One Trial Should Not Rewrite the Rules: Experts Challenge New Paclitaxel Guidelines

A provocative editorial published in CVIR Endovascular argues that the European Society for Vascular Surgery moved too fast when it revised its guidance on paclitaxel-coated devices in response to a single randomised trial. The editorial, written by an international group of vascular specialists, contends that clinical practice guidelines should evolve through comprehensive synthesis of the entire evidence base, not through reaction to one new study, however well conducted.

The trial at the centre of the dispute, SWEDEPAD 2, was a multicentre, participant-masked, registry-based randomised controlled trial run at 22 Swedish vascular centres. Patients with intermittent claudication, classified as Rutherford categories 1 to 3, who were undergoing infrainguinal endovascular revascularisation were randomised to receive either paclitaxel-coated or uncoated devices. The primary efficacy endpoint was the between-group difference in disease-specific quality of life at one year, measured with the six-item Vascular Quality of Life Questionnaire. All-cause mortality was assessed as a secondary outcome over a median follow-up exceeding seven years.

The ESVS focused update concluded that paclitaxel-coated devices should be used more selectively in patients with intermittent claudication, citing the absence of improvement in disease-specific quality of life at one year together with persistent uncertainty regarding long-term mortality. The editorial’s authors argue this represents a substantial shift from recommendations previously supported by more than fifteen years of randomised trial evidence, multiple patient-level meta-analyses, large national registries and extensive regulatory review, and that the update does not adequately explain why one study should outweigh that accumulated body of work.

A central methodological concern involves the interpretation of the neutral quality-of-life result. The authors emphasise that paclitaxel-coated balloons and drug-eluting stents were developed to inhibit neointimal hyperplasia, maintaining vessel patency and reducing repeat revascularisation. Their biological mechanism is expected to influence anatomical and procedural outcomes such as primary patency, freedom from restenosis and target lesion revascularisation, rather than directly determining patient-reported quality of life. Quality-of-life scores reflect a complex interplay of cardiovascular comorbidities, musculoskeletal disease, neurological impairment, physical activity, smoking behaviour, pharmacological therapy and patient expectations, and such instruments may lack the sensitivity to detect improvements arising from enhanced vessel durability, particularly over a follow-up of only one year.

Crucially, the editorial stresses that SWEDEPAD 2 was neither designed nor powered to demonstrate therapeutic equivalence between coated and conventional devices. A neutral result for the primary endpoint cannot be read as evidence that both strategies deliver identical clinical value; demonstrating equivalence requires a dedicated design with predefined non-inferiority or equivalence margins. Absence of evidence, the authors warn, is not evidence of absence. A trial failing to show superiority on one selected endpoint should be regarded as neutral for that endpoint, not extrapolated into a broader conclusion that the technology lacks clinical benefit.

The authors also highlight the pragmatic design of the Swedish trial as a potential source of heterogeneity. Patients in the intervention arm received a mixture of different paclitaxel-coated balloons and drug-eluting stents from various manufacturers, devices with distinct coating technologies, drug doses, excipients and pharmacokinetic profiles that cannot reasonably be regarded as biologically identical. Pooling such technologies into a single treatment category may attenuate measurable treatment effects, meaning a neutral overall result could reflect differences between individual devices rather than an absence of efficacy of paclitaxel technology itself.

On safety, the editorial takes issue with the guideline’s emphasis on excess mortality observed during the fifth year after intervention, even though the document acknowledges no significant difference in overall mortality between the treatment groups throughout follow-up. Because mortality was not the primary endpoint and the trial was not designed to detect differences in long-term survival, the authors argue any imbalance should be treated as an exploratory finding. An excess risk confined to one interval that does not translate into persistent separation of cumulative survival curves may reflect random variation, they note, and recommendations should rest on consistent effects observed across the entire follow-up period.

The broader mortality literature, in the authors’ view, does not support a causal link. Since the 2018 Katsanos meta-analysis raised the question, multiple independent research groups, regulatory authorities and professional societies have examined the issue using patient-level meta-analyses, large administrative databases, national registries and extended follow-up of randomised trials. No reproducible biological mechanism explaining delayed mortality from the relatively small quantities of paclitaxel delivered by endovascular devices has been identified; toxicological studies have not demonstrated systemic drug exposure at levels compatible with late lethal effects, and analyses of causes of death have found no consistent pathological pattern. Regulatory agencies, including the US Food and Drug Administration, have progressively concluded that available data do not support a causal relationship between paclitaxel exposure and excess late mortality.

Invoking consistency, a core domain of evidence frameworks such as GRADE, the authors argue that isolated findings which cannot be replicated generally reduce confidence rather than strengthen it. They caution that the guideline’s wording may create an unintended imbalance in the perceived benefit-risk profile, leading clinicians to infer a probable safety concern where the literature is more accurately characterised by uncertainty arising from inconsistent findings. Recommendations, they insist, must clearly distinguish proven risks, suspected risks and hypothesis-generating observations, because guidelines influence reimbursement policies, procurement and patient access to treatment.

The editorial closes with broader lessons for guideline methodology. Evidence synthesis differs fundamentally from evidence accumulation: a new randomised trial should not automatically supersede previous investigations merely because it is more recent, and committees should transparently document how new evidence is weighed against old, including management of financial and intellectual conflicts of interest among writing groups. The authors affirm that SWEDEPAD 2 is an important and welcome addition to the evidence base, but maintain that its findings should refine, not replace, a decade and a half of accumulated knowledge, with the strength of recommendations remaining proportional to the strength, consistency and reproducibility of the evidence behind them.

Subject of Research: Critical appraisal of the ESVS guideline update on paclitaxel-coated devices following the SWEDEPAD 2 trial

Article Title: When should clinical practice guidelines change? A critical appraisal of the ESVS focused update on paclitaxel-coated devices after SWEDEPAD 2

Article References: Fanelli, F., Holden, A., Ruffino, M. A., Muller-Hulsbeck, S., Haage, P., Hamady, M., Loffroy, R., Kobeiter, H., Wolf, F., O’Sullivan, G., Ratnam, L., & Morgan, R. (2026). When should clinical practice guidelines change? A critical appraisal of the ESVS focused update on paclitaxel-coated devices after SWEDEPAD 2. CVIR Endovascular, 9(1), Article 120. https://doi.org/10.1186/s42155-026-00775-8

Image Credits: AI Generated

DOI: 10.1186/s42155-026-00775-8

Keywords: SWEDEPAD 2, ESVS guidelines, paclitaxel-coated devices, intermittent claudication, clinical practice guidelines, mortality signal, quality of life, target lesion revascularization, evidence synthesis, GRADE, peripheral arterial disease, endovascular therapy