation, causing chromosome instability and altered gene expression, have been strongly associated with carcinogenesis. Due to the involvement of methylation in cancer, methylation profiles have been heralded as promising cancer biomarkers. Here, we present a primer design pipeline and an analysis workflow that we have developed to design and analyze custom methylation panels and detect methylation status.

An automated primer design pipeline for methylation sequencing has been developed, consisting of genome conversion, primer selection, amplicon tiling, and generation of optimal amplicons. Custom methylation panels can be designed using pre-converted genomes or reference genome sequences for any other organism which can then be converted. The pipeline has the capability to create custom targeted panels specific to any methylation sites of interest. The pipeline designs Ion AmpliSeq primers to enable high multiplexing and robust amplification of low abundance or degraded DNA.

Following the creation of a custom panel, a complete 3-day workflow has been developed, comprising bisulfite conversion, library construction, template preparation, sequencing and data analysis. This 3-day protocol offers manual or automated library options, low input (10-20ng DNA) and a flexible multiplexed approach with quantitative information at single base pair resolution. Sequencing is performed on the Ion GeneStudio S5 system. The bioinformatics analysis has been streamlined into a downloadable plugin performing alignment and DNA methylation calling for amplicons on both the Watson and Crick strands.

To evaluate the in silico performance of the primer design pipeline for targeted bisulfite sequencing, a custom methylation panel was created using a set of 48 oncology markers from the BLUEPRINT consortium. These markers were also used for the Ion AmpliSeq Methylation Panel for Cancer Research, which was compared to the custom methylation panel to evaluate the performance of the pipeline. Key metrics from in silico design such as total number of degenerate oligos, mean amplicon length and average Tm spread for the custom methylation panel are equal to or better than Ion AmpliSeq Methylation Panel for Cancer Research.

To assess the sequencing performance of the panel, two control gDNA samples were used. The expected average methylation status across all CpGs were >98% and <5% for the first sample and the second sample, respectively. The evaluation was also carried out with an equal mixture of these two samples. The wet lab testing of the custom methylation panel generated comparable results to the Ion AmpliSeq Methylation Panel for Cancer Research.

The primer design pipeline and 3-day workflow provide custom design of targeted methylation panels along with quantitative analysis of relevant oncology markers from low DNA input.

Citation Format: Zunping Luo, Loni Pickle, Andrew Hatch, Aren Ewing, Fiona Hyland, David Berman, Palak Patel, Mark Andersen. Custom primer design pipeline and analysis workflow for targeted methylation sequencing using NGS Ion AmpliSeq technology [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 158.

Abstract 5472: Development of customizable targeted RNA fusion panels using a novel automated high-multiplexing primer design strategy

Introduction

Gene fusions play an important role in oncogenesis and the progression of cancer. As important biomarkers, sensitive identification of gene fusions is critical to future oncology research. Next generation sequencing with Ion Ampliseq targeted enrichment enables simple, accurate and specific detection of relevant fusion isoforms. Here we introduce a novel automatic high-multiplexing primer design strategy that has the flexibility to develop customized Ampliseq fusion panels for any combination of fusion isoforms, scaling to panels that can detect thousands of isoforms in a single primer pool, which increases the sensitivity of fusion detection while decreasing the sample input required to as low as 10 ng.

Methods

The automated primer design pipeline takes a Gene-Transcript-Exon (GTE) file as input. Each record in the GTE file represents a unique RNA fusion isoform to establish an easy-parsing format for the pipeline. The pipeline locates the fusion breakpoint position, extracts gene sequences of every candidate fusion target and builds the fusion reference. Candidate amplicons are generated against the fusion reference. According to the design requirements of pool number and the conflicts among primer pairs, the pipeline performs pooling to minimize primer interactions. Finally, the pipeline generates an optimal set of amplicons strategically targeted for fusion junctions. The output files are used for downstream analysis with a fully automated analysis pipeline.

Results

The pipeline has been used extensively to develop high performing multiplex RNA fusion panels. The pipeline generates 175-base amplicons for use on formalin-fixed, paraffin-embedded (FFPE) samples or 120-base amplicons for use on cfRNA from blood samples. A single panel can include thousands of known fusion variants. This pipeline has been used to design the fusion assays contained in Oncomine Focus and Comprehensive assays, Oncomine Precision Assay, and others. Oncomine Comprehensive Assay v3 fusion panel was tested using the Ion GeneStudio S5 Sequencer; for example, testing on SeraCare fusion control confirms that all 14 fusion isoforms were detected with 100% accuracy. Testing on FFPE samples with known positive fusions confirms that the expected fusions including NTRK1, ERG, ETV1 and MET driver genes were also detected with 100% accuracy.

Conclusions

In summary, we have developed an automatic pipeline that can generate robust, comprehensive customized multiplex RNA fusion assays for targeted next-generation sequencing.

For research use only. Not for use in diagnostic procedures.

Citation Format: NA LI, Antonio Martinez-Alcantara, Aren Ewing, Rajesh Gottimukkala, Fiona Hyland, Seth Sadis. Development of customizable targeted RNA fusion panels using a novel automated high-multiplexing primer design strategy [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 5472.

Abstract 1326: A novel system that produces pre-qualified cancer NGS panels with customizable content

Introduction

Next-generation sequencing (NGS) is the preferred method to simultaneously characterize multiple relevant genetic variants in cancer samples. In addition to pan-cancer applications, researchers are increasingly interested in cancer type specific and custom solutions. We describe a flexible system for providing optimized, pre-tested primers for PCR-based NGS library preparation that generates cancer type specific panels with customizable content for use with FFPE tumor tissue samples.

Methods

A multifactorial scoring method was used to assign gene targets to 10 different cancer types for a total of approximately 300 genes. Fifteen to thirty genes most relevant to each particular cancer type were chosen for inclusion into core panels and an additional 15-50 genes were identified as supplemental content for each cancer type. The system is designed so that core panels can be modified by adding, dropping, or substituting any of the core genes with any of the other genes inventoried in the system. The assay system uses Ion AmpliSeq™ technology with manual or automated library preparation and sequencing on the Ion Torrent GeneStudio™ S5 sequencing platform. Ten ng of purified DNA per library pool (20 ng total) is used as input for library preparation. An automated tumor-only workflow for variant calling and sample quality reporting is provided within Ion Reporter. Streamlined access to reporting of variant relevance is enabled by Oncomine Reporter.

Results

Following primer design optimization and performance testing, core panels for Bladder, Colorectal & Pancreatic, Kidney, Liver, Melanoma, Prostate, Lymphoma (B-cell types), and Gynecological cancers, as well as a panel for BRCA1/2 and homologous recombination repair, were characterized with cell lines, commercial reference controls, and FFPE samples. Panel base uniformity was > 90% across all panels. Sensitivity for hotspot variants (both SNVs and indels) was 95% down to Minor Allele Frequency (MAF) of 5%. Positive Predictive Value (PPV) for hotspots variants (both SNVs and indels) was 99%. The sensitivity of CNV gain was 90%.

Conclusions

A novel system to provide high-quality NGS library prep reagents for pre-defined or customized cancer panels and representative performance data are described in detail.

For Research use only. Not for use in diagnostic procedures.

Citation Format: Steven Roman, Charles Scafe, Yun Zhu, Fernando Farfan, Brooke McKnight, Santoshi Bandla, Chenchen Yang, Yu-Ting Tseng, Xiaoping Duan, Jigar Patel, Natasha Arksey, Seth Sadis, Fiona Hyland. A novel system that produces pre-qualified cancer NGS panels with customizable content [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 1326.

Abstract 3305: Comparative analysis of RNA versus DNA as input material for IGH repertoire sequencing based detection of rare clonal B cells at a frequency of 10E-6

Background: Next generation sequencing of rearranged IGH chains has emerged as a reproducible and highly sensitive approach for detecting rare B cell clones, e.g. malignant B cell, in peripheral blood. Historically, efforts to evaluate the frequency of rare B cells by IGH chain sequencing have utilized DNA input given potential challenges in accurately quantifying template copy number from RNA data owing to B cell subtype specific variation in the expression of the B cell receptor. Hypothetically, however, RNA input based monitoring could be advantageous both owing to reduced input requirements and superior ability to detect B cell malignancies of plasmablast and plasma cell origin, where the BCR is robustly expressed. Here we compared the ability of RNA and DNA based IGH chain sequencing to detect two Burkitt's Lymphoma B cell lines (CA46 and GA10) at a frequency of 10E-6 from peripheral blood. We discuss advantages of either approach for detection of rare B cell clones.

Methods: IGH chain libraries were prepared using the Oncomine IGH-SR assay (framework 3 and joining gene multiplex primers) from gDNA or total RNA extracted from peripheral blood and spiked with Burkitt's lymphoma or CLL cell line gDNA or total RNA to a frequency of 10E-6 by mass ratio. Libraries were sequenced via the Gene Studio S5 followed by Ion Reporter analysis to identify clonotypes and evaluate B cell clone frequencies across samples. Automated downsampling analysis was used to confirm libraries were sequenced to saturation. Library preparation and analysis was performed in replicate to quantify sensitivity of detection.

Results: For each cell line, we prepared and sequenced (1) 30 libraries derived from amplification of 2ug gDNA spiked with 2pg cell line gDNA and (2) 10 libraries derived from amplification of 100ng RNA spiked with .1pg cell line total RNA. The Burkitt's lymphoma cell lines were detected in 10/30 and 8/30 gDNA libraries respectively, for CA46 and GA-10) consistent with the historical performance of orthologous DNA-based sequencing approaches. For RNA libraries, the Burkitt's lymphoma cell lines were detected in each library (10/10 and 10/10, respectively).

