FDA Rule Clears Medical Devices Without Human Testing


Laurie Kelly blinked awake from surgery at Hoag Hospital in Newport Beach, grateful for the cutting-edge technology that gave her the chance to beat breast cancer in a single day.

That June 2010 morning, Hoag doctors removed a tumor from her left breast. Then they positioned their brand-new Xoft Axxent device to deliver a high dose – 20 gray – of electron radiation to her surrounding flesh, eliminating the need for six weeks of treatment. Kelly’s healthy tissue was protected by the Axxent FlexiShield Mini, a malleable metal pad made of silicone-wrapped tungsten.

Six months later, with her ordeal apparently behind her, Kelly, 58, went in for a routine mammogram and MRI. Her stomach dropped when the film showed white specks about the size of salt grains inside her breast.

“My nightmare started all over again,” she said. “I thought, ‘Oh God, they found another spot.'”

But this time, cancer wasn’t the invader. Instead, tungsten particles from the shield had been driven into her breast during the procedure. Some of the metal subsequently migrated to her urine, medical records show.

“Even though the day I had my surgery the cancer was removed, something potentially worse was put inside of me. It was the most frightening thing I ever heard,” said Kelly, an Irvine psychologist.

Xoft voluntarily recalled the Axxent FlexiShield in February 2011, after two of the first hospitals to use the device – Hoag and Karmanos-Crittenton Cancer Center in Rochester Hills, Mich. – reported to the Food and Drug Administration that a metallic, powdery substance was found inside the breasts of 10 women treated with the device. One report described the metal as “apparently … debris left from the tungsten shield.” Since the recall, the number of patients reporting a metallic or powdery substance in their breasts has risen to 29.

The alarming outcomes brought new fuel to a longstanding debate over an FDA rule, known as 510 (k), that allows manufacturers to fast-track new medical devices to market without human testing.

The 1976 law was intended to provide quick approval of low- and moderate-risk medical devices. But after extensive lobbying by medical manufacturers, Congress changed the regulations to make it even easier to get clearance. Today, about 90 percent of the devices on the market are cleared for use through the 510 (k) process, FDA officials say.

Regulators, independent reviewers, doctors and consumer advocates agree that the 510 (k) process is seriously flawed. The process clears new inventions if they are “substantially equivalent” to devices already on the market. But there is a lot of leeway in that rule. The tungsten FlexiShield, for instance, was approved based on another product made from a different metal. In other cases, new inventions have been cleared even though they were similar to devices that proved faulty after their release.

“Patients and doctors would be shocked to learn how this is done,” said Diana Zuckerman, president of the National Research Center for Women and Families. “The 510 (k) program is the weakest and most nonsensical program in the FDA.”

Dr. Melvin Silverstein, medical director of Hoag Breast Care Center and Kelly’s surgeon, declined requests to be interviewed for this article, citing lawsuits against Hoag by Kelly and 12 other patients.

But Dr. Jack Cox, chief quality officer at Hoag, told The Orange County Register in a statement: “The physicians involved with the clinical trial took immediate action upon discovering the situation and notified all of their patients involved in the clinical research trial. In addition, as soon as the physicians and Hoag clinical trials staff became aware that there were concerns about the device, they immediately notified the FDA. … The highest quality patient care and safety are absolute top priorities at Hoag.”

Hoag also noted, in court filings, that they had warned patients in a written disclosure that the treatment they would receive was new and its safety was unproven.

Xoft, the shield manufacturer, and its parent company iCAD, declined requests for an interview.

In a response to the injury lawsuits filed in Orange County Superior Court, Xoft and iCAD have denied any negligence or liability. Among the possible defenses to the lawsuit they listed were that the injuries were the plaintiffs’ own fault or the fault of the hospital.

In a letter to hospitals last year, Xoft’s director of regulatory affairs wrote, “We have determined that the toxicity of these very pure tungsten particles is low as few health effects have been reported in humans.”

