Surgical and Endovascular Treatment of Extracranial Carotid Stenosis

Abstract

Background

Carotid endarterectomy (CEA) and carotid artery stenting (CAS) can be used to prevent stroke due to arteriosclerotic lesions of the carotid artery. In Germany, legally mandated quality assurance (QA) enables the evaluation of outcome quality after CEA and CAS performed under routine conditions.

Methods

We analyzed data on all elective CEA and CAS procedures performed over the periods 2009–2014 and 2012–2014, respectively. The endpoints of the study were the combined in-hospital stroke and death rate, stroke rate and mortality separately, local complications, and other complications. We analyzed the raw data with descriptive statistics and carried out a risk-adjusted analysis of the association of clinically unalterable variables with the risk of stroke and death. All analyses were performed separately for CEA and CAS.

Results

Data were analyzed from 142 074 CEA procedures (67.8% of them in men) and 13 086 CAS procedures (69.7% in men). The median age was 72 years (CEA) and 71 years (CAS). The periprocedural rate of stroke and death after CEA was 1.4% for asymptomatic and 2.5% for symptomatic stenoses; the corresponding rates for CAS were 1.7% and 3.7%. Variables associated with increased risk included older age, higher ASA class (ASA = American Society of Anesthesiologists), symptomatic vs. asymptomatic stenosis, 50–69% stenosis, and contralateral carotid occlusion (for CEA only).

Conclusion

These data reveal a low periprocedural rate of stroke or death for both CEA and CAS. This study does however not permit any conclusions as to the superiority or inferiority of CEA and CAS.

Arteriosclerotic lesions of the extracranial portion of the carotid artery are the cause of 10–20% of all ischemic strokes. The available methods of preventing carotid-associated stroke are optimal medical therapy, carotid endarterectomy (CEA), and stent-based carotid angioplasty (carotid artery stenting, CAS). Current national and international guidelines contain a strong recommendation for early elective CEA for patients with symptomatic 50–99% stenoses. CEA should also be considered for patients with high-grade asymptomatic carotid stenosis under certain conditions. CAS can be considered as an alternative to CEA for either symptomatic or asymptomatic stenosis, as long as the risk of complications is low (weak recommendation). For both CEA and CAS, the maximum permissible rate of periprocedural stroke and death is 3% for asymptomatic stenoses and 6% for symptomatic stenoses (1– 3).

CEA has been subject to legally mandated external quality assurance (QA) in Germany since 2003, and CAS since 2012 (4). The quality indicators of the QA program are intended to assess whether the procedures were carried out for the proper indications, as well as the type and extent of serious complications arising during hospitalization (5). The data obtained can be used to evaluate outcome quality under routine conditions (6, 7). They can also be used to test whether the results of randomized controlled trials (RCTs) apply to patients in a much larger, unselected collective.

Patients and methods

Patient population

This study is a retrospective secondary data analysis based on case-related data obtained from the mandatory external quality assurance for CEA and CAS (8).

Inclusion and exclusion criteria

All carotid revascularization procedures for clinically asymptomatic stenosis (indication group A, IG-A) and for clinically symptomatic stenosis (indication group B, IG-B) from the reporting years 2009–2014 (CEA) and 2012–2014 (CAS) were included. Procedures were excluded if they were from indication group C (emergency indication in progressive stroke or crescendo transient ischemic attacks, carotid aneurysms, coiling of the internal carotid artery, other special structural types of plaque with less than 50% stenosis, recurrent stenosis, and so-called tandem lesions) or IG-D (CEA or CAS as a component of combined procedures on the extracranial carotid artery [coronary arteries, peripheral vascular procedures, simultaneous intracranial PTA/stenting]). IG-A and IG-B account for more than 90% of all extracranial carotid procedures and consist of indications supported by high-level evidence and recommended in numerous guidelines (1– 3). The allocation to the indication groups was carried out in accordance with the generally accepted definitions for nationwide evaluations in Germany (9).

