Preprocedural risk stratification: Identifying an appropriate population for carotid stenting☆☆☆

Materials and methods 

During a 10-year period ending on December 31, 1998, 3061 carotid endarterectomies were performed at a single institution and entered prospectively into a computerized registry. The presenting symptoms were those of transient ischemic attacks (TIAs) in 742 patients (24.2%) and stroke in 341 patients (11.1%), whereas 1978 patients (64.6%) had asymptomatic, critically stenotic carotid lesions (Table I). Preoperative baseline comorbid conditions were recorded at the time of admission, as were the in-hospital complications of stroke, myocardial infarction, and death. Patients were subdivided into high-risk and low-risk subgroups, defining the high-risk patients as those with one or more of the following: coronary artery disease requiring angioplasty or bypass grafting within the 6 months preceding carotid repair, history of congestive heart failure (CHF), severe chronic obstructive pulmonary disease (COPD), or renal insufficiency with creatinine level greater than 3.0 mg/dL. Overall, 594 patients (19.4%) met the criteria for the high-risk subset, and 2467 patients (80.6%) were considered low-risk (Fig 1).

View Large Image Download to PowerPoint

Fig. 1. Proportion of patients by risk category.

Table I. Rate complications by indication for operation

	Indication	N	% Total	Composite	%	Death	%	Stroke	%	MI	%
	Asymptomatic	1978	64.6	66	3.3	19	1.0	32	1.6	23	1.2
	TIA	742	24.2	28	3.8	8	1.1	20	2.7	6	0.8
	Stroke	341	11.1	21	6.2	7	2.1	11	3.2	8	2.3
	Total	3061	100.0	115	3.8	34	1.1	63	2.1	37	1.2

Composite end point occurred when the patient experienced one or more of the solitary end points: death, stroke, or myocardial infarction (MI).

The rate of in-hospital stroke, myocardial infarction, and death, as well as the composite end point of one or more of these three events, were calculated for the high- and low-risk subsets. It should be noted that, in this prospective registry analysis, it was possible that minor postprocedure cardiac and neurologic events were not recorded; cardiology and neurology consultations were not mandated. Event rates were compared between groups with a χ2 analysis, assuming significance when the 2-tailed P value was less than .05.

Results 

The rate of the composite end point stroke/myocardial infarction/death was 3.8% (115 patients) in the total group of 3061 patients who underwent endarterectomy (Table I). As individual end points, the rate of stroke was 2.1%, myocardial infarction 1.2%, and death 1.1%. Patients with asymptomatic carotid lesions had a lower rate of perioperative morbid events compared with patients with stroke as the presenting symptom (3.3% vs 6.2%, P = .019). Patients presenting with TIA had a rate of perioperative stroke, myocardial infarction, or death that was intermediate between the two other groups (3.8%, P = .192 vs asymptomatic patients, P = .115 vs patients with stroke).

The rate of perioperative stroke/myocardial infarction/death was greater in those who had concurrent cardiac surgery and carotid endarterectomy than in those undergoing carotid endarterectomy alone, 8.5% vs 3.0% (P < .0005). Stroke alone also occurred with greater frequency in patients with combined cardiac/carotid procedures, 4.8% vs 1.7% (P < .0005), as did death, 5.3% vs 0.5% (P < .0001). By contrast, the frequency of perioperative myocardial infarction was not statistically dissimilar in patients with combined procedures (1.8%) versus those in whom a carotid procedure was performed alone (1.1%, P = .412).

Among patients who met the criteria for high risk, the composite end point stroke/myocardial infarction/death occurred in 7.4% during the index hospitalization (Fig 2, Table II).

View Large Image Download to PowerPoint

Fig. 2. Rate of perioperative events in low- and high-risk subgroups. MI , Myocardial infarction.

Table II. Rate of stroke, myocardial infarction, and death in low-risk and high-risk patients undergoing carotid endarterectomy

	Risk group	Indication	Total patients	% Total	D/S/MI	%	Death	%	Stroke	%	MI	%
	High risk	Asymptomatic	436	73.4	30	6.9	16	3.7	14	3.2	5	1.1
		TIA	107	18.0	9	8.4	6	5.6	4	3.7	3	2.8
		Stroke	51	8.6	5	9.8	4	7.8	3	5.9	2	3.9
		Total	594	19.4	44	7.4	26	4.4	21	3.5	10	1.7
	Low risk	Asymptomatic	1542	62.5	36	2.3	3	0.2	18	1.2	18	1.2
		TIA	635	25.7	19	3.0	2	0.3	16	2.5	3	0.5
		Stroke	290	11.8	16	5.5	3	1.0	8	2.8	6	2.1
		Total	2467	80.6	71	2.9	8	0.3	42	1.7	27	1.1

D/S/MI , Composite end point of death, stroke, or myocardial infarction.

