Mortality of large-artery complication (aortic aneurysm, aortic dissection, and/or large-artery stenosis) in patients with giant cell arteritis: A population-based study over 50 years

Abstract

Objective

To determine the mortality of giant cell arteritis (GCA) with large-artery complication compared with that of GCA without large-artery complication.

Methods

An inception cohort of 168 residents of Olmsted County, Minnesota, in whom GCA was diagnosed between January 1, 1950, and December 31, 1999, was followed up. Mortality in patients with incident large-artery complication (aortic aneurysm, aortic dissection, and large-artery stenosis) was determined and compared with that in patients in whom large-artery complication did not develop.

Results

No difference in survival was observed between the total group of patients with any type of large-artery complication and patients without large-artery complication or the general population. However, mortality was markedly increased in the 9 patients in whom thoracic aortic dissection developed (median survival 1.1 years [interquartile range 0.2–7.8 years]) compared with that in all other patients with GCA (P < 0.001). No difference in survival was observed between the group of patients with either aortic aneurysm and/or dissection (thoracic and/or abdominal aorta) and the group with GCA without large-artery complication. Survival of patients with GCA and large-artery stenosis was not different from that of patients with GCA without large-artery complication.

Conclusion

Thoracic aortic dissection in GCA is associated with markedly increased mortality. Overall, mortality in the whole group of patients with GCA with large-artery complication was similar to that in patients with GCA without large-artery complication.

Giant cell arteritis (GCA) is the most common vasculitis in populations with predominantly Northern European ancestry, with an annual incidence of 15–33 cases per 100,000 persons over age 50 years (1–5). GCA can affect the aorta and its primary and secondary branches and is not limited to the cranial arteries (e.g., the temporal arteries) (6–9). In fact, most patients with GCA appear to have giant cell aortitis at some time point (10–12). Aortic aneurysm (6, 13, 14), aortic dissection (7, 8, 13, 14), and large-artery stenosis of the arm or leg arteries (15–17) or cervical arteries (18) have been described in GCA. Large-artery stenosis affecting the arm or leg arteries can manifest as intermittent claudication (16), and large-artery stenosis affecting the cervical arteries can manifest as transient ischemic attack or stroke (18). Ample evidence suggests that these forms of large-artery disease are causally related to GCA.

Little information about the mortality of large-artery complication in GCA has been available from population-based studies. Such knowledge is important in order to assess the clinical significance and public health impact of large-artery complication associated with GCA. Previous studies examining the mortality of GCA have yielded conflicting results. Most studies, including those that are population-based, have suggested that the overall life expectancy of patients with GCA is normal (13, 19–22). One study even showed prolonged survival of patients with GCA 5 years following diagnosis, but not after 10 or more years (19). A recent Swedish study demonstrated increased mortality attributable to cardiovascular disease in patients with GCA (23). Nordborg and Bengtsson (24) reported an excess rate of deaths from vascular disorders within the first year after diagnosis of GCA, which they related to insufficient glucocorticoid therapy. In contrast, Nesher et al (25) reported increased mortality during the first year after diagnosis of GCA in a small referral cohort, which was believed to be attributable to complications of glucocorticoid therapy.

The association of GCA with aortic aneurysm and aortic dissection was explored by Evans et al (13) in a population-based study comparing the incidence of aortic aneurysm and dissection in 96 patients with GCA with that in the general population. The authors concluded that patients with GCA were 17.3 times more likely to develop thoracic aortic aneurysm and 2.4 times more likely to develop abdominal aortic aneurysm compared with the general population, based on established incidence data for thoracic aortic aneurysm and abdominal aortic aneurysm in the same population (26, 27). In that study, 6 of the 11 patients with thoracic aortic aneurysm died of thoracic aortic dissection. Increased mortality in patients with GCA compared with the general population could not be shown, possibly because of the limited size of their cohort. However, results of the study by Evans et al suggest that mortality in the subset of patients with large-artery complications of GCA, particularly those with aortic aneurysm and rupture, is increased compared with that in patients without disease-related large-artery complications.

This study is the first population-based assessment of the mortality of the entire spectrum of large-artery complications in GCA, including aortic aneurysm and/or dissection and large-artery stenosis, compared with that of GCA without large-artery complications.

PATIENTS AND METHODS

A retrospective cohort study was performed by patient record review for the inception cohort of residents of Olmsted County, Minnesota, in whom GCA was first diagnosed between January 1, 1950, and December 31, 1999. Only patients who were residents of Olmsted County for at least 1 year prior to the diagnosis of GCA were included. This prerequisite was important to avoid migration bias caused by a tendency for individuals who are sick to move to the vicinity of a major medical center.

