The Evidence Gap
Treatment, Quickly Vetted, Is Offered to Cancer Patients
After a surgeon removed a cancerous lump from Karen Medlock’s breast in November, he recommended radiation, a routine next step meant to keep cancer from recurring.
Fabrizio Costantini for The New York Times
The Evidence Gap
Articles in this series will explore medical treatments used despite scant proof they work and will consider steps toward medicine based on evidence.
Fabrizio Costantini for The New York Times
But he did not send her for the kind of radiation most women have received for decades.
Instead, the surgeon referred her to a center in Oakland, Calif., specializing in a newer form of treatment where radioactive “seeds” are inserted in the tumor site. It could be completed in only five days instead of the six weeks typically required for conventional treatment, which irradiates the entire breast using external beams.
To Ms. Medlock, it seemed an obvious choice. The newer treatment ”” given through a system called MammoSite ”” has been performed on about 45,000 breast cancer patients in this country since the Food and Drug Administration cleared it for use in 2002.
Only when Ms. Medlock, 49, sought a second opinion did she learn a startling truth: MammoSite is still highly experimental.
The MammoSite system is among the thousands of devices the F.D.A. lets onto the market each year after only cursory review and with no clear evidence that they help patients. Doctors are free to use those products as they see fit, without telling patients that the devices are not proved. And because the doctors are frequently paid more by Medicare as a way to compensate them for the extra time and expense of adopting new procedures, these unproven products can become widely adopted.
F.D.A. officials defend the quick-review process as a way to promote innovation. Because most new products are simply an improvement on an existing device, they say, there is rarely need for a full review.
Demanding lengthy study of such devices would be “very, very inappropriate and a waste of resources,” said Dr. Daniel G. Schultz, the director of the F.D.A.’s Center for Devices and Radiological Health.
The agency let MammoSite on the market on the basis of a study involving only 25 women that did not answer the fundamental question of how effective it is against breast cancer. Six years later, many cancer specialists say there is still no conclusive proof that it works as well as conventional radiation. The F.D.A. says it did require a label warning that the system had not been shown to be a substitute for conventional radiation.
Dr. Valery Uhl, the radiation oncologist who provided Ms. Medlock’s second opinion, outlined the evidence behind the available treatments. Ms. Medlock chose conventional external-beam radiation because of its well-documented record of success in preventing the return of cancer. To use an unproven therapy like MammoSite, Dr. Uhl said, “makes me really nervous.”
Critics say the F.D.A.’s process for reviewing medical technology, under which medical devices have become a $75 billion-a-year industry in this country, is often too lax. More devices, they say, should get the same scrutiny applied to new drugs. While that process is not perfect, a new drug is typically studied in hundreds or even thousands of patients before the F.D.A. will approve it as safe and effective.
But under the fast-track review for most devices, a product’s effectiveness is never directly established. Regulators simply determine if the device does what its maker says it does ”” in MammoSite’s case, that it delivers radiation ”” and whether it poses any undue safety risks.
“Nobody is looking to see whether they help patients,” said Diane C. Robertson, an executive with the ECRI Institute, a nonprofit group in Plymouth Meeting, Pa., that evaluates new devices for insurers and hospitals. “We’re never going to wisely allocate resources in health care unless we start to focus on what’s best for patients.”
In response to a Congressional request to study the effectiveness of the F.D.A.’s device-review process, the nonpartisan Government Accountability Office is expected to release a report next month.
Critics say that when the F.D.A. clears a device, the public may wrongly assume that the government has proof it is medically effective. F.D.A. approval has been “widely misinterpreted,” said Dr. Jay R. Harris, the chairman of radiation oncology at the Dana-Farber Cancer Institute and Brigham and Women’s Hospital in Boston.
Dr. Dennis R. Hill, the doctor who originally saw Ms. Medlock, scoffs at the notion that MammoSite is in any way experimental. “It is a proven method,” he said. He said the oncologist who told Ms. Medlock it was experimental was skeptical because she provides only traditional radiation.
But the oncologist in question, Dr. Uhl, says she has performed many radioactive-seed treatments but wants to make sure patients are fully informed about the range of options.