Clinical, morphological, and structural variables

The data were collected by the treating hospitals and included preprocedural data (patient characteristics, diagnostic procedures, degree of carotid stenosis, ASA class), intraprocedural data (technique of procedure and of anesthesia, stent material, procedure duration, neuro-monitoring, intraoperative completion studies, platelet inhibition), and postprocedural data (length of hospital stay after procedure, death, stroke, other local or general complications). In addition, the treating specialist department caring for the patient and the performance of neurological examinations before and/or after the procedure were recorded.

Primary and secondary endpoints

The primary endpoint of this study is the combined risk of stroke (ipsi- or contralateral) or death during hospitalization. A stroke is defined as any newly arising neurologic deficit due to cerebral dysfunction that lasts more than 24 hours, quantified according to the Modified Rankin Scale (mRS) (etable 2). A periprocedural stroke with a neurologic deficit that has resolved or is mild (mRS 0-2) at the time of discharge is classified as a mild stroke, and any stroke with mRS >2 is classified as a severe stroke. Death as an endpoint was registered administratively by the treating hospital. The secondary endpoints were the risk of in-hospital stroke or death alone and local or other complications during hospitalization.

eTable 2. Modified Rankin Scale (mRS) for the extent of disability after stroke.

Score	Extent of disability after stroke
0	No symptoms.
1	No significant disability despite symptoms; able to carry out all usual ‧duties and activities
2	Slight disability; unable to carry out all previous activities, but able to look after own affairs without assistance
3	Moderate disability; requiring some help, but able to walk without ‧assistance
4	Moderately severe disability; unable to walk without assistance and ‧unable to attend to own bodily needs without assistance
5	Severe disability. Requires constant nursing care, bedridden, incontinent.
6	Dead

Statistical analysis

The study protocol and data utilization were developed in collaboration with the AQUA Institute and approved by the German Joint Federal Committee (Gemeinsamer Bundesausschuss, G-BA) (8). Individual case data were stored at the AQUA Institute and accessed exclusively by controlled remote data processing in order to preserve confidentiality.

The patient characteristics and the procedural variables were analyzed with descriptive statistics. Time intervals were stated as medians with interquartile distances (IQ). Cerebral, local, and general complications were assessed separately for CEA and CAS. In addition, potential associations of clinically unalterable variables (age, sex, ASA class, neurological condition on admission, degree of ipsi- and contralateral stenosis) with the risk of in-hospital stroke or death were analyzed separately for CEA and CAS with a generalized linear mixed regression model including patient clustering in centers. The entire project was approved by the Ethics Committee of the Faculty of Medicine of the Technical University of Munich and was designed in accordance with current standards for secondary data analysis and observational studies (10, 11).

Results

182 033 cases were registered from 1 January 2009 to 31 December 2014; 142 074 cases of CEA and 13 086 cases of CAS met the inclusion criteria (figure 1). Most patients were male (CEA 67.8%, CAS 69.7%). The median age of patients undergoing CEA was 72 (IQ 65–77), and that of patients undergoing CAS was 71 (IQ 63–76). Among the CEA patients, more were in ASA class III than in classes I and II combined (68.0% vs. 29.4%). The data on clinical manifestations, degree of ipsi- and contralateral carotid stenosis, hospital caseload per year, and treating specialist department are given in eTable 3. The median postprocedural length of stay, in days, was 5 (IQ 2–6) after CEA and 2 (IQ 2–4) after CAS.

Figure 1.

Flowchart of the overall patient group and of patients included in the study (cases).

*¹ Crossover from PTA/stenting to open surgery (n = 68) and PTA or PTA/stenting as an adjunct to an intracranial procedure (n = 812),

*² Including patients with restenosis, simultaneous cardiovascular procedures, tandem lesions, carotid occlusion, etc. (for details, see Methods).

CAS, carotid artery stenting; CEA, carotid endarterectomy; PTA, percutaneous transluminal angioplasty; TIA, transient ischemic attack

eTable 1. Specific procedure codes for inclusion (quality assurance filters) for procedures performed in 2014 (adapted from year to year, and from 2012 onward for CAS as well), leading to automatic data entry for compulsory quality assurance (carotid revascularization module 10/2).