Conclusions: Here we demonstrate detection of B cell lines at a frequency down to 10E-6 from DNA and RNA input. Importantly, we find that RNA based IGH sequencing may significantly reduce input requirements for rare clone detection, potentially enabling routine detection of clones down to 10E-6 frequency from a single library.

Citation Format: Michelle Toro, Loni Pickle, Jayde Chang, Timothy Looney, Geoffrey Lowman, Mark Andersen, Fiona Hyland. Comparative analysis of RNA versus DNA as input material for IGH repertoire sequencing based detection of rare clonal B cells at a frequency of 10E-6 [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 3305.

The employment gap: the relationship between medical student career choices and the future needs of the New Zealand medical workforce

AIMS

To determine the career decision intentions of graduating doctors, and the relationship between these intentions and the predicted medical workforce needs in New Zealand in 10 years' time.

METHODS

A workforce forecasting model developed by the Ministry of Health (MOH) has been used to predict the proportion of doctors required in each medical specialty in 2028 in New Zealand. The future work intentions of recently graduated doctors at the Universities of Auckland and Otago were collected from the Medical Student Outcomes Data (MSOD), and compared with these predicted needs.

RESULTS

Between 2013 and 2017, 2,292 doctors graduated in New Zealand, of whom 1,583 completed the MSOD preferences section (response rate 69%). Of these only 50.1% had decided on a future medical specialty. The most popular were surgical specialties (26.2%), general practice (20.7%), and internal medicine (11.0%). Compared to the MOH workforce forecast model there appears to be insufficient interest in general practice at the time of graduation.

CONCLUSIONS

To shape the medical workforce to meet forecast needs, multiple stakeholders will need to collaborate, with a special focus on the early postgraduate years, as many doctors have yet to decide on specialisation.

Abstract 1701: Improvement of tumor mutation burden measurement by removal of deaminated bases in FFPE DNA

Tumor mutation burden (TMB) is a positive predictive factor for response to immune-checkpoint inhibitors in certain types of cancer. The Oncomine™ Tumor Mutation Load Assay, a 1.7Mb targeted next generation sequencing (NGS) panel, measures TMB and detects mutations in 409 cancer genes. The TMB values obtained using targeted sequencing are positively correlated with TMB measured by whole exome sequencing in NSCLC, colorectal, endometrial and gastric cancers. TMB from these tumor samples are correlated with other phenotypes associated with genomic instability, specifically microsatellite instability (MSI) and mutations involved in mismatch repair and cancer related genes. Analysis of the Oncomine™ TML Assay results with Torrent Suite and Ion Reporter software uniquely measures the degree of deamination of cytosines to uracils in fixed tissues. FFPE preservation methods can lead to significant cytosine deamination of the isolated DNA, resulting in decreased sequencing quality. In these samples, uracils are propagated as thymines and result in false C>T substitutions. To minimize the influence that excess deamination has on TMB results, we have incorporated a UDG enzyme treatment to eliminate damaged targets and improve usable TMB values of DNA from damaged FFPE tumor tissue. The Oncomine™ TML assay for TMB on Ion GeneStudio™ S5 in conjunction with MSI detection is informative and potentially predictive for the use of checkpoint inhibitors in multiple cancer types.

Citation Format: Warren Tom, Ruchi Chaudhary, Vinay Mittal, Dinesh Cyanam, Iris Casuga, Elaine Wong-Ho, Rob Bennett, Fiona Hyland, Seth Sadis, Janice Au-Young. Improvement of tumor mutation burden measurement by removal of deaminated bases in FFPE DNA [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 1701.

Abstract 5132: Assessing tumor mutational burden and profiling variants from FFPE samples using a PCR-based next-generation sequencing assay

Introduction: High Tumor Mutational Burden (TMB) has shown association with benefit from immune checkpoint blockade therapies. While TMB as computed from whole exome sequencing (WES) is still a gold standard, the high input material (tumor and germline DNA) requirement and complex bioinformatics refrains exploring this biomarker in individual labs. Herein, we develop a PCR-based targeted panel for computing TMB and detecting important variants from FFPE research samples.

Methods: A targeted panel was designed that covered 1.7 Mb of genomic region from 409 key cancer genes. First an UDG treatment is applied to repair samples with FFPE fixation error. Utilizing Ion AmpliSeq chemistry, the workflow requires very little (only 20 ng) input DNA. The assay enables a 3-day turn-around time from sample to the final report. The workflow enables < 60 minutes of hands-on time for automated library preparation and templating on a batch of 4 samples. Sequencing is performed on high throughput semiconductor sequencing platform to achieve sufficient depth (~1200x coverage) and accuracy. The analysis pipeline accompanies variant caller parameters optimized for high accuracy in variant detection. The workflow is tumor sample only, therefore, TMB (nonsynonymous somatic mutations/Mb) is calculated by removing germline variants using population databases. Eight NSCLC FFPE tumors were analyzed. Samples were also tested without UDG repair. Count of somatic C:G>T:A mutations below 15% allelic frequency was called deamination.

Results: In-silico analyses using 466 lung adenocarcinoma, 375 skin cutaneous melanoma, and 274 colon adenocarcinoma samples from TCGA MC3 dataset displayed high correlation between WES TMB and predicted panel TMB (r2 = 0.90 on lung, r2 = 0.96 on melanoma, and r2 = 0.98 on colon). Deamination reduced by average 85.67% (SD 16.87) on FFPE samples after applying UDG repair. Mean TMB of first six samples was 6.97 (SD 2.00), and TMB of last two samples was 43.58 and 87.14. Among two high TMB samples, first had gain-of-function and loss-of-function mutations in BRAF and NF1 genes and had 59.4% C:G>A:T somatic mutations relating to molecular smoking signature. On the second high TMB sample, the assay detected gain-of-function and loss-of-function mutations in FGFR3, NOTCH1, KRAS, HNF1A, and CREBBP genes and estimated 64.7% C:G>T:A somatic mutations at CpG sites consistent with deamination of 5-methylcytosine.

Conclusions: We have developed a workflow on the Ion Torrent sequencing platform with Ion AmpliSeq chemistry to estimate TMB from FFPE research samples. This solution will advance research in immuno-oncology.

Citation Format: Ruchi Chaudhary, Charles Scafe, Dinesh Cyanam, Vinay Mittal, Warren Tom, Janice Au-Young, Seth Sadis, Fiona Hyland. Assessing tumor mutational burden and profiling variants from FFPE samples using a PCR-based next-generation sequencing assay [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 5132.

Developing New Zealand's medical workforce: realising the potential of longitudinal career tracking

For over a decade, the Medical Schools Outcomes Database and Longitudinal Tracking Project (MSOD) has collected data from medical students in Australia and New Zealand. This project aims to explore how individual student background or attributes might interact with curriculum or early postgraduate training to affect eventual career choice and location. In New Zealand, over 4,000 students have voluntarily provided information at various time points, and the project is at a stage where some firm conclusions are starting to be drawn. This paper presents the background to the project along with some early results and future directions.

Peripheral blood TCRB chain convergence and clonal expansion following cytomegalovirus infection: Implications for the biomarker use of TCR-seq

134

Background: Human cytomegalovirus (CMV) is a common immune-evasive herpes family virus leading to lifelong asymptomatic infection in 50 to 80% of humans. The effect of CMV infection on the T cell repertoire may be relevant given interest in identifying T cell repertoire features predictive of response to checkpoint blockade immunotherapy (CPI) for cancer. Here we sought to identify features of CMV infection using TCRB profiling of peripheral blood (PBL) total RNA. Methods: Total RNA from PBL was obtained from 35 blood donors of known CMV status, then used for TCRB sequencing via the Oncomine TCRB-LR assay (amplicon spanning CDR1, 2 and 3) and the Ion Torrent S5. In parallel, we prepared libraries via the Oncomine TCRB-SR assay (CDR3 only). Data were used to identify TCRB repertoire features correlated with CMV status and compare repertoire features across the two assays. Results: T cell clone evenness was reduced in CMV positive individuals irrespective of age, predictive of CMV status (AUC = .86, p = 2E-4, Wilcoxon), and strongly correlated between LR and SR assays (Spearman cor = .96). TCR convergence was elevated in CMV positive individuals and uncorrelated with evenness (Spearman cor = -.03) such that the combination of convergence and evenness improved the performance of a logistic regression classifier (AUC = .93). Conclusions: We identify reduced T cell evenness and elevated TCR convergence as features of chronic CMV infection. CMV infection appears to significantly alter the T cell repertoire, suggesting that CMV status may be required for proper interpretation of T cell expansion in the context of CPI for cancer.

A scalable solution for tumor mutational burden from formalin-fixed, paraffin-embedded samples using the Oncomine Tumor Mutation Load Assay

Background

Tumor mutational burden (TMB) is an increasingly important biomarker for immune checkpoint inhibitors. Recent publications have described strong association between high TMB and objective response to mono- and combination immunotherapies in several cancer types. Existing methods to estimate TMB require large amount of input DNA, which may not always be available.