Olmsted County, Minnesota, is well suited for population-based epidemiologic studies. Its population is characterized by a balanced urban and rural distribution, is primarily of Northern European ancestry, and is representative of the white population in the US. Therefore, Olmsted County is particularly ideal for studying GCA, because GCA is a disease that affects predominantly persons of Northern European ancestry (1–5). Medical care for the entire county is provided almost exclusively by the Mayo Clinic and the Olmsted Medical Group. All medical information for each resident gathered at these facilities and by 2 now retired practitioners is collected, the diagnoses are indexed, and the records are linked centrally. The potential of this data retrieval system for use in population-based studies has been described previously (28).

This inception cohort included 166 patients who fulfilled the 1990 American College of Rheumatology (ACR) classification criteria for GCA (29) and an additional 2 patients with classic large-artery complication of GCA who did not formally meet the 1990 ACR classification criteria. Preliminary studies had identified 175 patients with GCA in this cohort, but the final cohort included 7 fewer patients (research authorization was withdrawn in 4 patients, 4 patients did not fulfill the residency criterion, 1 patient did not prove to have GCA at followup, and 2 additional cases of GCA fulfilling the 1990 ACR criteria of GCA were identified). Approval for the study was obtained from both the Mayo Clinic and the Olmsted Medical Group institutional review boards. No direct patient contact was involved. Only the records of patients who had given authorization for use of their medical records under Minnesota law were reviewed.

All medical records were reviewed by one investigator (DMN). For validation of the data abstraction, 10% of the records were reviewed by a second investigator (ELM). One hundred percent interobserver agreement in the classification of patients with both GCA and the particular kind of large-artery complication was assured. Classification questions were reviewed and adjudicated with a third investigator (GGH).

Mortality among patients in the inception cohort was determined according to the group of incident large-artery complication in GCA (aortic aneurysm, aortic dissection, large-artery stenosis, or no large-artery complication). The incidence of large-artery complications in this cohort is presented elsewhere in this issue of Arthritis & Rheumatism (30). Briefly, each patient had been assessed for any evidence of large-artery complication at any time. Incident large-artery complication was defined as a complication that was detected within 1 year prior to the diagnosis of GCA or thereafter. This avoided exclusion of patients in whom the manifestation of a large-artery complication of GCA occurred close to the time of diagnosis of GCA. A patient was classified as having an aortic aneurysm only when results of imaging studies were determined by a radiologist as being consistent with aneurysm, or when clear-cut evidence was present at the time of surgery or autopsy.

Imaging modalities included plain radiography, ultrasonography (including echocardiography), computed tomography, and magnetic resonance imaging. Physical examination findings alone were not sufficient to confirm the presence of an aortic aneurysm or large-artery stenosis. Both aortic dissection and aortic rupture were categorized as aortic dissection. Large-artery stenosis was defined as stenosis of any primary branch of the aorta with a size larger than that of the radial arteries, including the innominate, subclavian, axillary, brachial, common carotid, vertebral, and basilar arteries, or as stenosis of any unbranched secondary branch of the aorta, including the unbranched internal and external carotid arteries. Lower-extremity arterial stenosis was excluded unless there was classic angiographic evidence suggestive of vasculitis. Even clinically suspected lower-extremity large-artery GCA that appeared to respond to glucocorticoid therapy was not sufficient for a classification of large-artery complication unless it was supported by angiographic findings, because arteriosclerotic peripheral vascular disease of the lower extremities is common in elderly persons.

For each group, survival was estimated using the Kaplan-Meier method and log rank tests, comparing the survival in each group with large-artery complication with that in the group without any large-artery complication. Based on preliminary estimates, the number of patients with incident GCA in the respective incidence period was 175, and the number of patients with GCA with large-artery complication was thought to likely exceed 40. Given these numbers, we reached the following conclusion regarding our hypothesis: the study had 80% power to detect a 2.5-fold increase in the 5-year mortality of the entire group of patients with GCA and any form of large-artery complication compared with that in the group of patients with GCA without large-artery complication, given an estimated baseline 5-year mortality of 15% (21) and a significance level of 0.05 (2-tailed).

To assess for possible ascertainment bias related to the introduction and more widespread use of imaging studies over the decades and to increasing disease suspicion bias, results for the inception cohort of 107 patients in whom GCA was diagnosed between January 1, 1980, and December 31, 1999, were compared with those for the inception cohort of 168 patients in whom GCA was diagnosed between January 1, 1950, and December 31, 1999.

RESULTS

Kaplan-Meier plots for survival, by large-artery complication group, among patients with GCA are shown in Figures 1 and 2. For the combined group of patients with any kind of large-artery complication (n = 46 [27% of all patients]), there was no difference in survival compared with patients without any large-artery complication (n = 122, 73%) (P = 0.8) or compared with the general population of white persons living in Minnesota (P = 0.4). For the group with aortic aneurysm and/or dissection (thoracic and abdominal aorta; n = 30, 18%), there was no difference in survival compared with the group without large-artery complication (P = 1.0). However, among patients in whom thoracic aortic dissection developed (n = 9, 5%) mortality was markedly increased compared with that among patients without large-artery complications (P < 0.001).

Figure 1. Survival, by large-artery complication group, among patients with giant cell arteritis (GCA).

