Page 56 - Biomedical Engineering and Design Handbook Volume 2, Applications

FDA MEDICAL DEVICE REQUIREMENTS 35

the body, then the combo is a drug. This is easy in some cases. Take a stent, designed to keep an
artery open, which is also coated with a drug to help prevent restentosis within the artery. This has a
strong device orientation due to its physical effects. These have been regulated as devices. The
metered dose inhaler acts primarily through drug action on the lungs. The device is more of a delivery
method. So this was, and will continue to be, regulated as a drug. Some cases are less clear. The
implantable drug pump was regulated as a device, since all the new technology was electromechanical.
If such a new device came along today, would it be a device? The primary mode of action on the
patient seems to be the drug that is delivered. Would this shift such delivery devices to the Center for
Drug Evaluation and Research (CDER)? We will see as technology advances.
Regulatory professionals work to have a combination product reviewed by the center with which
they are most comfortable and which has the most favorable procedures. No matter what center con-
trols, they will commonly ask for consultative review by the other centers that have skills and expe-
rience related to the product. For example, in review of a drug-coated device by the Center for
Devices and Radiological Health (CDRH), it is common to have a review of the drug component by
CDER. However, all CDRH procedures apply as opposed to those of CDER. One can ask OCP for a
determination of which center will control. The more normal path is to discuss it in advance with the
center you prefer and submit it to them.

2.12 TESTING MEDICAL DEVICES ON NONHUMAN ANIMALS
AND LIFE FORMS

It is usual for FDA to require testing of devices on animal models before introduction into human
testing. This is generally referred to as preclinical or nonclinical testing. It should be noted that polit-
ical and ethical concerns continually press for reduction in testing on animals. However, it remains
a standard step in medical device development.
Preclinical testing is regulated by FDA under the regulations know as Good Laboratory Practices
∗
(GLP). GLP applies to any nonclinical testing studies that support or are intended to support a
research or marketing application to FDA. This is not just for human devices or drugs. GLP applies
to applications for food or color additives, animal food additives, animal drugs, biologics, and elec-
tronic products. Any nonclinical study used to support a PMA, 510(k), or Investigational Device
Exemption (IDE) is to be done under GLP.
Note that GLP does not apply to any testing, such as bench testing, that does not involve a live
organism. Any test that uses a test system involving an animal, plant, microorganism, or subparts
thereof is covered by GLP.
Research studies not intended to be submitted to FDA fall outside GLP. This raises a practical prob-
lem. Often research studies are done without GLP controls, only to find later that the data are needed in
an FDA submission. Then it must be argued and explained that the original purpose was for research but
that it is acceptable to use the data for a submission. It is often a financial trade-off to decide whether to
conduct early studies under GLP in order to have the data more easily available later for submission.
There are two general classes of controls in GLP: animal care and data integrity. FDA has joint
jurisdiction with the U.S. Department of Agriculture (USDA) over the care of the animals. Both
inspect, but USDA is usually viewed as the prime regulator of animal treatment. FDA has, of course,
a keen interest in the accuracy, completeness, and truthfulness of the data resulting from a GLP
study. This part of GLP has a three-part basis. First, GLP requires a protocol that lays out the pur-
pose of the testing, the end points to be tested, the data forms for collection, and the description of
the data to be collected. Second, there must be a quality assurance unit that oversees and audits the
study to make sure that the protocol and regulations are followed. Third, there must be a final report
presenting the data found, the study results, and any conclusions.
GLP studies are audited and investigated by FDA, just as human clinical trials are. Precision and
care must be exercised so that a GLP study produces data acceptable to FDA.

∗
21 C.F.R. pt. 58.