The discussion around the use of paclitaxel-coated and eluting devices continues at the 31st International Symposium on Endovascular Therapy (ISET; 27-30 January 2019, Hollywood, USA). Organisers hosted a paclitaxel town hall to consider whether or not the recent meta-analysis from Katsanos and colleagues—concluding that there is a higher risk of death at two and five years following the use of paclitaxel-coated balloons and stents in the femoropopliteal artery—was concerning enough to alter current standards of practice.
Seventy-nine per cent of the ISET audience did not believe that the findings from the Katsanos et al (Journal of the American Heart Association; JAHA) meta-analysis were sufficient for them to change practice. Although James Benenati (Miami, USA) drew attention to the interesting fact that 31% of respondents said they would use it less, with the remaining 69% answering that they will use these devices as they always have. This result may reflect physician concerns of litigation or a class action, an opinion voiced repeatedly by ISET attendees to Interventional News following the FDA’s recent announcement that it was launching its own investigation into the findings and the pausing of two ongoing paclitaxel trials (SWEDEPAD and BASIL III) until further investigation has been conducted.
Barry Katzen (Miami, USA) contextualised the discussion, saying: “It has been an unusual time over the last couple of months in our space, where a manuscript has come out in a highly reputable journal [JAHA] with some highly significant implications, that could potentially change the practice in our field. There has been a lot of discussion at meetings and… it seemed appropriate that we try and delve into this issue and its implications. We have heard all sorts of responses to this manuscript [the Katsanos meta-analysis], including pulling devices off shelves of inventory, and concerns about the use of devices associated with paclitaxel delivery.”
Paclitaxel dose not biologically relevant after 30 days
Providing a review of the pharmacology and paclitaxel’s mechanism of action, Juan Granada (New York, USA) reminded the ISET audience that paclitaxel is a cytotoxic drug and that the drug-coated balloon (DCB) pharmacokinetics depend on both the concentration and solubility of the paclitaxel. He explained that paclitaxel is a cytotoxic drug that may induce a constellation of toxicities that are usually dose and schedule dependent. However, he further explained that even at its highest dose, compared to systemic use, DCB use results in lower systemic exposure by several orders of magnitude. According to Granada, the transient multi-organ distribution of paclitaxel results in tissue concentrations approaching zero at 30 days. Granada told attendees that he wanted to emphasise this because, in his opinion, the paclitaxel concentration in tissues was “grossly overemphasised, and actually mistakenly overemphasised, by Katsanos”. Granada concluded that “systemic tissue concentrations remain below the therapeutic threshold and are unlikely to exert a sustained biological effect, especially at five years”.
Time-to-event data not used in Katsanos et al (JAHA) meta-analysis
In his review of the Katsanos et al (JAHA) meta-analysis, Jihad Mustapha (Grand Rapids, USA) noted that “metrics such as number of withdrawals, number of patients remaining at risk and timing of deaths were not considered in these calculations. Mortality rates drawn from time-to-event data are preferred in meta-analyses.” He stated that not using time-to-event data led to reporting of very different five-year mortality rates of patients treated with the Zilver PTX drug-eluting stent (DES; Cook Medical) versus controls compared with the original study.
Don’t start a “paclitaxel panic”
Gary Ansel (Columbus, USA) presented insights from the patient-level data from the IN.PACT DCB and Zilver PTX programmes. Ansel concluded that these data showed that there was no difference in mean paclitaxel dose by survival status, no difference in survival between paclitaxel dose level (low, mid or high) and paclitaxel dose was not identified as a predictor of mortality. Thomas Zeller (Bad Krozingen, Germany) stated that this is not time to start a paclitaxel panic: “Is it time to panic? No. One single bad publication can damage a valuable treatment modality. We have seen this with renal denervation, we have seen this with renal stenting. If papers are published in very big journals, they are taken as truth, and may really bias and impact out treatment of patients,” he said.
Zeller sought to differentiate the systemic effects of paclitaxel from the local effects. He presented a case of a 76-year-old female patient who was treated with the IN.PACT DCB (Medtronic) in 2011, and was still a patient five years later. An aneurysm grew to 5cm following treatment with the device. Zeller mused that while local vessel wall damage such as this has been reported, it is not understood in much detail. However, he concluded that “systemic toxicity with low dose paclitaxel release via DCB or DES seems to be unlikely but warrants further observation.”