Of the 250,000 women in this country who are found to have breast cancer each year, around 200,000 are candidates for radiation treatment if they choose to undergo a lumpectomy or partial removal of a breast. Most still get conventional radiation.
MammoSite proponents say that is because most doctors simply recommend the treatments they know best. “There is a natural bias for radiation oncologists to do what they have been doing,” said David Harding, an executive at Hologic Inc., the company in Bedford, Mass., that markets MammoSite.
Many prominent specialists, though, say the gold standard remains conventional radiation, for which breast cancer local recurrence rates are 3 percent or less at five years. Success in preventing cancer recurrence is measured in long intervals of 5 or 10 years or more, and there has been little long-term study of MammoSite.
Even a radiation oncologist who is a leading proponent of the MammoSite treatment, Dr. Frank A. Vicini, wants to know how it compares with traditional radiation. He is directing a national study of MammoSite’s effectiveness, but he cautions that it could take decades to conclude whether it should be used in lieu of conventional radiation.
“We have to make sure patients know we don’t have 30-year data,” said Dr. Vicini, the chief of oncological services for Beaumont Hospitals in Royal Oak, Mich. “We simply don’t.”
Dr. Frank Vicini, chief of oncological services at Beaumont Hospitals in Royal Oak, is leading a study comparing MammoSite therapy with traditional radiation but says a definitive answer could take decades.
An Array of Treatment Options
A System From the Past
The nation’s system for regulating medical devices was set up more than three decades ago, when devices played a much smaller role in medicine. A growing chorus of doctors, consumer advocates and health insurance executives say it is overdue for an overhaul.
The process has become “a barrier to evidence development,” said Dr. Winifred S. Hayes, whose firm, Hayes Inc., evaluates new health care technologies.
Although federal drug regulation dates back more than 100 years, medical devices did not come under the government’s purview until 1976, after Congress responded to deaths linked to the Dalkon Shield, an intrauterine contraceptive device.
Because the new law would not apply to devices on the market before 1976, Congress did not want makers of newer products to be at a competitive disadvantage. So lawmakers provided the quick review process for any new product deemed “substantially equivalent” to something already on the market. That expedited process became known as a 510(k) review under the relevant section of the law.
But critics say that what Congress intended as a way to let simple devices quickly enter the market has expanded so much that even critical therapies are cleared without enough research.
“It is supposed to be for the Band-aids of the world,” said Diana Zuckerman, president of the National Research Center for Women and Families, a Washington advocacy group. “The 510(k) process should have been used less and less. It’s being used more.”
But defenders of the F.D.A. process, including the officials in charge of it, argue that tighter gatekeeping could deprive patients of promising alternatives. And they contend that most new medical devices introduced each year involve minor modifications to mundane items like thermometers.
To be sure, sophisticated devices like a new artificial hip or a novel heart stent do go through the same evidence-based scrutiny given to new drugs. Of the 41 medical devices that went through that full review last year and in which the F.D.A. reached a decision, 27 received approval.
But during those same 12 months, the F.D.A. reviewed 3,052 devices under the more cursory 510(k) process and cleared 2,640 of them. Critics, though, say MammoSite is significantly different from anything on the market.
On one level, MammoSite was simply the latest version of brachytherapy, a technique that delivers radiation inside the body through seeds. Brachytherapy is widely used to treat prostate cancer because it allows high doses of radiation to be directed to a small area.
MammoSite was developed as a way to deliver brachytherapy for breast cancer, but more simply. For example, earlier versions of brachytherapy required multiple catheters.
With MammoSite, a surgeon inserts a single balloon catheter in the cavity from which the tumor has been removed. The catheter remains in place during the five days that radioactive seeds are inserted into the balloon, twice a day, using a computer-controlled system. After the treatments, the catheter and the balloon are removed, and there is no remaining radiation.
The main “substantially equivalent” product that MammoSite’s maker cited in its F.D.A. application was something called GliaSite, which itself reached the market in 2001 through a 510(k) review. Like MammoSite, it uses a catheter with a small balloon to deliver radiation.
But critics point out that GliaSite is aimed at a far different disease ”” advanced brain cancer ”” and it is often tried for patients who do not have the luxury of choosing another treatment. Hologic, which acquired MammoSite and was not involved in the original application process, declined to comment on the F.D.A. process.