CEA operation codes (procedures for inclusion)
Operations on blood vessels
5–	381.	00 to 03, 06	endarterectomy
	382.	00 to 03	blood vessel resection and reanastomosis
	383.	00 to 03	resection and replacement (interposition) of (parts of) blood vessels
	395.	00 to 03	patching of blood vessels
	397.	00 to 03	other plastic reconstruction of blood vessels
		0	extracranial arteries of the head and arteries of the neck
		00	carotid a., not further specified
		01	common carotid a. with carotid sinus
		02	extracranial portion of internal carotid a.
		03	external carotid a.
		06	vascular prosthesis
	393.		insertion of a different type of vascular shunt or bypass: extracranial head and neck arteries
		00	carotid a.
		01	carotid a. – carotid a.
CAS operation codes (procedures for inclusion)
Nonsurgical treatments
8–	836.	0 0h to 0n	angioplasty (balloon) percutaneous transluminal vascular intervention
	840.	0h to 0n 1h to 1n	percutaneous transluminal implantation of non-drug-eluting stents
	841.	0h to 0n 1h to 1n	percutaneous transluminal implantation of drug-eluting stents
	842.	0h to 0n 1h to 1n	percutaneous transluminal implantation of non-drug-eluting covered stents (stent-graft)
	843.	0h to 0n 1h to 1n	percutaneous transluminal implantation of bioresorbable stents
	845.	0h to 0n 1h to 1n	percutaneous transluminal implantation of uncovered Cheatham platinum (CP) stents
	846.	0h to 0n 1h to 1n	percutaneous transluminal implantation of covered Cheatham platinum (CP) stents
	848.	0h to 0n 1h to 1n	percutaneous transluminal implantation of drug-eluting covered stents (stent-graft)
	849.	0h to 0n 1h to 1n	percutaneous transluminal implantation of other uncovered wide-lumen stents
	84a.	0h to 0n 1h to 1n	percutaneous transluminal implantation of other covered wide-lumen stents
		0	one stent
		1	two stents
		h	carotid a., not further specified
		j	common carotid a.
		k	extracranial portion of internal carotid a.
		m	extracranial portion of internal carotid a. and common carotid a.
		n	external carotid a.

Approximately 90% of CEA procedures and 98% of CAS procedures were carried out under antiplatelet medication. 26.3% of CEAs were carried out under locoregional anesthesia. The use of a protection system in CAS varied depending on the indication group: 67.1% in IG-A and 48.4% in IG-B. Further data on procedure length and technique, stent types and design, neurophysiological monitoring, and intraopertive morphological completion studies are given in eTable 4.

eTable 4. Periprocedural variables in elective cases of surgical (CEA, 2009–2014) and endovascular (CAS, 2012–2014) treatment of asymptomatic and symptomatic carotid stenosis in Germany.

	n	%
Carotid endarterectomy (CEA)	142 074	100
Inhibition of platelet aggregation*1	106 653	90.4
Anesthesia technique
general anesthesia	101 522	71.5
locoregional anesthesia	37 355	26.3
combined/converted	3197	2.3
Surgical technique
CEA with direct suture	1765	1.2
CEA with patching	71 920	50.6
eversion CEA	60 297	42.4
other techniques*2	8092	5.7
Other technical aspects
intraluminal shunt placement	61 074	43.0
neurophysiological monitoring*3	51 945	36.6
intraoperative completion study (angiography, DUS)	69 561	49.0
Overall clamping time (median, minutes, Q25–Q75)*4	16 min (5– 27)	–
Overall clamping time only with shunting (median, minutes, Q25–Q75)*4	5 min (3– 7)	–
Overall clamping time without shunting (median, minutes, Q25–Q75)*4	25 min (18– 33)	–
Duration of procedure (median, minutes, Q25–Q75)*4	85 min (66–105)	–
Carotid artery stenting (CAS)	13 086	100
Inhibition of platelet aggregation	12 822	98.0
Type of intervention
angioplasty alone	441	3.4
stenting alone	810	6.2
angioplasty and stenting	11 835	90.4
Stent type
nitinol	9618	73.5
stainless steel	2553	19.5
other	474	3.6
Stent design
open-cell	4356	33.3
closed-cell	6554	50.1
semi-closed-cell	1416	10.8
other	319	2.4
Protection device	7901	60.4
Other aspects
neurophysiological monitoring*3	4968	38.0
intraprocedural lysis	175	1.3
duration of procedure (median, minutes, Q25–Q75)*4	45 min (40–60)	–

