Increased aortic arch calcification in patients older than 75 years: Implications for carotid artery stenting in elderly patients

Methods 

We reviewed the CT scans of consecutive patients who underwent thoracic CT scans between January and June 2006 for any clinically appropriate indication, including assessment of aortic dissection, study for pulmonary embolus, and assessment for a pulmonary malignancy. Study approval was obtained from the Institutional Human Investigation Committee. Patients younger than 40 years of age and patients who had recent thoracic aortic or cardiac valve surgery were excluded. A 64-slice CT scanner was used for this study (model VCT; General Electric Healthcare, Tampa, Fla). Non–contrast-enhanced, 1.0-mm-thick slices were selected to study the aortic arch calcium content; every fifth slice was reviewed to avoid overlap from large calcium deposits. The aortic arch was measured in its entirety, from the aortic root to the descending thoracic aorta at the level of the carina, in all patients. Calcium content was quantified by using the Agatston coronary artery calcium score grade (Vitrea 2.0 software; Vital Imaging, Minneapolis, Minn). The threshold for a calcified lesion is based on a CT density of 130 Hounsfield units over an area of 1 mm2 or larger.14

The aortic arch classification was determined by using two-dimensional and multiplanar image reconstructions. As a measure of arch length, aortic arch types were defined as the distance of the origins of the innominate to the left subclavian artery. A distance of less than 2 cm was considered a type I arch, a distance between 2 and 4 cm was considered a type II arch, and any aortic arch with greater than a 4-cm distance was considered a type III aortic arch.

Patient comorbidities were determined by review of the patients’ medical records. Chronic renal insufficiency was defined as a serum creatinine level of 1.6 mg/dL or higher. Other comorbidities recorded were coronary artery disease, chronic obstructive pulmonary disease, diabetes, hypercholesterolemia, and hypertension.

Simple regression and analysis of variance were used to determine the effect of patient age and arch classification on aortic arch calcium content. Parametric statistics were used because the calcium content grade is linear.14 P values were calculated by using χ2 or Fisher exact tests, as appropriate. Logistic regression was used for multivariable analysis. All tests were two tailed, and P values ≤.05 were considered statistically significant (StatView 5.0; SAS Institute, Cary, NC). Post hoc power analysis was performed with nQuery Advisor 4.0 (Statistical Solutions, Saugus, Mass).

Results 

The CT scans of 94 patients were analyzed, of which 53 (56%) patients were male and 41 (44%) were female (Table I). The mean age was 64 ± 1 years; 11 patients (12%) were 80 years of age or older (Table I). Using the Agatston coronary calcium score, we measured the aortic arch calcium content in these consecutive patients; representative pictures are shown in Fig 1. Aortic arch calcium content increased with patient age, with a positive correlation between these variables (r2 = 0.34; P < .0001; Fig 2, A). The mean calcium score increased by age decile (P < .001; Fig 2, B). Because very high levels of calcium were detectable only in the aortic arches of patients more than 75 years old (Fig 2, A), we compared the aortic arch calcium content in patients less than 75 years of age with that in patients 75 years of age or older. Significantly more aortic arch calcium was present in patients 75 years of age or older compared with patients less than 75 years of age (2458 ± 447 vs 145 ± 49; P < .001; Fig 2, C).

Table I. Patient demographics and comorbidities (n = 94 total)

	Factor	%
	Age (y)
	≤50	19%
	50-59	22%
	60-59	18%
	70-79	29%
	≥80	12%
	Male	56%
	Hypertension	71%
	Diabetes	28%
	Hypercholesterolemia	39%
	Coronary artery disease	29%
	Chronic obstructive pulmonary disease	10%
	Chronic renal insufficiency	4%

View Large Image Download to PowerPoint

Fig 1. Quantification of aortic arch calcium content. Vitrea software (version 2.0) was used to apply the Agatston coronary calcium score. A, The aortic arch of a 54-year-old patient without significant arch calcium. B, A representative aortic arch of a 77-year-old patient demonstrates diffuse aortic arch calcium content (arrows).

