Monday, November 24, 2014

FDA announces two-day public LDT Regulation meeting, January 8-9, 2015.

On November 24, 2014, the Federal Register published the FDA's announcement of a two-day public workshop on January 8-9, 2015, on its proposal to regulate laboratory developed tests.   The meeting announcement is available here.  The workshop will be held at the Natcher Center at the NIH.

The FDA posted transcripts webcasts  after the meeting, here.
The rest of the original blog continues...

Registration is first come first serve until December 12, 2014, at 4 pm, but may sell out earlier.  The FDA may "limit the participants from each organization."   A streaming webcast is expected to be available.

Oral presentations must be requested by December 12, and winners will be selected by December 17.  Presentation materials would be due by January 6, 2015.

Comments may be submitted before or after the meeting, up until February 2, 2015.

The meeting announcement is four pages long and contains additional commentary on the FDA's plans and goals.

The meeting agenda indicates that the FDA has been listening to stakeholders who are puzzled over how the FDA will handle "labeling" issues.   While the FDA has suggested that outside bodies (perhaps College of American Pathologists) might handle LDT reviews, medical devices are not merely classified as "low risk" and "high risk."  510(k) devices must have specific 510(k) predicates, or else a much more elaborate process of creating a de novo 510(k) category is involved.   Given the profusion of human gene tests, many LDTs that have never had a IVD equivalent may not have 510(k) predicates, even if they are otherwise low risk.

In addition, some molecular test FDA labels have quite convoluted "intended use" supplemented by equally impenetrable, or unguessable, "special conditions for use."    (Examples here).  How a CAP inspector approving LDTs could come up with novel labels of this complexity for each of a lab's LDT genetic tests seems far fetched.  Maybe...even the FDA didn't need to come up with such convoluted labeling in the first place.