MDDI Online is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

12 FDA Panel Questions on Paclitaxel-Coated Devices

While an FDA advisory panel struggled with data discrepancies this week, panel members did attempt to answer these 12 questions for the agency.

  • After two days of presentations regarding the death risk that may be associated with paclitaxel-coated balloons and paclitaxel-eluting stents on the market and in clinical trials, FDA learned a lot about but the agency still has a lot of work to do before any concrete decisions are made regarding the risk-benefit of the devices in question.

    While an FDA advisory panel struggled with data discrepancies last week, panel members did attempt to answer these 12 questions for the agency.

  • On day one of the two-day meeting (June 19, 2019) panel members attempted to address five of FDA's questions regarding the paclitaxel device issue. The slide above shows the first question which has to do with the presence or lack of a late mortality signal associated with paclitaxel-coated balloons and paclitaxel-eluting stents. FDA also wanted to know what the magnitude of that signal may be.

    The easy answer is yes, there is a late mortality signal associated with paclitaxel-coated balloons and paclitaxel-eluting stents. Tougher to answer is what the magnitude of that signal is and what to do about it.

    "There is a clear signal in aggregate, and it does require an aggregate or a meta-analysis to determine it," said Richard Lange, MD, chairperson of the panel and president of Texas Tech University Health Sciences Center El Paso. " ... The missing data is an issue. The data are what they are and we can analyze only what we can see. The adversity signal wasn't planned for, as Dr. [John] Somberg mentioned, it's late, it's a limited number of patients, on the other hand, mortality is always an important signal and we don't want to ignore it, as Dr. [Joaquin] Cigarroa mentioned."

    Somberg, program director of clinical research & clinical research and bioinformatics at Rush University in conjunction with the University of Chicago, also noted during the discussion that the data is not the sort of evidence to take the devices in question off the market, but it's also not the sort of evidence that should be disregarded. Somberg served as a temporary voting member of the panel.

    Col. Todd Rasmussen, MD, an attending vascular and endovascular surgeon at Walter Reed National Military Medical Center who also served as a temporary voting member on the panel, agreed that there is a signal. "However, with regards to magnitude, I find these findings to be statistically significant, but not practically significant."

    Screen capture from an FDA webcast of the panel meeting on June 19, 2019
  • For the second question, FDA asked the panel to discuss the strength of the evidence supporting a late mortality risk class effect. This question ignited a lengthy debate from panelists. As chairperson of the panel, Lange had the difficult job of summarizing the panel's comments for FDA.

    "The strength of the evidence is considered to be low, nobody at the panel feels like they can say one device is superior to another, nor do they feel that all the patients should be informed of the data, regardless what device it is."

    Screen capture from an FDA webcast of the panel meeting on June 19, 2019
  • When asked how the missing data impacted the panel's interpretation of the meta-analysis result, Lange began by summarizing what had already been discussed about the missing data.

    "People are disappointed there is missing data, and I'd take it a step further - and this hasn't been said - in some respects, it may not have been necessary," Lange said. "Kudos to Medtronic who went back and are now missing only 2.7% of their data, but that started at about 19%. Now I'm not picking on Medtronic, all of our industry sponsors have indicated that carrying these studies out longer there is missing data ... that makes it very difficult for us to interpret the data. So what we're left with is incomplete data, very good analyses, a signal, a smoking gun, but no bullet or dead bodies at this particular point. The missing data could make the data look better, could make the data look worse, all we can say is it's missing."

    Screen capture from an FDA webcast of the panel meeting on June 19, 2019
  • The fourth question dealt with subgroup analyses.

    "There's no particular subgroup signal that identifies who would or may not benefit from receiving a drug-eluting stent or balloon or is at risk of it, that would require a larger number of patients and more heterogeneous," Lange said in summarizing the panel discussion on this question. "And perhaps we need to look at it differently, instead of a particular risk, factor a risk group. That's the individuals with the highest risk, those with multiple co-morbidities, may or may not have a differential effect than those who have a much lower risk."

  • The fifth question dealt with the mechanism of the late mortality signal, which made some panelists uncomfortable.

    "People don't feel comfortable discussing a mechanism for a couple of reasons," Lange said. "One is there's not a lot of confidence in the data that were attribution of death, the study wasn't intended to do that, we're doing it retrospectively, there's not a clear definition across all different studies to do that. I don't think anybody feels very comfortable that we have actually attributed it to the right cause of death."

    That said, Lange acknowledged the following three hypotheses put forward by panel members.

    1. It's a systemic effect.

    2. It's local with a secondary effect.

    3. It's just amplifying a cause of death that would already be there.

    Screen capture from an FDA webcast of the panel meeting on June 19, 2019
  • The panel addressed questions six through 12 at the end of the second day of the meeting (June 20, 2019).

    The sixth question dealt with the potential paclitaxel dose/mortality relationship.

    "It's inconclusive based upon the data," Lange said, summarizing the panel discussion on question six. "It's complicated by patient variables including length of the lesion and there's a concern about high tissue levels."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019
  • For the seventh FDA asked the panel to weigh in on the pre-clinical animal data and whether any of that data provides mechanistic insight into the late mortality signal. The agency also wanted to know what new animal studies, if any, would be helpful.

    "None of the current animal models provide mechanistic insight," Lange said.

