Skip to main content
Premium Trial:

Request an Annual Quote

NEJM Articles Discuss FDA Regulation, Clinical Validation of Genomic Tests

NEW YORK (GenomeWeb) — A series of articles published online in the New England Journal of Medicine today highlights the promises and challenges of clinical genomic testing, in particular the regulation of such tests by the US Food and Drug Administration and their clinical validation, both in the context of breast cancer risk and general variant assessment.

Barbara Evans at the University of Houston Law Center and two colleagues at the University of Washington criticized the anticipated oversight of genomic tests by the FDA — through proposed regulation of laboratory-developed tests and of next-generation sequencing — drawing an immediate response from FDA representatives.

According to the authors, the 40-year-old medical device regulations under which the FDA seeks to regulate genomic tests and technologies are outdated and would not ensure their safety and effectiveness while fostering innovation at the same time.

One challenge is the premarket assessment of the clinical validity of variants — associations between a variant and a disease — because next-generation sequencing generates large numbers of variants. The FDA has proposed to use existing genetic databases for validation, but in the eyes of the authors that idea is "unduly optimistic" because such databases do not exist for the majority of variants.  

Requiring premarket review of clinical validity could lead to an unbundling of the testing process from the clinical interpretation, which could move offshore and potentially "diminish the safety of American consumers," they predicted.

Instead, they suggest, the FDA should consider a postmarketing surveillance model, which the agency already uses to monitor drug safety. "Premarket review is the wrong tool, and the traditional product-by-product regulatory focus of the FDA is myopic," they wrote.

While short-term postmarketing review already exist for medical devices, "genomic testing needs an ongoing, decades-long program of continuous learning to clarify both benefits and risks that are not yet known," they wrote.

To establish the clinical significance of novel variants, the FDA should help establish new large-scale data resources that provide access to both sequence and phenotypic data, possibly through public-private partnerships, they recommended, but this would require significant funding.

The authors are also critical of the FDA's intent to regulate the analytic validity of genomic tests, either individually or by issuing performance standards for laboratories. In particular, they wrote, it is unclear "how the standards-based approach proposed by the FDA would improve on the oversight that [the Centers for Medicare and Medicaid] already provides under CLIA."

However, CLIA gives laboratories considerable leeway in the processes they use, and the FDA could "add substantial value by standardizing what needs to be disclosed and communicated about laboratory processes and their potential effect on data quality," they wrote.

Overall, they concluded, "statutory reforms should focus on granting the FDA a correct package of legal powers, seed funding, and legal pathways to encourage public-private partnerships to develop and sustain data resources for the right regulation of genomic testing."

FDA representative agreed with some but not all of these points. "We disagree that the FDA will chill innovation because it lacks the correct set of statutory authorities," the FDA's David Litwack and colleagues wrote in a letter to the editor commenting on Evans' paper.

The agency has already proven that it is "not 'clinging' to traditional premarket review approaches" for novel genetic tests, for example by exempting the Illumina MiSeqDx from premarket review and by approving Myriad Genetics' BRACAnalysis CDx using a "flexible approach to permit the use of a defined process for the classification and reporting of variants."

Also, the FDA is already seeking to foster database development, they wrote, and unlike CLIA, it does not oversee laboratory operations but focuses "distinctly on tests."

Breast cancer risk prediction

While the clinical interpretation of many variants is still in flux, it may not be in the best interest of patients to provide tests where the risk associated with a variant is unknown, according to Douglas Easton of the University of Cambridge in the UK and a number of colleagues in the UK, the US, the Netherlands, Germany, Australia, and Canada.

In an article on gene-panel sequencing for breast cancer risk prediction, they proposed "that a genomic test should not be offered until its clinical validity has been established."

In contrast to Evans and others suggesting postmarketing studies to establish clinical validity, "we believe that failing to require the clinical validation of genomic tests before they are submitted for regulatory approval is likely to lead to substantial misuse of the technology," they wrote.

Following the 2013 US Supreme Court decision to invalidate claims by Myriad Genetics on the BRCA1 and BRCA2 genes, a growing number of laboratories started offering gene panel tests to assess breast cancer risk. According to the authors, these panels, some of which also cover other cancer types, analyze more than 100 genes in total, of which 21 are claimed to be associated with breast cancer.

However, assigning a risk to a variant in one of those genes is tricky, depending on what studies have been conducted. Also, translating relative risks into an individual's absolute risk for breast cancer poses challenges, because it depends on factors such as family history, age, and lifestyle.

Reviewing studies of genes included in breast cancer panels, the authors found that for some, there is insufficient evidence, so they "would caution against their use in the prediction of breast cancer risk."

"There is an urgent need for much larger, well-designed population- and family-based studies in diverse populations that will provide reliable estimates of risk for the purpose of counseling," they wrote, and collecting data from ongoing panel testing, with clinical data, will be helpful as well.

"Panel testing can make a useful contribution to prediction of a woman's risk of breast cancer, but end users need to be aware of the limitations of these panels," they concluded.

ClinGen updates

Elaborating on the theme of clinical validity and the interpretation of variants, Heidi Rehm of the Laboratory for Molecular Medicine at Partners HealthCare Personalized Medicine and colleagues provided an update on the National Institutes of Health's Clinical Genome Resource (ClinGen) program and the associated ClinVar database of the National Center for Biotechnology Information, which was launched two years ago.

As of earlier this month, ClinVar contained 145,311 unique sequence and structural variants across 22,864 genes, contributed by 314 submitters, among them clinical and research laboratories, locus-specific and aggregate databases, expert consortia, professional organizations, healthcare providers, and patients.

Of these variants, more than 118,000, or 81 percent, have clinical interpretations, though about a fifth of them are variants of uncertain significance "which highlights the additional work to be done," they wrote.

Eleven percent of the variants with clinical interpretations were submitted by more than one laboratory, and of those, 17 percent, or more than 2,000, have different interpretations from different submitters. For more than 400 of these variants, the clinical assessments are so different that they would likely lead to different medical decisions.

For that reason, ClinGen is working with laboratories to adopt new standards for interpreting genetic variants, which ask submitters to share the basis of their assessments of pathogenicity. "This collaboration has allowed laboratories to resolve differences in interpretation through expert consensus and application of these standardized methods," the authors wrote.

As of this June, ClinVar will also introduce a modified star-based system to define the type of evidence and review by which a variant has been assessed, ranging from zero stars for no assertion or documented method to four stars if an interpretation is endorsed by published practice guidelines. The new system will allow users to filter variants by review status.

In addition, ClinGen has assembled a number of working groups, focusing on standard procedures for the evaluation of genes, variants, genetic disorders, and phenotypes. For example, a new database called ClinGenKB will support expert curation of genes and variants and the resolution of conflicting interpretations.

Also, a growing number of clinical-domain working groups are focusing on individual disease areas, such as cardiovascular disease, hereditary and somatic cancer, metabolic disease, and pharmacogenomics, and an "actionability working group" is studying which genes are associated with specific interventions in asymptomatic patients.

Other working groups are tackling new informatics approaches for variant assessment, integration of electronic medical health records, and patient outreach through Genome-Connect, which allows patients to upload their own test results and phenotypic data and connects them to research studies and one another.

The FDA, the authors noted, "is now looking to ClinGen to provide a possible resource for the clinical interpretation of genetic variation."