CEA,carotid endarterectomy; CAS, carotid artery stenting; DUS, duplex ultrasonography

*¹ Platelet inhibition for CEA was documented only in the period 2010 to 2014.

*² Other techniques: carotid interposition graft; plastic reconstruction of the external carotid a.; carotid-to-carotid bypass

*³ somatosensory evoked potentials (SEP), electroencephalography (EEG), oximetry, other;

*⁴ median and interquartile distance (Q25 to Q75)

The primary endpoint (any stroke or death) was documented in the CEA group in 1.4% of the cases in IG-A and 2.5% of the cases in IG-B, and in the CAS group in 1.7% of the cases in IG-A and 3.7% of the cases in IG-B. The periprocedural death rates for IG-A and IG-B, respectively, were 0.5% and 0.8% after CEA and 0.4% and 1.5% after CAS.

Injuries of the cranial nerves (e.g., the vagus n. and the hypoglossal n.) arose in 1.2% of CEA cases and 0% of CAS cases; the rate of neck hematoma requiring reoperation or catheter-associated complications in the groin was 2.4% and 0.9%, respectively. The rate of documented myocardial infarction was 0.4% after CEA and 0.1% after CAS. Other serious complications were seen in 3.4% and 3.1% of cases respectively (table 1).

Table 1. Periprocedural mortality and periprocedural rates of cerebral, local, and general complications after elective surgical (CEA, 2009–2014) and endovascular (CAS, 2012–2014) treatment of asymptomatic and symptomatic carotid stenosis in Germany.

	Elective/Asymptomatic		Elective/Symptomatic
	n	%	n	%
Surgical treatment (CEA)	85 738	100%	56 336	100%
Cerebral complications and deaths
any stroke or death	1175	1.4%	1436	2.5%
any severe stroke (mRS >2) or death	769	0.9%	1057	1.9%
any stroke	783	0.9%	983	1.7%
any severe stroke (mRS >2)	377	0.4%	604	1.1%
death of any cause	392	0.5%	453	0.8%
Local complications
postoperative bleeding necessitating reoperation	1990	2.3%	1436	2.5%
cranial nerve lesion	1073	1.3%	652	1.2%
General complications
any*1	1925	2.2%	1941	3.4%
myocardial infarction*2	93/27 981	0.3%	67/17 975	0.4%
Endovascular treatment (CAS)	8360	100%	4726	100%
Cerebral complications and deaths
any stroke or death	144	1.7%	173	3.7%
any severe stroke (mRS >2) or death	83	1.0%	117	2.5%
any stroke	111	1.3%	129	2.7%
any severe stroke (mRS >2)	50	0.6%	73	1.5%
death of any cause	33	0.4%	44	0.9%
Local complications
Inguinal puncture site*3	66	0.8%	48	1.0%
General complications
any*1	159	1.9%	146	3.1%
myocardial infarction*2	6/5813	0.1%	3/3192	0.1%

*¹ other cardiovascular complications, deep venous thrombosis (pelvic or lower limb veins), pulmonary complications, other;

*² data available only for 2013 and 2014. *³ hemorrhage, hematoma, AV fistula, aneurysm, or other complication; CAS, carotid artery stenting; CEA, carotid endarterectomy; mRS, Modified Rankin Scale.

Associations between clinically unalterable variables and the relative risk of periprocedural stroke or death are shown separately for CEA and CAS in Figures 2a, b. The following variables were statistically significant for both CEA and CAS: increasing age (in any age group = 65 years), higher ASA class, and symptomatic rather than asymptomatic stenosis. Additional variables that were associated with a statistically significant elevation of risk for CEA alone were the presence of a 50–69% stenosis and contralateral carotid occlusion. Sex was not a risk factor for complications for either treatment modality. For all data on relative risk (RR), see Figures 2a, b.

