By Ann Neuer and Allison Proffitt
July 19, 2016 | On July 8, the Food and Drug Administration (FDA) released two pivotal draft guidances on the emerging role of next generation sequencing (NGS)-based in vitro diagnostics (IVDs). The technology is becoming the go-to method for quick determination of gene variants that might be causative for an array of diseases, such as breast cancer, colon cancer, and muscular dystrophy. Each draft guidance has a different purpose, but both emphasize using best practices and building quality into the processes of using NGS- based IVDs so they have clinical validity.
In the past, IVDs have been used to spot a limited number of predefined analytes for diagnosing specific conditions, but next-generation sequencing can measure millions of analytes in parallel related to many conditions. It’s a significant shift in diagnostics.
The expanded use of NGS in diagnostic testing reflects FDA’s efforts to comply with the Precision Medicine Initiative (PMI) announced by President Obama in January 2015, and meant to usher in an era of research, technology, and policies to accelerate individualized patient care. With involvement from the National Institutes of Health and the National Cancer Institute, PMI’s goals are innovation, safety, and effectiveness in test development to help ensure patients receive accurate and meaningful results. As described in both draft guidances, FDA has been focusing on optimizing flexible and adaptive regulatory oversight for NGS-IVD tests as part of its commitment to speeding adoption of precision medicine.
“It is a major step forward for FDA to recognize the incredible importance of public data sharing for NGS-based diagnostics,” says Heidi Rehm, Associate Professor of Pathology at Brigham & Women’s Hospital and Harvard Medical School. “The concepts described in the two draft guidances provide an important framework for clinical laboratories developing testing and for organizations looking to support publically accessible databases to assure quality patient testing.”
Public Human Variant Databases
One of the draft guidances is entitled Use of Public Human Genetic Variant Databases to Support Clinical Validity for Next Generation Sequencing (NGS)-Based In Vitro Diagnostics. Its focus is on the kinds of robust processes needed to ensure these databases are high quality sources of valid scientific evidence for FDA’s regulatory review of NGS-based tests.
A variant databases would be considered a source of valid scientific evidence to support an NGS-based test in a premarket submission if it meets four criteria, the draft guidance stated:
- It must operate in a manner that provides sufficient information and assurances regarding the quality of source data and its evidence;
- it must provide transparency regarding its data sources and its operations, particularly around how variant evidence is evaluated and interpreted;
- it must collect, store, and report data and conclusions in compliance with all applicable requirements regarding protected health information, patient privacy, research subject protections, and data security; and
- it must house sequence information generated by validated methods.
Among other recommendations on data security and procedures, FDA suggests that all public human genetic variant databases share their standard operating procedures (SOPs) for how variant data are aggregated, curated, and interpreted. Public databases should keep metadata including the names of the laboratories reporting the variant, the test and its technical characteristics used to detect the variant, and all evidence sources used to support variant interpretation.
The interpretation question is, of course, key. “FDA believes that use of publicly available decision matrices for variant interpretation that are based on rigorous professional guidelines is central to assuring that assertions from genetic variant databases constitute valid scientific evidence supporting the clinical validity of a test,” the guidance reads.
FDA recommends caution in assigning definitive assertions to variants. A variant shouldn’t be labeled “pathogenic” based on a single piece of evidence, or on weak evidence.
“Assertions that a particular genotype-phenotype association is clinically valid should generally involve multiple lines of evidence and, at a minimum, should identify a primary source of scientific evidence and other supporting evidence,” FDA recommends.
FDA warned that databases should not include any recommendations regarding clinical treatment or diagnosis, and stressed that when conflicts of interest arise, they should be disclosed as transparently and thoroughly as possible.
FDA plans to initiate a formal recommendation process for databases that meet all of the listed criteria, and expects the recognized databases to streamline premarket review of NGS tests. In some cases, FDA expects that data from a recognized database would suffice for premarket review of an NGS diagnostic test.
Database recognition would be voluntary, and “would not subject the database to FDA oversight, beyond that needed to retain the recognition.” Database administrators that want their databases reviewed for recognition would submit to FDA all of the metrics, SOPs, and documentation associated with the database. If recognition is granted, all of those materials would be made publically available on the database website.
In the draft guidance, FDA noted that the agency is open to enlisting third parties to help with the database review and recognition process.
The Importance of Standards
The second draft guidance deals with the Use of Standards in FDA Regulatory Oversight of Next Generation Sequencing (NGS)-Based In Vitro Diagnostics (IVDs) Used for Diagnosing Germline Diseases.
Standards in NGS-based testing for germline diseases are essential as they support analytical validity to streamline regulatory oversight of these tests, which are considered new types of medical devices that have not been previously classified by FDA.
In a statement sent to Diagnostics World, the FDA’s Precision Medicine team explains, “The use of recognized consensus standards allows for variations in development and validation, and accommodates the expected rapid evolution of NGS technologies. While there are no comprehensive consensus standards available for design and analytical validation of NGS clinical applications yet, the FDA plans to work with the genomics community to encourage standards developing organizations to initiate several key standards as soon as possible.”
The guidance is meant to provide recommendations for designing, developing, establishing quality metrics, and validating NGS-based tests for germline diseases. Germline diseases are those arising from either inherited or de novo germline variants, for example, cystic fibrosis, amyotrophic lateral sclerosis, some cancers, and Gaucher disease. FDA’s guidance is intended for targeted and whole exome sequencing tests—not whole genome sequencing—and the recommendations would not apply to screening tests, companion diagnostics, pre-implantation embryonic testing, or tumor sequencing.
Currently, all NGS-based tests are considered class III devices; there are no legally-marketed devices to serve as predicates for regulatory review. But if a test were to proceed through the regulatory process and be proven to be safe and valid, it could be classified as a class II device. If so, the regulatory approval process would be quicker, the test could possibly be exempt from the premarket notification process, and it could serve as a predicate for future tests.
In this case, standards become extremely important. “FDA believes that one approach for supporting the analytical validation of NGS-based tests may be through conformity with one or more FDA-recognized standards (if available) or special controls,” according to the draft guidance.
The draft guidance outlines information that should be included in a standard to be recognized by FDA including test design, test performance metrics, variant annotation procedures, and the return of results. Among other parameters, FDA specified that the sequencing platform for an NGS test should be disclosed, all bioinformatics software and pipelines should be detailed including any modifications, it should be clear if analyses were run remotely, and any databases used should be specified.
Test reports should include a list of pathogenic or actionable variants on the first page of the report, and the relationship between reported variants and the clinical presentation of the patient. If variants of unknown significance are reported, the guidance recommends clearly separating these from other variants, and “include[ing] a statement that their clinical relevance is not known.”
The 90 day comment period for both draft guidances is currently underway, ending on October 6, 2016, at 11:59 PM, Eastern Daylight Time.