Methods

In this study, we develop a method to estimate TMB using the Oncomine Tumor Mutation Load (TML) Assay with 20 ng of DNA, and we characterize the performance of this method on various formalin-fixed, paraffin-embedded (FFPE) research samples of several cancer types. We measure the analytical performance of TML workflow through comparison with control samples with known truth, and we compare performance with an orthogonal method which uses matched normal sample to remove germline variants. We perform whole exome sequencing (WES) on a batch of FFPE samples and compare the WES TMB values with TMB estimates by the TML assay.

Results

In-silico analyses demonstrated the Oncomine TML panel has sufficient genomic coverage to estimate somatic mutations with a strong correlation (r²=0.986) to WES. Further, in silico prediction using WES data from three separate cohorts and comparing with a subset of the WES overlapping with the TML panel, confirmed the ability to stratify responders and non-responders to immune checkpoint inhibitors with high statistical significance. We found the rate of somatic mutations with the TML assay on cell lines and control samples were similar to the known truth. We verified the performance of germline filtering using only a tumor sample in comparison to a matched tumor-normal experimental design to remove germline variants. We compared TMB estimates by the TML assay with that from WES on a batch of FFPE research samples and found high correlation (r²=0.83). We found biologically interesting tumorigenesis signatures on FFPE research samples of colorectal cancer (CRC), lung, and melanoma origin. Further, we assessed TMB on a cohort of FFPE research samples including lung, colon, and melanoma tumors to discover the biologically relevant range of TMB values.

Conclusions

These results show that the TML assay targeting a 1.7-Mb genomic footprint can accurately predict TMB values that are comparable to the WES. The TML assay workflow incorporates a simple workflow using the Ion GeneStudio S5 System. Further, the AmpliSeq chemistry allows the use of low input DNA to estimate mutational burden from FFPE samples. This TMB assay enables scalable, robust research into immuno-oncology biomarkers with scarce samples.

User involvement in digital health: Working together to design smart home health technology

Background

Public involvement adds value to numerous aspects of health research, yet few studies have attempted to evaluate its impact on research. Evidence of such impact is needed to develop recommendations for best practice and ensure adequate resourcing.

Aim

To evaluate public involvement within a large interdisciplinary Science, Technology, Engineering and Mathematics (STEM) research project that focused on digital health.

Methods

The evaluation was conducted with members of the project's Public Advisory Groups (PAG) and with researchers who had participated in involvement activities. Two questionnaires were designed based on a public involvement value systems and clusters framework.

Results

Responses from members of the PAG (n = 10) were mostly positive towards normative values, which include moral, ethical and political aspects of involvement in research, and towards values concerning the conduct of public involvement and best practices. Researchers’ responses (n = 16) indicated they felt that involvement was generally effective and increased the quality, relevance and generalizability of their work. However, their responses about the validity of involvement in research were varied. They also highlighted several challenges including how well public involvement impacted on research, how decisions made in the research might differ from the views generated from public involvement, and barriers to researchers’ participation.

Discussion and conclusion

Our evaluation suggests that members of the public and the researchers value involvement. However, there is a need to consider how to embed public involvement to an even greater extent in STEM contexts and a need to address any barriers for researchers’ own involvement.

Abstract 2267: A novel method for classification of microsatellite instability (MSI) using the Oncomine Tumor Mutation Load assay

Cancer-associated changes in short tandem repeats (STRs) throughout the genome have been shown to be predictive of response to immunotherapy treatment. Traditional assays efficiently rely on a small number of genomic loci to assign an MSI status, but are only effective in a small number of cancer types. Genome-wide profiling of aberrations in STRs has the potential to reveal a comprehensive signature of MSI and could allow for the association of response with samples that are labeled as microsatellite stable (MSS) by traditional assays.To facilitate the discovery of informative repeat sites across the genome, we have extended the Oncomine Tumor Mutation Load assay* to allow for the detection of variants in STRs. Our method addresses challenges of semiconductor sequencing of long STR elements including indel and partial incorporation signal errors. We developed an algorithm to avoid and correct these errors in flowspace, rather than base space, to ensure accurate signal assessment. Our pipeline is able to determine the optimal anchor locations used to identify the exact position of a repeat sequence within a read and then quantify the size, sequence, and number of reads associated with each allele.

An additional complication of accurately quantifying the length of an STR is the increased error rates of the polymerase during DNA amplification. By evaluating the similarity of every loci between tumor and normal pairs of an individual using a metric derived from information theory, we are able to account for both the sequencing and PCR associated noise to robustly identify the genome-wide changes. Furthermore by evaluating a set of 24 MSI-High and 24 MSS samples along with their paired normal samples, we created a novel method to identify a highly informative subset of STR loci. Creating a classifier using only these highly informative loci, we are able to recapitulate the traditional MSI assay with 95% accuracy on these initial samples. This general methodology can be further extended to a model where MSI status is predicted using only the tumor sample, without the need for a control from the same individual. Due to the flexibility of this method, we are able to examine an increased number of potentially informative loci and use the total mutation load along with mutation calls in the MMR genes to provide comprehensive, multi-dimensional genomic perspective of samples.

While this novel method has been developed and tested on the Oncomine Tumor Mutation Load assay, we have created a novel general approach that can be used on the output of any semiconductor sequencing assay. The extended information derived from individual mutations and the total mutation load, along with STR aberration information allows for a comprehensive evaluation of every sample and promotes the discovery of novel genomic components important for understanding the mechanisms of immune-oncology. *For research use only

Citation Format: Aleksandr Pankov, Sameh El-Difrawy, Warren Tom, Jeffrey Conroy, Sean Glenn, Sarabjot Pabla, Carl Morrison, Fiona Hyland, Simon Cawley. A novel method for classification of microsatellite instability (MSI) using the Oncomine Tumor Mutation Load assay [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 2267.

Abstract 580: A method for estimating mutation load from tumor research samples using a targeted next-generation sequencing panel

Introduction: Tumor mutation load predicts durable benefit from immune checkpoint inhibitors in several cancer types. Existing methods to estimate tumor mutation load have large input DNA and extensive infrastructure requirements and are associated with delays due to shipping precious biopsy samples to central laboratories. Herein, we demonstrate the ability of a targeted panel with fast turn-around time to estimate mutation load from tumor research samples using low input. Methods: We developed a single sample analysis workflow to estimate mutation load (mutation count per megabase) from FFPE and fresh frozen tumor research samples. The assay utilizes a PCR-based target enrichment panel that covers 409 genes and 1.7 Mb of genomic regions. The workflow requires only 10 ng of input DNA, and enables a 2.5-day turn-around time from sample to the final report. The workflow enables < 60 minutes of hands-on time for automated library preparation and templating on a batch of 8 samples. Sequencing is performed on high throughput semiconductor sequencing platform to achieve sufficient depth (~500x coverage) and accuracy. Our analysis pipeline calls variants at ≥5% allelic frequency with parameters optimized for high accuracy. The workflow is tumor sample only, with no matched normal sample required, and germ-line variants, along with background noise, are removed through application of filters based on variants prevalent in population databases. Results: In-silico analysis demonstrated that the predicted mutation counts associated with the covered regions of the targeted panel could effectively stratify responders and non-responders to immune checkpoint inhibitors in 3 separate cohorts with high statistical significance. The assay was applied to colorectal specimens previously typed for microsatellite instability (MSI) and distinguished samples that were MSI-H (median mutation count 94.5) and MSS (median mutation count 17.8). Matched tumor-normal analyses on 7 colorectal tumor samples showed that the single tumor analysis workflow detected somatic mutations with strong correlation (r = 0.97) with tumor-normal analysis. To assess reproducibility, we compared somatic mutation count in 6 library replicates of breast cancer cell line HCC1143 and observed low relative variability (CV=9.57%) with 19.17 average mutation count. The analysis report characterizes substitution type and context of somatic mutations, and highlights mutation signatures consistent with UV damage and spontaneous deamination of 5-methyl-cytosine, as well as FFPE deamination. Conclusions: We developed a simple analysis workflow on the Ion Torrent sequencing platform with an AmpliSeq panel to estimate mutation load from FFPE and fresh frozen tumor research samples. This solution will advance research in immuno-oncology.

Citation Format: Ruchi Chaudhary, Dinesh Cyanam, Vinay Mittal, Warren Tom, Janice Au-Young, Seth Sadis, Fiona Hyland. A method for estimating mutation load from tumor research samples using a targeted next-generation sequencing panel [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 580.

Abstract 4668: Evidence for antigen-driven TCRB chain convergence in the tumor infiltrating T cell repertoire of 85 research subjects with melanoma

Introduction

T cell convergence refers to the phenomenon whereby antigen-driven selection enriches for T cell receptors having a shared antigen specificity but different amino acid or nucleotide sequence. T cell recruitment and expansion within the tumor microenvironment (TME) may be directed by responses to tumor neoantigen, suggesting that elevated T cell convergence could be a general feature of the tumor infiltrating T cell repertoire. Here we evaluate evidence for T cell convergence within tumor biopsy research samples from a set of 85 subjects with melanoma.

Methods

Total RNA from 85 tumor biopsy research samples (non-FFPE) was extracted for use in long-amplicon TCRB chain sequencing (mean amplicon length of 330bp covering CDR1, 2 and 3) via the Ion AmpliSeq Immune Repertoire Assay Plus, TCRB. To evaluate T cell convergence within each biopsy, we searched for instances where TCRB chains were identical in amino acid space (shared variable gene identity and CDR3 amino acid sequence) but had distinct nucleotide sequences owing to N-addition and exonucleotide chewback within the V-D and D-J junctions of the CDR3. To provide context, we evaluated evidence for T cell convergence with T cell repertoires derived from healthy donor peripheral blood leukocytes (PBL).