Figure 2a.

The effect of clinically unalterable variables on the risk of periprocedural stroke or death after CEA, adjusted for platelet inhibition, pre- and postprocedural neurological examination, neuromonitoring, hospital caseload, operative technique, anesthesia technique, shunt placement, clamping time, and intraoperative completion studies. Patient clustering in hospitals was taken into account.

ASA, American Society of Anesthesiologists; CEA, carotid endarterectomy; NASCET, stenosis grading criteria of the North American

Symptomatic Carotid Stenosis Trial; Adj. RR, adjusted relative risk; CI, confidence interval

Figure 2b.

The effect of clinically unalterable variables on the risk of periprocedural stroke or death after CAS, adjusted for platelet inhibition, pre- and postprocedural neurological examination, neuromonitoring, hospital caseload, and use of a protection device.

Patient clustering in hospitals was taken into account.

ASA, American Society of Anesthesiologists; CAS, carotid artery stenting; NASCET, stenosis grading criteria of the North American

Symptomatic Carotid Stenosis Trial; Adj. RR, adjusted relative risk; CI, confidence interval

Discussion

Our collaboration with the AQUA Institute enabled us to analyze data from nearly all procedures that were carried out in Germany to treat extracranial carotid artery stenosis over the period of the study: 142 074 cases of CEA, 2009–2014, and 13 086 cases of CAS, 2012–2014. 18.7% of all elective procedures were carried out by the endovascular method (2012–2014). In an international comparison, the corresponding percentages in 2005–2010 were less than 5% in Switzerland, Hungary, Finland, Norway, Denmark, and the United Kingdom, approximately 11% in Australia and the USA, and 17.4% in Italy (12– 14).

Patient characteristics

The median age in our study was 72 years for patients undergoing CEA and 71 years for patients undergoing CAS; the percentage of patients who were male was 67.8% and 69.7% for the two types of procedure, respectively. We found in an earlier study that the median age of CEA patients has risen by about 2 years since 2003 (6). Similar age and sex distributions are seen in case registries from other countries (12, 13). Only in the USA is the percentage of female patients appreciably higher than in our study: 42.2% (CEA) and 40.9% (CAS) (14).

60.3% (CEA) and 63.9% (CAS) of cases involved an asymptomatic carotid stenosis. These figures are lower than in the USA (>80%) (14) and Italy (69%, only CEA), but much higher than in the United Kingdom (17%, CEA), Denmark (1.3%, CEA), and elsewhere (12).

In contrast to other registry studies, we found that the percentage of patients who were multimorbid (ASA classes III–V) was much higher for CEA (>70%) than for CAS (38.4%) (eTable 3) (13, 15– 22). It is unclear whether, in fact, fewer multimorbid patients were treated with CAS in Germany, or whether CEA patients in general tended to be assigned to higher ASA classes.

eTable 3. Characteristics of the patients undergoing elective surgical (CEA, 2009–2014) and endovascular (CAS, 2012–2014) treatment of carotid artery stenosis in Germany.

