Break Through the NGS Interpretation Bottleneck To Improve Patient Outcomes

Unlocking insights with NGS tertiary analysis software Navigating the somatic oncology interpretation bottleneck NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 2 Table of contents 3 Introduction 4 The burden of manual interpretation 5 Benefits of software-enabled variant interpretation 14 Considerations for software interpretation 17 Summary 18 References NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 3 Introduction The landscape of genomic research and clinical diagnostics in oncology has witnessed a remarkable trajectory over recent decades, in large part due to the advent of next-generation sequencing (NGS). Tests like comprehensive genomic profiling (CGP), a pan-cancer NGS assay that broadly assesses hundreds of actionable cancer biomarkers and genomic signatures. These sweeping assays are providing deep insights into the genomic drivers of cancer that have fueled groundbreaking targeted therapies that have been shown to improve patient outcomes, while exponentially generating more data.1-3 Amidst this wealth of genomic information lies a critical challenge— the accurate and time-efficient interpretation of complex NGS data in the context of oncology. Performing variant interpretation and report generation manually is a labor-intensive and time-consuming process, typically taking between 7-8 hours that can impact clinical decision making.4 A survey analyzing 170 responses from oncologists revealed that most respondents considered one week (46%) or two weeks (52%) an acceptable turnaround time to receive NGS results, yet 37% typically wait three or more weeks.4 Community oncologists who waited more than three weeks for results were more likely to initiate nontargeted treatment, contravening professional clinical guideline recommendations. This study demonstrates an urgent need to speed up variant interpretation and translate test results into actionable clinical decisions. Using specialized software that facilitates variant interpretation and reporting can significantly improve a lab’s efficiency, reducing the time and effort required to as little as ~30 minutes for faster clinical decision making (Figure 1).4 Designed for researchers and clinicians involved in the delivery of NGS-based somatic oncology testing, this eBook delves into the complexities of interpretation and highlights the critical need for innovative solutions to bridge the gap between raw data and actionable insights. of oncologists currently wait ≥ 3 weeks for NGS results4 37% Figure 1: Time to report with interpretation software—Software significantly reduces the time required to generate a report by automating complex processes, increasing efficiency, and reducing human errors to enable timely clinical decision making. 7-8 hours ~30 minutes Manual Software NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 4 Figure 2: The informatics interpretation bottleneck—Prioritizing pertinent variants, staying up to date with growing knowledge, updating past interpretations, standardizing interpretation across limited trained personnel, and other challenges results in significant increases in turnaround times per report and present a bottleneck for the typical state of manual, high-touchpoint methods of data interpretation. The burden of variant interpretation Oncology testing using NGS, from targeted to comprehensive genomic panels, exomes, and genomes, generates massive amounts of data that can represent a significant burden for labs needing to scale data analysis and interpretation. As testing volumes increase, so does the pressure to deliver meaningful and consistent results in a timely manner. Therefore, finding a balance between high-quality data interpretation and managing operational efficiency becomes increasingly crucial. Variant interpretation and reporting, coupled with an ever-increasing demand for testing, have the potential to create a significant bottleneck to lab operations (Figure 2). “If you are generating this massive amount of data, you need to interpret this in a meaningful way. With these thousands or even millions of different variants, even in intergenic regions, it’s very hard nowadays to reliably interpret them in a way that is clinically, directly applicable. Variant interpretation is still heavily based on human resources and manual curation. We need to come to a point where more and more of this part [of the workflow] can be automated to some extent.”6 Albrecht Stenzinger Director, Center for Molecular Pathology, University of Heidelberg Significant turnaround time Lack of trained personnel Limited ability to prioritize variants Difficulty staying current with expanding knowledge Arduousness of updating past interpretations Lack of standardization across personnel NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 5 Benefits