Development of aortic aneurysm/dilatation during the followup of patients with giant cell arteritis: A cross-sectional screening of fifty-four prospectively followed patients†
Presented in part at the 69th Annual Scientific Meeting of the American College of Rheumatology, San Diego, CA, November 2005, the 13th International Vasculitis and ANCA Workshop, Cancun, Mexico, April 2007, and the Annual European Congress of Rheumatology of the European League Against Rheumatism, Barcelona, Spain, June 2007.
Abstract
Objective
Giant cell arteritis (GCA) may involve the aorta. Retrospective studies have demonstrated a higher prevalence of aortic aneurysm among patients with GCA compared with the general population. We investigated the prevalence of aortic aneurysm in a cohort of patients with biopsy-proven GCA using a defined protocol and assessed whether persisting low-grade disease activity is associated with higher risk of developing aortic aneurysm.
Methods
Fifty-four patients with GCA (14 men and 40 women) were cross-sectionally evaluated after a median followup of 5.4 years (range 4.0–10.5 years). The screening protocol included a chest radiograph, abdominal ultrasonography scan, and computed tomography scan when aortic aneurysm was suspected or changes with respect to the baseline chest radiograph were observed. Clinical and laboratory data, corticosteroid requirements, and relapses were prospectively recorded.
Results
Twelve patients (22.2%) had significant aortic structural damage (aneurysm/dilatation), 5 of them candidates for surgical repair. Aortic aneurysm/dilatation was more frequent among men (50%) than women (12.5%; relative risk 3.5, 95% confidence interval 1.53–8.01, P = 0.007). At the time of screening, patients with aneurysm/dilatation had lower serum acute-phase reactants, lower relapse rate, and needed shorter periods to withdraw prednisone than patients without aortic structural damage.
Conclusion
There is a substantial risk of developing aortic aneurysm/dilatation among patients with GCA. Our data do not support that aneurysm formation mainly results from persistent detectable disease activity. Additional factors including characteristics of the initial injury or the target tissue may also determine susceptibility to aortic aneurysm/dilatation.
INTRODUCTION
Giant cell arteritis (GCA) is a granulomatous vasculitis affecting large and medium-sized vessels. The most common vascular symptoms of the disease (headache, jaw claudication, scalp tenderness) derive from inflammatory involvement of the craniofacial arteries, but other vascular territories may also be affected (1, 2).
Aortic inflammation in patients with GCA was first described in the late 1930s/early 1940s (3) and sporadically reported thereafter (4, 5). The prevalence of aortitis in GCA is unknown but appears to be remarkable. Systematic necropsy studies performed by Ostberg in 1972 disclosed aortic inflammation in 12 (92%) of 13 patients with GCA (4). Due to the lack of appropriate imaging techniques able to detect aortic inflammation in living individuals, the clinical relevance of aortic involvement has been neglected for years. Currently, computerized tomography or magnetic resonance imaging can detect thickening, increased mural contrast enhancement, and, possibly, edema in the aortic wall (6-8), but available data are still limited. 18F-fluorodeoxyglucose (FDG) uptake measured by positron emission tomography (PET) scan is emerging as a useful method to assess inflammatory activity in large vessels. In recent studies, increased aortic FDG uptake has been detected in approximately 50–60% of untreated patients, decreasing after 3–6 months of corticosteroid treatment (9-11).
Aortic inflammation appears to be frequent in GCA but remains asymptomatic unless structural damage leads to aneurysm, dissection, or aortic valve dysfunction. All of these events may have relevant clinical consequences and increase mortality in patients with GCA (5, 12). They may appear early in the course of the disease or, more frequently, as delayed complications.
The prevalence of aortic structural damage related to GCA is unknown given that the occurrence of aortic complications has only been evaluated in retrospective, chart-review studies encompassing long periods. Reported prevalences range from 9.5% to 18% (13-15). These studies include patients diagnosed over very extended periods (20–50 years) including times when awareness of GCA was lower, treatment delay was longer, recommended corticosteroid doses were lower, duration of corticosteroid regimens were more brief, and life expectancy was much shorter (3, 4). These factors may all influence both the intensity of aortic inflammation and the detection of clinically apparent complications.