    Michael Krucoff, MD, a professor of medicine/cardiology at Duke University Medical Center, director of the cardiovascular devices unit, and director of the eECG Core laboratories at Duke Clinical Research Institute in Durham, N.C. added to the discussion.

    "I definitely think that in this type of safety signal exploration that anything and everything we can learn from pre-clinical models we should maximize our efforts to adding knowledge rather than having to explore human risk to change ignorance into knowledge," said Krucoff, who served as a temporary voting member of the panel. "This is a particularly good example, potentially, for the kind of industry collaboration that we've seen over the past several days to think about working together rather than everybody doing their own animal models ... thinking about where or what we could learn that might be not necessarily a hair-raising mechanism might be quite reassuring."

    Bram Zukerman, MD, director of the office of cardiovascular devices at FDA's Center for Devices and Radiological Health, seemed to welcome that idea.

    "There are very important scientific issues to elucidate here but these are very expensive and time-consuming trials and the agency would be very interested in continuing to develop a project in this area with the industry," Zuckerman said.

    Eliot Levy, MD, a professor of medicine at National Institutes of Health Radiology and Imaging Sciences in Bethesda, MD, added that an animal model might be a good way to investigate drug interactions. Paclitaxel is known to have several highly toxic drug interactions, said Levy, who also served as a temporary voting member of the panel.

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019
  • For the eighth question, FDA asked the panel to discuss benefit-risk considerations related to paclitaxel-coated balloons and drug-eluting stents. To say that the question spurred a lot of debate would be an understatement.

    "You're not going to get a unanimous opinion, but I think what you've heard if I'm going to summarize, is there's a clear risk in terms of quality of life," Lange said. "There is some concern that the signal is weak but people are unwilling to trade death for what could be a marginal clinical benefit at three years but an acute, immediate benefit. But everybody said the patient should be involved, should be informed, fully, of the information we have, whether it's complete or incomplete, and should be involved in that decision-making process."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019.
  • In response to question nine on the slide above, the panelists agreed that there is data available right now but it needs to be cleaned up. There also are studies that have recently been initiated or studies that are not new but still ongoing that may provide the information needed to answer the question. 

    "I've heard the word collaboration from everybody on the panel and industry as well because I don't think anybody wants to tackle this or can tackle it by themselves," Lange said. "There's the Swedish study we can get data from as well. There are studies outside the U.S. but the FDA has to be convinced that they are good so we can include them in the analyses."

    Lange also said it is important to determine "what is the absolute number we need to get to the information we want? And if it requires a smaller randomized clinical trial to get us to that number than I would urge the FDA to do it with the least amount of information necessary. That's mortality, survival, how their treatment is going, and whether they are being revascularized."

    Lange also acknowledged a distinct divergence of opinion among panelists as to the value of registry data.

    "And we could talk about this all day long, you're not going to convince Dr. [Karla] Ballman or myself that that's the information," Lange said. "I think that represents the scientific community at large. There are people who are firmly in the registry camp and there are people who firmly say it cannot get this kind of information. For other information it's useful, but it can't answer this question."

    "At the end what we want is an answer," Lange continued. "At the end what we would hope from the industry and FDA is not to come back here two years from now or three years from now and say it's still in question and we don't know the answer because at that point we would demand a randomized controlled trial of 10,000 or 20,000 individuals."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019.
  • Question 10, as shown on the slide above, was a bit more straight forward for the panel to answer.

    "We're meeting because of a signal. Because of a concern, an honest and well-meaning concern of increased mortality. My opinion is the patients need to be informed of it," Lange said. "... If we weren't concerned about it the panel meeting would have ended yesterday and we wouldn't need to inform the patients about it. It's not definitive. We're going to get the answers to it, but until then, for a patient to make an informed decision, they need to have that information."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019.
  • Zuckerman had to clarify Question 11 because of some confusion among panelists. 

    "What we're asking is for a new PMA device, given what you said, you would wait until we have five-year safety results for a new paclitaxel drug-eluting stent?"

    Karla Ballman, PhD, a professor of Healthcare Policy and Research at Weill Cornell Medical College in New York, said she wouldn't recommend the agency wait five years to approve the device based on efficacy. But she said she would recommend FDA put teeth behind the post-market surveillance requirement and not let the company applying for approval get away with saying there is only long-term safety data on 75% of the trial patients or something like that.

    "One-year approval, five-year follow up with teeth, like real teeth ... [and make sure] the industry understands the importance of it so we're not debating this later on and the FDA really holding their feet to the fire," Lange said in his summary of the discussion on this question. "And I'm not quite sure how we do that legally, but morally it's imperative."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019.
  • And finally, Lange also summarized the discussion for the last question, shown on the slide above.

    "I would say it may be different if we're looking at a five-year increased mortality and we're looking at a group of individuals on dialysis that are likely to die in two or three years or those with chronic limb ischemia that if a 60% increased mortality at three years may be different just because of the longevity and risk of the patient," Lange said. "However, I would agree with Patricia [Daigle, the non-voting consumer representative on the panel] and agree with my colleagues that individuals receiving this, until we have a final determination at least, should have that information so that they can make an informed decision."

    Screen capture from an FDA webcast of the panel meeting on June 20, 2019.
500 characters remaining