The Cancer Letter asked David Challoner, emeritus vice president for health affairs at the University of Florida, to discuss the intricacies and effects of the FDA’s 510(k) medical device approval process, and the parties that influence it.
The process has come under scrutiny after laparoscopic power morcellation procedures were found to spread previously undetected sarcomas inside benign fibroids.
“Not only was the technology put to a new use without any real evaluation, but the signal from the clinical environment back to the regulatory environment that ‘there’s a problem,’ is a very low level signal in this particular case—until a wise doctor had it used on his unfortunate wife, and figured out what the hell was going on and blew the whistle,” said Challoner.
Challoner chaired an Institute of Medicine committee tasked by FDA and Congress in 2009 to review the 510(k) approval process.
“At the end of two years, the [IOM] committee unanimously recommended that the 510(k) process was—we didn’t use the word irrational, but we certainly could have—defective and should be replaced,” he said.
“We recommended that FDA go back to the drawing board and get all the players around a table and devise a safer way to get into the market than just a clearance process.”
Challoner spoke with The Cancer Letter’s Matthew Bin Han Ong.
Matthew Ong: How did you get involved in the debate on the 510(k) process?
David Challoner: I began as a full-time clinical researcher in endocrinology and found that science, health and health care policy was a nice avocation. I became involved in academic medical administration fairly early and went finally to the University of Florida as vice president for health affairs in 1982, which is, in our system, the chief executive officer for a large multi-college and hospital academic medical center.
I was fortunate enough to be elected as a member of the Institute of Medicine early in my career and was actively involved since the 1970s serving on or chairing study committees. I served as foreign secretary and I also on the National Research Council governing board.
So I have policy experience in Washington, D.C., and racked up a lot of miles on Delta between Florida and D.C. over 35 years.
In 2009, Harvey Feinberg, then the president of the IOM, asked if I would be willing to chair the study on recommendations for improving the 510(k) that had been requested of the IOM. It was a well-balanced committee—we had members who had been in the FDA, we had members who had been inventors, we had members who were lawyers for device companies and who were now retired. We had the full spectrum of points of view as members of the committee.
My only experience with the device industry had been during my time in Gainesville in the late 80’s to mid 90’s. I had been on the board of a device company in Miami, Fla., called Cordis Corporation, which was then purchased and is now a part of Johnson & Johnson. Thus I had some past experience in some of the issues with devices. That’s why they asked me to chair the committee.
I have learned a lot about 510(k) and the device industry since 2009. At the end of two years, the committee unanimously recommended that the 510(k) process was—we didn’t use the word irrational, but we certainly could have—defective and should be replaced. There is no mechanism by which the health of the public could be protected when moderate risk devices were only cleared to the market place, not approved, under 510(k). They are cleared to the market based on, in some cases, a multi-year, multi-device daisy chain of predicate devices.
The reason this whole thing was put in place the way it was in 1975 was because that was when devices were, for the first time, coming to the public attention, and there were some moderate-risk devices. Like for instance, the Dalkon Shield IUD in women, which were creating tremendous pelvic damage. Congress asked the FDA and the Department of Health and Human Services to put together a study to devise some way to regulate moderate-risk devices. Nobody is worried about a tongue depressor, which is a low-risk device. Everybody is worried about an implantable pacemaker, which gets treated like and requires pre-market evaluation analogous to drugs.
The device industry is very different from big pharma in terms of how they design and how they modify serially—and how they need, from their point of view, to get to market as quickly as possible after a device is designed and manufactured. There is a very strong pressure from the device industry to get into the market quickly.
That, in some way, led to the 510(k) process in 1975, which was designed originally only to last for two or three years to allow new designs and new devices to get into the post-1975 marketplace by comparison with predicate devices that had already been in the market prior to the new law. It’s designed as a transition.
The device companies loved it because it got their devices into the marketplace by a clearance process not by an approval process. They have been able to make it survive politically over the ensuing 40 years.
Until today, essentially.
DC: Yes. Our committee was put in place because there had been a series of failures of individual devices prior to 2009, and that’s where the political attention to the 510(k) process was coming from. And now the morcellator thing is just one more example of the clearance of a device for a use, not approval, based on predicates already in the market, that is, prior morcellators for other uses.