Corticosteroid treatment usually elicits satisfactory relief of symptoms as well as normalization of acute-phase reactants in patients with GCA. However, when corticosteroids are tapered, relapses are frequent and persistent mild to moderate elevation of inflammatory markers can be observed in a substantial proportion of patients in clinical remission, suggesting subclinical activity (16-18). Corticosteroid tapering and withdrawal are currently guided by assessment of clinical activity and acute-phase reactants, mainly erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Based on the reported finding of inflammatory lesions in surgical or necropsy specimens from patients with aneurysm or dissection, there is some concern regarding whether or not persistent subclinical activity may eventually lead to the development of these complications (4, 19, 20). To date, it is not known whether or not persistent disease activity or persistent elevation of inflammatory markers is associated with a higher risk of developing delayed complications such as aortic aneurysm or dissection.
The goal of our study was 1) to investigate the prevalence and distribution of aortic aneurysm/dilatation detected with a defined screening protocol in a series of 54 patients with biopsy-proven GCA who were prospectively evaluated and treated and 2) to investigate factors associated with the development of this complication, particularly whether persistent subclinical inflammatory activity or a smoldering/relapsing course is associated with higher incidence of aneurysm formation.
PATIENTS AND METHODS
Patient selection.
Between September 1995 and July 2001, 125 patients were diagnosed with biopsy-proven GCA at our department (Internal Medicine, Hospital Clínic, Barcelona, Spain). Seven were subsequently treated and followed at other departments/institutions, 16 died during followup, 5 were transferred to nursing homes for advanced dementia, and 38 were lost or had incomplete followup for a variety of reasons, including moving to other regions or not returning for periodic visits by study physicians after successful corticosteroid withdrawal. During the planned study period (2000–2005), 59 patients had already completed or would complete a prospective followup of at least 4 years and were considered eligible for aneurysm screening. This period was arbitrarily selected on the basis that aneurysm is considered to be a delayed complication. Five of the 59 patients declined participation due to advanced age or comorbidities and the remaining 54 agreed to participate and were included.
All patients were prospectively treated and followed by the investigators according to a defined protocol. All patients received an initial prednisone dosage of 1 mg/kg/day (up to 60 mg/day) for 1 month. Subsequently, prednisone was tapered 10 mg/week. Reduction below 20 mg/day was slower and individualized. A further reduction to a maintenance dosage of 10 mg/day was attempted over a 2-month period. If tolerated, reduction to 7.5 mg/day was attempted after 3 months and maintained for 3 additional months. A maintenance dosage of 5 mg/day was attempted for 6 months. If patients were asymptomatic with normal acute-phase proteins and ESR <40 mm/hour, tapering at an approximate rate of 1 mg per 3 months was attempted until discontinuation. If patients responded well but elevation of acute-phase proteins persisted, the maintenance dosage of 5 mg/day was maintained for 1 year before attempting withdrawal. If the ESR increased to >40 mm/hour, the corticosteroid dose was held for 2 months and if no clinical symptoms appeared, tapering was attempted again. Relapse was defined as reappearance of disease-related symptoms. Persistent malaise and anemia with elevation of acute-phase reactants were also considered relapses if they were not attributable to other causes after detailed evaluation and if they resolved after increasing steroids. When a relapse occurred, prednisone dosage was increased by 10 mg/day above the previous effective dose. Clinical findings and laboratory values at the time of diagnosis were prospectively recorded. These included ESR, CRP level, haptoglobin, α2-globulin, blood cell counts, and liver function tests by usual automatized systems.
Screening protocol.
Patients were screened once between 2000 and 2005 at their regular followup visits. Patients underwent a medical interview, complete physical examination, routine blood tests, and detection of serum concentration of proinflammatory cytokines (interleukin-6 [IL-6], tumor necrosis factor α, and IL-18). These were determined by immunoassay (R&D Systems, Minneapolis, MN) according to the instructions of the manufacturer. Chest radiography was performed in all patients and carefully compared with that performed at the time of diagnosis. When aortic dilatation or changes with respect to the baseline radiograph were suspected, a contrast-enhanced spiral chest computed tomography (CT) scan was performed. The diameter of the aorta was measured at 3 different levels (ascending aorta, aortic arch, and descending aorta). Significant aortic structural damage was considered when an aortic aneurysm was found (defined as focal dilatation of the aortic wall) or when the aortic wall was diffusely dilated with a diameter >4 cm in the ascending aorta or at least 4 cm in the aortic arch and descending aorta. The aortic diameter at the same levels was measured in 28 consecutively selected age- and sex-matched individuals who underwent a chest CT scan for melanoma or gastric cancer as routine followup. The abdominal aorta was evaluated by ultrasonography. Prednisone requirements and relapse rate were prospectively recorded in all patients.
