The SFA Discussion: What happens next?

I could not believe that a discussion regarding the SFA would continue this long but we could always find something to discuss. This month I wanted to examine why it is so hard to come to the realization that one device may trump another. From my perspective, I have had the esteemed honor to be the course director for an inaugural fellow’s course supported by industry this past week. The faculty I put together included some of the best and brightest who were, in fact, world-recognized for their ability and scientific achievements.

I gave a talk on the current data for the iliac and femoral territories, as another faculty member was assigned to discuss the infrapopliteal space. What we covered is exactly what we covered in the past year over and over again with the specifics we have focused on regarding lesion length and patency or the metrics of the PSVR. I discussed those trials that we should all know about given their significance to our data set and because implementing their findings into practice would be a good exercise in the practice of evidence-based medicine. We then discussed additional trials covering important topics such as debulking strategies. We discussed directional and rotational devices as well as atheroablative (laser) devices.

In combining the data to the best of my ability, I tried to compare trial designs and outcomes and suggested that in many cases the benefit of one device (drug-eluting stent) is mitigated by nearly the exact primary patency for bare metal stents in the same lesion lengths, albeit with different PSVR endpoints. Further, if one then compares the alternative therapies with these other trials, I believe we can also state that if the 12-month data are as compelling as the 6-month data for directional atherectomy, we may have a strong argument for following the “leave nothing behind” approach as a first step. I am clearly biased toward this therapy being one of the PI’s for the DEFINITIVE LE trial, but we need to see the data at 12 months before we start the discussion. What we cannot do is simply state that the data as previously presented in a terrible trial from directional atherectomy TALON trumps the further scientific pursuits that the DEFINITIVE trial did using an independent core lab both for angiographic and sonographic data and the additional adjudication of all events using an independent CEC (clinical event committee).

If we keep our scientific hats on for a short time, despite whatever bias you have such as “I like to stent” or “I like to drill,” etc., we can take the next logical and provocative step—test one device to another. Why not test atherectomy with stenting in the short, medium, or longer lesion subsets? Choose the stent; we now have 4 or so approved. Why not compare the stents to each other? Have the trial results translated to real-world experiences with a similar primary patency as was seen in the trials?

I think we need this information to make informed decisions for our patients and, better yet, for us to no longer be subscribers to the “that is yours and this is mine” minions of our little scientific turfs.


DEFINITIVE LE is exciting prima fascia, however upon deeper analysis, the single arm design without a comparable alternative trial in terms of enrollment criteria and PSVR-based outcomes data renders it less convincing. The bally hoo surrounding the 6 month results for DEFINITIVE LE somewhat diluted my overall enthusiasm for the trial in terms of intellectual honesty.

Don't throw a big party for your 6 month data when we all know that 6 month data for virtually every intervention looks great. That having been said, I am waiting with great anticipation for the 12 month results. Mainly because I hope that it anecdotally validates my preference for primary atherectomy. I invoke the description "anectdotal" because a "trial" without a control group is essentially anectdotal. I'd like directional atherectomy to have data support as my default choice, but I have to accept the data for what it is, non-randomized data without comparison.

Add new comment

Back to top