Diagnostics & Imaging Week

The FDA last week issued draft regulatory guidance for clinical laboratories and industry regarding so-called in vitro diagnostic multivariate index assays (IVDMIA) that use mathematical formulas to interpret "large amounts of gene and protein data" to produce results designed to help in both diagnosis and treatment decision-making.

An FDA spokesman wrote in an e-mail to Diagnostics & Imaging Week that the agency believes there is a "relatively small but growing number of tests in this area." And, judging from the guidance issued, the agency believes that greater oversight is needed. That could lead manufacturers to be required more often to seek a 510(k) or premarket approval (PMA) for its diagnostic tests.

"Regulation is always risk-based and so whether a 510(k) or a PMA for these devices [would be necessary] would depend on intended uses and risk profiles," the spokesman wrote.

The agency also noted it expected the new guidance to help ease any public concerns that such tests are safe and effective.

"More and more of these kinds of medical tests are being made available each year," said Daniel Schultz, MD, director of the Center for Devices and Regulatory Health. "It is important for the companies and labs making the tests to clearly understand the regulatory requirements in place so that the tests they develop are as safe and effective as possible."

In the draft guidance, the agency noted that "FDA is aware of some confusion about the regulation of IVDMIAS that are developed by, and used in, a laboratory. We believe this confusion derives in part from FDA's approach to regulation of laboratory-developed tests that use commercially available [analyte specific reagents] and other commercially available, FDA-regulated components."

Some of that confusion, the FDA acknowledged, comes as a result of the agency's Analyte Specific Reagent rule, which does not extend to tests developed in-house, or so-called "home-brew" tests developed by and used only in a particular laboratory.

Accordingly, the agency also issued guidance on ASRs in the form of "frequently asked questions."

The guidance defines IVDMIAs as "test systems that employ data, derived in part from one or more in vitro assays, and an algorithm that usually, but not necessarily, runs on software to generate a result that diagnoses a disease or condition or is used in the cure, mitigation, treatment or prevention of disease."

The agency said that classification of an IVDMIA "would depend on its intended use," and that it believes that "most IVDMIAs will be either Class II or III devices."

As an example, it cites that an indicator of a patient's risk of cancer would likely be classified as a Class II device, which generally requires a 510(k) clearance. However, even that same device, if the intended use is to predict whether a patient should have chemotherapy would require a PMA (premarket approval).

A PMA would require clinical study, and in such study, the companies can make no claims for the tests. Additionally, companies may be required to secure an investigational device exemption for tests in such studies.

"FDA recommends sponsors interact with the agency early and often in the development of these diagnostic assays and utilize appropriate scientific, medical and statistical expertise to assure that thresholds of safety and effectiveness are addressed," the agency wrote in its guidance.

The guidance also states that IVDMIAs are subject to the Quality System Regulation, but it noted that it "will work with laboratories" that make these devices and that also must comply with CLIA regulations. IVDMIAs makers also must comply with the Medical Device Reporting regulation, which would require them to report to the FDA any information regarding a device suspected of causing a death or serious injury.

One of the companies that offers such tests is Genomic Health (Redwood City, California). Its OncoType DX-21 gene panel breast cancer test, for example, was touted at the 28th annual San Antonio Breast Cancer Symposium last December, when study results were released.

The study evaluated the experience of four oncologists treating 68 early-stage, estrogen-receptor-positive breast cancer patients and found that knowledge of the Oncotype DX Recurrence Score, a score between 0-100 assigned based on a quantitative measurement of the expression of 21 genes, altered the adjuvant treatment administered to 25% of patients compared to physicians' original recommendations.

"Our findings clearly indicate that assessing a woman's individual Recurrence Score in addition to standard measures, such as patient age, tumor size and tumor grade, is critical to making well-informed treatment decisions," said Ruth Oratz, MD, clinical associate professor, NYU School of Medicine (New York), and lead author of the study, although she was at the Rocky Mountain Cancer Centers (Denver) when completing the study.

In a research note issued by Cantor Fitzgerald & Co. (New York), George Zavoica, PhD, said he "see[s] this as an opportunity for [Genomic Health] to shape public policy and the FDA's regulatory oversight for IVDMIAs in its favor."

"In our view, the FDA's action takes out some of the uncertainty overhanging [Genomic Health]," Zavoica wrote. "[The company] will be able to continue to sell Oncotype DX during the public comment period."

Cantor Fitzgerald maintained its "Buy" rating on Genomic Health.

Next: analyte-specific reagents.