Figure 2. Survival, by large-artery stenosis group, among patients with giant cell arteritis (GCA).

Survival of patients with GCA and large-artery stenosis (n = 21, 13%) was not different from that of those with GCA without large-artery complication (P = 0.6). Because of the small size of several subgroups (for subclavian/axillary/brachial artery stenosis, n = 6 [4%], for iliac/femoral artery stenosis, n = 1 [0.6%]), and for abdominal aortic dissection (n = 1 [0.6%]), survival of patients in these subgroups was not statistically assessed. Median survival of all patients, by type of large-artery complication, is shown in Table 1.

Table 1. Median survival of patients with GCA, by subgroup of large-artery complication*

Subgroup	No. of patients	Median survival after diagnosis of GCA, years	IQR	P†
* Some patients had >1 complication (see ref. 30). GCA = giant cell arteritis; IQR = interquartile range; NA = not applicable (due to limited sample size). † By log rank test, versus group without any large-artery complication.
No large-artery complication	122	10.9	5.2–18.1	NA
Any large-artery complication	46	10.6	3.1–20.1	0.84
Aortic aneurysm and/or dissection	30	9.6	2.1–20.1	0.99
Aortic aneurysm	25	14.1	5.0–20.7	0.33
Thoracic aortic aneurysm	13	12.3	5.0–20.7	0.43
Abdominal aortic aneurysm	16	14.1	4.0–20.7	0.55
Aortic dissection	10	1.6	0.2–7.8	<0.001
Thoracic aortic dissection	9	1.1	0.2–7.8	<0.001
Abdominal aortic dissection	1	7.8	NA	NA
Large-artery stenosis	21	10.6	4.9–20.4	0.55
Cervical artery stenosis	15	8.6	4.7–19.3	0.86
Subclavian/axillary/brachial artery stenosis	6	16.5	4.7–22.8	NA
Lower-extremity artery stenosis	1	20.4	NA	NA

Results were similar for both the entire inception cohort (representing 5 decades) and the inception cohort representing the most recent 2 decades (data not shown). The subgroup of patients in whom aortic dissection developed had markedly increased mortality compared with the other patients with GCA (P < 0.001) in both cohorts.

DISCUSSION

This was the first population-based study of a cohort of patients with GCA to determine the mortality of both aortic aneurysm and/or dissection and large-artery stenosis. No difference in survival was observed between the combined group of patients with any large-artery complication and either patients without large-artery complication or the general population of white persons living in Minnesota. Patients with thoracic aortic dissection were the only subgroup with significantly decreased survival compared with that of patients without large-artery complication. The latter finding was expected, because mortality in patients with thoracic aortic dissection is high, regardless of the presence of an underlying vasculitis. These results are concordant with those of most other epidemiologic studies examining the mortality of GCA (13, 19–22).

The high mortality in the group with thoracic aortic dissection is consistent with the observations of Nordborg and Bengtsson (24). Those authors reported an excess rate of deaths from vascular disorders mostly related to insufficient glucocorticoid therapy within the first year after diagnosis. The incidence of different causes of death in the current cohort was not further explored, because the cause of death was frequently insufficiently documented. Thus, no comment can be made regarding any possible increased mortality due to cardiovascular disease in patients with GCA in this cohort (23). The cause of death was most clearly documented in patients who died of aortic dissection resulting in aortic rupture.

Differences in survival between the groups at different time points were not separately analyzed, nor were sex-specific differences in survival (31, 32). However, sex was not predictive of any type of large-artery complication, as is presented elsewhere (30). This is in contrast to the findings of Graham et al (32), who reported decreased survival in GCA for women only. Bisgard et al reported decreased survival in GCA over a 10-year period after diagnosis but could not confirm a statistically significant decrease in survival when the analysis was restricted to women (31). The study by Graham et al was likely confounded by selection bias, because patients from the ophthalmology and neurology departments of a major referral hospital were assessed.

Because of the limited sample size, the study was underpowered to detect a <2.5-fold difference in mortality between the group with any kind of large-artery complication and the group without large-artery complication that may still be clinically important. The study had even less power to detect any difference in survival between subgroups of large-artery complications.

Even though ascertainment bias related to the introduction and more widespread use of imaging studies over the decades and to increasing disease suspicion bias was of concern, such biases did not appear to influence overall results and conclusions. The overall mortality and the mortality of subgroups of large-artery complication were similar for both the entire 5-decade inception cohort and the inception cohort representing the most recent 2 decades (data not shown). During the most recent 2 decades, ascertainment of both GCA and large-artery complication in GCA had reached a plateau, as is presented elsewhere (30). Although stroke is a known complication of cervical artery stenosis due to GCA, decreased survival in the group of patients with GCA and cervical artery stenosis was not demonstrated in this study.

Thoracic aortic dissection in GCA is associated with markedly increased mortality. However, the mortality of GCA with large-artery complication was similar to that of GCA without large-artery complication.