	CEA		CAS
	n	%	n	%
Overall (N)	142 074	100%	13 086	100%
Men	96 396	67.8%	9119	69.7%
Age (years; median, Q25–Q75)	72 (65–77)	–	71 (63–76)	–
Treated side (right)	71,379	50.2%	6539	50.0%
ASA class
classes I + II	41 751	29.4%	8069	61.7%
class III	96 638	68.0%	4773	36.5%
classes IV + V	3685	2.6%	244	1.9%
Clinical manifestations on admission*1
asymptomatic	85 738	60.3%	8360	63.9%
amaurosis fugax	9869	6.9%	798	6.1%
transient ischemic attack	20 453	14.4%	1354	10.3%
mild stroke (mRS 0–2)	14 391	10.1%	1351	10.3%
severe stroke (mRS 3–5)	8597	6.1%	779	6.0%
other manifestations	3026	2.1%	444	3.4%
interval from index event to treatment*1	9 days (5–17)		9 days (5–19)
Degree of stenosis on the treated side*2
<50%	2304	1.6%	326	2.5%
50–69%	6445	4.5%	771	5.9%
70–99%	133 325	93.8%	11 989	91.6%
Degree of stenosis on the untreated side*2
50–69%	25 510	18.0%	1419	10.8%
70–99%	18 851	13.3%	1561	11.9%
Occlusion	8187	5.8%	1243	9.5%
Annual hospital caseload*3
1st quintile	2770	1.9%	205	1.6%
2nd quintile	10 140	7.1%	787	6.0%
3rd quintile	21 162	14.9%	1273	9.7%
4th quintile	35 764	25.2%	3087	23.6%
5th quintile	72 238	50.8%	7734	59.1%
Neurological examination
before procedure	98 178	69.1%	9902	75.7%
after procedure	76 456	53.8%	8672	66.3%
Treating specialist department caring for patient
vascular surgery	76 807	54.1%	1502	11.5%
general surgery	52 273	36.8%	1031	7.9%
cardiac surgery	6594	4.6%	124	0.9%
cardiology/angiology	1057	0.7%	2586	19.8%
internal medicine	1080	0.8%	2791	21.3%
neurology	1961	1.4%	2615	20.0%
other	2302	1.6%	2437	18.6%
Hospital stay after procedure (days; median, Q25–Q75)*4	5 (4–6) days		2 (2–4) days

ASA, American Society of Anesthesiologists; CAS, carotid artery ?stenting; CEA, carotid endarterectomy; mRS, modified Rankin scale;

*¹ temporal interval; median and interquartile distance;

*² All degrees of stenosis are given according to the NASCET (North American Symptomatic Carotid Endarterecomy Trial) criteria;

*³ CEA was performed in 566 hospitals, divided in groups of 111–115 hospitals. The case?load quintiles, in cases per year, were: 1–10 (1 st quintile), 11–25 (2 nd quintile), 26–46 (3 rd qunitile), 47–79 (4 th quintile), and 80–734 (5 th quintile). CAS was performed in 366 hospitals; the caseload quintiles, in cases per year, were 1–2/3–6/7–12/13–26/ and 27–240;

*⁴ median and interquartile distance (Q25 to Q75).

Periprocedural stroke and death rates

Our study shows that the actual combined rates of periprocedural stroke and death in patients treated for asymptomatic and symptomatic stenoses—after CEA, 1.4% and 2.5%; after CAS, 1.7% and 3.7%—are well below the maximum permissible rates of 3% and 6%, respectively. Our data are based on the in-hospital stay. It should be pointed out, however, that registry studies from North America have shown that 20–40% of all strokes and deaths occurring within the first 30 days in fact take place after discharge from the hospital (cf. Limitations, below) (21, 23, 24).

The recommendations regarding CEA and CAS that were cited at the outset are based on multiple RCTs in which these two techniques were compared with each other, or with medical treatment alone, for the treatment of high-grade extracranial carotid stenosis. In particular, CEA for symptomatic high-grade stenosis has been shown to reduce the absolute risk of stroke by 10–20% in 5 years compared to medical treatment alone; the comparable figure for asymptomatic high-grade stenosis is only approximately 6% in 5 years. In view of recent improvements in the primary prevention of stroke with drugs (platelet inhibitors, statins), the indication for CEA or CAS in the treatment of asymptomatic carotid stenosis is now being critically reassessed (25).

A combined analysis of the four non-industry-sponsored multicenter RCTs comparing CEA and CAS (the SPACE, ICSS, EVA3-S, and CREST trials; Carotid Stenosis Trialist Collaboration [Table 2]only for symptomatic stenosis) (26, 27) has shown that CEA is associated with a lower combined rate of periprocedural stroke and death. Subgroup analyses showed that this is particularly the case among patients over age 70 (28). The long-term probability of a new ipsilateral stroke is low after either of the two procedures (28). CAS is associated with a lower rate of periprocedural myocardial infarction and a markedly lower rate of periprocedural cranial nerve dysfunction (26, 29).