Manufacturers Defend System

Medical device manufacturers defend the 510 (k) process, saying less than 1 percent of the estimated 3,000 to 4,000 devices cleared each year are recalled because of concerns about death or serious injury.

“It’s got a good safety record,” said David Nexon, senior executive vice president of the Advanced Medical Technology Association, an industry lobbying group also known as AdvaMed. “You can’t always identify problems before the marketplace.”

Peter Barton Hutt, a former FDA counsel who helped craft the 510 (k) statute, also defended the process in an interview with the Register.

“There’s no proof that it is a broken system. I’ve yet to hear anybody suggest a better approach,” said Hutt, a Washington, D.C., lawyer. “Any system run by human beings is going to occasionally misfire.”

But when a 510 (k) device fails, the toll on patients and families can be devastating. Some examples:

  • Certain hip implants in which the ball and socket are made of metal are shedding chromium and cobalt, in some cases damaging the tissue around the device, according to FDA reports. Hundreds of thousands of these devices have been implanted internationally from manufacturers such as DePuy, Zimmer and Biomet. The FDA has received more than 16,000 adverse reports and has ordered additional, post-market monitoring by the manufacturers. DePuy, a division of Johnson & Johnson, recalled its ASR device in August 2010, after implantation in 93,000 people worldwide. About 9,000 adverse reports and 6,000 lawsuits have been filed against the DePuy device alone. The hip devices were supposed to last 15 years, but in some cases are failing after two or three. “DePuy decided it was in the best interests of patients to voluntarily recall the ASR Hip System,” the company says on its website. Zimmer and Biomet have not recalled their devices.
  • Vaginal mesh used to treat incontinence and repair a condition called prolapsed uterus has been blamed for nearly 4,000 reports of serious injury or complications, including infections and damage to internal organs, FDA documents show. The mesh has been found in some cases to be too rigid for the constant movement of a woman’s pelvic area, and the production materials are not always harmonious with the bacteria found in the human vagina. Three deaths were reported to the FDA between 2008 and 2010. There are dozens of manufacturers of vaginal mesh, but the four biggest defendants in more than 600 lawsuits are Johnson & Johnson, Boston Scientific, C.R. Bard and American Medical Systems. The manufacturers insist that surgical mesh remains a valuable option for treating prolapsed uterus and incontinence. “These products offer a safe and effective alternative to non-mesh treatment options,” Boston Scientific said in a statement.
  • An infusion device for insulin pumps built by Roche generated 500 adverse reports to the FDA because the delivery tube kept bending and kinking, causing under-delivery of insulin. Some patients became hyperglycemic, some developing sores, severe vomiting and other complications of diabetes from using the Accu-Chek FlexLink Plus infusion set. The company recalled the device in March 2011, less than a year after it was cleared for sale. The device was cleared based on its similarity to another Accu-Chek product. Roche has advised customers to stop using the FlexLink product and consult their doctor.
  • Bivona tracheostomy tubes made for children by Smiths Medical of Gary, Ind., were found to be breaking and dislodging, causing the need for emergency tube changes. The product was recalled by the manufacturer on Feb. 23. The FDA said users had trouble detaching accessories from the tube and used excessive force, causing the breathing tube to dislodge. The results could be fatal if another tube could not be immediately found. Smiths Medical noted that if the accessory is correctly connected to the affected tube, there should be no problem disconnecting.
  • Some devices classified as low-risk don’t even get 510 (k) scrutiny, such as the Octopus Nuvo Tissue Stabilizer by Medtronic Perfusion Systems, a device used to stabilize the heart for bypass procedures. The Octopus Nuvo fell apart in two surgeries in 2010, with pieces tumbling into the patients’ chest cavities. The pieces were retrieved by surgeons, and the patients were not reported injured. However, the device was voluntarily recalled in September 2010. The FDA said there was a severe danger of broken pieces causing fatal damage to the chest cavity and heart. Medtronic stressed that no other item in the Octopus family of products is affected.

