SAPPHIRE: Stenting and Angioplasty With Protection in Patients at HIgh Risk for Endarterectomy

Luis Gruberg, MD, FACC


November 25, 2002

Presenter: Jay Yadav, MD, Cleveland Clinic Foundation, Cleveland, Ohio

Several prospective randomized trials have supported the benefit of carotid endarterectomy (CEA) over medical therapy alone for the treatment of significantly obstructed carotid lesions. The North American Symptomatic Carotid Endarterectomy Trial (NASCET)[1] demonstrated the superiority of endarterectomy over medical treatment for symptomatic carotid stenosis ≥ 70%. More recently, the Asymptomatic Carotid Atherosclerosis Study (ACAS)[2] showed a statistically significant reduction in stroke incidence after carotid endarterectomy in asymptomatic patients with a carotid stenosis of ≥ 60%.

Just as standard stenting procedures reduced the need for surgery, clinicians hoped that employing a minimally invasive approach to manage carotid lesions would also lessen the need for surgically based CEA procedures. Elective carotid stenting was initiated in 1994 by the Endovascular Interventional Team at the University of Alabama at Birmingham. In addition to preventing the surgical wound-related complications of CEA, an endovascular approach would eliminate the need for general anesthesia, and could have several distinct applications, especially in patients who are considered to be at significantly high risk for CEA.

However, carotid stenting has faced a great deal of criticism, particularly from the surgical community. By virtue of its proximity to the brain, the procedure requires precise technical skill. The procedure has been faulted because of the associated increased risk of stroke, resulting from released particle debris at the treated site that travels up the bloodstream to the brain. Distal embolic protection devices have emerged as a means to reduce the incidence of stroke during stenting by collecting and trapping particle debris during the reperfusion procedure. The approach had not previously been widely studied in carotid artery stenting patients.


SAPPHIRE (Stenting and Angioplasty With Protection In Patients at High Risk for Endarterectomy) was a randomized, multicenter (29 sites) trial that compared carotid artery stenting with the use of distal protection to CEA in patients at high risk for surgical treatment. Patients in the stent group were treated with a Precise nitinol self-expanding stent and the AngioGuard distal protection device (Cordis, a Johnson & Johnson company).

The study was unique in that both surgeons and interventional cardiologists had to meet certain procedural criteria to participate in the study. Surgeons were required to have ample surgical expertise with CEA procedures, defined as having performed an average of 30 CEAs annually (procedural range 15-100) with a corresponding low rate of major complications (stroke, death, or myocardial infarction [MI] < 1%). Likewise, interventionalists required experience with percutaneous carotid intervention, performing an average of 64 annual interventions (20-700) with low complication rates (< 2% for stroke and TIA).

Enrollment Criteria and Randomization

Patients recruited in the trial required a ≥ 50% stenosis in the common or internal carotid artery, assessed by ultrasound or angiography in symptomatic patients, and ≥ 80% in asymptomatic patients with 1 or more comorbidity criteria. Consensus agreement by a multidisciplinary team of neurologists, surgeons, and interventionalists was required for enrollment into the randomization arm of the study.

A total of 723 patients were enrolled in the SAPPHIRE trial. Consensus was achieved in 307 patients who were randomized to either stenting (n = 156) or CEA (n = 151). Patients who did not undergo randomization entered a stent or surgical registry. The stent registry consisted of 409 patients who were refused by the surgeon for surgery, and the surgical registry included 7 patients refused by the interventionalist. Both registries were completed in February 2002.

In June 2002, the randomization arm was stopped prematurely due to slow enrollment, which investigators attributed to the fact that both patients and physicians were reluctant to undergo or perform CEA.

Primary and Secondary Endpoints

The primary endpoints of the trial included a clinical composite of death, stroke, and MI at 30 days; and major adverse cardiac events at 30 days, plus death and ipsilateral stroke between day 31 and up to 12 months after the procedure. Secondary endpoints included:

  • Patency, defined as < 50% restenosis by ultrasound at 48 hours, 6 months, 1, 2, and 3 years postprocedure.

  • Disabling stroke at 30 days and 6 months.

  • Composite of major adverse clinical events.

  • Safety assessment of the distal protection device .

Researchers also assessed quality of life and cost-effectiveness.