of software-enabled variant interpretation Data analysis for NGS typically includes three steps: primary, secondary, and tertiary analysis. After libraries have been prepared and sequenced, Real-Time Analysis (RTA) software on board the sequencing system performs primary analysis and provides base calls and associated quality scores. Next, sequencing reads are aligned and assembled, providing the full sequence for a sample, and then variants are called, enabling variant calling at the secondary analysis step. This step typically reveals thousands to millions of variants, which are then interpreted in the final step of NGS data analysis called tertiary analysis (Figure 3). Call variants Secondary analysis Tertiary analysis Prepare libraries Sequence Analyze Primary analysis Call bases Interpret variants Perform quality control Annotate and prioritize Classify and interpret Generate report Pulling out relevant, meaningful variants from the thousands of possibilities can be hampered by a time- and labor-intensive variant interpretation workflow that involves four steps: variant quality control and filtering, annotation and prioritization, classification and interpretation, and report generation (Figure 4). Figure 3: Standard NGS workflow—Library preparation and sequencing are followed by data analysis, which typically includes base calling, variant calling, variant interpretation and reporting. Figure 4: Variant interpretation workflow—Variant interpretation, also referred to as tertiary analysis, involves, at a high level, four steps. NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 6 Variant QC Comprehensive quality control (QC) is a fundamental step for NGS data analysis and should be performed before the data are used for interpretation.7 It is important to determine if the coverage of the tested genes is sufficient to detect or rule out specific mutations that could be critical for diagnosis or treatment recommendations. Depending on the variant class, different QC metrics are important. For example, the number of paired and split reads can help ensure the quality of detected fusions. A state-of-the-art interpretation software imports variant calling data (VCF files) and associated QC metrics directly from secondary analysis software and can enable several important features for variant QC: • Sorting and filtering of variants based on multiple parameters, including variant allele frequency (VAF), variant quality scores (QUAL), strand bias (SB), and number of paired or split reads8 • Setting thresholds that automatically highlight certain metrics as PASS or FAIL • Visualizing raw and other data to assist with QC via coverage plots and VAF plots • Flagging sequencing artifacts, for example, by inspecting a data set of samples accumulated in a laboratory • Reviewing aligned sequencing reads using The Integrative Genome Viewer (IGV) or other genome browser for visual inspection of called variants9 Advantages of software-assisted variant QC include: Time savings resulting in increased efficiency and capacity Accuracy improvements leading to reduced risk of reporting false positives Consistency of results among analysts executing a standardized workflow NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 7 Variant annotation After filtering out variants not passing QC, identified variants are annotated with functional information for interpretation. Genomic databases that are used to facilitate variant interpretation by cataloging whether a variant has previously been reported as associated with disease or a drug response are called “knowledge bases.” These knowledge bases, trusted and adopted by clinicians and researchers worldwide, contain varying information ranging from biological and functional data for translational research to clinical applications, eg, the OncoKB precision medicine knowledge base from Memorial Sloan Kettering Cancer Center, CIViC, and The Jackson Laboratory Clinical Knowledgebase (JAX-CKB) (Table 1).9,10 Variants that are not yet described in knowledge bases can be evaluated with the help of computational tools aiming to predict variants’ impact on protein function, eg CADD, REVEL, SIFT, SpliceAI, and others. Several such tools use artificial intelligence (AI) algorithms to enhance their performance.11–13 Multiple tools are typically used simultaneously for variant annotation to increase confidence in the results. Using knowledge bases and other data sources while automating the querying process streamlines variant annotation, enabling rapid and systematic retrieval of information. Annotation functionalities offered within a commercial interpretation software can include: • Access to multiple sources and knowledge bases in one view • Updates to annotation sources at a regular cadence • Prediction of functional impact of variants using computational tools • Ability to sort, filter, and