Recruitment proved to be a major challenge for the STEVECO (Stent versus conservative treatment in patients with deep venous obstruction) randomised controlled trial (RCT), prompting discussion on how best to randomise patients in future trials, the ethics of doing so, and possible alternatives to RCTs.
Jorinde van Laanen (Maastricht University Medical Center, Maastricht, The Netherlands) presented results of the STEVECO trial at this year’s European Vascular Course (5–7 March, Maastricht, The Netherlands).
While venous stenting is “basically common practice” and supported by various guidelines, the presenter noted, the problem is that the evidence available on venous stenting is limited. Most reports focus on stent safety and patency, with only three studies—according to a literature review van Laanen referenced—reporting on quality of life. “Are we really helping the patients? Are the clinical symptoms improving?” she questioned.
The STEVECO trial, sponsored by Optimed, was designed as a multicentre, randomised study. Van Laanen detailed that patients had to be treated conservatively for one year before they could be included in the trial, which had a primary outcome of VEINES-QoL/Sym score at 12 months and secondary outcomes of general quality of life, clinical improvement and stent patency.
“The trial had a lot of challenges,” van Laanen remarked, with randomisation of patients being the most notable. She explained that patients present to the Maastricht centre after tertiary referral because they have been struggling with symptoms for a long time. “These patients have already been treated conservatively and are sent to expert centres for treatment, so they want a treatment, an intervention,” the presenter continued. When the investigators tried to randomise patients—with the possibility that they would be assigned to the conservative arm for at least one more year—she noted that patients often refused to participate.
As a result of this and other challenges, van Laanen detailed that the investigators stopped recruitment at 63 patients out of a target of 130, with 21 in the control group and 42 in the intervention group.
At EVC, the presenter reported an eight-point VEINES-QoL score advantage for the stent group at one year, as well as a six-point VEINES-Sym benefit.
In terms of secondary outcomes, she revealed no significant difference in general quality of life—as assessed by the EQ5D questionnaire—and a significantly diminished pain score (11 points) in the stented group. Finally, she detailed a significant reduction in venous clinical severity score (VCSS) but not in Villalta score.
The patency results, van Laanen added, were comparable to other studies, with a primary patency of 91% and secondary patency of 97%.
The presenter stressed that there are some key questions that must be asked going forward. “How should we do this in the future?” van Laanen asked the audience at EVC. “If we really want to prove stenting is better than conservative treatment, how are we going to include patients in an RCT when we are all doing the treatment all of the time?”
Randomisation: Ethics and alternatives
“You need to be applauded for trying this,” panellist Stephen Black (Guy’s and St Thomas’ Hospital, London, UK) remarked in the discussion following van Laanen’s presentation on the STEVECO trial. He stressed that recruiting for a venous RCT is not an uncommon problem, noting BEST-PTS is “struggling” and C-TRACT is “at risk of failing to recruit”. He also referred to past trials, highlighting that ATTRACT and CAVA both took 10 years to recruit.
Considering what investigators need to do differently in the future, van Laanen suggested that “maybe the only option” would be to offer stenting exclusively to patients who participate in an RCT. Alun Davies (Imperial College London, London, UK) expressed his support for this idea, telling the audience it is “exactly what happened in the UK” with the EVAR 1 and EVAR 2 trials. “The only way you could get a device was by actually going into the trial,” he said.
Black remarked that the ethics of randomised trials are “quite difficult,” expressing the opinion that “we need to find alternatives to RCTs”. These trials are “exceptionally difficult to run” and “exceptionally expensive,” he said, adding that he is not sure the narrative that an RCT is the only level of acceptable evidence is completely correct.
Looking ahead, he underlined the potential of artificial intelligence (AI) and other new technologies in looking at large datasets. “There are opportunities there to run studies that provide you with exactly as robust a level of evidence, without having to randomise patients,” he explained. “I suspect the only way we are going to get around ethics is really looking at study design in a more robust fashion.”