We recommended that FDA go back to the drawing board and get all the players around a table and devise a safer way to get into the market than just a clearance process. The pre-market evaluation would have a rationale for use and careful review of engineering and manufacture. And, more rapid means for adverse effects to be reported from the marketplace back to the FDA so we get an early warning signal when a device was doing something unexpected.
That’s part of the problem here, with the morcellator. Not only was the technology put to a new use without any real evaluation, but the signal from the clinical environment back to the regulatory environment that there’s a problem is a very low level signal in this particular case, and was hard to detect. Until a wise doctor had it used on his unfortunate wife, figured out what the hell was going on and blew the whistle.
That is a sort of quick summary of our report. We said 510(k) going into the marketplace makes no sense. That needs to be revised. We also said to the FDA, you really have to make your signals from the marketplace back to the regulatory system more sensitive. We said this on the basis of the studies of the devices that had failed before 2010.
The FDA, on its own, has been active in trying to make its surveillance system better. There are some particular medical societies with special interests like the orthopods with hips that are also undertaking improved surveillance.
The problem on the surveillance side is that everybody has got a reason not to report an unexpected or weak finding. The nurse in the operating room may have been told, ‘We don’t know exactly why this device did this, but it probably doesn’t mean anything. We have never seen this before.’
The doctor’s going to say, ‘God, I am busy, I don’t want to talk to the clinic administrator about this.’ The clinic administrator is going to ask his or her lawyer, and the lawyer is going to say, ‘Maybe you don’t have to report it, it doesn’t look like it does.’
And the device company sure as hell doesn’t want to report something unless they are sure—and their lawyers are damn sure—and they’d get in trouble if they didn’t report it. So you have a whole set of negative incentives before a report gets to the FDA about a problem.
That’s not to say that everyone is weaseling out, it’s just the way the system works. It’s not a system that magnifies or speeds signals on the way to the FDA, it’s a system that at every level would diminish the message getting to the FDA.
So in the specific case of the morcellator, you are saying that there are two main points that are not addressed by 510(k) that would obviously lead to something like this, because of a non-reporting of any kind of adverse outcomes. You are saying that the first point is that they need to change the way they evaluate pre-market clearance for devices such as the morcellator, and the second part is to require—within the process—reporting of adverse outcomes and patient follow-up. Would that be accurate?
DC: Yes. Improve the systems. Basically, the requirements are there, the systems are faulty. You want my hypothesis about the best of all possible worlds, how things might work five to 10 years from now?
Three things:
Number one: We have going in to place a uniform electronic medical record. We have early examples of large systems like Kaiser and the [Veterans Affairs] medical system in which universal electronic medical records already exist. And slowly, especially with the help of the new legislation, we’ll be trying to put in place universal electronic medical records for everybody over the course of the next decade.
Number two: The FDA has already begun to put a unique device identifier on every device that is made and to require that it be done by the manufacturer. So that, just like you scan a box of macaroni when you leave the grocery store at the clerk’s desk, every device is going to have a unique identifier on it that will tell the device, the date, the manufacturer—every detail. It will be tracked in every detail as only computers can. When something is put in you, or an X-ray machine is used on you, that unique device identifier will be in your electronic medical record.
Number three: We’ve got the new science and technology of big data that has been developed in particular by our intelligence agencies such as the CIA and the NSA as they scan communications data for clues in ways that I don’t understand. It’s amazing what quantities of data can be scanned and clues sought. We have seen a little bit of this in medicine, where Google flu has scanned and looked at symptom searches. They have been able to pick up early-on flu epidemics simply by questions that are being asked on Google. That’s a primitive example of the use of big data analyses to solve medical questions.
So you put those things together—electronic medical records, unique device identifier, and big data analyses—and you have a means by which the public’s health can be protected and the industries’ interest in getting a device into the marketplace early could both be dealt with.
You could make sure a device was engineered properly on the approval side, and FDA could let devices into the market that made sense early because the understanding would be that something like the morcellator complication would be picked up as an early-warning signal out of the big data analyses of the electronic medical record.