prioritize variants based on annotated information • Data visualization, eg, map known pathogenic variants on protein domains By simultaneously accessing various knowledge bases and performing regular updates, annotation features within an interpretation software offer several key benefits: Comprehensive insights, providing thorough information on variant associations with diseases and therapies for larger tests by integrating the breadth of coverage of multiple sources Accurate information, ensuring trusted results with additional QC by comparing variant records between different sources side by side Novel variant interpretation, going beyond known variants by harnessing multiple functional predictors to decipher the impact of novel variants on protein function Up-to-date insights, keeping pace with expanding knowledge, benefitting from software–enabled new information alerts Increased efficiency, reducing the time to results by focusing on the interpretation of variants pertinent to the case NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 8 Table 1: Commonly used knowledge bases and other sources Population databases Update frequency Genome Aggregation Database (gnomAD) Variable International Genome Sample Resource Variable Cancer databases The Cancer Genome Atlas Program (TCGA) Variable Catalogue of Somatic Mutations in Cancer (COSMIC) Four times a year Clinical Interpretation of Variants in Cancer (CIViC) Daily Oncology Knowledge Base (OncoKB) Approx. monthly The Jackson Laboratory Clinical Knowledgebase (JAX-CKB) Weekly ClinVar Monthly National Library Clinical Trials Daily Online Mendelian Inheritance in Man (OMIM) Daily Scientific literature PubMed Daily Functional impact prediction tools PhyloP Variable PhastCons Variable SpliceAI Variable VarSEAK Variable NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 9 Variant interpretation After variant annotation, variants are classified into one of five categories with respect to their pathogenicity (also called oncogenicity in the context of cancer). This categorization is based on the level of evidence that a variant is oncogenic following established guidelines (Table 2). After classification, typically only oncogenic and likely oncogenic variants are further considered. At the next step, variants are assigned to a tier based on their clinical significance and actionability, according to published standards and guidelines. To determine a variant’s clinical significance, the variant needs to be evaluated for: • Association with sensitivity, resistance, or toxicity to a specific therapy • Inclusion criterion for clinical trials • Association with disease prognosis • Association with cancer diagnosis It is common for labs to skip the variant classification step and determine actionability right away for well-studied variants for which information on clinical significance is readily available. Interpretation and reporting of variants associated with increased hereditary risk of cancer is an additional step in the variant interpretation process that labs may consider implementing. Variant interpretation in this case would follow established guidelines for pathogenicity of germline variants.14 An optimized variant interpretation solution streamlines the application of various guidelines, enabling rapid and high-throughput variant classification and tier assignment, which is particularly valuable when dealing with a large volume of variants. Also, software can store and reuse variant interpretations for new reports, accelerating data interpretation. Variant interpretation functionalities offered within a state-of-the-art interpretation software include: • Precalculation of variant classification and actionability tiers • Enablement of guided manual interpretation based on the latest guidelines • Customization of variant tiers and other aspects of variant interpretation, eg, to meet regional needs • Storage and reuse of variant interpretations for new reports • Assignment and tracking of user actions NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 10 Consistent classification, producing standardized and reproducible results that reduce subjectivity across analysts by applying predefined algorithms and guidelines, eg, oncogenicity calculators In the context of clinical decision making, the features of a modern interpretation software minimize the laboratories’ burden and offer value through: Ease of maintenance, with alerts about new information and full history of interpretation, review, and approval actions for previously interpreted variants Localized content, addressing regional needs with customizable tiering categories, local clinical trials, and the ability to translate to local languages Extensive time savings, from precalculating variant classification and reusing past interpretations, freeing up analysts’ time to focus on more unique and