To the Editor: Gurm and colleagues (April 10 issue)1 concludethat carotid stenting is not inferior to carotid endarterectomyin high-risk patients. This conclusion may be misleading. First,patients were at high risk not for stroke but for surgery. Second,in the absence of controls, the best strategy could have beenmedical therapy alone. A pooled analysis of the AsymptomaticCarotid Atherosclerosis Study2 and the Asymptomatic CarotidSurgery Trial3 shows that in patients with asymptomatic carotidartery stenosis (71% of patients randomly assigned to a studygroup in the Stenting and Angioplasty with Protection in Patientsat High Risk for Endarterectomy [SAPPHIRE] study) who did notundergo immediate surgery, the risks after a mean follow-upof 3.16 years were 8.6% for any stroke and 14.3% for any causeof death, which are similar to the risks observed in the stentinggroup of the SAPPHIRE trial after 3 years of follow-up (9.0%and 18.6%, respectively). Finally, if the subjects were at highrisk for surgery, which was therefore not an appropriate optionfor most patients, a noninferior technique cannot be recommendedeither.
Clotilde Balucani, M.D. Charlotte Cordonnier, M.D., Ph.D. Lille University Hospital 59037 Lille, France clotilde_balucani{at}hotmail.com
References
Gurm HS, Yadav JS, Fayad P, et al. Long-term results of carotid stenting versus endarterectomy in high-risk patients. N Engl J Med 2008;358:1572-1579. [Free Full Text]
Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;273:1421-1428. [Free Full Text]
Halliday A, Mansfield A, Marro J, et al. Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomised controlled trial. Lancet 2004;363:1491-1502. [Erratum, Lancet 2004;364:416.] [CrossRef][Web of Science][Medline]
To the Editor: How many patients had atrial fibrillation inthe SAPPHIRE study, which compared carotid artery stenting withendarterectomy? We would expect a high prevalence of atrialfibrillation because of relevant risk factors in the study population,including older age, hypertension, coronary heart disease, heartfailure, and smoking.1 Differences in the prevalence of atrialfibrillation might have contributed to differences in the incidenceof stroke. Were the use of antiplatelet agents, the intensityof medical therapy, and bleeding complications registered? Howwould the authors explain the rates of stroke that were higheramong asymptomatic patients (10.3% in the stenting group and9.2% in the endarterectomy group) than among symptomatic patients(6.0% in the stenting group and 8.7% in the endarterectomy group)?How would they explain the decreasing divergence of the survivalcurves at the 3-year follow-up?
Claudia Stöllberger, M.D. Josef Finsterer, M.D., Ph.D. Krankenanstalt Rudolfstiftung 1030 Vienna, Austria claudia.stoellberger{at}chello.at
References
Stewart S, Hart CL, Hole DJ, McMurray JJ. Population prevalence, incidence, and predictors of atrial fibrillation in the Renfrew/Paisley study. Heart 2001;86:516-521. [Free Full Text]
To the Editor: Gurm et al. report long-term follow-up data fromthe SAPPHIRE trial, which showed no significant difference betweencarotid artery stenting and endarterectomy in outcomes at 3years. In their study, 14.4% of the patients in the stentinggroup and 29.9% of the patients in the endarterectomy groupwere lost to follow-up. It has been suggested that a rate ofloss that is less than 5% probably leads to little bias, whereasa rate of loss that is greater than 20% potentially poses seriousthreats to validity.1 Therefore, the influence of loss biason the interpretation of these results is substantial and shouldbe considered. Moreover, a previous report2 indicated that 310patients received the assigned treatment. When they comparedcarotid artery stenting with endarterectomy at 3 years, theauthors calculated all the percentages on the basis of the 334patients who were randomly assigned to a study group originally.We would like to know why the authors did not include an analysisbased on patients receiving the assigned treatment.
Jin-Tai Yu, M.D. Lan Tan, M.D. Qingdao Municipal Hospital Qingdao 266071, China dr.tanlan{at}163.com
References
Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. New York: Churchill Livingstone, 1997.
Yadav JS, Wholey MH, Kuntz RE, et al. Protected carotid-artery stenting versus endarterectomy in high-risk patients. N Engl J Med 2004;351:1493-1501. [Free Full Text]
The authors reply: We agree with Balucani and Cordonnier thatthere are no data from randomized trials to support or refutethe need for revascularization in addition to best medical therapyin patients with severe carotid stenosis who are at high surgicalrisk. As we noted in our article, the SAPPHIRE trial randomlyassigned only patients who were referred by their physiciansfor carotid revascularization to undergo either surgery or protectedstenting if a surgeon and an interventionalist thought thatboth procedures could be performed with an acceptable risk ofprocedural complications. These data cannot be extrapolatedto all patients with carotid stenosis, and the need for treatmentin each patient must be individualized on the basis of a carefulassessment of the patient's risk of stroke with medical therapyor revascularization as well as his or her expected long-termsurvival.
With regard to the comments of Stöllberger and Finsterer,29 patients in the endarterectomy group (17%) and 14 in thestenting group (8%) had a history of atrial fibrillation. Inthis group, two patients who underwent endarterectomy had ipsilateralstrokes (on day 2 and day 87), as did two in the stenting group(on day 0 and day 182). Among patients with no history of atrialfibrillation, there was no significant difference at 3 yearsbetween stenting and surgery with regard to the cumulative incidenceof death (18% and 20%, respectively), stroke (8% and 9%, respectively),or the prespecified major secondary end point (23% and 26%,respectively). The details of antithrombotic therapy duringlong-term follow-up were not recorded, but there were no reportedmajor bleeding events after 30 days. We believe that the smallnumbers of patients and events in the subgroups of patientswith asymptomatic and symptomatic disease preclude any meaningfulconclusions, although we agree with the concern about possiblyhigh rates of events among asymptomatic patients who receivedeither treatment. We believe that the survival curves shouldbe interpreted as being not significantly different, with thechanges after 1 year indicating only three more ipsilateralstrokes and one less death from a neurologic cause in the stentinggroup.
Yu and Tan are concerned about data being presented on an intention-to-treatbasis rather than on an as-treated basis. We believe that theimportant safety issues in this study require an intention-to-treatanalysis.1 We also included the estimates of events using survivalmethods to partially correct for incomplete follow-up. The resultsat 3 years, however, were similar when the population was analyzedon an as-treated basis for the cumulative incidence of the majorsecondary end point (25% in the stenting group vs. 28% in theendarterectomy group) and ipsilateral stroke (6% in the stentinggroup vs. 5% in the endarterectomy group).
Hitinder Gurm, M.D. University of Michigan School of Medicine Ann Arbor, MI 48109-5853
Donald E. Cutlip, M.D. Harvard Clinical Research Institute Boston, MA 02215 don.cutlip{at}hcri.harvard.edu
References
Lachin JM. Statistical considerations in the intent-to-treat principle. Control Clin Trials 2000;21:167-189. [Erratum, Control Clin Trials 2000;21:526.] [CrossRef][Web of Science][Medline]
Previous
