There was real excitement two weeks ago among some FDA officials and drug and medical device lobbyists when reports came out that Senator Charles Grassley (R-IA) might leave his position as leader of the Senate Finance Committee to take over Arlen Specter’s spot as ranking member of the Judiciary Committee. Grassley has been a tenacious critic the Food and Drug Administration’s oversight of food, drug and medical device safety. He’s also fought to curtail industry payments to physicians and researchers who conduct trials on new drugs and devices.

 As it turns out, the cheering was premature. The FDA—and its new leadership—will still have to answer to Chuck Grassley, at least for the near future: The Judiciary position was given to Senator Jeff Sessions (R-Ala), although Grassley is likely take over leadership at the end of next year.

The new administration has promised to make many changes at the  FDA over the next few years—including bumping up funding by 14% in 2010 and hiring more manpower to carry out the regulatory duties proscribed by law. But one area where the agency really needs to clean house is in the oversight of medical devices.

The Center for Devices and Radiological Health (CDRH) is facing a firestorm of criticism over the way it regulates so-called Class III devices, those that carry the highest risk of harm to patients. In the last year alone, the press has reported on several alarming cases of new Class III devices being approved that have not met standards for safety or effectiveness. These reports accuse top officials at the CDRH of disregarding scientific evidence provided by their own researchers and in some cases, the recommendations of advisory committees. Instead the officials bowed to pressure from the device industry to speed up approval of some Class III devices, waiving the randomized trials required by the traditional pre-market approval (PMA) process and using a lower standard to determine safety and effectiveness. The controversial approvals involved knee implants, heart valve replacements, implantable defibrillators and devices to treat stress incontinence in women; all of which are surgically implanted in the body.

These press reports, along with a study from the Government Accounting Office that criticized the review process for devices, point to deep-seated problems in the device center. And in January, nine scientists wrote a letter to the Obama transition team, calling on the new administration to take action on medical device oversight. The scientists, whose names have not been made public, had this to say:

To really understand the problems at the Center for Devices and Radiological Health, you have to go back to 1976 when the Medical Device Amendments originally gave FDA oversight over devices. The agency “grandfathered” in devices that were already on the market; promising to review their safety and effectiveness in the future. The Amendments also created an alternate pathway for FDA approval called 510(k; it  was really meant to provide expedited approval for Class I and II devices that didn’t need as stringent a review process as drugs or riskier devices. Under this process, rather than having to show that their device is safe and effective through randomized clinical studies, device makers need only show that their device is "substantially equivalent" to one that is already on the market; a so-called "predicate device."

 In theory, all Class III devices are supposed to undergo a pre-market approval (PMA) process that includes randomized clinical trials. But the FDA started granting 510(k) status to these riskier devices that were considered substantially equivalent to older devices on the market before 1976. The process continues to this day, with newer devices being granted expedited approval because they are considered “substantially equivalent” to current 510 (k) devices.

According to Diana Zuckerman, President of the National Research Center for Women and Families, “Unfortunately, over time the definition of “substantially equivalent” was changed to include almost any product for the same medical condition. The FDA is now using the 510k process for 98% of the medical devices that they review. As a result, new products, using new materials, or a new mechanism, made by a different manufacturer, are being reviewed as if they were a mere tinkering improvement over previously sold products. In fact, it doesn’t even matter if the previously sold product was subsequently found to be unsafe or ineffective and is no longer for sale. There are medical devices on the market today that were approved as ‘substantially equivalent’ to products that were subsequently recalled for safety reasons.”

The GAO report found that the FDA uses this expedited review process far too often in approving Class III devices, putting patients at risk. The report maintains that for at least 20 types of devices, including certain pacemakers, automatic external defibrillators, and artificial hips, the FDA is allowing American patients to be exposed to minimally- reviewed medical devices. The GAO recommends that "the FDA expeditiously take steps to issue regulations for class III device types currently allowed to enter the market via the 510(k) process."

  It’s not hard to see why the FDA, chronically understaffed and under-funded, took the expedited route to approving devices. According to the GAO report,  FDA generally makes decisions on 510(k) submissions faster than on pharmaceutical submissions (the agency reviews and decides on 90% of 510(k) submissions within 90 days, compared to 90% of original pharmaceutical submissions within 295 days.) The estimated cost to the agency for a 510 (k) submission was about $18,200 in 2005, vs. a whopping $870,000 for a drug submission. Of course, medical device manufacturers have similar time/cost concerns when choosing how to apply for approval.