Results

Sequencing of melanoma biopsies yielded an average of 6029 clones per sample. 11 of 85 samples yielded fewer than 100 clones and were eliminated from downstream analysis. Convergent T cell receptors were identified in 68/74 (92%) of tumor infiltrating T cell repertoires having greater than 100 detected clones. The frequency of convergent rearrangements was approximately 50-fold greater in the melanoma-infiltrating T cell repertoire than healthy PBL samples (p<.001).

Conclusions

These data suggest that T cell convergence may be a common feature of the melanoma infiltrating T cell repertoire. Convergence was more frequently observed within the TME than T cell repertoires derived from healthy PBL, consistent with elevated antigen-driven T cell selection within the TME. The extent to which convergence is a feature of the TME in other cancers is not yet known. T cell receptor convergence may be driven by T cell responses to tumor neoantigen within the TME. In such case, in silico identification of convergent T cell receptors by long-amplicon sequencing may serve as an approach for rapid identification of antigen-specific T cell receptors for future therapeutic use.

For research use only.

Citation Format: Timothy J. Looney, Sean Glenn, Sarabjot Pabla, Jeff Conroy, Carl Morrison, Alice Zheng, Lauren Miller, Elizabeth Linch, Denise Topacio, Geoff Lowman, Fiona Hyland, Mark Anderson. Evidence for antigen-driven TCRB chain convergence in the tumor infiltrating T cell repertoire of 85 research subjects with melanoma [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 4668.

Estimating mutation load at 5% LOD from FFPE samples using a targeted next-generation sequencing assay

28

Background: High tumor mutation load is a biomarker which positively correlates with response to immune checkpoint inhibitors. Current methods to estimate tumor mutation load often require large amounts of DNA. Herein, we demonstrate the ability of a targeted panel to estimate mutation load from FFPE samples using low input amount of DNA. Methods: We developed a single-sample analysis workflow to estimate mutation load (somatic mutation count per Megabase (Mb)) from FFPE and fresh frozen tumor research samples. The assay utilizes a PCR-based target enrichment panel that covers ~1.7 Mb. Our workflow requires only 10 ng of input DNA, and enables a 2.5-day turn-around time from sample to the final report. The workflow enables < 60 minutes of hands-on time for automated library preparation and templating on a batch of 8 samples. Sequencing is performed using a high throughput semiconductor sequencing platform to achieve sufficient depth (~500x coverage) and accuracy. Our analysis pipeline calls variants with optimized parameters on the tumor sample only, with no matched normal sample required, and applies filters to remove germ-line. Results: An in-silico analysis using 21,000 exomes from COSMIC demonstrated the panel can achieve high sensitivity (>=93%) and specificity (>99%) necessary to stratify high and low mutation burden samples. Matched tumor-normal analyses on 14 lung and colorectal samples showed that our single tumor analysis workflow detects mutation load with strong correlation (r=0.8699) with tumor-normal analysis. Through dilution experiments on engineered control we learned that the workflow provides accurate estimates of mutation load, predicting ± 7% of expected mutation load on >=20% tumor content. To assess reproducibility, we compared mutation load in library replicates for a cohort of 8 samples (FFPE and fresh frozen tumors, engineered control, and NIST cell-lines) and observed high correlation (r=0.9906) among replicates. Conclusions: We developed a simple workflow on the Ion Torrent sequencing platform to estimate mutation load from FFPE and fresh frozen tumor research samples. This solution will advance research in immuno-oncology.

Long-amplicon TCRβ repertoire sequencing to reveal human T-cell receptor variable gene polymorphism: Implications for the prediction and interpretation of immunotherapy outcome

129

Background: Human T cell antigen receptors play a critical role in protective immune responses but are also implicated in autoimmune disease and immune-mediated adverse events during immunotherapy. The antigen specificity of the T cell receptor is determined in part by the sequence of the CDR and Framework regions encoded by the TCRB variable gene. Previous studies of population sequencing data indicate that current antigen receptor allele databases, such as IMGT, fail to capture a significant portion of human variation. Here we use long-amplicon multiplex sequencing of rearranged TCRB receptors to validate putative novel human variable gene alleles previously recovered from 1000 genomes data. Methods: TCRB rearrangements were amplified from cDNA from 85 Caucasians undergoing treatment for melanoma using AmpliSeq-based multiplex Framework 1 and Constant gene primers to produce ~330bp amplicons. Samples were sequenced using the Ion Torrent S5 530 chip to produce ~1.5M raw reads per sample. Ion Reporter was used for clonotyping and identification of variable gene sequences absent from the IMGT database. Putatively novel sequences were compared to those reported in the Lym1k database of alleles recovered from 1000 genomes data. Results: We identified 15 novel variable gene alleles that are absent from the IMGT database and result in amino acid changes to the CDR or Framework regions of the TCR. Typically, a single individual was found to be heterozygous for a novel variant, though we note two instances where multiple individuals possessed a novel variant. We also identified novel variable gene alleles that are absent from the Lym1k database, potentially due to challenges in inferring receptor alleles from short-read population sequencing studies. Conclusions: We find evidence for significant human diversity in TCRB variable gene alleles beyond what is currently represented in the IMGT database. TCRB sequencing using multiplex Framework 1 and Constant gene targeting primers is ideally suited for studying the role of TCRB polymorphism in autoimmune disease and immune-mediated adverse events during immunotherapy.

Abstract 1631: Sequencing the circulating and infiltrating T-cell repertoire on the Ion S5TM

T-cell repertoire (TCR) sequencing by next-generation sequencing (NGS) is a valuable tool for building a deeper understanding of the adaptive immune system. As immunotherapies, particularly T-cell dependent therapies, show increasing potential in treating cancer, the ability to gain a detailed, unbiased view of the TCR becomes imperative for biomarker discovery, immune response to treatment, and study of tumor microenvironments. A key question the field seeks to understand is the relationship between circulating T-cells and infiltrating T-cells at the tumor site. Here, we present a novel approach for TCR sequencing using the Ion S5 ™ sequencer which leverages simplified library construction workflows and offers a more complete characterization of the entire V(D)J region of TCRB. This method can leverage mRNA as input, minimizing requirements in starting materials and focusing sequencing to productive TCRB arrangements. This approach targets the constant (C) and the FR1 regions, minimizing the potential for primer bias and greatly increasing the phylogenetic information content compared to techniques that exclusively characterize the CDR3 domain. Our results show that the observed circulating T-cell repertoire size is approximately 2 orders of magnitude higher than the infiltrating T-cell repertoire. Accordingly, while it is difficult to fully capture the complete repertoire of circulating T-cells due to its vast diversity, we show that it is possible to reliably capture the complete infiltrating T-cell repertoire with as high as 10 samples on the Ion 530 ™ chip. Replicate sequencing runs of infiltrating T-cells offers correlation of ~0.9, indicating that the results were reproducible, and the samples were sequenced to appropriate depth. In summary, we believe that this workflow will allow researchers to more routinely characterize the infiltrating T-cell repertoire and offers the field a better understanding of the impact of repertoire diversity on tumor elimination.

Citation Format: Geoffrey Lowman, Elizabeth Linch, Lauren Miller, Denise Topacio-Hall, Timothy Looney, Alex Pankov, Yongming Sun, Xinzhan Peng, Mark Andersen, Fiona Hyland, Ann Mongan. Sequencing the circulating and infiltrating T-cell repertoire on the Ion S5TM [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 1631. doi:10.1158/1538-7445.AM2017-1631

Abstract 5363: Measuring gene expression at the tumor microenvironment: A comparison between nCounter PanCancer Immune Profiling Panel and Oncomine Immune Response Research Assay

The tumor microenvironment, especially infiltrating T lymphocytes and inflammatory molecules, is believed to be highly relevant to the tumor’s sensitivity to cancer checkpoint blockade therapy. At the same time, the exact markers that are predictive of response for each therapeutic agent are still the subject of active investigations. To address the need for better understanding of the effect of different T cell subsets, antigen presentation, and tumor killing, gene expression profiling presents an attractive means to simultaneously evaluate the tumor microenvironment and cancer cells. In this study we compare the results and performance of the nCounter PanCancer Immune Profiling Panel and the Oncomine Immune Response Research Assay, both of which are designed to measure the expression of genes indicative of an immune response and potential immune-editing activities by tumor cells. The nCounter panel detects gene expression by counting unique probes that hybridize target mRNA, while the Oncomine panel employs NGS to sequence and count reads derived from the targets. While both panels are designed to work with FFPE samples, The nCounter panel expects 100 ng of unamplified RNA, while the Oncomine panel requires only 10 ng total RNA with its AmpliSeq technology. The two panels share 254 common genes, which constitute the basis for this comparison. Across 12 cancer samples (breast, H&N, melanoma, NSCLC, and RCC), results show that the Oncomine panel offers 20% higher dynamic range, thereby providing more robust readouts about the differences among samples. More importantly, with virtually no background noise, the underlying NGS technology provides an absolute zero readout for non-expressing genes which significantly improves the sensitivity for detecting low expressing genes, as their presence can be confirmed by as few as two reads. The two technologies show moderate correlation (R ~ 0.7), with the Oncomine panel more strongly correlating with qPCR (R ~ 0.9). Finally, when clustered using all genes on the panel, only the Oncomine panel provides clear stratification of cancer types, thus allowing the panel to be used for tissue type confirmation in addition to evaluating the immune response.