Like many other devices that had received fast-track clearance and then injured patients, the Axxent FlexiShield was recalled by manufacturer Xoft, now owned by New Hampshire imaging giant iCAD, after hospitals sent adverse event reports to the FDA.

In a letter to hospitals dated Feb. 3, 2011, Steve Lin, Xoft’s director of regulatory affairs wrote, “When Intraoperative Radiation Therapy patients return for follow-up as part of their usual patient care, it is very likely that mammography and/or breast MRI exams will show evidence of small tungsten particles that were shed from the FlexiShield Mini during the IORT procedure.”

The letter added, “We have determined that the toxicity of these very pure tungsten particles is low as few health effects have been reported in humans.”

Thirteen women have since filed suit in Orange County Superior Court against Xoft, iCAD and Hoag, claiming personal injury, medical negligence and product liability.

Last year, Kelly made the painful decision to have a surgeon remove both of her breasts in an effort to rid her body of the tungsten. Even so, the metal continues to show up in high levels in her urine, records show, and she’s experienced a variety of health maladies.

“At least with the cancer, we knew we could go in and cut it out,” Kelly said. “With the tungsten, we’re shooting in the dark. We’re all lab rats.”

Kelly said she feels let down by the FDA, the device maker, the hospital and her surgeon.

“I always thought the FDA was there to protect people,” Kelly said. “Science requires investigation. You test things out. … Instead, through their lack of oversight, they put people at tremendous risk.”

The Laws

When Congress passed the Food, Drug and Cosmetics Act in 1938, little attention was given to medical devices, which consisted mostly of things like tongue depressors and stethoscopes. Devices went mostly unregulated until 1976, when Congress approved the “Medical Device Amendments” in reaction to thousands of deaths and injuries – 8,000 from the IUD birth-control device alone.

Lawmakers concluded that devices should not be subject to the same level of time-consuming scrutiny as drugs. Scientists could change one aspect of a device and still have the same device. If they changed one molecule of a drug, they had a new drug. So Congress adopted a program in which devices existing in 1976 would be grandfathered in and new devices would be judged according to their risks.

High-risk devices would receive the most scrutiny, called “pre-market approval,” which include human trials. Low- and moderate-risk devices would be reviewed through the 510 (k) program, according to whether they were “substantially equivalent” to devices already on the market. Industry officials say that if a device is deemed substantially equivalent, then it is “no less safe” than an existing device.

Congress expected that eventually most devices would be new and subject to the highest level of scrutiny or “pre-market approval” and that 510 (k) would ultimately be phased out, said Dr. David Challoner, a retired medical professor and chairman of a recent Institute of Medicine study critical of 510 (k).

“Of course, that is not what happened. We have the vestigial appendage for what was a transitional market process,” Challoner said in an interview. “What happened is that industry, with the help of Congress, has kept this more convenient clearance process in place.”

In 1997, the review process was further modified by Congress to adopt the “least burdensome” principle. That calls for reviews to be done in a way that involves the smallest investment of time, effort and resources on the part of the manufacturer and the FDA. Industry groups argued that the principle increased efficiency, which would lead to greater safety and effectiveness.

Sharon Segal, vice president of technology at AdvaMed, the industry booster group, said “least burdensome” means working smarter, not harder.

“There’s no lowering of the standard of safety and effectiveness,” Segal said. “The premise is to make the process more efficient.”

But some critics say the “least burdensome” principle was an effort to clip the FDA’s wings.

“The ‘least burdensome’ wording has always resulted in shoddy reviews with a lack of safeguards,” said Zuckerman, the consumer advocate. “If you go too far with least burdensome, you end up with people dying.”

Over the years, investigators at the Government Accountability Office have criticized the FDA for allowing some high-risk devices to be cleared through the less rigorous 510 (k) process. A loophole had allowed some Class III, or high-risk devices, that were in use prior to 1976, and their successors, to be cleared through 510 (k). The GAO concluded in a 2009 report that too many devices were going through the fast-track process when they required more intense testing.