Table III. Rate of stroke, myocardial infarction (MI), and death in patients undergoing carotid endarterectomy alone and in combination with cardiac procedures, subgrouped with respect to the presence or absence of risk factors (coronary artery procedures within 6 months, history of CHF, severe COPD, or creatinine level greater than 3.0 mg/dL)

	Cartoid/cardiac	Risk group	Indication	Total patients	% Total	Composite	%	Death	%	Stroke	%	MI	%
	Cartoid alone	With risk factors	Asymptomatic	130	66.7	6	4.6	3	2.3	2	1.5	1	0.8
		TIA	42	21.5	2	4.8	1	2.4	0	0.0	1	2.4
		Stroke	23	11.8	2	8.7	1	4.3	0	0.0	1	4.3
		Total	195	6.4	10	5.1	5	2.6	2	1.0	3	1.5
		Without risk factors	Asymptomatic	1542	62.5	36	2.3	3	0.2	18	1.2	18	1.2
		TIA	635	25.7	19	3.0	2	0.3	16	2.5	3	0.5
		Stroke	290	11.8	16	5.5	3	1.0	8	2.8	6	2.1
		Total	2467	80.6	71	2.9	8	0.3	42	1.7	27	1.1
		All	Asymptomatic	1672	62.8	42	2.5	6	0.4	20	1.2	19	1.1
		TIA	677	25.4	21	3.1	3	0.4	16	2.4	4	0.6
		Stroke	313	11.8	18	5.8	4	1.3	8	2.6	7	2.2
		Total	2662	87.0	81	3.0	13	0.5	44	1.7	30	1.1
	Cartoid with cardiac	With risk factors	Asymptomatic	66	78.6	9	13.6	4	6.1	5	7.6	2	3.0
		TIA	11	13.1	2	18.2	1	9.1	2	18.2	1	9.1
		Stroke	7	8.3	2	28.6	2	28.6	1	14.3	0	0.0
		Total	84	2.7	13	15.5	7	8.3	8	9.5	3	3.6
		Without risk factors	Asymptomatic	240	76.2	15	6.3	9	3.8	7	2.9	2	0.8
		TIA	54	17.1	5	9.3	4	7.4	2	3.7	1	1.9
		Stroke	21	6.7	1	4.8	1	4.8	2	9.5	1	4.8
		Total	315	10.3	21	6.7	14	4.4	11	3.5	4	1.3
		All	Asymptomatic	306	76.7	24	7.8	13	4.2	12	3.9	4	1.3
		TIA	65	16.3	7	10.8	5	7.7	4	6.2	2	3.1
		Stroke	28	7.0	3	10.7	3	10.7	3	10.7	1	3.6
		Total	399	13.0	34	8.5	21	5.3	19	4.8	7	1.8
	Total	3061	100.0	115	3.8	34	1.1	63	2.1	37	1.2

Composite end point is defined by the occurrence of perioperative stroke, MI, or death.

Among the group undergoing combined cardiac and coronary procedures, patients with a history of CHF, severe COPD, or renal insufficiency with serum creatinine level greater than 3.0 mg/dL, stroke, death, or myocardial infarction occurred with significantly greater frequency than in those who did not manifest one of these baseline comorbidities. The rate of the composite end point rose to 15.5% in this subgroup of patients, compared with 6.7% in patients undergoing combined procedures who did not manifest one of the comorbidities (P = .020).

Discussion 

Carotid endarterectomy is considered to be a remarkably safe procedure with well-proven benefits in symptomatic and asymptomatic patients with high-grade stenotic lesions.3, 4 Among 326 patients in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) who underwent carotid endarterectomy for symptomatic stenotic lesions of 70% or greater diameter reduction, the risk of stroke or death was 5.5% within 30 days of operation.5 There was a 17% absolute risk reduction in the 2-year risk of ipsilateral stroke when compared with patients treated medically. Among 1087 patients in the NASCET trial who underwent endarterectomy for lesions between 50% and 69% diameter reduction, the 30-day ipsilateral stroke rate was 5.6% within 30 days of operation.6 In this group with more moderate stenoses, there was a 6.5% absolute risk reduction (15.7% vs 22.2%, 29% relative risk reduction) in the 5-year risk of ipsilateral stroke compared with patients treated medically. The Asymptomatic Carotid Atherosclerosis Study (ACAS) documented similar benefits of carotid endarterectomy in patients with asymptomatic carotid stenoses of greater than 60%.7 The risk of perioperative morbid events was extremely low in the ACAS population, with a 30-day ipsilateral stroke rate of 1.8% (0.6% attributable to angiography and 1.2% to the operative procedure) and a mortality rate of 0.1%. The risk of the composite end point (perioperative stroke, death, or ipsilateral stroke) during a period of 5 years was 5.1% for surgically treated patients versus 11.0% for those treated medically. Carotid endarterectomy was associated with a 53% reduction in the risk of the primary end point during 5 years of follow-up.