Citation Format: Ann Mongan, Warren Tom, Janice Au-Young, Aleksandr Pankov, Gauri Ganpule, Fiona Hyland. Measuring gene expression at the tumor microenvironment: A comparison between nCounter PanCancer Immune Profiling Panel and Oncomine Immune Response Research Assay [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 5363. doi:10.1158/1538-7445.AM2017-5363

Abstract 3567: Sequencing the human TCRβ repertoire on the Ion S5™ System

Next-generation Sequencing (NGS) is proving an important tool in increasing understanding of the human immune system, and thereby cancer immunology. αβ-T cells are the primary constituents of human cell-mediated adaptive immunity. The antigen specificity of each αβ-T cell is encoded in the 500-600 bp transcript encompassing the variable portion of the rearranged TCRα and TCRβ subunits, which can be read via NGS in a process termed repertoire sequencing. Until now, the main challenge the field faces is the lack of a technology that can provide a contiguous read of 600 bp to minimize the complexity of designing bias-prone primers and informatics challenges of stitching short reads. Here we leverage the long read capability of Ion 530™ chip to comprehensively sequence all three CDR domains of the TCRβ chain. The Ion 530™ chip offers greater than 15 M productive reads, allowing a multiplex of 2-4 samples with sufficient coverage for most repertoire profiling studies. Initial testing with Leukocyte total RNA demonstrates that this multiplex PCR assay produced repertoires that were much more similar to data derived from 5’RACE protocol than the commonly used BIOMED2 primer set. This result suggested that the use of long reads minimizes bias by allowing targeting of less variable regions. To further assess the performance of the assay, we designed a model system of 30 plasmid controls containing common human T-cell CDR3 sequences. Each plasmid was amplified individually and sequenced to confirm the detection of a single clonal population. Analytical sensitivity of the assay and accuracy of the accompanied analysis solution were further evaluated by spiking in plasmid concentrations from 10 pg to 0.0001 pg (5 million to 50 copies) in a background of 100 ng cDNA reverse transcribed from leukocyte total RNA. Results showed the assay offers linearity over 5 orders of magnitude of decreasing input concentration. In summary, we have demonstrated a NGS workflow for TCRβ sequencing that offers multiplex flexibility on Ion S5 with sample to answer in less than 48 hours. For Research Use Only. Not for use in diagnostic procedures.

Citation Format: Denise S. Topacio-Hall, Tim Looney, Yongming Sun, Lauren Miller, Elizabeth Linch, Geoffrey Lowman, Lifeng Lin, Mark Andersen, Fiona Hyland, Ann Mongan. Sequencing the human TCRβ repertoire on the Ion S5™ System [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 3567. doi:10.1158/1538-7445.AM2017-3567

Abstract 5364: A targeted NGS solution to evaluate gene expression signature of the tumor microenvironment from 40 NSCLC FFPE and matched fresh frozen samples

Cancer cells and their surrounding non-malignant cells, including immune cells, signaling molecules, stromal and extracellular matrix, create the tumor microenvironment (TME). The composition of this TME plays important roles in tumor progression, evading growth suppressors and activating metastasis. However, the regulatory mechanism and function of each constituent remains poorly understood. With several checkpoint blockade therapy studies, the presence of PD-L1 has been reported to be a promising marker to predict positive response. Current IHC methods to measure PD-L1 are subjective and highly variable. A higher-throughput and standardized solution that can systematically measure gene expression of cells present in the TME has emerged to be a more desirable alternative. Here, we applied the OncomineTM Immune Response Research Assay to measure the expression of 395 genes in non-small cell lung cancer (NSCLC) samples from 40 matched FFPE and fresh frozen sample types. This assay leverages NGS technology to sequence and count reads derived from the original transcript. With an input requirement of 10 ng of total RNA, libraries were generated, templated on the Ion ChefTM and sequenced on the Ion S5TM System. Results showed that, despite small input amount, the expression profiles of FFPE and fresh frozen samples are highly correlated with an average correlation greater than 0.9. We selected 22 genes out of the panel to validate expression with qPCR using FFPE samples. These genes were selected to cover a range of low, medium, and high expressors per our NGS data. Again, we observed a strong correlation (R ~ 0.9) between NGS and qPCR data. Approximately 80% of the 40 samples show moderate to high expression of CD8+ T cell cytokines, IFNG and TNFa. We further found that the expression of CD8A and CD8B are highly correlated with CD4, suggesting the co-presence of both cytotoxic and helper T cells. High correlation among CD4, FOXP3, TGFB1, and IL2RA (CD25) also suggests that their expression can be used as markers for the presence of Treg cells. We conducted a differential expression analysis between a group of samples (n=8) with high percentage of surrounding and infiltrating lymphocytes and another group (n=5) with low stromal content but devoid of infiltrating lymphocytes. Interestingly, we found a large number of genes which annotated as markers for infiltrating lymphocytes (CTSS, CXCR4, CD37, SRGN, FCER1G, SAMHD1, and GZMA) are significantly up-regulated in samples with high percentage of surrounding and infiltrating lymphocytes. In summary, this study highlights the robustness of using a targeted panel to understand the composition and regulatory mechanism of the TME and tumor immune response.

Citation Format: Yuan-Chieh Ku, Warren Tom, Yongming Sun, Alex Pankov, Tim Looney, Fiona Hyland, Janice Au-Young, Ann Mongan. A targeted NGS solution to evaluate gene expression signature of the tumor microenvironment from 40 NSCLC FFPE and matched fresh frozen samples [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 5364. doi:10.1158/1538-7445.AM2017-5364

Abstract 3994: Using NGS to characterize 40 NSCLC tumor with gene expression and tumor infiltrating T cell repertoire profiling from FFPE and matched Fresh Frozen samples

The tumor microenvironment (TME) is made up of stromal cells, immune cells, signaling molecules, and blood vessels surrounding the tumor cells. It has emerged as a key factor in multiple stages of cancer progression, immune-escaping, and disease progression. The composition and activity of TME co-evolve with tumor cells and may likely affect how the cancer responses to immunotherapy. A clear understanding of the functional effect and evolution of the TME necessitates a comprehensive approach to identify key immune cells as well as to characterize signaling and inflammatory (immune-editing) activities at the tumor site. Here we are describing the use of a targeted NGS panel to detect aggregated gene expression and a novel AmpliSeqTM approach to profile the relative abundance of different T cell clones at the TME. Specifically, we measured the expression of 395 relevant genes that capture interferon and chemokine signaling, T and B cell activation, checkpoint pathway, tumor proliferation, and antigen presentation. By looking at the expression of markers specific to different effector cell types, this gene panel offers a high-level view of the composition of different lymphocyte infiltrates. Complementarily, the TCR sequencing assay provides an estimate of T cell diversity and therefore offering a different dimension of the immune response. We studied 40 NSCLC tumor samples, of which we have matched FFPE and fresh frozen specimens. With a targeted panel, we could detect expression of transcripts present as few as 50 copies in 10 ng of total RNA as input. Across 40 NSCLC samples, we were able to measure expression of many low expressing cytokines such as IL2, IL21, IL23, IFNG and TNF. We observed strong co-expression pattern among genes involved in type II interferon signaling, indicating that they’re informative of T cell activation. More specifically, we found strong correlation between CD8 expression and other T cell co-stimulatory receptors (CD28, CD80, CD86, CD40), suggesting that expression of these genes can be reliable surrogates for the protein counterparts as markers for CD8+ T cells. TCRβ was sequenced for each of the matched fresh frozen sample in duplicates. Clonotype abundance of replicates was highly correlated with each other, indicating that the assay was reliable, and the samples were sequenced to appropriated depth. We identified 2000-10,000 unique clones in each tumor sample; with the diversity index moderately correlated with the percentage of tumor infiltrating lymphocytes provided by pathology review. Together, these two assays provide a convenient means to characterize T cells and their activities in combating tumor cells, thereby offering insights into how that tumor may respond to a particular immunotherapy.

Note: This abstract was not presented at the meeting.

Citation Format: Ann Mongan, Geoffrey Lowman, Yuan-Chieh Ku, Timothy Looney, Lifeng Lin, Xinzhan Peng, Fiona Hyland. Using NGS to characterize 40 NSCLC tumor with gene expression and tumor infiltrating T cell repertoire profiling from FFPE and matched Fresh Frozen samples [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 3994. doi:10.1158/1538-7445.AM2017-3994

Comprehensive detection of ctDNA variants at 0.1% allelic frequency using a broad targeted NGS panel for liquid biopsy research

e23065

Background: Advances in non-invasive tumor biomarker research have shown that tumor cells release fragments of DNA called circulating tumor DNA (ctDNA) into peripheral blood. Somatic mutations representing the tumors could be successfully detected from isolated ctDNA, providing new potential for tumor sample assessment in addition to traditional tissue biopsy methods. However, the low amount of ctDNA in the blood, which can be less than 1% allelic frequency, presents significant challenges for reliable variant detection with NGS assays. Improvement of sequencing accuracy at low allelic frequency is a critical factor in the implementation of NGS in ctDNA liquid biopsy research. Methods: We demonstrate the technical feasibility for a sample-to-variant NGS workflow that utilizes a broad multi-gene panel to survey a comprehensive list of variants relevant to multiple tumor types for liquid biopsy research. The method includes novel library preparation and analysis reporting for Ion Torrent™ sequencing platforms. 20ng of input cell-free DNA was subjected to the library generation protocol. Prepared libraries were templated on Ion Chef™ and sequenced on Ion S5™. Results: We successfully optimized an NGS workflow that enables the simultaneous examination of more than 360 driver and resistance hotspot mutations in a single-pool assay panel, achieving high sensitivity and specificity with limit of detection at 0.1% allelic frequency. The targeted regions span genes and variants relevant to multiple tumor types for comprehensive variant detection across high-value content reviewed by industry experts and researchers. Sequencing on the Ion S5™ delivered > 95% on-target reads and uniform amplification across targeted regions with deep sequencing depth ( > 40,000x). The workflow is compatible with single or multiple pooled samples on Ion Torrent™ sequencing chips. Conclusions: We demonstrate the ability to accurately detect high-value variants implicated in multiple tumors at 0.1% allelic frequency on Ion Torrent™ NGS. (For Research Use Only. Not for use in diagnostic procedures.)