In an investigative article about how Menaflex, a disc-shaped implant used to replace damaged meniscus cartilage in the knee, ended up on the market, the Wall Street Journal provides a telling example of the how the medical device industry can manipulate the  approval process:  “Political Lobbying Drove FDA Process”

 In this case, the company, ReGen Biologics, first tried to get its device approved through the pre-marketing approval process. But an FDA advisory panel looked at data from the company’s randomized trials and recommended that the implant not be approved. ReGen came back to the FDA  with an application to review Menaflex again, but this time through the 510 (k) route—arguing that the knee implant was substantially similar to a device already approved for use in shoulder surgery.
 

Scientists working in several different areas of the FDA, including the head of the agency’s science and engineering laboratory, agreed that Menaflex was too different from this other device (the most obvious difference being that the knee is weight-bearing and the shoulder is not…) to be approved under 510 (k).

 But after intense lobbying from the company—including calls from a Congressional delegation from New Jersey (ReGen’s home state), the FDA cleared the device for marketing on December 18 under the expedited process.

This same kind of pressure led to a recent approval of a computer-aided detection (CAD) system that is used along with digital mammography. Over the last several years, the FDA has approved three CAD systems for mammography, despite independent research that has found the technology to be little better than visual inspection of films and more likely to detect false positives.

 In January, the FDA approved the iCAD SecondLook Digital Computer-Aided Detection System even though scientists at the agency recommended against approval, charging that the company had not sufficiently tested the device. According to internal documents from disgruntled agency scientists obtained by Gardiner Harris
at the New York Times , “An internal review said the risks of the iCAD device include missed cancers, ‘unnecessary biopsy or even surgery (by placing false positive marks) and unnecessary additional radiation.’”

 It appears that the FDA again ignored its own scientists; this time after receiving pressure from Representative Christopher Shays, a Connecticut Republican who lost re-election in November. In the documents, Shays is described as calling a supervisor at the Center for Devices and Radiological Health to express concern about the fate of the iCAD system. The device is designed to be used with digital mammography equipment made by Fujifilm Medical Systems—a Stamford, Conn.-based company that sits in the heart of Shay’s former district.

Independent research shows that the FDA scientists were right in originally questioning the benefit of CAD devices in mammography. In an April 2007 article in the New England Journal of Medicine, (“Influence of Computer-Aided Detection on Performance of Screening Mammography”) researchers found that:

“[A]mong large numbers of diverse facilities and radiologists, the use of computer software designed to improve the interpretation of mammograms was associated with significantly higher false positive rates, recall rates, and biopsy rates and with significantly lower overall accuracy in screening mammography than was nonuse.”

 The authors reported that for every time that the CAD software marked a real cancer, the radiologist had to consider about 2,000 additional false-positive markers. They also estimated that for each additional woman diagnosed with breast cancer with the help of CAD, 156 are recalled for false positives and 14 had unnecessary biopsies.

 Despite “tentative evidence of clinical benefits,” the FDA-approved technology is reimbursed by Medicare and according to the authors, “has been incorporated quickly into mammography practices” and “used in the screening of millions of healthy women.” They estimate that if all mammography facilities adopt CAD, the annual cost of mammograms in the United States could increase 18 percent, or an additional $550 million nationwide.

In the end, it is clear that the problems at the Center for Device and Radiological Health are endemic. Oversight is too lax, the industry has far too much influence over the approval process and the 510(k) loophole is clearly putting the public at risk. It has gotten to the point where medical devices are approved without clear evidence of medical or cost benefit.

 The overriding issue, according to Dr. Peter Lurie, deputy director of the Health Research Group at Public Citizen, is that the agency isn’t always using science-based information to make their decisions about which devices to approve. To do a better job, the FDA has to require clinical trials for all medical devices that could potentially harm patients and consumers. The agency also needs to establish an appropriate definition of “substantial equivalence.” They should revert to the original intent of the 510(k) process: “the review of products that are substantially equivalent in terms of intended treatment, form, what they are made of, mechanism, and function.”

  And finally, the problems with medical devices highlight similar conflict of interest issues that permeate all medical research. Researchers need to conduct their studies without the interference of industry,  and FDA officials must do a better job of supporting them. That means instituting clear rules about financial disclosure and limiting the waivers offered to advisory board members. If, as the device industry charges, these changes will slow innovation—perhaps that won’t be such a bad thing after all.