No signal for cancer
As the Katsanos meta-analysis indicated a signal for increased mortality risk at two years following the use of paclitaxel devices in femoropopliteal lesions, since its publication there has been interest in the causes of increased mortality at two years, which remain to be elucidated. Juan Parodi (Buenos Aires, Argentina) asked the panel if there was a link with cancer. However, William Gray (Wynnewood, USA) said this was not supported as a cause of death, referring to patient-level data presented at the Leipzig International Course (LINC; 22-25 January, Leipzig, Germany) that showed no signal for cancer.
Also critiquing the meta-analysis, Andrew Holden (Auckland, New Zealand) told delegates that study investigators need to consider the way patients are randomised in trials. However, he began by saying that “I think we need to congratulate Katsanos; I think it takes some courage and, while there are some reasonable concerns about the methodology, at least this has got us considering some very important findings.”
He continued: “Overall, though, I think we are struggling, because the finding is illogical, in that, as we know, the drug dose is much lower than the systemic dose of paclitaxel, the duration is very limited, and given that very limited duration, why would the mortality be different at one year and different at two years? There is also no mechanistic explanation, which we need to find.
“One point I would like to raise, having done a lot of these randomised trials, is the mechanism of randomisation. I think we really need to look at this, as well as looking at the overall use of meta-analyses, and particularly the abuse of meta-analyses we have seen in recent years. What actually happens when you randomise a patient is essentially you go online, and you have some demographic data, primarily lesion length, lesion complexity, patient age and gender—but there is very little data on cardiovascular risk factors. The way the programmes are run, they tend to favour the control arms slightly over the investigational arm. What you do not want in a trial is to find that lesions were more simple in the investigational arm, and what that means is, it is likely that patients in the investigational arm have a higher incidence of risk factors than those in the control arm. We have seen that time and time again. I think we need to evaluate the way we randomise patients, because essentially I think we are randomising patients in trial arms that have slightly more comorbid findings, and that may result in more mortality.”
Holden concluded by saying: “I am not convinced by the finding. I am convinced by the superior patency that we get with drug-eluting technologies, and until we get more information, in New Zealand anyway, we are going to keep using these devices.”
Patient consent at the forefront of interventionalists’ minds
Through the audience discussion, it was evident that patient consent for procedures involving a paclitaxel-eluting or coated device is now at the forefront of many physicians’ minds. Speaking from the panel, Richard Neville (Fairfax, USA) said: “This has had more of an impact on my patients than any paper I can recall recently. I actually had a patient come to the office recently with her family, who had heard about this, and asked if I was going to use ‘that balloon that will kill’ their father. I do not think I have seen a publication that I can recall in recent memory, despite many of the advanced devices that we all use, that has created quite the controversy and the awareness that this has had in patients.”
Holden reiterated this patient response, commenting how he had also seen patients bring up concerns based on the results of the meta-analysis.
Describing how, upon seeing the meta-analysis for the first time, he had felt conflicted about whether or not to inform his patient of the findings, Michael Jaff (Boston, USA) asked the ISET audience if they would have mentioned the publication to a patient who would potentially be receiving treatment involving a paclitaxel-eluting or –coated device. Over half of the audience (52%) indicated that they would not have informed their patient, 29% of respondents said they were “uncertain”, and 19% of patients said “yes”, they would tell their patient. When asked by an audience member which option he chose, Jeff informed delegates that he had told the patient that he “just seen this alert on his phone”, but that he “needed to read more about it”. He reassured his patient that “this was not an urgent situation” and said to give him more time to review this—Jaff is in fact due to see this same patient the week following the ISET meeting, and is hoping to use the clarity gained through discussion to guide his practice going forward.
Review of patient-level data at upcoming conferences
As reported in TCTMD, Granada described how the next steps involve reviewing the existing data: “Everyone is reviewing again all the experimental data, to look at the systemic exposure of these devices. I think that is going to be number one.”
VIVA [Vascular Interventional Advances] has assembled an invite-only special session of the Vascular Leaders Forum series on the topic “Drug elution in peripheral arterial disease (PAD): A critical analysis from a multidisciplinary consortium” (1-2 March, Washington, DC, USA). Following this analysis and discussion, the Charing Cross symposium (CX; 15-18 April, London, UK) will have a highlight session to independently review whether or not paclitaxel usage is beneficial or detrimental to the patient, which aims to have the final word on this controversy.