FDG uptake assessment by PET scan.
To assess whether aneurysm development could be related to detectable subclinical inflammation, FDG uptake was evaluated by PET scan in 11 patients with GCA (7 with and 4 without aortic aneurysm or dilatation, all confirmed by CT scan), in 4 age- and sex-matched controls randomly selected among patients undergoing evaluation for cancer staging, and in 3 patients with noninflammatory thoracic aortic aneurysm who were scheduled for surgery. In these latter patients, the noninflammatory nature of the aortic aneurysm was confirmed by histopathologic examination after surgical repair. Two of the patients had aortic aneurysm secondary to myxoid degeneration of the aortic valve and 1 had severe atherosclerosis. Funding limitations precluded extension of PET scan study to the entire series.
After a fasting period of 6 hours and after verifying a blood glucose concentration <120 mg/ml, 370 MBq of FDG was injected intravenously and PET/CT was performed with a Biograph (Siemens Medical Solutions, Enlargen, Germany). Whole-body images from the base of the skull to mid-femur were acquired 50 minutes after the radiotracer injection. CT parameters were 50 mA, 130 kV, and 8-mm sections. Iterative reconstruction was performed and attenuation correction was based on CT. Attenuation-corrected and nonattenuation-corrected images were evaluated by 2 independent investigators. The maximum standard uptake value (SUV) and the median SUV were obtained from a zone of interest drawn on sagittal slices over the thoracic aorta. The study was approved by our local ethics committee and all patients gave informed consent.
Immunohistochemistry and gelatin zymography.
Serial 4–6-μm cryostat sections from a surgically removed aortic segment from a patient with GCA were air dried and fixed with cold acetone. Sections were incubated with a polyclonal rabbit anti-human matrix metalloprotease 2 (MMP-2; Chemicon, Temecula, CA) at 1:500 dilution or a mouse monoclonal anti-human MMP-9 (clone GE-213; Chemicon) at 1:1,000 dilution. Immunoglobulins obtained from the same species as the primary antibodies were used as negative controls at the same concentrations. Immunodetection was carried out with an HRP-labeled polymer conjugated to secondary antibodies (EnVision kit from Dako, Carpinteria, CA).
Elastic fibers were stained with 1% Shikata's orcein (Scharlau Chemie, Barcelona, Spain) in 70% ethanol. Gelatin zymography of tissue extracts from a normal temporal artery, a temporal artery with active GCA lesions, and a surgically excised GCA-related thoracic aortic aneurysm was performed as described (21).
Statistical analysis.
Mann-Whitney U test and Student's t-test, when applicable, were applied to quantitative data. Kruskal-Wallis test was used for multiple comparisons. Fisher's exact test was used for contingency tables. The time required to achieve a stable maintenance prednisone dosage <10 mg/day and the time until definitive corticosteroid withdrawal were analyzed by the Kaplan-Meier survival analysis and compared by the log rank test.
RESULTS
Prevalence and characteristics of aortic structural damage in patients with GCA.
Changes in the screening chest radiograph led to the performance of a chest CT scan in 28 (52%) patients. Significant structural abnormalities in the thoracic aorta were confirmed in 11 patients. In the remaining 17, suspected changes observed in the radiograph were positional or due to aortic elongation or hiatal hernia. No thoracic aortic aneurysm was found among controls and only 2 had an ascending aorta diameter >4 cm. Aortic diameters among individuals considered not to have aortic dilatation tended to be higher in patients with GCA at the level of the descending aorta when compared with controls (median 2.5 cm, range 2.1–3.6 versus median 2.3 cm, range 2–2.8; P = 0.018). No significant differences were found in the other segments. This finding indicates that a low degree of structural damage leading to slight diffuse dilatation is common in patients with GCA.
Ultrasonography revealed abdominal aortic aneurysm in only 1 patient. Overall, 12 (22.2%) patients developed significant structural aortic damage (aneurysm or dilatation) during a median followup of 5.4 years (range 4–10.5 years).