complex cases NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 11 Standardization of somatic variant classifications per Belgian ComPerMed15 • Pathogenic • Likely pathogenic • Variant of uncertain significance (VUS) • Likely benign • Benign Oncogenicity/pathogenicity guidelines Actionability guidelines Joint consensus recommendations of the Association for Molecular Pathology (AMP), American Society of Clinical Oncology (ASCO), and College of American Pathologists (CAP)16 Tier 1: Variants with strong clinical significance Tier 2: Variants with potential clinical significance Tier 3: Variants of unknown clinical significance Tier 4: Variants deemed benign or likely benign Joint recommendations of Clinical Genome Resource (ClinGen), Cancer Genomics Consortium (CGC), and Variant Interpretation for Cancer Consortium (VICC)8 • Oncogenic • Likely oncogenic • VUS • Likely benign • Benign European Society of Medical Oncology (ESMO) Scale for Clinical Actionability of Molecular Targets (ESCAT)17 ESCAT Tier I: Alteration–drug match is associated with improved outcome in clinical trials ESCAT Tier II: Alteration–drug match is associated with antitumor activity, but the magnitude of the benefit is unknown ESCAT Tier III: Alteration–drug match suspected to improve outcome based on clinical trial data in other tumor types or with similar molecular alteration ESCAT Tier IV: Preclinical evidence of actionability ESCAT Tier V: Alteration–drug match is associated with objective response, but without clinically meaningful benefit ESCAT Tier X: Lack of evidence for actionability Joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP)14 • Pathogenic • Likely pathogenic • Uncertain significance • Likely benign • Benign Food & Drug Administration (FDA)18 Level 1: Companion diagnostics Level 2: Cancer mutations with evidence of clinical significance Level 3: Cancer mutations with potential clinical significance Table 2: Published guidelines on interpreting somatic oncology variants from professional societies NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 12 Generate report The last step of NGS tertiary analysis involves report generation, where the results of interpretation are compiled into a comprehensive and informative document (Figure 6). Figure 6: Typical report generated with commercial interpretation software—Results are organized into a concise structured format that should be easy to understand for an oncologist. Common report sections include: (1) Case, patient, and sample information, providing key details to uniquely identify the sample and the patient; (2) Report summary, including main takeaways from report findings; (3) Positive findings, providing identified variants, associated drugs, and diseases, as well as classification and actionability tier; (4) Negative findings, listing genes that were in scope of testing but did not return any findings; (5) Relevant clinical trials, matching patient biomarkers, location, and other criteria; (6) Biomarker details, including summaries for positive findings and other additional details; (7) Report signoff section, including date and name for report approval. 1 Case, patient, and sample information 2 Report summary 3 Positive findings 4 Negative findings 5 Relevant clinical trials 6 Biomarker details 7 Report signoff NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 13 Software-enabled variant reporting addresses the burden of organizing test results into a PDF document manually by: • Ensuring the completeness and accuracy of a report by automatically incorporating all relevant variant interpretations and references • Providing a visually intuitive report, making it convenient for the oncologist to find specific information easily • Tracking changes made to reports after they are issued and storing a full history for each report in case of modifications or errors The report contains the identified genetic variants, their associations with therapies and diseases, clinical significance, and any other relevant information, organized and presented in a structured format (Figure 6). Insights in the report are used to help guide clinical decision making, eg, developing a treatment plan or recommending enrolling a patient in a clinical trial. A modern interpretation software offers users several features for report generation functionality. These enable users to: • Populate reports with interpreted variants automatically • Provide different report templates for different tests offered by the lab • Customize report templates per lab requirements • Generate a report summary automatically • Store and retrieve past reports and track changes made to them NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 14 Considerations for software implementation Currently, there are multiple commercially available interpretation and reporting software options. Labs should carefully