Paclitaxel-Coated Devices’ Late Mortality Signal Not Well Understood, FDA Panel Hears
Posted 20 June 2019 | By
Following Wednesday’s debate on missing data necessary for paclitaxel-coated devices’ mortality assessments, the US Food and Drug Administration (FDA) advisory committee on Thursday offered recommendations on how to use what is in place to understand mortality signals.
Limited data on dose effect, toxicity and other potential factors have posed barriers for FDA and industry to achieve a complete understanding of paclitaxel- or drug-coated balloons (DCBs) and paclitaxel- or drug-eluting stents (DESs) in peripheral arterial disease (PAD) patients, according to the panel. The panel deliberated on recommendations to help address the lack of key data elements. This came after the panel generally agreed with the agency's conclusion that a signal associated with an increase in the late mortality rate exists through five years for DCBs and DESs.
The late mortality signal, however, is not yet fully understood. To maximize the totality of the data that is maturing, panel members discussed whether to make greater use of more registries for DCBs and DESs. Some showed support for registry use to improve on data already available from FDA-mandated post-market approval studies, citing the urgency for more data and limited resources. Others recommended iterating ongoing trials as registries tend to be prone to bias and provide limited short-term data on acute adverse events.
Despite the diverging opinions on using registries, panel deliberations made clear that there is need for a collaborative approach across all stakeholders to leverage the current amount of data and pragmatically generate the missing data deemed necessary to answer remaining questions on the magnitude and causality for the increased long-term mortality rate in PAD patients treated with DCBs or DESs compared to patients treated with uncoated comparator devices. The panel members reached consensus on the need for obtaining additional mortality data, long-term follow-up in future studies, as well as improving cause of death adjudication.
“Just because we do not have data to show a dose response…does not mean one does not exist,” said panel member Michael Krucoff, professor of medicine and cardiology at Duke University Medical Center. “Again, we get in the conundrum mentioned
yesterday,” he said of studies stratified by dose for outcomes. “Subgroups mean smaller denominators but the hazard” remains that “just because we can’t see” a mortality association “does not exonerate our ignorance of whether or not it is there.” The panel's "conundrum" that came up on Wednesday was related to different numbers in analyses presented by FDA and the five manufacturers marketing DCBs and DESs in the US.
The panel was generally in agreement with FDA’s conclusion during late-Wednesday deliberations that a signal associated with an increase in late mortality exists, despite manufacturers’ pushback. Yet panel members said that they lacked confidence in the totality of the data to assess the magnitude of the signal and attribute causality due to the large amount of missing data and inconsistencies in the analyses.
FDA launched a January probe partly based on a meta-analysis published
in the Journal of the American Heart Association,
followed by an analyses the agency created
for recommendations to health care providers in March. Yet the discussion on the concept of ascertainment bias that challenged the meta-analysis as well as the inconsistencies and limitations in the analyses made reaching the decisions FDA had requested difficult.
Panel members did not believe there are enough available data to suggest clear relationships based on specific subgroups, toxicity or device class either. “They also needed the data evaluated in aggregate and did not believe that any specific device should be excluded,” said FDA lead reviewer Eleni Whatley. Whatley added that “the panel believed that the amount of missing data” for the three DCBs and two DESs that received approval from 2012 to 2018 "was unfortunate and any previous and future attempts to gain missing data are important.”
Data limitations also challenged the panel's decisions, which are intended to inform the potential for additional regulatory actions, on labeling, changes to study designs and other use indications. The devices in question include Bard’s LUTONIX DCB, Medtronic’s IN.PACT DCB, Philips’ Spectranetics Stellarex 035, Cook Medical’s Zilver PTX DES and Boston Scientific’s ELUVIA paclitaxel-eluting system.