Delegates voted overwhelmingly against the notion that there was a demonstrable danger in any organ of the body attributed to circulating paclitaxel. These polling results, with a 67% majority, came after vascular pathologist Elena Ladich (Hollywood, USA) presented insights from new analysis on the effects of paclitaxel in the body. A CX Symposium Highlight Session explored the ramifications and possible resolutions four months on from the Katsanos et al meta-analysis published in the Journal of the American Heart Association (JAHA). The session was chaired by Roger Greenhalgh and moderated by CX Executive Board members Gunnar Tepe, Andrew Holden and Thomas Zeller. Experts, representatives of regulatory bodies (US FDA and UK MHRA) and other stakeholders scrutinised the current controversy surrounding paclitaxel-eluting devices over the course of four hours.
Additional polling at the close of the session revealed that 85% of CX delegates believed “no danger has been demonstrated in any organ of the body that is attributable to circulating paclitaxel”. This conviction was further consolidated, when attendees were asked: “Have we established that there are substantial levels circulating in arteries used for reducing restenosis”—68% of respondents said no.
Ladich, whose presentation closed the session, spoke on the effects of paclitaxel on the artery wall, drawing from her practice with human autopsies. She explained that a proportion of the drug and excipient coating of paclitaxel-eluting devices are known to embolise distally. While she says she “does not know the answer”, Ladich stresses the importance of asking whether or not we should be concerned with “the potential shower of drug into the end organ”. In her experience, “rare case reports have described aneurysms, localised hypersensitivity, downstream vasculitis and panniculitis following drug-coated balloon treatment”. However, she also notes that “to date, there are no published reports of autopsy findings or specific organ toxicity in humans treated with drug-coated balloons”. Additionally, to her knowledge there are “no causal links or mechanisms to explain the reported association of paclitaxel in the lower limbs and patient mortality”. Ladich clarified that despite this lack of evidence in both the literature and her experience, “this does not mean a potential damaging effect to organs does not exist”.
The Highlight Session began with a summary of the key messages from last month’s Vascular Leaders Forum (VLF; 1–2 March, Washington, DC, USA) by Gary Ansel (Columbus, USA). He told CX delegates that a central discussion point was whether a “protective benefit” with plain percutaneous transluminal angioplasty (PTA) could explain the higher mortality seen with paclitaxel devices in the meta-analysis. Ansel commented there was “no mechanistic signal to date” for such a benefit but, potentially, risk factors may be more frequently identified—and thus modified—in PTA patients because they are seen more often and undergo more interventions than those who receive paclitaxel devices.
“This [peripheral arterial disease] population dies at a very high rate during the five years after a procedure. We should be focused on how we stop these patients from dying. From my standpoint, these patients should be followed much more closely for their general medical condition. We have to make sure these patients are not having infections or different oncology problems. We need to change this mortality rate with much more aggressive risk factor modification,” Ansel noted.
The lack of mechanistic signal was one main criticism levelled at the Katsanos et al meta-analysis findings. Defending his methodology and conclusions, Konstantinos Katsanos (Patras, Greece) told CX delegates that the increased risk of death reported in the meta-analysis following the application of paclitaxel-eluting devices in the femoropopliteal artery consistent and significant across multiple statistical analyses. “We are mathematically 99% certain of this significance”, Katsanos stated emphatically, pointing to results of a trial sequential meta-analysis. He added, “every trial is only as good as its methodology. […] The key message here is not a smoking gun—it is the consistency of the findings. You see here the consistency of the mortality signal, and this I would say is the main take-home message.”
Mårten Falkenberg (Gothenburg, Sweden), investigator of the SWEDEPAD registry-based randomised controlled trial, reported on the status of the trial which ceased enrolment in the wake of the JAHA meta-analysis. He told the CX audience that despite keeping up with the ongoing international conversation, the investigators were still trying to determine the best course of action, but no obvious consensus has emerged. Answering the question, “Do you agree with BASIL-3 and SWEDEPAD to recommend stop using paclitaxel products at this moment?”, a slight majority of 52% agreed. Furthermore, a similar split was seen as 53% voted for the motion “Do you suspect that the finding of Dr Katsanos is correct?”.
Jonathan Michaels (Sheffield, UK), chair of the steering committee for the BASIL-2 and BASIL-3 trials, spoke about the halting of enrolment into the latter trial, which was set to expand on BASIL-2’s investigation of paclitaxel treatments for claudicants by examining the therapy in patients with critical limb ischaemia (CLI). However, BASIL-3 enrolment ceased following the Katsanos et al meta-analysis in order to review the ethics of continuing to treat patients with this drug. Michaels queried if there was an “innocent explanation” for the mortality signal. Like Ansel, he suggested a “plausible” reason for the observed increased mortality may be related to a greater follow-up frequency in patients treated with PTA, for whom potential comorbidities may be more likely to be diagnosed. Meanwhile, both Michaels and principal investigator Andrew Bradbury (Birmingham, UK) expressed their hopes to restart recruitment, although the unanswered questions surrounding paclitaxel present “serious challenges” for the trialists. Speaking from the floor, Bradbury further argued that a distinction “must be made” between claudicants and CLI patients as the risk profile varies drastically between the two. “For claudicants, treatment is about enabling the patient to walk further.” With CLI, Bradbury added, “the risk weigh-up” instead hinges on improving quality of life in end-stage peripheral arterial disease (PAD), for which effective treatment may be worth a potential increased late mortality risk.