The device industry would be happy, and the public health could be protected—but that’s not where we are right now.
How would things have been different if the 510(k) process had been done right?
DC: First of all, I am not a surgeon. Certainly, I am not in that ballgame. I am also not fully aware of what the earlier uses were of the morcellator devices. There were some other previous uses, but what happened in this case, I think, was it was fairly easily adapted to gynecologic use. It was, through the predicate process, it was microsized to a smaller and smaller blender. One instrument was the predicate for another was the predicate for another and then with the use of robotic microsurgery, it sort of—exploded is too strong a word—but widespread use appeared fairly quickly.
The FDA, at some point, might have asked the questions that they are now being forced to ask, but given the 510(k) process as it is currently in law, it’s a clearance.
Get it into the marketplace. It’s similar enough to this predicate device which is already in the marketplace which is similar enough to the predicate device which was in the marketplace four years before that. It just sweeps into the marketplace without the adverse effects being thought through and certainly without a system by which the adverse effects, especially a low signal effect, is going to feed back into the currently existing adverse reporting system.
This situation we find ourselves in now is a combination of both things, the failure of the 510(k) and the failure of the reporting systems.
Why do you think the process was not changed earlier? Do you think it’s because there was no real catalyst?
DC: Let me give you my assessment of the political environment in which this has occurred. We know how big pharma does things. We also know that big pharma is wealthy and because of that wealth, has political power. But it’s primarily on the east coast and the west coast. So they maybe have 10 senators who look after their welfare, even though they contribute to a bunch of others, obviously.
The device industry is structurally extraordinarily different. The device industry has a shop in the district of almost every representative in the House of Representatives—I’m exaggerating a little bit now—even in rural Oklahoma.
The head of orthopedic surgery at his community hospital has designed an orthopedic screw for a broken elbow, and is manufacturing it in his garage, there in rural Oklahoma. The device industry has a widespread, hometown, know-each-other-by name, contributed-to-his-or-her-campaign constituency in the House. Their lobbying is personal, active, vocal and nationally organized in a couple of different national device organizations.
They can kill any piece of legislation in the House that they don’t like. In a way that big pharma can’t even. There are big players in the device industry in Minneapolis such as Medtronic, and St. Jude in Memphis, but it is primarily a small-company, national enterprise with widespread geographic influence. It will be interesting to see if Medtronic will still have the same influence as a foreign company, which they are trying to pull off.
That’s simply to say that our report given to the FDA in 2011 fell not only on deaf ears, but fell on a very noisy counter power play by the device industry to keep the 510(k) process in play. So that’s the politics.
The next natural question in that case would be, do you think there is sufficient traction on the issue we have at hand, open power morcellation, to compel the relevant authorities to change the 510(k) process? Do you think this has what it takes, and what else needs to happen before that can happen?
DC: A whole series of incidents from 1975 to 2009 were what created the political momentum for the study that I chaired to be requested of the IOM in the first place. And there have been failures of joints and cardiac leads, since then.
But the morcellation story has got an interesting set of public relations legs that some of these others did not. This is simply yet again another story of a public health failure of a device clearance process that may attract enough interest back to the issue—to our recommendations in 2011—so that the FDA will sit down and rethink this process with all the players around the table.
Do you think that the FDA hearing will in any way address the 510(k) process?
DC: I think the 510(k) process will clearly be part of the process at that hearing. I have submitted a brief statement about our study, and our assessment of the weaknesses of the 510(k) process.
I think that, given the momentum that appears to be building, and the understanding that the primary provocateur has both a medical side to this issue and has developed the understanding of the importance of the regulatory side of the causations here. He seems to be set on making sure the 510(k) process is part of the discussion. I think it will be.
Does that advisory committee have any authority whatsoever to speak on the 510(k) process, and if they do, would the FDA listen to them?
DC: I don’t know. My understanding is the committee is comprised of an OBGYN advisory group. Will that committee itself do that? That is hard for me to predict. But there is no other than what Hooman is doing himself to make sure it’s part of the discussion. And there may or may not be on that roster someone who really understands 510(k). I just don’t know. But it’s certainly going to be part of the public discussion now.
What kind of political legs it’ll have after that, I don’t know.
