New Guidelines Suggest Tiered Approach to Standardize Interpretation and Reporting of Sequence Variants in Cancer

Aconstant refrain within the diagnostics industry is a need for standardization. Rapidly developing technology and testing methods, the call for interoperability, and the vast amount of data and information producible via molecular testing makes this even more important. The capabilities of next generation sequencing have provided both opportunities and challenges as clinical laboratories interpret and report results of cancer-related sequencing tests.

Now, the Association for Molecular Pathology (AMP) has released guidelines developed by an industry working group that recommend standard classification, annotation, interpretation and reporting for somatic sequence variants in cancer. The AMP was joined in this effort by the American College of Medical Genetics and Genomics (ACMG), American Society of Clinical Oncology (ASCO), and College of American Pathologists (CAP).

"Cancer genomics is a rapidly evolving field so the clinical significance of any variant in therapy, diagnosis, or prognosis should be reevaluated on an ongoing basis by all of the key stakeholders," said one working group member, Marina N. Nikiforova, MD, Professor of Pathology at University of Pittsburgh Medical Center, in a statement announcing the release of the guidelines. Nikiforova is also the 2016 AMP Clinical Practice Committee Chair and adds, "These new recommendations resulted from the successful ACMG, AMP and CAP efforts on germline variant interpretation and were additionally informed by the diverse perspectives expressed at the ASCO, AMP and CAP Genomic Roundtable stakeholder discussions."

Working group gathers industry professionals
Recognizing the need for standardization to maximize the benefits of cancer gene sequencing, AMP gathered industry experts in a working group to review literature, data, laboratory surveys and input from industry via public meetings. The result is a report, "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" (Guidelines). The Guidelines will be published in the January 2017 issue of The Journal of Molecular Diagnostics but were released online by AMP in December.

"We worked diligently to ensure the cancer genomics community was well represented and it is our hope that we will soon see the widespread adoption of these guidelines leading to improved communication between molecular pathologists, oncologists, pathologists, and most importantly, patients," said Marilyn M. Li, MD, AMP Member and Chair of the Working Group in a statement. Li also serves as Vice Chief of the Division of Genomic Diagnostics and Director of Cancer Genomic Diagnostics at Children's Hospital of Philadelphia. To gain insight into the real world operations of varying laboratories the working group solicited surveys from AMP members. The surveys focused on technical issues and reporting issues. They received 67 responses on technical issues and 44 for reporting issues. Survey responses revealed variation among NGS techniques used and the annotation and reporting of variants.

The U.S. Food and Drug Administration has similarly led industry-wide discussion of standards for NGS interpretation and reporting. Recognizing the vast amount of information that can be yielded from NGS, the difficulty in interpreting results of NGS testing and lack of evidence in some cases linking genetic variants to specific diseases, a recent FDA workshop discussed how best to make use of the results of NGS testing—based on patient and provider preferences. The goal, according to the announcement of the workshop, is "to learn when results are generated in a CLIA-compliant laboratory, which results are of importance to patients and providers, how these results should be returned and how much and what types of evidence supporting interpretation of those results is necessary."

Need for standardization grows with testing capabilities
The Guideline document explains the significant volume of information regarding genetic variations that NGS yields and the clinical benefits in terms of diagnosis and treatment, concluding "therefore, it is imperative to unify the interpretation and reporting of molecular results among laboratories performing these tests."

The guidelines declared standardization is needed for the following quality of care related reasons: "accurate reporting of tumor response to targeted therapy; establishment of national guidelines for patient care; and collaborative institutional clinical trials, thereby supporting the need for standardization among laboratories performing these tests."

The authors conclude "It is our hope that the guidelines presented herein will achieve widespread use in the cancer genomics community and engender significant improvements in the practice of genomic testing and precision care for cancer patients."

Tiered classification system
The Guidelines share several recommendations for laboratories involved in such NGS variant testing for cancer. To categorize variants, the guidelines recommend using a four-tiered system that categorizes somatic variants according to their impact on clinical care—using "currently available evidence."

• Tier I would be variants with strong clinical significance based on evidence regarding FDA approved therapy and "well-powered studies with consensus from experts in the field."

• Tier II covers variants of potential clinical significance based on evidence from FDA approved treatments but for different tumor types or investigational therapies, small published studies having some consensus, and preclinical trials or case reports without consensus.

• Tier III includes variants of unknown clinical significance when there is no "convincing published evidence of cancer association" and the variant isn't "observed at a significant allele frequency in the general or specific subpopulation databases, or pan-cancer or tumor-specific variant databases."

• Tier IV is for benign or likely benign variants when there is a lack of published evidence of cancer association and the variant is "observed at significant allele frequency in the general or specific subpopulation databases."

With regard to the evidence relied upon to categorize the variants, the authors note "It is important to recognize that molecular genetics in cancer is a rapidly evolving field; therefore the level of evidence in therapy, diagnosis, and prognosis should be continuously evaluated based on evolving research data and modified accordingly." The authors discuss levels of evidence and indicate specific sources can shift between evidence categories as more becomes known about the source.

Other observations and recommendations
With regard to variant identification and annotation, the guidelines recommend laboratories pay close attention to the bioinformatics pipeline used and emphasizes the need for validation of that pipeline. The authors recommend against in silico prediction algorithms being "used as the sole evidence for variant classification or clinical decision making."

When relying on databases to review genomic information, the guidelines caution laboratories to fully understand the database in terms of its content, how it is aggregated, the limitations of the database, how genome information is annotated, the source of the data and data quality of the diagnosis provided. The report then surveys various types of databases and their usefulness with recommendations for evaluating the data within the databases. The authors also address internal laboratory databases and "urge clinical laboratories to contribute their well-curated variants to public variant databases to facilitate accurate interpretation of somatic variants."

In discussing how test results should be reported the Guidelines offer a few tips for laboratories:

• Contain all necessary information for patient and physician explaining what was done, results and factors influencing interpretation as well as what the test doesn't reveal;

• Keep reports "short, simple, and to the point";

• Make sure most critical information is provided clearly, conspicuously, and at the start of the report so it is "seen and understood";

• Include standard and colloquial nomenclature;

• Don't limit reports to just positive findings;

• Consider using tables to provide information about "genomic coordinates, genome build and transcript reference sequence";

• Explain distinction between somatic and germline alterations;

• Clearly indicate date of issue for the report;

• Indicate method, assay performance characteristics and "critical quality metrics for assay run";

• Don't just list gene names but explain what was tested; and

• Provide results in format capable of being incorporated into the electronic health record (pdfs or other formats that have to be scanned are not ideal).

Takeaway: Industry tackles complexities of somatic variant testing and key organizations provide consensus on standards for interpreting and reporting results to assist patients and clinicians in decision-making.