evaluate their choice of software for implementation to be sure that it meets their specific needs. There are several key considerations for implementing tertiary analysis software, including: accuracy of content, approach to curation, usability, customization, report readability and integrations, and data security. Accuracy of content Accuracy of content refers to the correctness and reliability of the scientific and clinical information used for variant annotation and prioritization.21 Factors of time, with regard to the timeliness of the integrated knowledge, and geography, are crucial contributors of accuracy. In view of changing guidelines by different medical societies, different drug approval timelines around the world, and the rapidly changing nature of science, any knowledge sources integrated within a software need to be updated regularly to reflect advancements in clinical research and disease management. It is favorable for knowledge sources within a variant interpretation software platform to be updated as frequently as once a month. Approach to curation The approach to curation consists of how variants are prioritized, annotated, and interpreted.22 In principle, the more knowledge bases or resources included, the more comprehensive the annotation will be. Missing information from some important knowledge bases will result in the annotation being incomplete and potentially misleading. Too much information could include aspects that are irrelevant for annotations and make it challenging for lab personnel to filter out the most critical information for clinical practice.4 Therefore, it is critical to have balanced information captured in the annotations by including the right choices of knowledge bases. Ideally, all knowledge bases that are desired by the lab are included in a single tertiary analysis software interface for streamlining interpretation, updating knowledge bases, and avoiding fees for accessing multiple commercial knowledge bases. NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 15 In addition, the criteria a software uses for evidence to support an assertion of clinical actionability for a given variant is key to facilitating standardized practice in precision medicine. Professional societies have published guidelines on interpreting somatic oncology variants, such as recommendations from the FDA and a joint consensus from AMP, ASCO, CAP, ACMG, and ESMO (Table 2). However, these guidelines have some degree of differences in their classification systems.21 Labs should carefully evaluate the differences in guidelines that a software adopts to determine the most suitable approach for their own population needs.23 Usability of software Usability, the intuitiveness and user friendliness of a software, can be an important consideration for labs. An optimal user experience can promote efficiency, especially for lab staff with minimal bioinformatics expertise. Software usability covers a wide range of criteria, such as setup and maintenance, flexibility for customization, and data management.21 Customization capabilities Interpretation software that offers customization options is highly desirable for institutions with varying clinical needs, in part based on regional differences. For example, a variant could be classified as Tier I according to standard guidelines because it has a matched approved targeted therapy; however, the targeted therapy may not be available in the local hospital formulary due to reimbursement reasons. Hence, it may be appropriate to downgrade the variant to a lower tier. In such cases, a software that allows information to be presented originating from the lab’s prior case interpretation is critical to ensure relevance to their local practice. System integration The report should also be easily connected to, and displayed in, the in-house lab information system (LIS) and electronic health record (EHR). Findings should be easily pulled out from other results so that oncologists do not have to log on to multiple platforms to access critical information. The variant interpretation and reporting software should easily integrate with software necessary for other steps within the NGS workflow, including LIMS and secondary analysis solutions, enabling seamless data flow from sequencing to interpretation. Data management and connectivity to the lab’s workflow are paramount. NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 16 Data security and deployment options Maintaining the privacy of highly sensitive personal information analyzed and stored in any interpretation software is fundamental and requires enterprise-level protection. When evaluating commercial interpretation software, consider three elements of data protection features: (1) product-specific security and privacy functionality, (2) regulatory compliance, and (3) relevant deployment options.24,25 Institutions