“Although Congress envisioned that class III devices would be approved through the more stringent (pre-market approval) process, and the Safe Medical Devices Act required that FDA establish a schedule for doing so, this process remains incomplete,” the GAO said in January 2009.

The program has become so convoluted that an 18-month study by the Institute of Medicine, an independent, nonprofit organization that seeks to provide unbiased scientific advice, concluded last summer that 510 (k) should be scrapped and rebuilt because it does not ensure safety and effectiveness.

“The 510 (k) process cannot be transformed into a premarket evaluation of safety and effectiveness as long as the standard for clearance is substantial equivalence,” the study concluded. “In practice, the assessment of substantial equivalence generally does not require evidence of safety and effectiveness.”

The report also chided the FDA for being lenient with violators and for having inadequate databases of previously cleared devices. “Because data systems are inadequate, the FDA does not have the ability to trace 510 (k) decisions,” the report said.

Challoner, who chaired the study, said, “The current program is so defective that it cannot be perfected.”

The FDA has no plans to scrap the program but has initiated measures intended to make it more rigorous in some spots and more streamlined in others. For instance, the FDA is updating its appeals process for devices that don’t receive clearance and is clarifying the amount of information needed for each submission.

Zuckerman doesn’t think the FDA is going far enough. She and other consumer groups say the 510 (k) program should require clinical testing for all implanted or life-saving devices.

“I’m just amazed an implant, like a hip or a knee, is allowed to be sold without being tested in a human being,” she said. “Patients shouldn’t pay for unproven devices that are not even being studied.”

Jeffrey Shapiro, an industry lawyer in Washington, D.C., disagrees.

“If you want to review the safety and effectiveness of every device starting from scratch, you’re not going to have any medical devices,” Shapiro said. “The device industry would grind to a halt.”

While the actual 510 (k) reviews are completed in about 90 days, the time it takes between submission and clearance has grown to as much as 161 days, according to the latest GAO study in February. A pre-market approval can take as long as 627 days, the GAO says.

David Gollaher, president of California Healthcare Institute, an industry think tank, testified in September before a congressional subcommittee on health that FDA reviewers have become overly sensitive to possible risks because of criticism from consumer groups and are overlooking the potential benefits.

“Increasingly companies have encountered an attitude of risk aversion at the FDA that is expressed in demands for more and more data and larger clinical studies,” Gollaher told the subcommittee. “As a result, drug and device development costs have skyrocketed.”

Matthew Jenusaitis, president of OCTANe, a high-tech development group in Aliso Viejo, said the need for safety should be balanced against the cost of losing jobs.

“Any device that fails is a bad thing, but at the same time, these devices have the ability to save lives and stimulate the economy. We’re at a risk in the United States of losing this industry,” Jenusaitis said.

Congress has responded with 10 new bills that legislators say will make the pre-market process more efficient. All the bills were introduced Oct. 14 by members of the House Energy and Commerce Committee after robust lobbying by the industry. Consumer advocates say the new legislation would just weaken regulation.

“The industry keeps saying we’re going to stifle innovation, but it’s not innovative if it’s not safe and does not work,” said Lisa McGiffert, manager of the Safe Patient Project for Consumers Union, a national consumer advocacy group that has been raising alarm about 510 (k).

Zuckerman added: “I want jobs in my community as much as anyone, but I don’t want jobs for people making devices that are going to break inside my mother’s body.”

Rep. Rosa DeLauro, D-Conn., says lawmakers should be working to keep dangerous products from slipping through.

“The 510 (k) device approval process must be reformed – as it stands today, the process is critically flawed. I have long called for the FDA to review this process and reform it to better protect patients,” DeLauro said in an email.

A bill introduced this year by Rep. Edward Markey, D-Mass., would allow the FDA to turn down products that are based on previously cleared devices with known safety defects. The new versions would no longer be eligible for clearance simply because they are “substantially equivalent” to the defective versions.

“If an automobile is recalled for a major safety problem, we wouldn’t allow future models to repeat this same flaw, and the same should be true for the medical devices used in our bodies,” Markey said.