Tracking the interplay between circulating and tumor-infiltrating T cells using AmpliSeq-based Ion Torrent TCRβ immune repertoire sequencing

e14518

Background: TCRβ immune repertoire analysis by next-generation sequencing is emerging as a valuable tool for research studies of the tumor microenvironment and potential immune responses to future cancer immunotherapy. Here we describe a multiplex PCR-based TCRβ sequencing assay that takes advantage of the exceptionally low base-call error rate and long read capability of the Ion S5 530 chip. The S530 chip provides > 15M reads, allowing for sample sequencing in multiplex. The research assay utilizes AmpliSeq technology and Framework 1 and Constant gene targeting primers to provide comprehensive coverage of TCRβ CDR 1, 2 and 3 from RNA input with a hands-on time of less than 30 minutes. CDR region sequences reveal the antigen specificity of the receptor and may be genetically modified to increase TCR affinity for tumor antigen. Methods: We evaluated assay performance by sequencing TCRβ rearrangements from normal donor peripheral blood leukocyte (PBL) cDNA that had been spiked with 30 reference rearrangements taken from the literature. We then used our assay to evaluate the extent of clonal overlap between matched tumor infiltrating lymphocyte (TIL) and peripheral blood leukocyte repertoires in an individual with squamous cell carcinoma of lung. Results: Results from sequencing of spike-in reference rearrangements indicate that the assay is accurate and sensitive over 5 logs of input template amount while showing minimal amplification bias. Sequencing of matched PBL and TIL repertoires revealed that a subset of the PBL repertoire (8%) consisted of clones also found in TIL. Technical replicates showed high concordance (r > .96) in the frequency of detected clones, indicating that the results were reproducible and samples were sequenced to an appropriate depth. Conclusions: In summary, these data suggest that AmpliSeq-based multiplex PCR and Ion Torrent sequencing provide unbiased, reproducible, scalable, complete, and accurate information for immune repertoire research sequencing applications.

Verification of targeted gene expression profiling panel for identifying biomarker signatures for immunotherapy research

e23208

Background: Immunotherapy has led to an unprecedented and long-lasting response in susceptible populations. Despite the therapeutic potential of the treatment, identifying biomarkers and stratifying populations that are likely to respond has been a challenge. While gene expression profiling has previously been successfully used to stratify individuals, there exist limitations with the prevalent technologies. In particular, full transcriptome gene expression estimates use limited biological material to measure the concentrations of thousands of uninformative genes and often lacks the depth required to accurately measure expression of lowly-expressed genes. These low-expressing genes may be critical to the identification of a signature associated with susceptible population. Methods: To efficiently measure the expression of the genes potentially informative of an immunotherapy response, we developed a high-throughput targeted gene expression solution measured with our RNA Ion Oncomine™ Immune Response Research Assay panel* containing 395 genes. This panel provides information about the expression of genes involved in tumor checkpoint inhibition, as well as markers of T cell signaling pathway, interferon signaling, tumor infiltrating lymphocytes (TIL). Results: We used publicly-available TCGA data to characterize the complexities of estimating unbiased gene expression from a targeted panel and developed a solution using a new normalization procedure that allows for accurate comparisons of samples within cancer types. Furthermore, we verified that lowering the RNA input amount or changing the assay operator does not contribute to a large variation in the gene expression estimates; each only accounts for less than 10% of variance of the average gene when the assay is compared across biological samples. Conclusions: Creating a panel that achieved high reproducibility and accurate expression estimates of key immune response genes allowed us to accurately separate high and low TIL samples within squamous and adenocarcinoma samples, emphasizing the utility of the panel to biomarker immunotherapy research. *For Research Use Only. Not for use in diagnostic procedures.

Estimating tumor mutation burden using next-generation sequencing assay

e14529

Background: High tumor mutation burden is a promising biomarker shown in some cancer types to predict positive response to immune checkpoint inhibitors. We show the ability of a targeted cancer research panel to estimate tumor mutation burden per megabase. Methods: We present a single sample analysis workflow for estimating tumor mutation burden from FFPE research samples. Our assay utilizes a PCR-based target enrichment panel that interrogates 409 known key cancer genes covering ~1.7 megabase of genomic space. Our customized workflow requires only 20 ng of input DNA, and enables a 2 day turn-around time from sample to the result. The ease of our workflow enables less than 60 minutes of hands-on time for automated library preparation and templating on a batch of 8 samples. Next-generation Sequencing is performed using high throughput semiconductor sequencing platform to achieve sufficient depth (~500x coverage) and accuracy. Our custom analysis pipeline calls variants with optimized parameters on the tumor sample only, with no matched normal sample required, and applies filters to remove germ-line variants and background noise. Results: Through in silico analysis performed on The Cancer Genome Atlas (TCGA) data we demonstrate that the panel achieves high sensitivity ( > 90%) and specificity ( > 95%) necessary to separate high and low mutation burden samples. Our workflow provides clear separation between allele ratio of somatic and germ-line variants. Our filters consistently eliminate ~98% of germ-line variants from the set of all variants called in single sample analysis workflow. Evidence from tumor-normal analyses on matched tumor and normal samples suggests that our single sample analysis, on the tumor sample only, detects somatic mutations with high sensitivity and specificity with residual of < 3% germ-line variants. Our pipeline identifies mutational signatures consistent with specific mechanisms such as spontaneous deamination of 5-methyl-cytosine, as well as base-damage from FFPE processing. Conclusions: A simple workflow has been developed on the Ion Torrent sequencing platform to estimate per megabase somatic mutational burden from a single tumor FFPE sample. This solution can help advance research in immuno-oncology.

Principles and Recommendations for Standardizing the Use of the Next-Generation Sequencing Variant File in Clinical Settings

A national workgroup convened by the Centers for Disease Control and Prevention identified principles and made recommendations for standardizing the description of sequence data contained within the variant file generated during the course of clinical next-generation sequence analysis for diagnosing human heritable conditions. The specifications for variant files were initially developed to be flexible with regard to content representation to support a variety of research applications. This flexibility permits variation with regard to how sequence findings are described and this depends, in part, on the conventions used. For clinical laboratory testing, this poses a problem because these differences can compromise the capability to compare sequence findings among laboratories to confirm results and to query databases to identify clinically relevant variants. To provide for a more consistent representation of sequence findings described within variant files, the workgroup made several recommendations that considered alignment to a common reference sequence, variant caller settings, use of genomic coordinates, and gene and variant naming conventions. These recommendations were considered with regard to the existing variant file specifications presently used in the clinical setting. Adoption of these recommendations is anticipated to reduce the potential for ambiguity in describing sequence findings and facilitate the sharing of genomic data among clinical laboratories and other entities.

Abstract B17: Validation of targeted gene expression profiling panel for identifying biomarker signatures of immunotherapy responders

Cancer immunotherapy has led to an unprecedented, long-lasting response in populations susceptible to the therapies. Despite the therapeutic potential, identifying biomarkers and stratifying populations that are likely to respond has been a challenge. Gene expression profiling has previously been successfully used to stratify individuals based on survival and treatment characteristics, but there exist limitations with the prevalent technologies. In particular, full transcriptome gene expression estimates use limited biological material to measure the concentrations of tens of thousands uninformative genes and often lack the depth required to accurately measure expression levels of lowly-expressed genes. These genes may be critical to the identification of a signature associated with immunotherapy responders. To efficiently measure the expression of the key genes potentially informative of an immunotherapy response, we developed a high-throughput targeted gene expression solution measured by our RNA Ion Oncomine Immune Response Profiling panel* containing 395 genes. This panel provides information about the expression of genes involved in tumor checkpoint inhibition (including CTLA4, PD-1, PD-L1, OX-40, 4-1BB, TIM3, LAG3) and other targets such as CSF1R, and IDO1, as well as additional markers of T cell signaling pathway, interferon signaling, and markers of tumor infiltrating lymphocytes (TIL). We used publicly available TCGA data to demonstrate the need and develop a solution for a new normalization procedure that allows for accurate comparisons of samples within various cancer types. Furthermore, we verified a linear and unbiased estimate of fold change in our assays across mixing concentrations of a cell-line titration experiment. Finally, by achieving a high correlation (r &gt; .99) of technical replicates, along with robust expression estimation even at low input amounts (10 ng RNA), our panel offers a valuable solution for biomarker research in cancer immunotherapy.

*For research use only. Not for use in diagnostic procedures.