should look for the following key security and privacy features when considering a commercial software: • Full audit trails to ensure accountability for all actions and objects • High-level data encryption both “in transit” (TLS 1.2 at a minimum) and “at rest” (AES-256) to maintain data confidentiality • Fully automated data management, retention, and performance of regular data integrity checks to protect against data loss • Multifactor authentication and exceptional login policies for enhanced security Any interpretation software of choice should be developed in accordance with industry best practices and relevant standards under an enterprise-level Quality Management System (QMS). Institutions operating in regulated environments are often required to maintain compliance with global and regional data protection laws, and should identify vendors whose QMS aligns with the following certifications: • International Organization for Standardization (ISO) 27001 • ISO 13485 • General Data Protection Regulation (GDPR) • Health Insurance Portability and Accountability Act (HIPAA) Institutions should examine how software options will be hosted; this could include on-premises server or cloud-based deployment options. In general, laboratories that choose to interpret and store their NGS data locally should expect higher investment times for staying up to date with security and compliance requirements. In contrast, cloudbased solutions offer a scalable, cost-effective deployment option for institutions looking to shorten turnaround times. If a cloud-hosted solution is chosen, the vendor’s cloud service provider should be considered, as data storage and handling policies may differ between global and regional providers. NGS TERTIARY ANALYSIS M-GL-01601 v1.0 | 17 Discovery and translational research use cases Variant interpretation and reporting software can also be used in a research setting, enabling laboratories to capture critical knowledge on variant associations with biological processes, disease origins, and biomarker potential. To enable such use cases at scale, researchers should consider a solution with additional functionalities to those previously described: • Aggregation and querying of structured multimodal data sets • Building and mining cohorts of molecular and clinical research data • Collaboration and sharing of data within a secure, enterprise-level workspace • Customization and automated generation of a research report • Visualizations for exploring genome view, DNA and RNA coverage plots, VAF, distributions, and more Clinical laboratories focused on test development, discovery, and translational research use cases may also consider using variant interpretation and reporting software, specifically for the following: • Biomarker development studies—selecting genes for a panel, or understanding the performance of new emerging biomarkers in a retrospective cohort • Assay utility studies—estimating the number of findings in a test and comparing with alternative solutions • Methodology comparison studies—benchmarking elements of a clinical report for cost effectiveness and assessing test settings such as impact of coverage, read length, and filtering of findings • Longitudinal studies—for example, tracking tumor mutation burden (TMB) and occurrence of new variants in samples with time and its association with disease progression Summary Now more than ever, labs need optimized software solutions that will decrease the time and effort required to extract and report meaningful biological insights from genomic data. When evaluating commercially available software options, labs should consider each step of the interpretation and reporting workflow and the features each software option offers. Labs should select the software that best meets their needs and overcomes the informatics bottleneck to deliver a streamlined, simplified, accurate, and consistent workflow for variant interpretation. M-GL-01601 v1.0 | 18 UNLOCKING ACTIONABLE INSIGHTS References 1. Tsimberidou AM, Hong DS, Wheler JJ, et al. Long-term overall survival and prognostic score predicting survival: the IMPACT study in precision medicine. J Hematol Oncol. 2019;12(1):145. Published 2019 Dec 30. doi:10.1186/s13045-019-0835-1. 2. Singal G, Miller PG, Agarwala V, et al. Association of Patient Characteristics and Tumor Genomics With Clinical Outcomes Among Patients With Non-Small Cell Lung Cancer Using a Clinicogenomic Database [published correction appears in JAMA. 2020 Feb 4;323(5):480]. JAMA. 2019;321(14):1391-1399. doi:10.1001/jama.2019.3241. 3. Schwaederle M, Zhao M, Lee JJ, et al. Impact of Precision Medicine in Diverse Cancers: A Meta-Analysis of Phase II Clinical Trials. J Clin Oncol. 2015;33(32):3817-3825. doi:10.1200/JCO.2015.61.5997. 