Implications of ACST-2 for Asymptomatic Carotid Stenosis

This transcript has been edited for clarity.

John M. Mandrola, MD: Hi, everyone. This is John Mandrola from theheart.org | Medscape Cardiology. I'm pleased to be with Dr Richard Bulbulia, who is a consultant vascular surgeon at the University of Oxford and co-PI of the ACST-2 trial, which was designed to study carotid artery stenting vs carotid endarterectomy, and was a hotline presentation at the European Society of Cardiology meeting. Welcome, Dr Bulbulia.

Richard Bulbulia, MD: Hi, John. I'm pleased to be here.

Mandrola: Would you tell us the main findings of this very important trial in vascular surgery?

Bulbulia: I think there were two main findings. The first concerns procedural hazards. What we found was that the rates of death and disabling stroke within 30 days of the procedure were similar for both carotid surgery and carotid stenting, at around 1%.

Considering long-term efficacy, the 5-year disabling stroke rates were comparable for both surgery and stenting, at a rate of around 0.5% per year.

Mandrola: Tell us exactly who these patients were. As far as I can see, it was a fairly long trial.

Bulbulia: It was a labor of love. It ran for around 13 years to recruit these patients from around 130 sites across 30 countries, predominantly European. Thanks to the hard work of our network of collaborators across Europe and wider afield, we randomized 3625 patients in whom carotid intervention was considered necessary by the managing clinician at these local sites to either surgery or stenting.

We had quite broad eligibility criteria. The inclusion and exclusion criteria were really left, in large part, to the treating clinicians at these individual sites. It was this streamlined design where we integrated trial processes into routine clinical practice that allowed us to complete the largest carotid surgery trial today.

Mandrola: Can you say anything about important subgroups related to the primary endpoint? Were there any particular findings there?

Bulbulia: In our Lancet paper, which is available by open access, we did some subgroup analyses on the main possible criteria that might influence outcome, and there were no significant subgroup abnormalities or differences seen in the trial. We're always a bit careful about subgroup analyses. ACST-2, as I said, is the largest-ever carotid surgery trial to date, but it's not really large enough to look at subgroups in a meaningful way. You do run the risk of getting data-dependent findings, which aren't particularly reliable.

We are collaborating closely with all the other major carotid surgery and carotid stenting trials, and we do plan to do an individual patient data (IPD) meta-analysis, pulling the results of ACST-2 with the asymptomatic population in CREST and ACT I. Perhaps with that IPD meta-analysis, which will give us around 6000 patients, we might have the ability to look at subgroups in a more statistically robust fashion.

We did look at the results of ACST-2 in the context of the total available randomized evidence that existed before ACST-2, and we found remarkable consistency with our results when pooled with the other asymptomatic — and indeed, symptomatic — trials of surgery vs stenting. That was a tabular meta-analysis, and obviously the IPD meta-analysis will allow us to do things in a little bit more of a sophisticated fashion.

Mandrola: I suspect that there's been quite a bit of iteration in carotid artery stenting over the years, but it seems like in the past carotid artery stenting may have been reserved for higher-risk patients. What's different now about that technique compared with endarterectomy?

Bulbulia: The nice thing about doing a trial over 13 years is that we probably captured and possibly reflect the evolution of carotid artery stenting. We noticed new stent designs being brought in as the trial went on. There were new embolic protection devices that were used as well, and clinician experience, both in the selection of patients and technical skills in performing these procedures, almost certainly evolved over time.

In contrast, carotid surgery, which has been around for almost 70 years, is a pretty established technique. Probably there have been few changes in that over time. The important thing is that these changes in carotid artery stenting may well impact on the procedural risks — the stroke, death, and myocardial infarction rates seen within 30 days of the procedure — but they may not be terribly important in the long-term durability of surgery vs stenting.

Indeed, that was the most important comparison, I think, that we can get from a trial like ACST-2. If you're looking at secular changes in carotid stenting approaches, and indeed carotid stenting outcomes, it's probably wise not to look at a trial like ACST-2 which, although we have 3600 patients, is dwarfed in comparison to the large mandated national registries that exist in countries like Germany, where they have the outcomes of tens or hundreds of thousands of carotid procedures carefully recorded in these registries.

If you're looking at secular changes over time, go to the registries. Don't go to a relatively small trial where practice and patients probably may not reflect more widespread clinical practice. Indeed, the event rates and ACST-2 may well not be comparable or generalizable to routine clinical practice in the 2020s.

Mandrola: I want to get back to the generalizability of it. The phrase that really stuck out in my mind is "Patients were thought to require intervention." Now, as you know, there's some variability across countries about the treatment of patients with asymptomatic carotid stenosis. In the US, we do many interventions for asymptomatic disease but maybe medical therapy predominates in other countries. Just to be honest, I would have loved to see a medical therapy arm. I wonder how you would respond to that.

Bulbulia: Thanks, John. I did some background research on you before this interview, and I'm aware of your views on asymptomatic carotid intervention.