Citation Format: Aleksandr Pankov, Yongming Sun, Yuan-Chieh Ku, Warren Tom, Jianping Zheng, Timothy Looney, Janice Au-Yong, Fiona Hyland, Ann Mongan. Validation of targeted gene expression profiling panel for identifying biomarker signatures of immunotherapy responders. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2016 Oct 20-23; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2017;5(3 Suppl):Abstract nr B17.

Somatic mutation burden in cancer samples determined by targeted next generation sequencing

15

Background: High somatic mutation burden in tumor tissues is associated with the presentation of neoantigens that promote immune responses particularly in the context of immune checkpoint therapies. Herein, we characterize the ability of targeted cancer research panels to generate estimates of somatic mutation burden. Methods: Somatic mutation data from > 8000 cancer samples obtained from The Cancer Genome Atlas (TCGA) was curated and standardized, and the number of single nucleotide variants (SNVs) in exonic regions of each sample determined. Next, the number of SNVs associated with target regions of two Ion AmpliSeq cancer panels (Oncomine Comprehensive Assay [OCA, 146 genes, 0.35 MB]; Comprehensive Cancer Panel [CCP, 409 genes, 1.7 MB]) was likewise determined and the frequency of mutation counts in the exome and the panel target regions was compared. Mutation counts of samples containing truncating mutations in mismatch repair (MMR) and other DNA repair genes were characterized. A facile workflow with less than 60 minutes of hands-on time was developed to estimate mutation counts for a batch of 8 samples using the Ion Chef for automated library preparation and templating followed by sequencing on the Ion S5. Results: The sensitivity of targeted panels in estimating somatic mutation burden was positively correlated with panel size. The area under the Receiver Operating Characteristic (ROC) curve showed that CCP had > 90% sensitivity and > 95% specificity to differentiate high and low mutation burden based on informatics analysis of TCGA data. As expected, truncating mutations in MMR genes were associated with higher somatic mutation counts in colorectal tumor tissue. Using data generated from OCA and CCP, we characterized a set of filters that provided a good estimate of somatic mutation counts when applied to a tumor-only workflow. Conclusions: A simple workflow was developed on the Ion Torrent sequencing platform to estimate somatic mutation burden in cancer samples. The methods described herein will help advance research in immuo-oncology.

Abstract 2042: A next-generation sequencing-based sample-to-result pharmacogenomics research solution enables both SNV and CNV detection at once

Pharmacogenomics (PGx) is the study of genetic variations in terms of their response to drugs. Variations in gene sequence or copy number may result in complete loss of function, partial decrease or increase in enzyme activity, or an altered affinity for substrates, which may in turn significantly impact a drug's efficacy. PGx studies are increasing in significance as precision medicine is becoming a reality in standard practice. Different technologies have been developed to measure the sequence variation and copy number variation (CNV) in the PGx genes. Among them, a complete sample-to-result PGx workflow solution using the QuantStudio™ 12k Flex Real-Time PCR System is the most notable high throughput solution and has broad adoption by advanced PGx laboratories. Both PGx SNP/INDEL genotyping assays on OpenArray™ plates and copy number analysis on 384-well plates can be performed on the QuantStudio™ 12k Flex System. Integrated analysis software translates genotyping and copy number assay results into star allele genotypes for ease of interpretation. Recently we have developed a next generation sequencing (NGS) based PGx research solution with increased flexibility on the assay targets and combined detection of SNP/INDEL genotyping and CNV using Ion AmpliSeq™ technology for low to medium throughput laboratories. With a highly multiplexed PGx research panel, we can profile a set of 136 genetic markers in 40 known PGx related genes and CYP2D6 copy number variation in a single reaction using Ion Torrent™ semiconductor sequencing. The number of genetic markers can be customized easily based on the user need. To systematically compare these two end-to-end PGx workflows, we collected buccal swab samples from 20 individuals and performed both QuantStudio™ based assays and PGM™ based Ion AmpliSeq™ PGx research assay on them. Both systems generated high quality results. Compared with OpenArray™ plate genotyping results and 384-plate CYP2D6 copy number assay results from the QuantStudio™ system, the Ion AmpliSeq™ PGx research solution demonstrated &gt;99.9% genotyping concordance, 100% CYP2D6 gene CNV concordance, &gt;99.7% reproducibility, &lt;0.2% no-call rate. The Ion AmpliSeq™ PGx solution enables flexible and integrated SNV & CNV detection for both standard genotyping practice and sophisticated exploratory research needs.

Citation Format: Melvin S. Wei, Zhoutao Chen, Shann-Ching Chen, Manimozhi Manivannan, Emily Zeringer, Sunali Patel, Toinette Hartshorne, Guoying Liu, Fiona Hyland, Mark Andersen. A next-generation sequencing-based sample-to-result pharmacogenomics research solution enables both SNV and CNV detection at once. [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 2042.

Abstract 5272: Cloud-based informatics enables the design and analysis of massively multiplex custom gene fusion panels for next-generation sequencing on FFPE RNA samples

Gene fusions, a combination of two genes, comprising their coding and/or regulatory sequences, are caused by structural rearrangements in DNA or in RNA transcripts. Many gene fusions are strong driver mutations in neoplasia, and are important in understanding basic biology, interaction with targeted therapy, and research into risk stratification and outcomes.

Next-generation sequencing enables sensitive, specific and precise detection of particular fusion isoforms for defined gene pairs. Massively multiplex Ampliseq gene fusion assays enable enrichment of fusion transcripts using as little as 10 ng of RNA extracted from FFPE samples. Sequencing on Ion Torrent instruments reveals the full sequence of the gene fusion, for precise definition of the breakpoint and the expressed exons or promoter regions of both genes.

We developed cloud-based software to support the design of a custom Ampliseq gene fusion panel, comprising 1 to 1,000 fusion isoform assays and any gene expression assays for normalization.

We extensively mined the scientific literature on fusions and the COSMIC database to identify over 1000 fusion isoforms. We rigorously curated this data using automated and manual methods, including mapping, confirmation and correction of reported sequence to obtain genomic coordinates, identification of breakpoints, annotation of exon junctions, and selected wet lab testing. We created a database containing over 1000 high quality curated and annotated fusion isoforms, including 70 ALK, 60 RET, 26 ROS1, and 21 NTRK1 fusions. We designed Ampliseq primer pairs for each of these fusions using advanced assay design and pooling algorithms, such that all fusion and gene expression assays can be multiplexed into 1 or 2 compatible pools.

Assays can be selected by gene or gene pair; detailed information about each assay selected includes isoform, genes, exon numbers, and links to COSMIC and to relevant publications.

We developed cloud-based analysis software to analyze the BAM file resulting from amplification and sequencing of custom Ampliseq fusion panels on an Ion Torrent sequencer. This analysis leverages the rich annotation information from the assay design. The reads are mapped to a custom reference sequence tailored to the custom Ampliseq fusion assay, and applying an optimized algorithm to select confidently mapped reads based on read length and overlap with each gene of the gene pair based on the reference and annotated breakpoint. Gene fusions are detected based on the total number of fusion reads and optionally frequency, and on the properties of those reads. Software QC steps for total number of mapped reads, number of reads for gene expression controls, and elimination of cross-talk artifacts result in a highly sensitive and specific detection of fusions, with LOD below 1%. Fusion results for any or all samples can be viewed, annotated, filtered, and visualized, and exported.

Citation Format: Fiona Hyland, Rajesh Gottimukkala, Efren Ballesteros, Heinz Breu, Yuandan Lou, Scott Myrand, Michael Hogan, Kelli Bramlett, Guoying Liu, Seth Sadis. Cloud-based informatics enables the design and analysis of massively multiplex custom gene fusion panels for next-generation sequencing on FFPE RNA samples. [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 5272.

Abstract 3622: Complete workflow for detection of low frequency somatic mutations from cell-free DNA using Ion Torrent™ platforms

Research detecting of somatic mutations in circulating cell-free DNA (cfDNA) using research blood samples from subjects previously diagnosed with cancer provides a potential non-invasive approach to monitor cancer status and evaluate cancer evolution in the future. However, most of the existing mutation detection methods show insufficient sensitivity to detect cfDNA mutations since only small amount of mutant gene fragments, derived from tumor cells, is present in a large amount of normal circulating DNA background.

We demonstrated a complete workflow that includes blood collection, cfDNA isolation, library preparation, sequencing, and data analysis to enable detection of rare DNA variants in blood plasma samples. Blood samples were collected using Streck™ DNA tubes followed by plasma preparation and cfDNA isolation using MagMAX™ Cell-Free DNA Isolation Kit. Library preparation was performed using Oncomine™ lung cfDNA kit. Barcoded libraries were pooled and sequenced on Ion Torrent™ Next Generation Sequencing Platforms. Sequencing data was analyzed in Torrent Suite™ using variantCaller-cfDNA plugin. ∼150 biomarkers relevant to non-small cell lung cancer were interrogated in one sequencing run.

We demonstrated detection sensitivity at 0.1% frequency using engineered mutants that were spiked into control DNA samples. The workflow was tested on a set of research samples from matched tumor FFPE and blood plasma collected from research subjects with non-small cell lung cancer (NSCLC). About 1 mL of plasma was processed using the workflow described above. RecoverAll™ Multi-Sample RNA/DNA Isolation Workflow was used to isolate DNA from FFPE samples, followed by library preparation, sequencing and data analysis using the same workflow described above. Summary of variant calls from matched cfDNA and FFPE tumor samples are presented here. Results indicate high sensitivity of the workflow and expected levels of concordance between variants detected in the two types of research samples.