Patient Enrollment and Clinical Characteristics

Among patients in the randomized group, a significantly higher number of patients in the stent arm had undergone previous coronary artery bypass graft (CABG) and also had higher history rates of cardiovascular disease compared with patients randomized to CEA (P < .05; Table 1).

Table 1. SAPPHIRE: Baseline Clinical Characteristics of Randomized Patients
(n = 156)
(n = 151)
Age (yrs) 73 73
Female (%) 33 31
History of CVD (%)* 85 74
Symptomatic (%) 32 29
S/P CABG (%)* 43 31
S/P PTCA (%) 34 24
Diabetes (%) 24 27

CABG = coronary artery bypass graft; CEA = carotid endarterectomy; CVD = cardiovascular disease; PTCA = percutaneous transluminal coronary angioplasty
*P < .05

The baseline characteristics of patients from the stent registry were very similar to the baseline characteristics of those who were randomized. Previous CEA or post-radiation treatment were 2 of several high-risk characteristics that influenced the decision to refuse surgical treatment to some patients (Table 2).

Table 2. SAPPHIRE: High-Risk Surgical Characteristics by Treatment Arm
  Randomized Group
(n = 307)
Stent Registry
(n = 409)
Post-radiation treatment (%) 7 17
Previous CEA (%) 25 41
High cervical ICA (%) 6 17
Common carotid lesion below clavicle (%) 0 4
More than 1 criterion (%) 25 37

CEA = carotid endarterectomy; ICA = intracranial aneurysm

30-day Results

Although individual endpoints of death, stroke, and MI were lower in patients randomized to carotid artery stenting, the differences did not reach statistical significance (Figure 1).

Figure 1. SAPPHIRE: Individual components at 30 days.

Despite the fact that the study was stopped prematurely, the combined endpoint of death/stroke/MI was statistically significantly lower in patients randomized to stenting vs CEA (5.8 vs 12.6, P = .047; Figure 2).

Figure 2. SAPPHIRE: 30-day composite endpoint (death/stroke/MI).

Analyses of each endpoint for each group (symptomatic and asymptomatic patients) found no significant difference in outcome between stenting or CEA for any of the composite or individual components of the primary endpoint. This was due to the small patient population studied. On a numerical basis, however, both symptomatic and asymptomatic patients randomized to stenting had better outcomes (Table 3).

Table 3. SAPPHIRE: Stenting vs CEA in Symptomatic and Asymptomatic Patients
  Stenting CEA
Symptomatic patients (n) 48 39
Primary composite endpoint* (%) 4.2 15.4
Death (%) 0.0 5.1
Stroke (%) 2.1 7.7
MI (%) 2.1 5.4
Death/stroke (%) 2.1 10.3
Asymptomatic patients (n) 104 98
Primary composite endpoint (%) 6.7 11.2
Death (%) 1.0 1.0
Stroke (%) 4.8 5.1
MI (%) 2.9 7.1
Death/stroke (%) 5.8 6.1

*P = .13; P = .33

Two of the secondary endpoints (transient ischemic attacks [TIAs] and major bleeding) were similar between the 2 groups, but cranial nerve injury, another secondary endpoint, had a significantly higher incidence in patients randomized to the surgical arm (P < .01; Figure 3).

Figure 3. SAPPHIRE: rates of TIA, major bleeding, and stroke at 30 days.

The rates of occurrence of secondary endpoints were similar in the stent registry (TIAs 5.4%, major bleeding 11.5%, and no reported cranial nerve injury).


SAPPHIRE is the first randomized study to compare carotid artery stenting with distal embolic protection to CEA in only high-risk patients. This was also one of the first studies that relied upon an interdisciplinary approach to determine eligible (or ineligible) candidates for therapy, and surgical ineligibility was based specifically on the judgment of the surgeon.

Overall, patients randomized to carotid stenting with embolic protection had a significantly lower 30-day rate of major adverse cardiac events (death/stroke/MI) compared with patients randomized to CEA. In addition, there was a favorable trend for stenting in all individual endpoints in both symptomatic and asymptomatic patients and in the stent registry. This trial also provides valuable information on the rates of surgical complications in a large group of patients that had been excluded from previous surgical trials.

Follow-up and analyses of secondary endpoints remain ongoing for these patients, and 1-year results are expected by mid-2003.


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.