Questions from the SAPPHIRE trial

The Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial was an industry-sponsored, randomized study designed to address the question of equivalence between carotid artery stenting (CAS) and carotid endarterectomy (CEA) in a group of patients designated as being at high risk for complications during CEA. Although the results of this trial have been widely publicized, a number of serious concerns have been raised about the study regarding early termination because of recruitment problems, the large number of recurrent stenoses (20%), which is uncharacteristic of any surgical series of endarterectomies, and the definitions of high-risk patients.

The medical definitions of high-risk patients are based on exclusion criteria for the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and the Asymptomatic Carotid Atherosclerosis Study (ACAS). The exclusion criteria for these studies were based on concerns over long-term follow-up, not on any evidence that the criteria implied increased perioperative risk. In fact, several studies have demonstrated that excellent surgical results can be obtained in the high-risk medical group.1-3 Therefore, one may then question whether the actual selection criteria used for the SAPPHIRE trial reflect the realities of a clinical practice.

When the trial results were finally published, a number of additional questions were raised, including:

(1) Why were more than 50% of the patients (70% of whom were asymptomatic) unsuitable for endarterectomy, whereas less than 2% were unsuitable for stenting? What does this say about the perspective of the investigative centers?

(2) Why did 5% of CEA patients fail to achieve “target artery revascularization”? This is highly unusual for any surgical series and raises questions about either patient selection or operative results.

(3) The stenting group had a higher incidence of preprocedural coronary revascularization, and all of these patients received periprocedural clopidogrel (Plavix). How did these factors affect the 30-day myocardial infarction rate, which was the major difference in study outcome between the two groups?

(4) Finally, given the results of both stenting and endarterectomy in this series, how can the authors recommend either approach from their experience for patients with asymptomatic carotid stenosis?

Since SAPPHIRE’s publication, additional data have come out, raising concerns about the widespread application of carotid stenting. These include the Imaging in Carotid Angioplasty and Risk of Stroke (ICAROS) study in which 7.1% of patients with echo-lucent plaque had a stroke, a rate that was not influenced by use of embolic protection devices,4 as well as the lead in results of the Carotid Revasculari-

zation Endarterectomy versus Stenting Trial (CREST) in which the stroke rate after stent-ing was higher than 5% in patients older than age 70 and higher than 12% in patients older than age 80.5 The increased risk of stroke after CAS that is associated with advanced age, which was one of the risk factors in the SAPPHIRE study, may shed some light on the high complication rates seen in the SAPPHIRE investigation. Taken together, these recent publications on carotid stenting cause one to continue to regard the procedure with a healthy concern.

Advocates of stenting have used the SAPPHIRE trial results to justify a broader application of carotid stenting to clinical situations that have not been properly investigated. The fallacy of this approach is highlighted by the case report by Gurm and Yadav (page 32) in which an asymptomatic, high-grade carotid stenosis was stented prior to coronary artery bypass graft (CABG) surgery. Although this may ultimately prove to be appropriate therapy, there are no data to justify this approach at present. The appropriate treatment of combined coronary artery disease itself is unsettled. After over a decade of studying this problem, my colleagues and I have concluded that the increased complication rates associated with combined CEA and CABG surgery are primarily a reflection of the comorbidities of the patients with combined disease, not the addition of the CEA.6-8 Put another way, in a patient having CABG surgery, adding CEA, particularly for an asymptomatic stenosis, does not seem to increase morbidity or mortality. No data on CAS in this patient group are available (pre-CABG surgery). Stenting may be associated with increased risk in these patients, given their 5% to 8% incidence of aortic arch, and it does

not avoid an open operation (CABG surgery). Unless carotid stenting can reduce stroke risk in these patients, its only effect would seem to be to delay the coronary bypass by 6 to 8 weeks, as in the case reported by Gurm and Yadav. Clearly, further studies are needed.

There is no doubt that carotid stenting has a role in the overall treatment of carotid bifurcation stenosis. At present, its clearest role is in “anatomic high-risk” patients. These patients constitute approximately 5% of the overall endarterectomy population.1 Data suggest that stenting may be inferior

to endarterectomy in patients with echo-lucent plaque or advanced age.4,5 SAPPHIRE notwithstanding, CAS has not been shown to be superior to well-performed endarterectomy in properly selected patients. The clearest lesson from SAPPHIRE is that asymptomatic patients should not have interventions of any sort if the anticipated stroke and death rate exceeds 5%, which was the case in the SAPPHIRE trial.

Conclusion

The reason the morbidity and mortality rates reported with CEA in the SAPPHIRE trial are so much higher than those reported from single centers cannot be determined. Perhaps it has to do with patient selection and surgeons deferring any interventions in asymptomatic patients who are truly thought to be at high risk. Clearly a trial comparing stenting, endarterectomy, and medical therapy in asymptomatic patients, as well as a trial comparing stenting plus coronary bypass with concurrent endarterectomy and bypass is the logical next step.