In this study, we developed a highly sensitive and reliable research workflow to detect rare somatic mutations in circulating cfDNA samples. Significant overlapping of mutations discovered in FFPE tumor and cfDNA samples suggests that this workflow may be used to monitor tumor dynamics in NSCLC and potentially other tumors in the future.

Disclaimer: For research use only. Not for use in diagnostic procedures.

Citation Format: Jian Gu, Dumitru Brinza, Ann Mongan, Richard Chien, Dalia Dhingra, Fiona Hyland, Kelli Bramlett. Complete workflow for detection of low frequency somatic mutations from cell-free DNA using Ion Torrent™ platforms. [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 3622.

Abstract 3941: Novel biomarkers and multiplexed NGS to stratify FFPE NSCLC by tumor infiltrating lymphocytes and histopathology phenotypes

There is increasing evidence supporting the association of tumor infiltrating lymphocytes (TIL) and drug sensitivity of solid tumors. In particular, primary and meta-analyses have reported a positive correlation between TIL level and outcome in advanced non-small cell lung cancer (NSCLC) treated with checkpoint inhibitors such as PD-1 and PD-L1. Recent trials with immunotherapies have started including TIL assessment in the study protocol in recognition of this metric as a predictive and prognostic biomarker. TIL levels are typically quantified by visual assessment of H&E-stained tumor sections. While this approach is generally accepted as the standard, there is an increased recognition that visual assessment of H&E-stained slides lacks precision and is relatively subjective (Salgado R, 2015; Schalper KA, 2015). Furthermore, as investigators are often also interested in measurements of additional biomarkers such as IFNg as well as the drug targets, a gene panel approach offers a convenient solution to objectively quantify expression levels of these informative markers. Here we report the discovery and verification of a unique gene expression signature that is capable of stratifying FFPE samples of NSCLC tumors by TIL levels and histopathology phenotypes (adenocarcinoma vs. squamous cell carcinoma). Gene expression was measured by an RNA Ion AmpliSeq Gene Expression research panel* containing 200 assays. Each research sample was measured with replicates at library generation step and sequencing step. Technical replicates were found to have &gt;0.99 correlation among each other. Assays on the panels were also found to be robust with respect to low input amount (1-10 ng RNA). *For Research Use Only. Not for use in diagnostic procedures.

Citation Format: Ann Mongan, Sophie Rozenzhak, Geoffrey Bien, David Chi, Hiroyoshi Nishikawa, Fiona Hyland, Jim Godsey. Novel biomarkers and multiplexed NGS to stratify FFPE NSCLC by tumor infiltrating lymphocytes and histopathology phenotypes. [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 3941.

Abstract 5267: Gene fusion database to create custom panels: Enabling detection of fusion transcripts and gene expression assays

Gene fusions play an important role in tumorigenesis and are increasingly recognized as important entities for the diagnosis and treatment of hematological malignancies and solid tumors. Fusion events generate a hybrid mRNA transcript comprising sequence from multiple otherwise distinct genes. Oncogenic fusion events often involve tyrosine kinases or transcription factors, leading to aberrant growth signaling, making these events potentially attractive drug targets. For instance, targeted therapies such as known tyrosine kinase inhibitors are currently approved to treat ALK fusion positive Non-Small Cell Lung Carcinoma (NSCLC) patients. Detection of known gene fusion events is an important part of genomic characterization which can inform patient diagnosis. Current methods for fusion detection include chromosome banding analysis (CBA), fluorescence in situ hybridization (FISH), and reverse transcription polymerase chain reaction (RT-PCR). New developments in next-generation sequencing (NGS) enable the efficient and simultaneous assessment of multiple gene fusion targets with high sensitivity. To enable researchers to design their own custom panels and assess a set of gene fusions of interest, we developed a comprehensive RNA gene fusion database. Oncology researchers now have the capability to create custom panels from this comprehensive database which includes over 1,000 well annotated and optimized gene fusion assays and over 20,000 gene expressions assays.

To build this comprehensive gene fusion database, we identified breakpoint information for 1,178 well annotated fusions described in publications and in the COSMIC and NCBI databases. We prepared a target RNA sequence for each breakpoint using transcript sequences from the Ensembl database. We used a proprietary primer designer to generate candidates for each fusion target amenable to the AmpliSeq™ product line requirements. Quality control was performed throughout the design process to identify the best primer set for each target, to avoid primers overlapping common germline SNPs, potential primer/primer or primer/amplicon interactions, or off-target or wild-type amplifications.

With this comprehensive database we provide a complete range of solutions available on ampliseq.com. Making use of the AmpliSeq™ technology, researchers now have the capability to create their own custom fusion panel and place the order within an hour. These custom panels are used with AmpliSeq™ Library reagents and Ion Torrent™ sequencing platforms for targeting next-generation sequencing. The analysis solution is provided through the Ion Reporter™ (IR) software package. Custom fusion panel workflows in IR are used to analyze sequencing data coming from the custom panels, which includes visualization of fusion transcripts and gene expression levels in a heat map feature.

Citation Format: Efren Ballesteros-Villagrana, Jeoffrey Schageman, Kelli Bramlett, Paul Williams, Scott Myrand, Guoying Liu, Fiona Hyland, Seth Sadis. Gene fusion database to create custom panels: Enabling detection of fusion transcripts and gene expression assays. [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 5267.

Abstract 3959: Detection of somatic mutations at 0.1% frequency from cfDNA in peripheral blood with a multiplex next-generation sequencing assay

Background:

Effective blood screening for tracking of recurrence and resistance of tumors may improve outcomes in the future. Research studies suggest that virtually all tumors carry somatic DNA mutations, and these may serve as biomarkers that also can be tracked in blood. One of the sources containing tumor DNA in blood is circulating cell-free DNA (cfDNA). Tumor DNA comes from different tumor clones, and its abundance in plasma can be very low at critical stages such as early recurrence or development of resistance. Hence, there is great interest in being able to detect mutation biomarkers at very low frequency from cfDNA for detection and characterization of tumor clones.

Method:

We present a research use only analysis workflow for peripheral monitoring that enables detection of low frequency DNA variants in blood.

We developed an analysis algorithm that models errors accumulated during amplification and sequencing, and accurately reconstructs sequence of original DNA molecules based on multiple next generation sequencing reads. The reads contain genomic sequence and an adaptor that allows identification of reads originated from the same DNA molecule. We then developed a variant calling method that uses accurately reconstructed sequences to enable sensitive and specific detection of somatic mutations to 0.1% allele ratio. We demonstrate the analysis in control and archived cfDNA research samples.

We used a next generation sequencing assay that allows interrogation of ∼150 biomarkers relevant in lung from COSMIC and Oncomine™ databases, and de-novo variant detection at ∼1,700 genomic positions in 11 genes implicated in non-small cell lung cancer (NSCLC).The assay delivers &gt;95% on target reads and highly uniform amplification across targeted cfDNA molecules.

Results:

We tested the limits of variant detection in controlled dilution series and in cfDNA. First, we diluted engineered AcroMetrix™ Oncology Hotspot Control plasmids into background GM24385 genomic DNA down to 0.1% frequency, and then fragmented into fragments with average size of 170bp. The Acrometrix sample contains ∼45 common tumor mutations interrogated by our assay. Next, we used 0.1% Horizon's (HD780) cfDNA reference sample that contains 8 mutations at our hotspot positions including two large insertion and deletion variants of size &gt;10bp. Finally, we performed analytical verification of variant detection performance in cfDNA using a dilution series of normal blood samples.

We achieved &gt;95% sensitivity with &gt;20ng input DNA and &gt;90% sensitivity with ∼20ng input DNA and &lt;1 false call per sample for variants in hotspot positions present at frequency 0.1%.

Conclusions:

This new lung cfDNA analysis workflow may facilitate researchers to study relevant NSCLC biomarkers at 0.1% frequency in cfDNA. Analysis is compatible with lower frequency variant detection, but will require higher input DNA amount and higher sequencing coverage.

Citation Format: Dumitru Brinza, Ann Mongan, Richard Chen, Dalia Dhingra, Jian Gu, Janice Au-Young, Fiona Hyland, Kelli Bramlett. Detection of somatic mutations at 0.1% frequency from cfDNA in peripheral blood with a multiplex next-generation sequencing assay. [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 3959.

Extensive sequencing of seven human genomes to characterize benchmark reference materials

The Genome in a Bottle Consortium, hosted by the National Institute of Standards and Technology (NIST) is creating reference materials and data for human genome sequencing, as well as methods for genome comparison and benchmarking. Here, we describe a large, diverse set of sequencing data for seven human genomes; five are current or candidate NIST Reference Materials. The pilot genome, NA12878, has been released as NIST RM 8398. We also describe data from two Personal Genome Project trios, one of Ashkenazim Jewish ancestry and one of Chinese ancestry. The data come from 12 technologies: BioNano Genomics, Complete Genomics paired-end and LFR, Ion Proton exome, Oxford Nanopore, Pacific Biosciences, SOLiD, 10X Genomics GemCode WGS, and Illumina exome and WGS paired-end, mate-pair, and synthetic long reads. Cell lines, DNA, and data from these individuals are publicly available. Therefore, we expect these data to be useful for revealing novel information about the human genome and improving sequencing technologies, SNP, indel, and structural variant calling, and de novo assembly.