DE Strandness Jr., MD, D.Med(Hon)
Professor of Surgery
Department of Surgery
Seattle, Washington 98195
With the interest in carotid stenting and the controversies surrounding this issue, it is clear that randomized trials will be mounted and indeed demanded. In fact
there are currently in this country at least three clinical trials that I am aware of and perhaps there are even more. The three that I am aware of are the
Schneider/Boston Scientific, the Guiding high-risk trial and finally the CREST NIH supported trial. We are the ultrasound reading center for each of these trials.
The protocols have been worked out and I am not in a position to comment on them or their implementation. At the outset what are the rules for an ultrasound reading
center and what is our role?
(1) In this century as in the latter part of the last, duplex scanning is now recognized by most clinicians in this country as the first test to be done when
carotid disease is suspected;
(2) This screening test if properly done can be used as the initial screen for entrance into ~ clinical trial as long as the degree of diameter reduction required
is agreed upon at the outset;
(3) For most symptomatic trials the accepted cut-off points for randomization are either ~ 50% diameter reduction or as shown in the large clinical trials, the greatest
benefit of intervention occurs with those lesions that exceed the >70% diameter reducing category. The expected reduction in stroke risk occurs when the degree of
diameter reduction exceeds 80%;
(4) Duplex criteria for the 70% diameter reducing stenoses and higher are easily verified as long as the testing is properly done;
(5) Vascular laboratory performance is a key issue here and has to be taken into account. The certification process with the Intersocietal Commission on the
Voluntary Accreditation of Vascular Laboratories has had a big impact on quality but even with this accreditation process, ongoing quality control is essential;
(6) The greatest uncertainty with regard to the degree of narrowing centers about the 50~ lesions. This is also true with arteriography where the intraobserver
variability even with calipers approaches 20%;
(7) For a clinical trial that includes carotid endarterectomy, there is good evidence that operation can be performed in at least 90% of the cases when the high-grade
lesions (>70%) are the target population. There are numerous studies showing that this can be done safely and without any increase in the complications. It also avoids
the low but real risk of a stroke from the angiographic procedure;
(8) Since the 50% cut-off point seems to be a reasonable on~one would well encourage angiography for this group of patients where the duplex scanning suggested the
stenosis was >50% but <70%. Given the above considerations, I can review some of the problems that we have encountered in monitoring and suggest some solutions.
One of the most important factors is to be sure that the studies are done in a uniform and acceptable fashion. Some of the problems are as follows:
(1) The ultrasonographers and staff are often accessories to the study simply performing a service function;
(2) The coordinators may not have a close enough relationship with the endovascular therapist or the surgeon. This is a critically important matter;
(3) The studies themselves may not be uniform due to changes in staff and lack of interest. This is seen often as sloppy testing and even sloppier reporting even
to the point of leaving out critical data from the reporting forms;
(4) Some of these problems have been to some degree overcome when they are addressed upfront and monitored on a continuous basis.
What can the ultrasound studies do and how might they aid in the evaluation of the role of stents? Obviously-the key issue with any form of intervention is the
morbidity and mortality outcome. What can ultrasound contribute?
(1) We can document stent position in the short and long-term;
(2) We can document thrombosis of the stent;
(3) We can document the development of myointimal hyperplasia and estimate its severity.
What we can't necessarily do;
(1) Use ultrasound findings alone as clinical endpoint;
(2) We may not be able to predict long-term outcome on the basis of a single study except perhaps when the internal carotid artery is or becomes occluded. We then have
sufficient data to indicate likely outcome;
(3) Imply that recurrent lesions even the >80% are necessarily associated with an adverse clinical outcome even though we know historically that there will be a
higher incidence of a neurological event.
Finally and most importantly it is important to realize that no single diagnostic test is perfect and all have variability that must be taken into account.
|
