Medical Device Daily Washington Editor

WASHINGTON — FDA recently released its second device-specific guidance for the year, but the most recent guidance was not the subject of the kind of speculation that made the rounds for the drug-eluting stent (DES) development guidance the agency published last month (Medical Device Daily, March 28, 2008). It also is much shorter than the DES guidance, coming in at 28 pages vs. the 89 pages devoured by the DES document.

Still, the guidance for total artificial discs has its own quirky government pedigree, given that the document is dated April 11 but did not appear on FDA's web site until this past Friday, April 18.

The motivation for the guidance is also difficult to discern, given that the most recent approval for a total artificial disc replacement sailed through the advisory committee at a near-unanimous vote of 7-1 when Medtronic (Minneapolis) brought the case for its Bryan cervical disc to the advisory committee (Medical Device Daily, July 19, 2007).

FDA recommends that a sponsor's functional animal studies should "perform all testing on sterilized components of the final device design," but will go along with studies of particulate matter "using a small-animal model (e.g., rabbit)." The objective of this study would be to "evaluate the local and systemic responses," and the guidance recommends that such testing include deliberate dosing of "10 million particles for the high-dose group" in the event that "wear particles cannot be identified and located" in animal models with the device.

Particulate matter was a consideration in the advisory panel hearing for the Bryan cervical disc, but the firm's voluminous data essentially laid those concerns to rest as demonstrated by the panel's 7-1 vote to recommend approval.

Later in the guidance, the agency indicates that a sponsor "should perform all durability and wear testing in a physiological solution (e.g., bovine calf serum) and extract the wear debris ... for characterization." FDA makes reference to two standards published by ASTM (formerly known as the American Society for Testing and Materials) for collection and testing of wear debris, specifically ASTM F561 and ASTM F1877, and recommends debris testing every million cycles.

One of the questions FDA had for the advisory panel last July was that of the reliability of device fixation and expulsion, which was prompted by the fact that the Bryan relied on no screws to affix the device to the vertebrae. However, this guidance says rather little about such considerations, stating only that "FDA may recommend you evaluate the risk of subluxation and expulsion in more than one loading direction."

On the subject of mechanical data, the guidance suggests that in the absence of "an adequate rationale, all mechanical testing should represent a worst-case construct design of the total system" rather than of individual components. FDA indicated that it would like to see "static and dynamic mechanical testing," and for most designs, "axial compression and compression shear testing."

Perhaps the most vital portion of the guidance, however, is the section covering clinical trials.

The guidance says that because an artificial disc may be a novel design and/or material, "protocols for total artificial discs may vary in scope from a feasibility or pilot study to a pivotal study" in order to work toward a PMA. Still, the guidance says that while "uncontrolled studies may be appropriate in some cases for feasibility studies ... FDA recommends ... a multi-center, randomized, prospective, concurrently controlled clinical trial" for pivotal trials.

Sponsors that see such trials as impractical or unnecessary to support the safety/efficacy argument can invoke other study designs, but FDA cautioned that any other design will have to, among other things, "minimize inherent biases." A sponsor who wants to employ a meta-analysis of data from the literature will have to "provide a rationale that supports pooling of patients into a single control cohort from [those] published studies." Any sponsors eyeing this path to approval should "consult with the review branch about literature controls."

An IDE's data on inclusion criteria should describe the status of adjacent spinal levels "both radiographically and clinically," the guidance says. FDA recommends that the inclusion criteria stipulate that any case of scoliosis impose a spinal displacement of less than five degrees. A sponsor that is open to the idea of including multiple disc placements in such a study is urged to "include a statistically significant sample size" for this group and an analysis that stratifies this group from the single-disc portion of the study.

FDA's perspective on patients with scoliosis is based in part on the notion that "axial, translational and rotational forces of the spine" in non-adult patients with scoliosis "may lead to early failure" of the device. Similarly, patients with moderate to severe spondylolisthesis are deemed inappropriate for such a study, and a sponsor will have to justify inclusion of patients with a mild case of the disease. The agency also sees reason to "contraindicate total artificial disc replacement for patients undergoing revision surgery for failed fusions used to previously treat pseudoarthrosis."

Among the exclusion criteria the agency expects to see in a proposed pivotal trial is the group of patients with less than 5 mm of disc space, although the agency parenthetically suggests that sponsors "identify subjects with a remaining disc space of 5 mm-7 mm because it may be appropriate to modify this criterion based on the particular intended use or mechanics of your device." Other exclusion criteria include patients with other spinal conditions that may be exacerbated by increasing the patient's range of motion, and patients who are on "chronic medication affecting bone metabolism," such as synthetic corticosteroids and osteoclast inhibitors. The list of exclusion criteria includes eight other elements.

Patients who might be excluded strictly for the sake of simplifying the study include those "with multiple involved levels in the spine" and anyone who has had "prior surgery at the level you plan to study." Inclusion of this latter group may require a written justification for doing so.