A brief description of the abnormalities detected is shown in Table 1. In 5 patients surgery was recommended because of the size of the aneurysm or resulting aortic valve insufficiency. Two of these patients underwent successful surgical repair of the aneurysm. One patient refused intervention. In the remaining 2 patients, surgery was eventually declined because of advanced age and comorbidities. Incidentally, the screening protocol led to the discovery of a thoracic hydatid cyst in 1 patient, lung cancer in 1 patient, hypernephroma in 1 patient, and ovarian mucinous cystadenoma in 1 patient.
| Patient | Aortic characteristics |
|---|---|
| 1 | Diffuse dilatation of aortic root and ascending aorta, maximum diameter of 5.7 cm. Moderate aortic insufficiency secondary to dilatation. Surgical repair declined because of age and concomitant diseases. |
| 2 | Diffuse dilatation of thoracic aorta with maximum diameter of 6 cm in ascending aorta. The aneurysm was surgically repaired. The histology showed moderate inflammation in adventitia and scattered inflammatory foci in the media layer. |
| 3 | Aneurysm of ascending aorta and aortic root with maximum diameter of 7.3 cm and severe aortic insufficiency. Surgical repair refused because of age and concomitant diseases. |
| 4 | Dilatation of ascending aorta with maximum diameter of 5 cm and important dilatation of aortic arch. Moderate aortic insufficiency. The patient refused surgical repair. |
| 5 | Aneurysm of ascending aorta with maximum diameter of 5.8 cm. Moderate aortic insufficiency secondary to dilatation. The aneurysm was surgically repaired. The histology showed moderate atherosclerosis with moderate chronic inflammation in the intima and adventitia. |
| 6 | Aneurysm of ascending aorta with maximum diameter of 5 cm. |
| 7 | Aneurysm of ascending aorta with maximum diameter of 4.8 cm. |
| 8 | Aneurysm of abdominal aorta (5.1 × 3.1 × 2.9 cm). |
| 9 | Dilatation of the ascending aorta (4.5 cm). |
| 10 | Dilatation of the ascending aorta (4.2 cm). |
| 11 | Dilatation of the aortic arch (4 cm). |
| 12 | Dilatation of the aortic arch and the descending aorta (4 cm). |
Clinical findings associated with the development of significant aortic structural damage.
No significant differences in age, duration of followup, or initial clinical manifestations were found between patients with and without aortic structural damage. The prevalence of traditional cardiovascular risk factors did not differ among patients with or without aortic structural damage except for hypercholesterolemia, which, surprisingly, was more frequent among patients who did not develop aortic structural damage (relative risk [RR] 0.29, 95% confidence interval [95% CI] 0.081–1.062, P = 0.021). In our series, significant aortic structural damage was detected in 50% of men but only 12.5% of women (RR 3.5, 95% CI 1.529–8.014, P = 0.007) (Table 2). Interestingly, patients who later developed aortic aneurysm/dilatation tended to have lower concentrations of acute-phase reactants at the time of diagnosis compared with patients who did not develop significant aortic damage (Table 3). When the overall intensity of the acute-phase response was evaluated combining clinical and analytical abnormalities as reported (22), aneurysm/dilatation was significantly more frequent among patients with a weak systemic inflammatory reaction (RR 1.7, 95% CI 1.166–2.626, P = 0.046) (Figure 1A). This was unexpected given that patients with strong acute-phase response usually have more resistant disease (22).
| Altered aorta (n = 12) | Normal aorta (n = 42) | P | |
|---|---|---|---|
| Sex, male/female | 7/5 | 7/35 | 0.007 |
| Age, median (range) years | 76 (70–89) | 79 (63–91) | NS |
| Followup, median (range) years | 5.4 (4–8.5) | 5.5 (4–10.5) | NS |
| Duration of symptoms, median (range) weeks | 10 (2–52) | 16 (1–104) | NS |
| Cranial symptoms | |||
| Headache | 75 | 83 | NS |
| Jaw claudication | 25 | 50 | NS |
| Scalp tenderness | 25 | 57 | NS |
| Ischemic events | 0 | 19 | NS |
| Systemic symptoms | |||
| Polymyalgia rheumatica | 42 | 52 | NS |
| Fever | 25 | 43 | NS |
| Weight loss | 58 | 52 | NS |
| Vascular risk factors | |||
| Smoking | 8 | 5 | NS |
| Hypertension | 75 | 79 | NS |
| Diabetes | 17 | 12 | NS |
| Hypercholesterolemia | 17 | 57 | 0.021 |
- * Values are the percentage unless otherwise indicated. Vascular risk factors have been determined at baseline or during proper followup. NS = not significant.