Table 2. Complications in the first 30 days after CEA or CAS for symptomatic carotid stenosis: pooled data from the four large-scale RCTs comparing CEA and CAS (SPACE, CREST, ICSS and EVA3-S, Carotid Stenting Trialist Collaboration).

	CAS	CEA	Absolute difference
Stroke or death in the first 30 days (26, 29)	7.2%	4.1%	–3.1%
Any stroke in the first 30 days (26, 29)	6.9%	4.0%	–2.9%
Death of any cause in the first 30 days (26, 29)	0.9%	0.4%	–0.5%
Effect of the interval from the index event to the treatment*1 (31)
any stroke/death after treatment <7 days after index event	8.4%	1.3%	–7.1%
any stroke/death after treatment >7 days after index event	7.1%	3.6%	–3.5%
Effect of age (28)
any stroke/death for age <70 years	4.4%	3.6%	–0.8%
any stroke/death for age ≥ 70 years	10.2%	4.1%	–6.1%
Myocardial infarction	0.5%	1%	+0.5%
Cranial nerve lesion	0.4%	5.7%	+5.3%
Serious wound hematoma*2 (26)	0.7%	1.9%	+1.2%

*¹ adverse events occurring in the 30 days after CEA or CAS; *² based on the European RCTs (SPACE, ICSS, EVA 3-S)

CAS, carotid artery stenting; CEA, carotid endarterectomy; RCT; randomized, controlled trial

Current guidelines recommend treating symptomatic carotid stenoses within 14 days; in our study, the actual median interval from the neurological index event to the treatment was 9 days for both CEA and CAS. In an earlier study on CEA alone, we showed a reduction of this interval from 28 days to 8 days (median) over the period 2003–2012 (6). As the risk of recurrent cerebral ischemia is highest in the first 14 days (30), further measures should be taken so that symptomatic carotid stenosis can be treated even earlier than at present. In this context, we note that, as revealed by the CSTC study (table 2), the periprocedural rate of stroke and death after procedures carried out within one week of the index event was markedly higher for CAS than for CEA (31). In a further analysis of German registry data, we confirmed the low complication rate of early CEA (32).

Reasons for the higher stroke rate after CAS for symptomatic stenosis (1.2% higher than after CEA, in Germany as elsewhere) may include a higher rate of embolism during passage of the catheter through the arteriosclerotic lesion, the performance of the procedure despite morphologically unsuitable lesions (e.g., severe calcification, unstable plaques), and inadequate experience of the specialist performing the procedure.

The superiority or inferiority of CEA, CAS, or medical treatment alone for asymptomatic carotid stenosis is currently being studied in multiple ongoing trials (CREST-2, ECST-2, long-term data from SPACE-2). The treatment risk associated with CEA and CAS for this indication is lower than in symptomatic stenosis.

The complication rates of CEA that we found in this study were lower than those in the abovementioned RCTs; this may have been due to optimized perioperative management. We showed in a recent supplementary study that perioperative platelet inhibition (RR: 0.83; 95% CI: [0.71; 0.97]), local anesthesia (RR: 0.85; 95% CI: [1.22; 2.01]), and intraoperative imaging to check the adequacy of the final result (angiography: RR: 0.80; 95% CI: [0.71; 0.90], ultrasonography: RR: 0.74; 95% CI: [0.63; 0.88]) are associated with a significantly lower perioperative risk of stroke (6, 33). In comparison, the use of a protection system in CAS has been found to lower the complication rate by 35% (RR: 0.65; 95% CI: [0.43; 0.77]) (34).

Associations between clinical variables and treatment risk

Our study shows that the risk of periprocedural stroke or death increases from age 65 onward. This effect is stronger for CAS than for CEA, presumably because of the influence of age on the risk of stroke after endovascular therapy, rather than because of an increase in periprocedural mortality (35).

We found, in accordance with earlier findings from the USA, that the patient’s ASA class and the presence of symptomatic stenosis was associated with a higher risk of treatment complications after either CEA or CAS (36); this effect, too, was more marked with CAS than with CEA. We conclude that the potential indication for a revascularization procedure in a severely ill patient should be considered on an individual basis.