The F.D.A. says MammoSite was hardly a radically new technology, because radiation is a standard therapy in treating breast cancer and brachytherapy is an established technique. F.D.A. officials also cite MammoSite’s label as evidence they have warned doctors the treatment is unproved.
“At the end of the day, doctors have to read what’s on the label and make a clinical decision,” Dr. Schultz said.
Still, critics of the 510(k) system point to the sometimes disastrous consequences of letting a device be widely used without adequate study.
Under a 510(k) review, for example, the F.D.A. in 1996 cleared Protegen, a synthetic sling implanted under a woman’s bladder to prevent stress incontinence. No clinical research was required to see how it would work, and Protegen was recalled in 1999 because of a high rate of complications.
Dr. Schultz, the F.D.A. official, said that even when using the fast-track review process, the agency always considers potential risks to patients.
MammoSite’s proponents say its main advantage is that the shorter course of treatment and reduced exposure to radiation mean more women will consider radiation. Without some form of radiation, the alternative is complete removal of the breast through a mastectomy. Many women and doctors are persuaded by this argument.
But critics, say some of MammoSite’s popularity is a result of the relatively high reimbursements paid by insurers.
Surgeons traditionally do not play a role in a patient’s radiation therapy. But with MammoSite, they can earn several thousand dollars when they insert the balloon catheter that the radiologist uses to administer the radioactive seeds. That money, critics say, encourages some surgeons to recommend the treatment.
What’s more, Medicare initially set the total payments for MammoSite, including the surgeon’s fee, by one estimate at nearly $20,000 ”” almost twice the amount paid for conventional radiation. That was in keeping with Medicare’s practice of offering more generous payments for new treatments to encourage investment in equipment and training involved in a new procedure.
In keeping with Medicare’s practice, though, the payments have shrunk in recent years. Hologic says Medicare is expected to pay around $15,000 for the treatment next year, compared with $12,000 for conventional radiation. Commercial insurers typically follow Medicare’s lead on reimbursements.
In at least some cases, it is the patient who pushes for MammoSite. After Ida Daugherty, 59, was found to have breast cancer last May, a doctor advised her to consider the MammoSite treatment after her operation. But the specialist she consulted advised against it, telling her, “I don’t feel the results are in.”
After Ms. Daugherty looked at the available research, though, she chose to get MammoSite treatments in July at Beaumont, where Dr. Vicini works.
“I really didn’t want to do six weeks,” Ms. Daugherty said. She worried that the radiation might weaken her bones enough to cause a broken rib ”” a possible side effect.
The debate now among cancer specialists is whether there is enough evidence to feel comfortable recommending the treatment to women who are not part of a clinical study.
In its marketing materials, Hologic refers to five-year data indicating that women seem to do as well with MammoSite as with conventional therapy. But the cited study involved only 43 patients enrolled in the first clinical trial of the device, 25 of whom were included in the evidence provided the F.D.A.
“We all know there is no sufficient evidence to offer it outside a trial because no one knows whether it is equivalent,” said Dr. Silvia Formenti, the chairwoman of radiation oncology at New York University Medical Center.
Dr. Formenti is wary of pushing any new treatment too aggressively because she recalls another formerly popular breast cancer treatment: high-dose chemotherapy coupled with a bone marrow transplant. In the 1990s, tens of thousands of women with advanced breast cancer were subjected to the toxic effects of that treatment before discovering that it was no more effective than conventional therapies.
Other doctors see no need for further study of MammoSite, saying the practical evidence is already compelling.
“I’m a big believer and proponent and practitioner of brachytherapy,” said Dr. Bradley R. Prestidge, the chief executive of the Texas Cancer Clinic, in San Antonio, which says it has performed more than 600 MammoSite treatments, the most in the country.
Of those 600 patients, Dr. Prestidge said, only three women have had a cancer recurrence, a lower rate than in his patients with conventional radiation.
The clinical trial being led by Dr. Vicini of Beaumont is meant to provide more than anecdotal evidence. The study will look at different techniques, including MammoSite, in which radiation is delivered to only part of the breast. He says the goal is to study 4,300 patients and results are expected in 2015.
“Doing the right thing is collecting the data,” Dr. Vicini said.