FDA committee recommends clearance for GE’s digital mammography unit

The full-field digital mammography landscape shifted with breathtaking speed last month when the Food and Drug Administration’s Radiological Devices Panel of the Medical Devices Advisory Committee unanimously recommended clearance for GE Medical Systems’ unit, Senographe 2000D. If the FDA’s Center for Devices and Radiological Health takes the panel’s recommendation—which it most likely will, according to FDA sources—GE could have clearance for its unit by the end of this month.

The panel made its recommendation at a public meeting in Rockville, MD, Dec. 16, only weeks after GE submitted a PMA for the system. Waukesha, WI-based GE met with the FDA to discuss its application in the early weeks of October, submitted it Oct. 29, and received expedited review. GE’s clinical data for its application included 662 patients, 641 of whom were diagnostic patients. The remainder were patients with known cancer. The studies were conducted at Massachusetts General Hospital, the University of Colorado, the University of Massachusetts, and the Hospital of the University of Pennsylvania.

The panel’s counsel to approve the PMA came with three conditions, however. The committee requested that GE refine the PMA’s executive summary to emphasize differences between the application’s study population, which was enhanced by patients with known cancers, and a true screening, or diagnostic population. The submission was based on clinical trial results from hard-copy reading; the committee requested that GE deploy workstations as soon as possible for clinical testing. Finally, the panel also requested that the company submit an amendment to its PMA for soft-copy use of Senographe, for which the FDA would expedite its review.

“The FDA’s original protocol (for digital mammography clearance) was to compare film-screen, which is only available in hard copy, with digital, which meant taking digital images and putting them on hard copy,” said Scott Donnelly, GE’s vice president of global technology. “The recommendation of the panel is to approve (Senographe for hard copy use), but they’re emphasizing that the power of the machine can only be fully exploited in soft copy.”

The company concedes that it was planning to take the 510(k) route, and that it changed its strategy after meeting with the FDA in October. Both GE and the FDA were interested in facilitating the technology’s commercialization, Donnelly said.

“(Our application) was quick, and we had to do a huge amount of extra work, including a more extensive accounting of the device’s characteristics, manufacturing data, and physical audits of the manufacturing facilities,” he said. “The time (in which this was) converted from an intent to do a 510(k) to a PMA is unheard of in the industry. We met with the FDA and agreed that the data collected to date was sufficient to prove the safety and efficacy of the machine. So the motive from both parties was to get the technology to market as soon as possible.”

In its informal communication to vendors this fall, the agency suggested the PMA as an alternative to the 510(k), and emphasized that a post-approval screening study comprising as many as 20,000 patients could be required. The panel reviewing GE’s application strongly recommended against this kind of study, according to Robert Doyle, executive secretary of the Radiological Devices Panel and a scientific reviewer in the FDA’s radiology branch.

“During the discussion, the panel emphasized that for us to require a post-market study would be a waste of resources, Doyle said. “They concluded that (GE) had proven that the device is safe and effective.”

The panel made this particular point in part because of approximately 7000 analog and digital images that have been gathered by the U.S. Army using GE equipment. Although these films were not included in GE’s PMA, the panel held the opinion that because of the existence of this large pool of films—which could be interpreted by a wide range of readers and used to support the PMA—a post-approval study would be unnecessary, Doyle said.

The industry and clinical community respond. Despite the fact that digital mammography’s long haul to regulatory approval may soon end, industry reactions to the panel’s recommendation were mixed. Other manufacturers and members of the clinical community continue to be frustrated by the FDA’s suggestion that vendors pursue the PMA.

At the meeting, Fischer Imaging presented its views to the panel, arguing that the PMA is not the appropriate method of approval for digital mammography, since it effectively puts digital mammography systems into a Class III device category, which are the highest-risk devices regulated by the FDA. Fischer’s letter reminded the agency that in 1998, an advisory panel recommended that clinical trials were not necessary to clear the technology, and concluded that a features analysis would be sufficient to establish digital’s equivalence to screen-film. The FDA has yet to publish public draft guidance on how it would like vendors to determine substantial equivalence.

In its comments, Fischer suggested that digital mammography’s clearance be categorized as “an issue of scientific controversy” and referred to a third party, such as the National Cancer Institute or the Radiological Society of North America. In addition, the company urged the panel to approve GE’s application as a 510(k), not a PMA, provided it met the policy the agency professed in a 1999 vendor communication that studies supporting digital mammography must establish the technology’s sensitivity and specificity.

The panel also received some criticism on the clinical side. In a summary of the meeting’s proceedings sent to the International Digital Mammography Development Group, Dr. Daniel Kopans, director of breast imaging at Harvard Medical School, wrote that GE’s data demonstrated equivalency between digital and screen-film mammography. Kopans criticized the FDA’s comparison of full-field digital technology to automated cervical slide reader technology, which is not like digital mammography but like computer-aided detection technology, according to Kopans. The PMA’s problematic aspects were acknowledged at the meeting, Kopans said: Although the FDA emphasized that amendments to vendors’ PMAs would get expedited review, the agency confirmed that these amendments would go through a rigorous review process.

Other vendors must decide how to proceed. How are other digital mammography vendors responding to the panel’s recommendation? Trex Medical of Danbury, CT, which has experienced more than its share of tribulations from the regulatory process, plans to submit a PMA.

“The FDA has made it clear that a PMA is what they will support and recommend. We’re going about pulling that together,” said Bill Webb, CEO. “We’ve maintained an active dialogue with the FDA and believe that we can (prepare a PMA) in a way that won’t cause us too much delay. We know what the best digital approach is, and we think when our digital product comes out, it will have better image quality than what we’ve seen so far.”

Firms such as Fuji Medical Systems and Siemens Medical Systems continue to move forward with their own digital mammography efforts. Fuji is in discussions with the FDA about a regulatory approval application for its digital mammography products, according to Bill Cioffi, the Stamford, CT, firm’s national marketing manager for conventional imaging systems. Siemens is evaluating its course of action, according to Anne Marie Keevins, manager of the Iselin, NJ-based company’s special products division.

“We began clinical trials at Medical College of Virginia in October. We had previously planned those trials, but we’ve certainly set them up to be able to fulfill PMA requirements,” Keevins said. “Until the FDA comes out with more specific guidelines, we’re conducting clinical trials in a way conducive to the PMA.”