Meanwhile, more complaints have come from within the FDA itself, from reviewers who say they are overworked, underpaid and saddled with incomplete submissions from manufacturers. There is high turnover, with almost half of the review staff having four years or less experience. Front-line supervisors have three years or less, according to the FDA.

The FDA is negotiating with the industry and Congress to raise the fees that help fund the program. Currently, large companies pay $4,000 per submission, while smaller companies pay $2,000. Consumer groups say the large companies should be paying more than 10 times what they pay now. The industry has tentatively agreed to double the fees.

Case Study of a Device

Like other medical devices cleared by 510 (k), Xoft’s Axxent radiation system has a lengthy family tree.

The system, as well as others on the market, uses a portable machine wheeled inside the operating room to deliver radiation to a cancer site in a single dose. The process is known as Intraoperative Radiation Therapy, or IORT.

With the Xoft device, electron radiation is delivered after the tumor is removed but before the incision is closed. A syringe-like applicator, with an inflatable balloon at the end, is used to guide the radiation to the treatment area. A miniature X-ray tube inside the applicator, operating at 50 kilovolts, generates the radiation. A shield is used to protect healthy tissue and organs from receiving harmful radiation.

The treatment can be delivered in less than 20 minutes, rather than the standard six weeks of post-operative radiation treatment.

FDA records describe the history that led to clearance of the radiation system and shield.

In April 2005, Xoft submitted its application for the Axxent Electronic Brachytherapy System for the radiation of breast cancer, based on the earlier clearance of three different devices. The company described the system as “very similar” to devices that were cleared between 1995 and 2004. One of those devices, the Photoelectron Corporation Photon Radiosurgery System, was cleared in 1999 based on three previous radiation systems. The FDA cleared the Xoft system in December 2005.

In February 2009, Xoft submitted an application for the FlexiShield, described as a flexible pad made of tungsten and silicone. It called its shield “substantially equivalent” to Arplay Medical Lead Blocks, which were cleared in 2001. That device is made of lead with mounts of screws or rubber. Arplay’s device had been cleared based on yet another device – Aktina Photo Beam Blocking System, cleared in 1998. The Aktina application was based on the 1994 clearance of the Medtec 360 Degree Rotating1/2Beam Block.

The FDA cleared the Xoft shield in June 2009. Documents do not reflect any concern by FDA reviewers that the FlexiShield was made of a different metal than its predecessors – tungsten instead of lead.

It’s unclear exactly why Hoag chose the new Xoft system for its foray into IORT.

Dr. Alice Police, a Costa Mesa surgical oncologist, said she participated in meetings at Hoag to discuss purchase of a radiation system. She advocated for a system made by Carl Zeiss Meditec, based on a 10-year trial in nine countries that followed more than 2,000 women after treatment.

Police said negotiations were under way with Zeiss when suddenly the plan changed without explanation.

“The next thing I knew, they had canceled their negotiations with Zeiss and chosen the Xoft (system),” Police said. “I was dismayed to hear they had gotten a system with no data.”

Police was so upset she decided to lease the Zeiss system, for $8,000 a month, for use in her surgery center, rather than use the Xoft system at Hoag.

After recalling the FlexiShield, iCAD applied in May 2011 for clearance for a replacement shield, this one made of stainless steel. That application was based on the 1999 clearance of a shield made by Photoelectron Corp. This time, FDA documents show, iCad tested the shield for shedding and found none occurred during “worst case clinical usage.”

The FDA cleared the Axxent Radiation Shield – Rigid for use in July 2011.

ICAD is continuing with its Xoft IORT radiation treatment, using the new shield, launching a 10-year study enrolling 1,000 patients at 50 study sites throughout the United States and Europe.

Living With Tungsten

Kelly learned she had an aggressive form of cancer in May 2010 after discovering a lump in her breast. She was referred to Silverstein, medical director of Hoag Breast Care Center. When she saw him, he told her she was a good candidate for IORT.

“He was just totally excited about this new technology,” Kelly recalled. “He was like a super fan.”