| Altered aorta | Normal aorta | P | |
|---|---|---|---|
| At diagnosis | |||
| ESR, mm/hour | 87 ± 24 | 95 ± 28 | NS |
| CRP, mg/dl | 6.5 ± 5.7 | 10 ± 9.3 | NS |
| Haptoglobin, gm/liter | 3.2 ± 1.27 | 4 ± 1.73 | NS |
| Hemoglobin, gm/dl | 117 ± 18 | 110 ± 15 | NS |
| Alkaline phosphatase, units/liter | 227 ± 80 | 283 ± 212 | NS |
| GGT, units/liter | 66 ± 107 | 48 ± 51 | NS |
| Proteins, gm/liter | 68 ± 8 | 68 ± 7 | NS |
| Albumin, gm/liter | 34 ± 5 | 35 ± 5 | NS |
| α2-globulin, gm/liter | 8.4 ± 2.5 | 10.2 ± 3.6 | NS |
| Platelet count, × 109/liter | 292 ± 52 | 341 ± 110 | NS |
| At screening | |||
| ESR, mm/hour | 18 ± 9 | 34 ± 14 | 0.001 |
| CRP, mg/dl | 0.9 ± 1.4 | 1.1 ± 0.8 | NS |
| Haptoglobin, gm/liter | 1.38 ± 0.62 | 1.73 ± 0.51 | NS |
| Hemoglobin, gm/dl | 141 ± 17 | 128 ± 12 | 0.005 |
| IL-6, pg/ml | 13 ± 8 | 24 ± 39 | NS |
| TNFα, pg/ml | 30 ± 19 | 32 ± 13 | NS |
| IL-18, pg/ml | 290 ± 132 | 288 ± 159 | NS |
- * Values are the mean ± SD unless otherwise indicated. ESR = erythrocyte sedimentation rate; NS = not significant; CRP = C-reactive protein; GGT = gamma glutamyl transpeptidase; IL-6 = interleukin-6; TNFα = tumor necrosis factor α; IL-18 = interleukin-18.
Systemic inflammatory response and clinical outcome in patients with giant cell arteritis with or without aortic structural damage. A, The proportion of patients with aortic structural damage was higher among patients with weak systemic inflammatory response (SIR; relative risk 1.7, 95% confidence interval 1.17–2.63, P = 0.046). Weak SIR was defined as the presence of ≤2 of the following: erythrocyte sedimentation rate ≥85 mm/hour, hemoglobin <110 gm/liter, fever >37°C, and weight loss >3 kg. Strong systemic inflammatory response was defined as the presence of 3–4 of the above items (20). B, Aortic structural damage did not preferentially occur in relapsing individuals. In contrast, aortic structural abnormalities tended to be more frequent among patients in sustained remission. C, Percentage of patients requiring ≥10 mg of daily prednisone over time. D, Percentage of patients requiring prednisone treatment over time.
At the time of screening, all patients were in clinical remission. Twenty-seven were in stable remission without therapy and 27 still required low doses of corticosteroids (median dosage 3.75 mg/day, range 1.25–12.5). No significant differences in clinical outcome during followup were observed between patients with and without aortic abnormalities. Contrary to what was expected, patients with aneurysm/dilatation did not show a smoldering or relapsing course that might indicate stronger persisting inflammatory activity. In fact, as shown in Figure 1B, aneurysm/dilatation tended to be more frequent among patients who did not have recurrences compared with those who had a relapsing course (RR 2.9, 95% CI 1.214–7.965, P = 0.05). No significant differences in cumulated prednisone dosages during the first year (mean ± SD 6.3 ± 1.3 gm versus 6.2 ± 1.8; P = 0.86) or in the time required to reach a maintenance daily prednisone dosage <10 mg were observed between patients with or without significant aortic structural damage. However, patients with aortic structural damage needed shorter periods to withdraw prednisone therapy than patients without aortic structural damage (Figures 1C and 1D). At the time of screening, no significant differences in proinflammatory cytokine concentrations were observed between patients with or without aortic structural damage. Nevertheless, patients with aortic structural damage had significantly lower ESR (P = 0.001) and higher concentrations of hemoglobin (P = 0.005) than patients without significant aortic structural damage (Table 3). Although these data should be confirmed in larger series, they suggest that persistent subclinical inflammatory activity is not the major determinant of aneurysm formation in patients with GCA and that other factors may be involved.
FDG positron emission tomography.