Local complications of CEA

In this study, the local complications of CEA included hemorrhage or hematoma requiring reoperation (2.4% of cases) and documented cranial nerve lesions (1.3% of cases). Comparable complication rates in a British study were 4.2% and 2.1%, respectively (37). In the literature, cranial nerve lesions (e.g., of the hypoglossal nerve) are described in as many as 5.6% of cases. The rate of cranial nerve injury persisting 10 months after the procedure is 0.7% in the literature (38).

The rate of documented perioperative myocardial infarction was very low for both CEA (0.3–0.4%) and CAS (0.1%). In a recent review, the 30-day rate of myocardial infarction was found to be somewhat higher than this (0.9% for CEA, 0.7% for CAS) (39). The systematic detection of myocardial ischemia (with biomarkers, ECG, etc.) is not a part of the quality assurance program from which the data that we analyzed were derived. Thus, the true rates of both clinically evident and subclinical myocardial infarction may be somewhat higher than the rates that we observed.

Limitations

This study did not include any conservatively managed patients with extracranial carotid artery stenosis, nor did it include adequate information about the differential indications for CEA or CAS (e.g., the morphology of the stenosis, the aortic arch configuration, the desires of patients and referring physicians, the profile of the treating hospital, or interdisciplinary case discussions).

In this, as in all observational studies, association cannot be taken to imply causation. The procedures were generally documented by the treating specialist himself or herself, and there was no external audit. The Federal Joint Committee delegated to the Institute for Quality Assurance and Transparency in Health Care (Institut für Qualitätssicherung und Transparenz im Gesundheitswesen, IQTIG) the tasks of checking the data for validity and analyzing the structured dialogue carried out in the individual German states in case of discrepant calculations. The last report on the structured dialogue of 2015 (concerning data obtained in 2014) yielded a total of 223 such discrepancies nationwide, 16 (7.1%) of which turned out to indicate an actual qualitative deficit (5).

The findings of our study cannot be directly compared with those of 30-day-endpoint studies, as our study dealt only with the period of inpatient treatment. Pre- and postprocedural neurological examination of the patient is recommended in the German and Austrian guideline on carotid stenosis but is not carried out in all patients (3).

Our study may also have been affected by selection bias (indications for revascularization, choice of treatment modality) and information bias (with respect to the quality of data acquisition and documentation).

Overview

This study is a secondary data analysis for the evaluation of outcome quality after the treatment of extracranial carotid stenosis by surgery (CEA) or an endovascular procedure (CAS) in Germany. The complication rates of both CEA and CAS were found to be lower than the recommended maximum permissible rates of 3% for asymptomatic and 6% for symptomatic carotid stenosis. With either technique, the risk of a complication of treatment increases with age, ASA class, and the presence of symptomatic rather than asymptomatic stenosis. Sex has no effect on risk. Ipsilateral 50–69% stenosis and contralateral carotid occlusion are associated with a higher operative risk in CEA. This study does not permit any conclusion as to the superiority or inferiority of CEA and CAS. The indications for a revascularization procedure and the choice among the available treatment modalities (conservative, CEA, CAS) should still be considered individually for each patient.

Key Messages.

• Legally mandated documentation enabled the analysis of a nearly complete dataset of all cases of carotid endarterectomy (CEA) and carotid artery stenting (CAS) in Germany (a total of 155 160 cases in the period of the study).

• The data do not allow a statistical comparison of the two techniques.

• As determined from these data, the periprocedural combined risk of stroke or death after CEA in Germany is 1.4% for asymptomatic and 2.5% for symptomatic carotid stenosis; the corresponding figures for CAS are 1.7% and 3.7%.

• Approximately 60% of the patients who underwent elective CEA or CAS were clinically asymptomatic. In view of recent improvements in medical primary prevention, the preventive effect of revascularization procedures is now being re-evaluated in controlled trials.

• Pre- and postprocedural neurological examination of all patients undergoing these procedures is recommended.