Medical Device Daily Washington Editor

GAITHERSBURG, Maryland – The effort to reclassify bone growth stimulators (BGSs) undertaken by RS Medical (Vancouver, Washington) seemed to progress fairly well despite a large party in opposition at the June 2 meeting of CDRH's orthopedic and rehabilitation devices panel. The two sides traded blows on a number of points and the first round of voting by the panel seemed to favor the petitioner's position, albeit by a narrow margin.

In the end, however, a difference of opinion on engineering between an FDA engineering expert and two orthopedic surgeons on the panel derailed the effort, leaving makers of this class of devices in need of a premarket approval (PMA) to get a new machine to market regardless of any similarities with existing technology.

The effort to reclassify bone growth stimulators included only two modes of field generation, namely capacitive coupling (CC) and pulse electromagnetic field (PEMF). At present, the five PMAs in existence for these machines cover both these technologies as well as a combination configuration known as CMF. The two indications applied for were treatment of established, non-union fracture secondary to trauma and as an adjunct to lumbar spine fusion surgery. The five existing PMAs cover both these indications along with treatment of congenital pseudoarthrosis and as an adjunct to cervical fusion surgery in patients at high risk for non-fusion.

The opposition to reclassification included officials from Healthonics (Bethesda, Maryland) and dj Orthopedics (Vista, California), now known as DJO (Medical Device Daily, June 5, 2006). Both firms manufacture bone growth stimulators.

Representing DJO, Gary Friedlaender, MD, professor of orthopedics at Yale University School of Medicine (New Haven, Connecticut), said that the proposed reclassification of BGS machines was “unwarranted, potentially problematic and risky” because the 510(k) process was inadequate to deal with risk when a “direct biological effect was involved.”

Friedlaender said that public health would be “much better served by providing clinically significant endpoints other than X-ray” radiographic measures and that the mechanisms of action of BGSs were only partially understood.

Barbara Boyan, PhD, a professor of biomedical engineering at Emory University (Atlanta) and a co-founder of OsteoBiologics (San Antonio) also advocated for not reclassifying BGSs, insisting that the petition “does not meet FDA's regulatory requirements” in part because the petition covers two modalities and two indications.

James Ryaby, PhD, vice president of Orthologic (Tempe, Arizona) and an adjunct professor of tissue engineering at Arizona State University (also Tempe) argued that the application did not fulfill FDA's request for “a range of technical specifications” and that “seemingly minor changes to these waveforms” can have meaningful impacts. He said that the difference between a wave generated with 14 and 15 hertz could influence calcium uptake and that there exists “no adequate public database” defining the required signal to achieve the desired effect.

In rebuttal, Robert Sheridan of Sheridan Consulting (Wilmington, North Carolina) answered one of Ryaby's complaints by noting that the FDA does not require the kind of device specifications Ryaby discussed if the devices “do not differ significantly in purpose.” Sheridan pointed out that the FDA had not required new clinical trials for the reclassification of pedicle screws, which are also used for spinal fusion, and that the agency had already concluded that bone-growth stimulators employing CC and PEMF sources “do not differ significantly” as they employ “essentially the same mode of action” and “the same risk.”

Sheridan said that the petitioner recognizes that a change in signal can impact the process of bone healing, but insisted that the review process under 510(k) would ensure that “the public will not be exposed” to unsafe devices because of design specification requirements under quality systems regulations.

Edmund Frank, MD, a professor in the neurological surgery department at Oregon Health & Science University (Portland), said that the petition included 41 studies for a total of 6,500 subjects and that 28 of those were prospective studies, a feature that the opponents of classification claimed was wanting. Of the non-union studies, outcome measures were said to have included radiological and clinical definitions of union, and some of the studies included controls as treated by sham devices.

Frank said that the 29 clinical studies for treatment of non-union fractures included more than 5,300 subjects and that 19 of those were prospective studies.

Christine Brauer, PhD, of Brauer Device Consultants (Bethesda, Maryland), addressing risk and Class II controls for the petitioner, argued that risks associated with BGMs are “typically transient” in reference to the risks of electrical shock, burn and skin irritation/allergic reaction. She added that general controls for devices, such as design validation and specific inputs and outputs generated by a specific design were sufficient to deal with many of the concerns associated with the 510(k) review process.

Michael Janda of the orthopedic devices branch at CDRH addressed a number of issues in connection with current BGS devices, including adverse events. These included two reports of death in connection with BGSs, both of which involved patients with implanted cardioverter defibrillators (ICDs). No causal connection was made between the BGS and these two deaths, however.

One of the panel members, Sanjiv Naidu, MD, professor of orthopedics at Pennsylvania State College of Medicine (Hershey), questioned whether the limited data presented justified the reclassification, and was among those calling for extensive clinical trials to support any new applications via 510(k).

Another panel member – albeit a non-voting one – Pamela Adams of Etex (Cambridge, Massachusetts) argued that it probably did not make sense to impose “a higher standard” of proof than was required of the PMA holders.

Cedric Walker, PhD, a professor of biomedical engineering at Tulane University (New Orleans), said that “good engineering design” can prevent burns and that most of those were due to problems with the battery pack rather than due to the actual stimulus. However, he made the case that to be on the safe side, it might be best to avoid placing a PEMF device on a patient with an ICD.

As the panel addressed the formal questions dealing with the application, it appeared that the request would pass with the proviso that a 510(k) applicant would conduct a clinical study of its device to ensure safety and efficacy. However, the question of the reproducibility of the waveform began to take its toll.

In the first go-round on the question of whether special controls were sufficient to provide the necessary reassurances, the majority of the panel voted aye, but panel member Choll Kim, PhD, assistant professor of spine surgery at the University of California, San Diego argued that the waveform question had not been dealt with satisfactorily and a second vote, prompted by the nagging question of how much clinical trial liability to recommend, resulted in a failing vote on the waveform reproducibility question.

Adams commented after the vote that “I regret that the decision has gone this way,” and that given the proposed controls, these devices should be dealt with via the 510(k) process because the FDA needs the time to deal with “higher-risk devices.”

Sheridan told Medical Device Daily that while “it would have helped” to have had more corporate backing for the reclassification effort, he nonetheless was of the opinion the principal problem was not that of sufficient industry interest in reclassification. “I don't think the panel understood” that a clinical trial would have been more than adequate to deal with the concerns raised by the opponents of reclassification, Sheridan remarked.