SAPPHIRE trial under scrutiny


As part of the International Society of Endovascular Specialists (ISES) General Assembly there was a series of debates during the Special Session. One of these debates put the SAPPHIRE (a clinical investigation of Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy) trial under close scrutiny. The 30-day interim findings of the SAPPHIRE trial were reported in the previous issue of Vascular News (issue 16, November 2002, page 1).

During the ISES Special Session, Jay Yadav, Director of Peripheral and Carotid Interventions at the Cleveland Clinic Foundation and SAPPHIRE trial investigator, represented the SAPPHIRE trial results and Bruce Perler responded with a surgical perspective. Yadav took the audience through the trial design and said that there has to be consensus about when to randomise patients to compare carotid endarterectomy and carotid stenting for high-risk patients. Those who were considered optimal stenting candidates went into a stenting registry, which were some 409 patients, and those considered prime surgical candidates went into a surgical registry (seven patients). The rest, 307 patients, went into the randomised arm of the study. 156 received Cordis’ Precise Nitinol Self-expanding Stent – in combination with the Angioguard or Angioguard XP Emboli Capture Guidewire. The remaining 151 patients underwent carotid endarterectomy. At 30-day follow up, the major adverse event (defined as death, stroke or myocardial infarction) rate for the randomised stented group was 5.8% versus 12.6% for the endarterectomy-treated patients. These initial results were statistically significant, supporting the use of carotid stenting.

Yadav pointed out that the centres involved in the trial were experienced with at least 15 endarterectomies each year and the mean was 30 endarterectomies per year. He also highlighted that this was the first randomised control trial comparing carotid artery stenting with emboli protection to surgery. However, he said that the long-term event rates have yet to be seen.

Bruce Perler responded by saying “extrapolating and misrepresenting trial data for marketing purposes is dangerous.” He felt that the inclusion criteria were biased in favour of carotid stenting and presented more risks for carotid endarterectomy, such as recurrent carotid stenosis. He could not understand why the primary endpoints of the SAPPHIRE trial included stroke, death, and very surprisingly, myocardial infarction. He also highlighted a third problem – the randomisation process. Perler highlighted the statistic that 60% of patients refused randomisation. He also questioned the experience of the surgeons, pointing out that 50% of the SAPPHIRE surgeons carried out less than 30 carotid endarterectomies a year. Perler said that he does not consider these experienced centres. He questioned the fact that a composite outcome, including myocardial infarction, was needed to make the initial results significant.

Perler also stated that two-thirds of the patients had asymptomatic carotid disease and that the results of both carotid endarterectomy and carotid stenting were “unacceptably high” (SAPPHIRE 30-day asymptomatic results: Death/CVA 5.8% (stent); 6.1% (CEA)).

Perler concluded by saying: “SAPPHIRE is not a watershed trial, nor the nail in the coffin for carotid endarterectomy. The randomisation process was flawed and the carotid endarterectomy results were significantly worse than surgeons’ stated experience.”

Frank Veith in discussion highlighted the fact that asymptomatic carotid disease is a very benign condition and really questioned whether there should have been any operations on that number of patients who have asymptomatic carotid disease. If he were a patient, in the light of the initial results from SAPPHIRE, Veith said that he would not have anything done. You are better just leaving the disease alone was his advice.

Hugh (Jock) Beebe joined the attack on SAPPHIRE, by saying that the endpoint should be the prevention of stroke not to show that a procedure is safe. He, therefore, claimed that the whole basis of the SAPPHIRE trial is wrong because the trial is looking at the wrong end points.