4. Austin-Tse CA, Jobanputra V, Perry DL, et al. Best practices for the interpretation and reporting of clinical whole genome sequencing. NPJ Genom Med. 2022;7(1):27. Published 2022 Apr 8. doi:10.1038/s41525-022-00295-z 5. Mileham KF, Schenkel C, Bruinooge SS, et al. Defining comprehensive biomarker-related testing and treatment practices for advanced non-small-cell lung cancer: Results of a survey of U.S. oncologists. Cancer Med. 2022;11(2):530-538. doi:10.1002/ cam4.4459. 6. Illumina. As Use of Large Cancer Panels and Whole-Genome Sequencing Expands, Illumina Plans to Future-Proof NGS Interpretation and Reporting. Genome Web. Published March 28, 2023. Accessed September 12, 2023. 7. Jobanputra V, Wrzeszczynski KO, Buttner R, et al. Clinical interpretation of whole-genome and whole-transcriptome sequencing for precision oncology. Semin Cancer Biol. 2022;84:23-31. doi:10.1016/j.semcancer.2021.07.003. 8. Horak P, Griffith M, Danos AM, et al. Standards for the classification of pathogenicity of somatic variants in cancer (oncogenicity): Joint recommendations of Clinical Genome Resource (ClinGen), Cancer Genomics Consortium (CGC), and Variant Interpretation for Cancer Consortium (VICC). Genet Med. 2022;24(5):986-998. doi:10.1016/j.gim.2022.01.001. 9. Robinson JT, Thorvaldsdóttir H, Winckler W, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29(1):24-26. doi:10.1038/ nbt.1754. 10. Malone ER, Oliva M, Sabatini PJB, Stockley TL, Siu LL. Molecular profiling for precision cancer therapies. Genome Med. 2020;12(1):8. doi:10.1186/s13073-019-0703-1. 11. Jaganathan K, Kyriazopoulou Panagiotopoulou S, McRae JF, et al. Predicting Splicing from Primary Sequence with Deep Learning. Cell. 2019;176(3):535-548.e24. doi:10.1016/j.cell.2018.12.015 12. Sundaram L, Gao H, Padigepati SR, et al. Predicting the clinical impact of human mutation with deep neural networks [published correction appears in Nat Genet. 2019 Feb;51(2):364]. Nat Genet. 2018;50(8):1161-1170. doi:10.1038/s41588-018-0167-z 13. Gao H, Hamp T, Ede J, et al. The landscape of tolerated genetic variation in humans and primates. Science. 2023;380(6648):eabn8153. doi:10.1126/science.abn8197 14. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-424. doi:10.1038/gim.2015.30. 15. Froyen G, Le Mercier M, Lierman E, et al. Standardization of Somatic Variant Classifications in Solid and Haematological Tumours by a Two-Level Approach of Biological and Clinical Classes: An Initiative of the Belgian ComPerMed Expert Panel. Cancers (Basel). 2019;11(12):2030. Published 2019 Dec 16. doi:10.3390/cancers11122030. 16. Li MM, Datto M, Duncavage EJ, et al. Standards and Guidelines for the Interpretation and Reporting of Sequence Variants in Cancer: A Joint Consensus Recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists. J Mol Diagn. 2017;19(1):4-23. doi:10.1016/j.jmoldx.2016.10.002. 17. Mateo J, Chakravarty D, Dienstmann R, et al. A framework to rank genomic alterations as targets for cancer precision medicine: the ESMO Scale for Clinical Actionability of molecular Targets (ESCAT). Ann Oncol. 2018;29(9):1895-1902. doi:10.1093/annonc/ mdy263. 18. U.S. Food & Drug Administration. CDRH’s approach to tumor profiling next generation sequencing tests fact sheet. https://www. fda.gov/media/109050/download. Accessed September 5, 2023. 1.800.809.4566 toll-free (US) | +1.858.202.4566 tel techsupport@illumina.com | www.illumina.com © 2024 Illumina, Inc. All rights reserved. All trademarks are the property of Illumina, Inc. or their respective owners. For specific trademark information, see www.illumina.com/company/legal.html. M-GL-01601 v1.0 19. Strom SP. Current practices and guidelines for clinical next-generation sequencing oncology testing. Cancer Biol Med. 2016;13(1):3-11. doi:10.28092/j.issn.2095-3941.2016.0004 20. Koboldt DC. Best practices for variant calling in clinical sequencing. Genome Med. 2020;12(1):91. doi:10.1186/s13073-020- 00791-w. 21. Song J, Hussaini M. Adopting solutions for annotation and reporting of next generation sequencing in clinical practice. Pract Lab Med. 2020;19:e00154. doi:10.1016/j.plabm.2020.e00154. 22. Pandey KR, Maden N, Poudel B, Pradhananga S, Sharma AK. The curation of genetic variants: difficulties and possible solutions. Genomics Proteomics Bioinformatics. 2012;10(6):317-325. doi:10.1016/j.gpb.2012.06.006. 23. Perakis SO, Weber S, Zhou Q, et al. Comparison of three commercial decision support platforms for matching of next-generation sequencing results with therapies in patients with cancer. ESMO Open. 2020;5(5):e000872. doi:10.1136/esmoopen-2020-000872. 24. Yaung SJ, Pek A. From Information Overload to Actionable Insights: Digital Solutions for Interpreting Cancer Variants from Genomic Testing. J Mol Path. 2021; 2(4):312-318. doi.org/10.3390/jmp2040027. 25. Aziz N, Zhao Q, Bry L, et al. College of American Pathologists’ laboratory standards for next-generation sequencing clinical tests. Arch Pathol Lab Med. 2015;139(4):481-493. doi:10.5858/arpa.2014-0250-CP.