The publication of the NASCET and ACAS data was associated with a dramatic increase in the number of patients undergoing carotid endarterectomy, although a causal relationship can only be suspected. In Medicare beneficiaries, the frequency of the procedure decreased from 61,273 per annum in 1985 to 46,571 in 1989, rising to 108,275 in 1996 after the appearance of the data from the randomized trials.1, 8 At the same time, the mortality rate in these patients decreased from 3.0% in 1985 to 2.5% in 1989 and to 1.6% in 1996.

The excellent outcomes of the randomized trials cannot, however, be directly applied to the total population of patients undergoing carotid endarterectomy. In fact, there exist data to suggest that the results of the trials cannot be generalized to patients undergoing endarterectomy at the hospitals that participated in the studies. Hsia et al8 reported that the 30-day mortality rate in Medicare beneficiaries undergoing carotid endarterectomy in highvolume urban teaching hospitals of the type that participated in the clinical trials was only 1.0%, compared with greater than 2.5% overall. In a study of 113,000 Medicare patients who underwent carotid endarterectomy during the conduct of the NASCET and ACAS trials (1992 to 1993), Wennberg et al9 noted that the perioperative mortality rate was 1.4% in hospitals participating in the trials and 1.7% in hospitals that did not participate in the trials. The rate of perioperative death rose to 2.5% in lowvolume nontrial hospitals where fewer than seven carotid endarterectomies were performed yearly. These relatively high complication rates are in direct contrast to the much lower mortality rates observed in the patients entered into the trials (0.1% in ACAS and 0.6% in NASCET). These findings suggest that eligibility criteria were sufficiently strict that patients in the NASCET and ACAS represented a small subset of the total population of patients undergoing carotid endarterectomy, comprising a subgroup with the lowest frequency of baseline comorbid conditions and, likely, the lowest rate of perioperative adverse events. These observations also question the widely held belief that carotid endarterectomy is a trivial procedure with an almost nonexistent rate of perioperative mortality.

The data from the Cleveland Clinic registry offer an explanation for the findings of Wennberg et al.9 Patients in the multicenter trials of carotid endarterectomy were likely similar to the low-risk group of patients undergoing carotid repair at the Cleveland Clinic. In fact, the mortality rate of 0.2% in 1542 low-risk asymptomatic patients treated with carotid endarterectomy is remarkably similar to the ACAS mortality rate of 0.1%. Similarly, the mortality rate was 0.5% in 925 symptomatic patients undergoing carotid endarterectomy at the Cleveland Clinic, almost identical to the 0.6% mortality rate observed in the NASCET trial.

These observations have implications in the design of trials aimed at assessing carotid stenting procedures. Periprocedural morbidity is likely to be highest during the relatively early evolution of stent design and stenting treatment paradigms. At a time when the safety and efficacy of carotid stenting remain unproved, investigations seem most appropriately undertaken in the population of patients with less ideal outcomes following the standard therapy of carotid endarterectomy. The current study sought to identify comorbid conditions that increase the risk of endarterectomy; there exist a variety of anatomic features that are also associated with poorer outcome. These include such variables as contralateral carotid occlusion,6 recurrent carotid lesions,10 and a history of radiation therapy to the neck.11 In addition, authors such as Sundt et al12 and, more recently, Rockwell et al13 have outlined other clinical and anatomic factors that are associated with an increased risk for poor outcome after carotid endarterectomy. These issues may be of equal or greater importance than medical comorbidities and should be the topic of subsequent investigations.

Whereas it may be unrealistic to assume that the results of carotid stenting will attain clinical equipoise with endarterectomy in all patient subgroups, it seems quite reasonable to predict that this potentially less invasive procedure will result in a lower complication rate in medically compromised, high-risk patients. If equivalence of carotid stenting to endarterectomy cannot be demonstrated in this group of patients with less ideal outcomes after carotid endarterectomy, it is unlikely that stenting can compete successfully with endarterectomy in a lower risk group. These considerations argue for focusing on an evaluation of stenting in a select subset of the population with carotid disease, a subset in which standard therapy is associated with a high rate of complications on the basis of significant baseline medical comorbidities.