View Large Image Download to PowerPoint

Fig 2. Increased aortic arch calcium content correlates with increasing patient age. A, Positive correlation between age and aortic arch calcium content by linear regression (r2 = 0.34; P < .0001). B, The mean calcium score for patients increases by decade in a highly significant fashion (P < .001). *Significant difference compared with patients younger than 50 years of age (P = .03 [70-79 years] or P < .001 [≥80 years]; Scheffé post hoc test). C, Increased arch calcification in patients 75 years or older compared with those less than 75 years old (*P < .001).

Because elderly patients may develop elongated and tortuous arches, we examined whether arch type was also associated with calcium content. We identified 80 patients with type I arch morphology, 14 patients with type II aortic arches, and no patients with type III arches. Patients with type II aortic arches had a higher calcium content (2028 ± 546; n = 14) compared with patients with type I aortic arches (712 ± 191, n = 80; P = .01; Fig 3). As expected, a significant proportion of aortic arch elongation was present in patients 75 years or older compared with younger patients (35.5% vs 4.7%; P = .0002).

View Large Image Download to PowerPoint

Fig 3. Aortic arch elongation is associated with a higher calcium content compared with nonelongated arches. *Significant difference (P = .01; unpaired t test).

Post hoc power analysis showed that this study had more than 99% power to detect the differences in arch calcification between patients younger than 75 (n = 31) and patients 75 or more (n = 63) years old. In addition, the study had more than 99% power to detect differences by using logarithmically transformed data.

Because both patient age and arch morphology are associated with aortic arch calcium content, we used multivariable logistic regression to determine the effect of these and other patient risk factors on aortic arch calcification. Patient age was the only variable that was significantly associated with aortic arch calcium content (odds ratio, 82; P = .004; Table II). There was no statistically significant effect of other patient comorbid conditions on aortic arch calcium content, including diabetes and renal failure (Table II).

Table II. Effect of patient age, arch type, and associated comorbidities on aortic arch calcification

	Independent variable	P value	OR	95% CI
	Age ≥75 y	.004*	82	5-772
	Chronic renal insufficiency	.98	<0.01	0
	Coronary artery disease	.98	0.98	0.08-12
	Chronic obstructive pulmonary disease	.91	0.78	0.01-43
	Diabetes	.27	0.20	0.01-3.4
	Hypercholesterolemia	.65	0.56	0.04-7
	Hypertension	.49	2.4	0.19-30
	Aortic arch type II	.75	0.60	0.03-13

OR, Odds ratio; CI, confidence interval.

Discussion 

We used an established, noninvasive quantitative coronary calcium scale to determine the amount of calcium in the segments of the aortic arch that are traversed and potentially manipulated during CAS.14 We showed that patients 75 years of age or older have significantly more aortic arch calcification compared with younger patients. In addition, type II arch morphology is associated with increased age and calcification. These findings suggest that increased aortic arch calcium content and arch elongation may be used as markers of increased potential for embolization during endovascular manipulation of the aortic arch.

Our results are consistent with a recent report examining 18 patients older than 80 years undergoing CAS; the authors reported a high prevalence of anatomic risk factors in this elderly patient population.15 Although no major differences in outcomes were reported in this single-institution experience, aortic arch calcification (P = .045), but not arch elongation, was reported to be more prevalent in elderly compared with younger patients.15

Multiple reports have documented an increased periprocedural stroke risk with CAS in octogenarians. An increased stroke risk in elderly patients is likely to prevent widespread applicability of CAS in these patients. A single-center report of the early experience with CAS reported an overall death or major stroke rate of 1.6% in 182 patients; however, the rate of neurologic complications in octogenarians was 25%, compared with 8.6% in patients younger than 75 years (P = .042).5 These early results were confirmed in a more recent direct comparison of patients 75 years of age or older undergoing either CAS (n = 53) or carotid endarterectomy (n = 110); the 30-day stroke rate was significantly higher in the group undergoing CAS (four minor and two major strokes; 11.3%) compared with the group undergoing carotid endarterectomy (no minor and two major strokes; 1.8%; P < .05).3

The recent report of the results from the German CAS registry confirmed a persistently increased periprocedural risk with CAS in octogenarians compared with younger patients.1 CAS was aborted more frequently (6.9% vs 2.2%; P < .001), and the procedural time was longer (median time, 45 vs 40 min; P < .008) in octogenarians compared with younger patients. The in-hospital death or stroke rate was higher in octogenarians compared with younger patients (5.5% vs 3.2%; P = .032). It is interesting to note that the death or stroke rate was increased in patients in their 70s (4.2%) compared with patients in their 50s (1.4%) or 60s (2.5%).1 These increased stroke rates in septuagenarians are consistent with our finding that calcium content is increased in these patients (Fig 2, B).