He told her he’d used the technology at USC, where he worked before joining Hoag.

Kelly liked the idea of missing less time from work and the appeal of a shorter recovery. She researched IORT and USC and saw that the 10-year international clinical trial using a system by Zeiss was nearly complete.

“It never occurred to me to ask, ‘Is this the same machine?’ I was so frantic to get the surgery,” she said.

Kelly was the first Hoag patient to undergo the lumpectomy and radiation in June 2010, not realizing that the device was in fact made by a rival company.

“I felt very lucky that I lived in a community that had such cutting-edge technology and I got to be part of it,” Kelly said.

Six months later, she returned to Hoag for a follow-up mammogram. She was told she needed an MRI for more detailed images. Tiny particles were “snowing” all over the film of her left breast tissue, where the shield had been placed.

Kelly underwent a biopsy, but the particles, spread over 9 centimeters of her breast, couldn’t be identified by the local lab.

Hoag sent the sample to the renowned Mayo Clinic in Minnesota, which determined that it was tungsten. One of the tissue samples contained 207 micrograms per gram of tungsten, lab records show. The second contained 67,264 micrograms per gram.

The metal subsequently migrated through her body and showed up in her urine. A lab report from May 2011 indicates the tungsten level measured in her urine at 26 micrograms per gram of creatinine, a waste product that is analyzed for foreign substances. A July 2011 urine test shows a tungsten level of 19 micrograms per gram.

Although Xoft has maintained that the tungsten particles will not harm the women, not everyone agrees.

Doctor’s Data, the Illinois-based lab that examined Kelly’s urine, calls any level above 0.4 micrograms/gram “potentially toxic.” Long-term industrial exposure to tungsten has been associated with “hard metal lung disease” and lung cancer, the lab reported.

Kelly found a Canadian scientist, Koren Mann of McGill University in Montreal, who researches tungsten. Their conversation did not ease her fears.

“I don’t think it is inert at higher levels,” Mann said in an interview, adding that more studies need to be done. “If you look at higher doses, there may be toxic effects, changes to the immune system. It may damage the DNA. There isn’t a lot of information.”

With tungsten in her flesh, Kelly feared it would be impossible for doctors to detect a recurrence of cancer because of the obstruction on the screening films. Although the tungsten appeared in only her left breast, she worried the foreign substance in her body could cause more cancer. For those reasons, she opted to undergo a double mastectomy in April 2011.

“I decided I don’t want to deal with any of this again,” she said. “It was too painful.”

Kelly said no one from Hoag ever apologized. She said she felt she had no choice but to go to a lawyer in search of answers and protection for herself and other women. She filed the suit using a pseudonym.

“In the beginning, if Hoag had just stood up and said this is a terrible, terrible thing that’s happened; we will endeavor to do anything to help you, I never would have considered suing them,” she said.

According to a Feb. 25, 2011, letter from iCAD to Silverstein, the manufacturer offered to pay for costs of biopsies, blood and urine testing, MRIs and other treatment for patients who had tungsten in their bodies. But Kelly and her attorney, Jeffrey A. Milman, say iCAD has ignored their requests to pay her medical bills.

Kelly says she has experienced a variety of health problems that she believes are tied to tungsten. Tungsten is expelled through urine and for a while urination was extremely painful for her, though doctors found no signs of a bladder infection.

She’s also worried that the faulty shield means her heart was exposed to a high dose of radiation. She feels persistent fatigue.

“I just don’t have my energy back,” she said. “I don’t do hiking. I don’t do a lot of things. I used to love to cook. Now I go to work, and I’m too tired to cook when I get home.”

In January, Kelly underwent an 11-hour reconstructive surgery. Afterward, she developed a staph infection and had to be hospitalized again for a week and take intravenous antibiotics for a month. When her body is strong enough, she will have a final reconstructive surgery.

“I’ll never be without a lot of scars. It’s not like before when it was a cancer scar,” Kelly said, “and that’s a badge.”