番茄社区

WASHINGTON -- There is a safety signal with paclitaxel-eluting drug-coated balloons (DCBs) and stents that should be discussed with patients considering this therapy for peripheral artery disease, a panel of stakeholders and experts concluded in an informal vote.

Ten of the 14 panelists voted that, yes, such a signal exists after considering the totality of the data presented by industry and academic researchers throughout the day at the Cardiovascular Research Technologies (CRT) meeting here. The other four responded with "I don't know."

Since the Smoking Gun

The panelists convened at the CRT town hall session in response to a meta-analysis published in December finding among those who got paclitaxel DCBs and drug-eluting stents (DES) placed in lower-limb arteries compared to controls. The excess mortality appeared to grow worse over time, up to a 93% higher relative risk at 5 years in this study led by Konstantinos Katsanos, MD, of Patras University Hospital in Rio, Greece.

In February, a group defended the safety of Medtronic's IN.PACT Admiral DCB in their own meta-analysis -- one that later turned out to have errors.

At the same time, a team of investigators surfaced corrections to 5-year data on the Zilver PTX in which the authors admitted that they had flipped their results and should have reported that the Zilver PTX DES was associated with greater 5-year mortality over percutaneous transluminal angioplasty. These deaths cannot be attributed to the device, however, the group maintained.

More recently at VIVA's , an FDA representative announced that preliminary results of the agency's investigation appear to confirm the original meta-analysis.

Uncertainty in the Data

Session moderator Jeffrey Popma, MD, of Beth Israel Deaconess Medical Center in Boston, admitted he was among those who could not decide whether such a safety signal really exists. He couldn't be sure of the meta-analysis' validity -- not because of shoddy statistical work on the part of the researchers, but because "the underlying structure of the data may not be right," he said.

Trials included in that meta-analysis suffered from issues ranging from medication non-adherence, loss to follow-up, and lots of cross-overs.

Even so, "the body counts are in the wrong direction and I voted yes" that the safety signal is there, said Roxana Mehran, MD, of The Mount Sinai Hospital in New York City.

She and fellow panelists, nonetheless, voted 8-6 that the labeling of the devices in question should remain unchanged in the face of the uncertainty surrounding the data. A vote on whether patients should be informed about the paclitaxel device controversy before getting such a procedure done came out near-unanimously yes (with one "I don't know").

During the CRT session, industry representatives individually presented patient-level data from their research programs showing no uptick in mortality with paclitaxel-eluting devices.

The panel voted 9-1 that this should all be put together in a statistically-robust pooled analysis.

Separately, VIVA Physicians is taking on its own independent meta-analysis of the available patient-level data. Panel members disagreed over whether the emphasis of future investigations should be on intention-to-treat or as-treated analyses.

Searching for Biological Plausibility

Besides the questions on data quality, the panel questioned the biological plausibility of paclitaxel causing the observed deaths in such a way as to yield a risk difference at 2 years but not at 1 year.

After all, "paclitaxel is not a slow damaging compound. It causes damage immediately," said Aloke Finn, CVPath Institute in Gaithersburg, Maryland, who said that one would expect to see increased mortality earlier on.

A chemotherapy drug, paclitaxel is cytotoxic and is thought to lead to positive remodeling through the apoptosis of smooth muscle cells. It can have the toxic effect of neutropenia, but there have been no excess neutropenia-related deaths in the DCB and DES studies, Finn said.

Katsanos reported that his group has detected higher rates of cardiac, GI, infectious, and respiratory adverse events at 1 year despite the mortality difference not being evident at that point. These are clusters that "seem to be consistent with systemic cytotoxic side effects" of paclitaxel, he told the audience.

Finn presented his hypothesis that the mortality can be explained by undertreatment of some in the paclitaxel-receiving arm of the trials.

"DCB drug delivery is inefficient because most of the paclitaxel does not make it to the target site and is lost in the body. Overall, total doses of DCBs to the patient is much lower than required to cause paclitaxel toxicity," according to Finn. "We must look for other causes."

Comment