The interim results from the lead-in phase of the US Multicenter Carotid Revascularization Endarterectomy vs Stent Trial (CREST) pose a strong caveat against undertaking CAS in octogenarians. In this report of 749 patients, octogenarians had a 12.1% rate of periprocedural adverse events, an 8.1-fold increased risk compared with younger patients (P < .0006).2 It is interesting to note that the periprocedural stroke risk jumped even for patients 70 to 79 years old, with 5.3% events in patients aged 70 to 79 years, a 3.3-fold increased risk compared with younger patients. These results are consistent with the previously described single-center German registry,1 as well as our observation of a substantial increase in aortic arch calcium content starting at age 75 (Fig 2, A). Furthermore, in the CREST trial, age, but not internal carotid artery tortuosity or symptomatic status, was the only factor associated with increased periprocedural stroke risk after CAS.2

One reason for the high complication rate observed in the lead-in phase of the CREST trial and single-center reports could be the steep learning curve needed for CAS. Some groups with high-volume CAS experience have suggested that with increasing technical expertise, the stroke risk in octogenarians can be overcome.16, 17 For example, over a 5-year period, 1222 CAS procedures in 1053 patients were performed at a single high-volume center in Italy.16 No difference (P = .40) was noted in the 30-day periprocedural stroke risk in those less than 80 years of age (n = 1078; 2 fatal strokes and 1 minor stroke; 1.11%) compared with those 80 years or older (n = 144; 3 deaths, 3 major strokes, and 6 minor strokes; 2.12%). As opposed to the lead-in phase CREST trial report, a higher frequency of vessel tortuosity was noted in the older group.2 Moreover, these investigators found increasing aortic arch calcification in the older group. Such advocacy for CAS in octogenarians acknowledges the potential risk from increased aortic arch calcification, but the authors suggest that this technical difficulty can be managed by highly skilled operators.16

This study is limited by its retrospective nature and needs confirmation with a prospective correlation of arch calcium content with outcome after CAS. However, by examining unselected, consecutive patients presenting for a wide variety of diagnoses, we believe that this study may be generally applicable to elderly patients. This study is also limited by the resolution of the CT scanner and may have a different outcome as technology continues to improve; in particular, the ability to discriminate between finer amounts of calcium may be important or change the age threshold associated with early calcium deposition. However, the significance of small amounts of calcium in the arch is not clear. An advantage of using the Agatston calcium scoring system is its availability in Vitrea software workstations, its ease of use, and its reproducibility14; however, as additional calcium scoring systems are developed, other scales may reveal subtle differences in calcification that may have different prognostic significance. In addition, because only 4% of our cohort had chronic renal insufficiency, it is likely that more patients with this diagnosis are needed to confirm the influence of chronic renal disease on arch calcification. Finally, our definition of arch classification may not be sensitive to minor amounts of tortuosity and needs confirmation with larger series.

It is still generally believed that aortic arch calcification continues to be a relative contraindication to CAS.18 It is possible that patients whose arches are laden with large amounts of calcium have a significant risk for emboli. Early identification of these arches at higher risk for embolic complications may allow planning of alternative techniques associated with less manipulation of the arch, such as the use of soft-tip guide catheters rather than stiff sheaths or use of the transcervical approach. These maneuvers could potentially decrease the periprocedural risk of stroke. It is not clear whether performance of CAS in high-volume centers by highly skilled operators is of increased benefit in comparison to performance in low-volume centers, as has been demonstrated for carotid endarterectomy. Aortic arch calcification represents a patient-specific factor, in addition to operator experience and carotid plaque morphology, that may prove to have an independent influence on outcomes after CAS. Preprocedural determination of aortic arch morphology and calcification may help to determine which elderly patients are at increased risk for stroke during CAS and help to further stratify periprocedural risk in elderly patients.

Author contributions