Yes, you're right. There's a huge variation in practice. Geographically, rates go from 0% in Denmark to really quite high rates in North America. The ideal sweet spot is somewhere probably mid-Atlantic, I imagine. You're right — ACST-2 did not seek to answer that question. That's not to say it's not an important question, it just wasn't a question that we set out to answer.

There was some discussion at the very beginning of ACST-2 about having a third arm, but we decided against that, really, for grounds of the impact that would have on recruitment. As you pointed out, it took 13 years to recruit these 3600 patients. If you throw in a third comparison, that just becomes a very difficult sell to patients. It's quite hard for patients to embrace the uncertainty of surgery vs stenting. If you add in an extra arm — surgery vs stenting vs medical therapy alone — it could really be quite confusing.

Indeed, the experience of our German colleagues in SPACE-2 shows that they failed to recruit to that trial, largely on the grounds of complexity. Really, you need a big, simple trial to answer a question reliably, and that's what ACST-2 sought to do.

I think if you look at the totality of the randomized evidence of intervention plus medical therapy vs medical therapy alone, that still does come down very favorably on selected intervention for asymptomatic patients, provided you can do these procedures safely with a low procedural morbidity and mortality rate.

My colleagues, Tom Brott, George Howard, and others, are doing CREST-2. I would encourage American clinicians to randomize patients to that trial. We'll pool the results of CREST-2 with ACST-1, ACAS, and the VA trial, and perhaps more clarity will come about that. I'm pretty certain that asymptomatic carotid intervention still has a role for selected patients.

I would like to point out that in this trial, the annual risk of these patients having a disabling or fatal stroke following successful carotid surgery or stenting was 0.5% per annum. If we accept that successful carotid surgery or stenting will halve the risk of a long-term stroke, the background rate of these patients, had they not had a procedure, is probably 1% per annum for a fatal or disabling stroke. That's not negligible. If you say you've got 10 or 15 years of reasonable quality of life, that's quite a burden of risk to carry around because we know that successful surgery or stenting now halves that risk.

Mandrola: Where do these patients with asymptomatic carotid stenosis come from? Are these from screening facilities, screening programs, or just routine practice?

Bulbulia: Routine practice, really, John. In most of continental Europe and certainly in the United Kingdom, there's no appetite or market for routine carotid artery screening. Most of these patients came via the clinic. Possibly a contralateral carotid stenosis in a patient who previously had a symptomatic carotid intervention. Possibly patients who had some carotid artery imaging as part of a workup for neurologic symptoms that were actually not carotid related, such as posterior circulation events or stroke mimics.

Indeed, around one fifth to one third of these patients will have had some form of prior neurologic event in the run-up to surgery, either contralateral or noncarotid related. It's perhaps not right to call these entirely asymptomatic — they're not recently symptomatic — but that wasn't quite as catchy a title as ACST.

Mandrola: Fair enough. As far as generalizability goes, I know you made comments in the paper about the skill of the operator or the surgeon. I'm sure that in ACST-2 you've got extremely good operational technique. Does that play a role in generalizability of these data?

Bulbulia: Absolutely, John. I think the procedural risks we see in ACST-2 may well not be generalizable to sites outside of a trial or to contemporary or future surgical practice or stenting practice. The long-term effects on stroke reduction, and in particular, the proportional effects on long-term durability of these procedures, probably do generalize both to the trial today and to future clinical practice.

The take-home message for me is that the long-term efficacy of ACST-2 is the unique result for this trial. The procedural hazards we see in ACST-2 are consistent with those you see in registries. If you want to know what procedural risks are in the 2020s or 2030s, don't look to ACST-2, which is already historic as far as procedural risks go, but go to large, contemporary registries.

Mandrola: Finally, as far as the translation of this, if most patients can be fixed with a less-invasive stenting procedure and there's no real difference, they might vie for that. How would you answer that question?

Bulbulia: The readout for me for patients and clinicians is that this is a good, reassuring result. It's always good to have more than one way to treat a problem. Now, we have carotid surgery and carotid stenting, which have been shown to be broadly comparable both in terms of procedural risks and long-term efficacy.

I think there will be strong patient preferences that come into play, but also there will be patient characteristics that might push patients and doctors toward one treatment or another. In those circumstances, patients, surgeons, and centers can be reassured that both procedures are reasonable ones to consider, and both can be achieved, safely, and with good long-term durability.

I would also put in a plug for the long-term follow-up of ACST-2. We're going to carry on following up this population for a further 5 years, which will give us a median follow-up of 10 years and some follow-up out to 20 years for this large clinical trial. We can do that at low cost, directly with patients via mailed questionnaires.

I think it's that long-term result which will be really key, because often the Achilles heel in endovascular treatments is long-term durability. That's a question that ACST-2 with long-term follow-up is uniquely placed to answer.

Mandrola: I want to re-emphasize that for the listeners. That's really a remarkable thing about trials like this because so many of the cardiovascular trials have 2 to 5 years of follow-up. Anything more than 10 years is remarkable and laudable for sure.

Bulbulia: Thank you, John.

Mandrola: Thank you very much for being with us. I really appreciate it. Congratulations on your trial.

Bulbulia: My pleasure. Thank you, John.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.