I>Medical Device Daily Washington Editor
GAITHERSBURG, Maryland — The definition of "first-of-kind device" is not the same for everyone, but a combination of 510(k) components that brings together materials that have not previously been mated constitute a new device for FDA. That was one lesson from Tuesday's meeting of the orthopedic and rehabilitative devices advisory committee, which reviewed the PMA for the CoMplete acetabular hip system made by DePuy Orthopedics (Warsaw, Indiana).
The other lesson is that a well-engineered clinical trial along with outstanding safety and efficacy data can swiftly sway advisory committees as DePuy showed, snaring a unanimous vote for approvability. While the panelists expressed concerns about precipitates from the novel combination of materials – previous hip systems have all used the same materials for the interfacing components – those concerns were allayed by what one panelist said was a PMA study that "was remarkably well done."
The CoMplete, so named because it consists of a ceramic-on-metal (CoM) acetabular configuration, includes a femoral head made of Biolox ceramic, made by CeramTec (Plochingen, Germany). The other component, likewise cleared by FDA according to agency documents, is DePuy's Pinnacle Ultamet acetabular insert, a cobalt-chromium-molybdenum alloy manufactured with 28mm and 36mm inner diameters, originally intended to be mated with the company's titanium alloy outer shell, also from the Pinnacle line of products.
One of the sticking points where both the panel and FDA were concerned was that of precipitate matter from wear, but the loss of 85 patients during the 24-month follow-up on the pivotal trial also imposed some drag on the application.
Tara Shepherd, FDA's lead reviewer, addressed the issue of cleared components combined to form a PMA device by remarking, "FDA clears systems, not individual components." She said bench testing showed that the steady-state wear at between 1.5 and 5 million cycles for the CoM configuration at a diameter of 36mm averaged .02 cubic millimeters of debris, better than for metal-on-metal (MoM), which was said to have averaged more than half a cubic millimeter, but not quite as clean as ceramic-on-ceramic (CoC), which the agency's documents said produced .0075mm of debris.
Shepherd noted that a check of ionized material released into serum returned the same pattern as for the steady-state wear study. CoC returned the best numbers, followed by the CoM. However, the distance between the CoM and MoM configurations were not as wide as the panel would have liked, a fact that came up during discussions in the afternoon.
Linda Terry-Choyke, medical officer for the orthopedic joint branch at the Center for Devices and Radiological Health, said the prospective study with concurrent controls enrolled 194 subjects for the study arm and 196 as controls, who received MoM systems.
Part of the motive for a fairly large device study was that the company included three different acetabular shells in the study arm, although a slight majority of those shells were from the company's 100 series of shells, a plain shell with only one opening for fixation. Also well represented in the study arm was the Sector shell.
Terry-Choyke told the panel that the safety data showed that "intraoperative complications were twice as high in the control" arm, but that generally speaking, "these intraoperative complications are not unique to this type of surgery." Post-operative complications were higher in the study group, but not to a level of statistical significance.
At 24 months, acetabular radiolucencies numbered the same for both groups – 21 – but femoral radiolucencies were higher in the controls, 60 to 52. No revisions or removals were associated with radiolucencies to date, Terry-Choyke said, although longitudinal data suggest "early signs of loosening."
Perhaps the most important remark Terry-Choyke made came at the end of her discussion, when she said, "this PMA represented a well-run clinical trial," and that the trial's endpoints "attained statistical significance."
The panel did not wait until the end of the meeting to hash out some of the questions regarding a post-approval study (PAS), with some discussions of the term of follow-up running to a decade. However, some panelists noted that "there are some confounding reasons why it may be difficult to reach 85% compliance" in a PAS with a follow-up period of 10 years, with issues such as transportation and relocation serving as barriers. Other issues might be financial, including a patient's insurance coverage.
As noted earlier, the panel was uncomfortable with the metal ion data from pre-clinical testing, and voted unanimously that "this device should carry with it the same indications as metal-on-metal bearings," summarized panel chairman Jay Mabrey, MD, of Baylor University Medical Center (Dallas). FDA also asked the panel whether it was sufficiently comfortable with the data for the two shell types and the two different shell and head sizes. Mabrey summarized, "the panel is unanimous in its support of the two cups and different size heads." The panel had no problem with safety or efficacy, either.
As is often the case with advisory committees, this one flirted briefly with the notion of a randomized, controlled trial for the PAS, but Danica Marinac-Dabic, PhD, chief of the epidemiology branch at the Center for Devices and Radiological Health, informed the panel that FDA is not fixated on randomization, although she expressed a preference for controls.
"We are looking for novel ways to integrate the findings from other studies" to provide a control cohort for PASs, Dabic told the panel, adding, "there are ways to perform the math analyses" of completed studies that will make such a cohort appropriate for comparison.
After further discussion, Mabrey said, "the opinion of the panel is that setting up a new control group would be overly burdensome" and that the needed data can be obtained by other means. As for the suggested duration of follow-up to 10 years, he said, "the opinion of the majority of the panel that this length of time should be extended" beyond five years, adding, "five years, for any total joint, does not give enough data on its efficacy."
Serum levels should be tested for metal ions, Mabrey said, but conceded that "at this point, there do not appear to be any standards on how to interpret those." He also said that the procedure is problematic for surgeons "because it's highly specialized and very expensive." Company officials indicated a willingness to follow out to 10 years and to conduct metal ion analyses.
Pam Plouhar, DePuy's VP for clinical affairs, told Medical Device Daily that sponsors of other total hip replacement systems have obtained reimbursement codes, but she could not address a reimbursement strategy because "it's not approved yet." As for whether a follow-up study should include tests of urine as well as serum to evaluate renal capture of ions, she said "I think we're all interested" in that aspect of particulate matter capture, but that the company will have to work out those details with FDA.
Mark McCarty, 703-268-5690; firstname.lastname@example.org