Responding directly to Bradbury’s comments concerning CLI, Zeller noted that teasing out an association of increased mortality would be more difficult in a CLI patient cohort, due to the “higher background noise” caused from greater overall mortality rate in this patient population compared to less ill patients. Bradbury further said “we cannot assume that Katsanos’ findings applies to CLI patients.” Expressing his personal view, he told attendees that the BASIL and SWEDEPAD trials “represent a unique opportunity to settle the question with regard to CLI”. Indeed, speaking to this point, US FDA representative Misti Malone (Silverspring, USA) suggested that regulatory bodies may be able to make such a distinction between claudicantion and CLI, and pointed out that the FDA has currently not recommended the halting of paclitaxel trials.
The methodology and statistical validity of the meta-analysis outcomes were comprehensively described by statisticians and regulatory authorities. Overview of the meta-analysis’ statistical analysis was provided by Jon Deeks (Birmingham, UK), statistician of the BASIL-3 trial, and Sue Duval (Minneapolis, USA), one of the two independent, academic statistical advisors to the VIVA independent patient data (IPD) review. Deeks suggested that many of the studies used in the meta-analysis by Katsanos et al were “inadequate” in terms of the raw data, and that the reporting of the trials was “suboptimal”. However, Deeks acknowledged “statisticians are never satisfied with the data”. Polling following this statistical discussion showed a slight shifting in favour of believing that there is statistical evidence currently indicating a real increased mortality from the use of paclitaxel on a balance of probabilities: 57% to 43%.
Malone gave the FDA position, and announced the IPD review panel to take place on 19–20 June this year. Hazel Randall (London, UK) of the UK Medicines and Healthcare products Regulatory Agency (MHRA) relayed the working process of the expert advisory group created last month, which is due to release its final recommendations for paclitaxel use “in the next few weeks”. Both regulatory bodies received considerable support from audience polling, with 68% believing the authorities were “reacting appropriately”. This result was quite decisive in comparison to the less unifying answers of previous polling.
Lessons from the paclitaxel pioneers and other disciplines
CX delegates heard from Ulrich Speck (Berlin, Germany), who spoke of his early work to enable the drug to coat the balloon, resulting in the plethora of devices available today. It was evident from Speck’s account that from the outset, the cell-killing effects of paclitaxel were known. “Of all the drugs tested in the swine model, none were as effective as paclitaxel”, Speck stated.
Speaking on the pharmacokinetics of paclitaxel, Lindsay Machan (Vancouver, Canada) gave a description of the early application to stents, which was incorporated to address the issue of in-stent restenosis. The choice of paclitaxel was reliant upon its observed “powerful” ability to inhibit cytokine-induced migration of cartilaginous cells, which Machan explained cannot grow when more than 2mm away from a capillary.
Providing an oncological perspective, Erica Mayer (Boston, USA) detailed the use of paclitaxel as a primary therapy in cancer care. Notably, Mayer referred to paclitaxel as “one of our most commonly used treatments” in breast cancer patients, and even reported its safe, effective use in the second and third trimester of pregnancy. “We think it is safe enough to give to our pregnant patients”, Mayer told the CX audience. Furthermore, she pointed to the significant difference in dosing: a higher-dose drug-coated balloon reportedly providing 0.55% of an average dose given in a weekly breast cancer regimen over 12 weeks.
Finally, Bruno Scheller (Homburg/Saar, Germany) outlined the lessons learned from paclitaxel use in coronary arteries, comparing the drug to its alternatives, sirolimus and everolimus. Referring to a similar controversy over a decade ago, in which a potential association with increased mortality was discussed for paclitaxel in the coronary arteries, Scheller said this 2006 “firestorm” is the “perfect blueprint” for the current paclitaxel debate. Ultimately, however, Scheller reported that the coronary community concluded that the stent design and coating technology is “more important” than drug selection in these arteries, and drug-eluting devices are in continued use.
Commenting on these lessons, Greenhalgh summarised that the early users of paclitaxel were “fully aware that the drug was cytotoxic, is has been used in oncology since”, iterating that its use in vascular applications was “always known to damage the artery wall in some way, which discouraged the restenosis process”. At this stage, 85% of the audience agreed with this sentiment, answering the poll: “Do you think that those who used paclitaxel in the early days were quite aware that paclitaxel kills cells?”. A follow-up question elaborated: “If you accept that the pioneers knew that paclitaxel would kill cells, do you think that they always aimed to have a balance between achieving less restenosis in the patients’ interest without increasing mortality?”. The result of this query was 72% in favour, capturing the audience support for the idea that there may exist a trade-off in the use of paclitaxel. “On the one hand,” Greenhalgh summarises, “paclitaxel contributes to greater patency in the lower limbs, but on the other hand, there is an alleged increase in death.” Extensive discussion followed on whether there exists a causative action as the drug circulates within the body, or whether this is “merely an association”. Overall, there was near unanimous support from speakers, audience and the regulatory bodies on the position that while an association may exist, further investigation is required.