None of the patients examined showed remarkable FDG uptake by aortic tissue, as has been reported in active disease (9-11). Accurate measurement of maximal and median SUV did not show significant differences between patients with GCA-related aneurysm and patients with GCA with no aneurysm, controls, or patients with noninflammatory aneurysm (Figure 2). The intensity of uptake was much lower than that reported in active patients and similar to that found in patients in remission or patients with atherosclerotic lesions. Although the number of patients examined was small, these findings suggest that, in appropriately treated patients, the development of aneurysm is not mainly related to major differences in persistent, detectable, local inflammatory activity.
18F-fluorodeoxyglucose aortic uptake in patients with giant cell arteritis (GCA) according to the presence or absence of aortic aneurysm/dilatation. A, Positron emission tomography scan of a patient with GCA-related aneurysm (left) compared with a patient with noninflammatory aneurysm (right). B, Measurement of standard uptake value (SUV) maximal (Max) and median (Med) in patients with GCA and aortic aneurysm (GCA+, AA+), patients with GCA with a normal aorta (GCA+, AA−), patients with noninflammatory aneurysm (GCA−, AA+), and control individuals (GCA−, AA−).
Histopathologic examination of aortic specimens.
Surgically removed specimens showed inflammatory infiltrates in the adventitial layer in both 2 patients with GCA and 3 patients with noninflammatory aneurysm. Scattered inflammatory foci were seen in the media only in patients with GCA and in the patient with severe atherosclerosis. No dense granulomatous lesions or giant cells were observed. Remaining foci of inflammatory cells immunostained positive for MMP-9 and MMP-2 (Figure 3A). MMP-2 expression by vascular smooth muscle cells was also observed. Elastic fibers were markedly disrupted in areas with remaining inflammatory cells, but also in many additional areas devoid of inflammatory infiltrates (Figure 3B). Gelatin zymography of tissue extracts revealed MMP-9 gelatinolytic signal in the temporal artery with active inflammatory lesions, whereas in a normal temporal artery and in a GCA-related aneurysm MMP-9 gelatinolytic signal was faintly detectable. MMP-2 gelatinolytic signal was observed both in active GCA lesions and in GCA-related aneurysm (Figure 3C).
Histopathologic examination of the aortic wall in a patient with giant cell arteritis (GCA)–related aneurysm. A, Hematoxylin/eosin (H/E) staining showing scattered chronic inflammatory foci in the media and serial sections showing matrix metalloprotease 2 (MMP-2) and MMP-9 expression by inflammatory cells (magnification × 200). B, Lower magnification (magnification × 40) covering a wider area and showing a paucity of inflammatory infiltrates; orcein staining of elastic lamellae in a serial section displaying multiple foci of disruption in areas devoid of inflammatory infiltrates. C, Gelatin zymography of tissue extracts (80 μg/lane) from a normal temporal artery (TA); a TA with active, treatment-naive GCA lesions; and a GCA-related thoracic aortic aneurysm (AA). MMP-9 is only detected in active GCA lesions whereas active MMP-2 can be detected both in active lesions and in the GCA-related aneurysm. ProMMP = Promatrix metalloprotease.
DISCUSSION
Systematic screening of a cohort of 54 patients with biopsy-proven GCA demonstrated that 12 (22.2%) patients had significant aortic structural abnormalities (aneurysm or diffuse dilatation) after a median followup of 5.4 years. Thoracic aneurysms were much more frequent than abdominal aneurysms, as reported in retrospective studies.
Our screening method, chosen on the basis of its reasonable cost:benefit ratio and feasibility in general clinical practice, relied on a careful examination of a chest radiograph and an abdominal ultrasound, which may have reduced sensitivity. It is possible that performing echocardiography in patients with aortic murmurs, as recently suggested by Bongartz and Matteson (7), would increase sensitivity. Systematic screening with more sensitive imaging techniques such as CT scan would have probably revealed a higher prevalence of subtle aortic structural abnormalities, although perhaps not always clinically relevant. Despite the potential limitations of the screening method applied, the prevalence of aortic structural damage observed is higher than that reported in retrospective studies over a much more extended period. Based on the size of the aneurysm or the resulting aortic valve insufficiency, 5 patients (9.2% of the global series and 42% of those with aneurysm or dilatation) were considered candidates for surgery. The development of aortic structural damage is, therefore, a major health threat in the outcome of patients with GCA, with a potentially increasing impact given the growing life expectancy of elderly persons in developed countries.
A relevant question arising from the recognition of aortic aneurysm/dilatation as a major and frequent complication of GCA is whether aortic structural damage appears as a consequence of the initial injury or develops progressively due to persisting, low-grade inflammatory activity. Concerns about the potential development of aortic damage as a consequence of persisting low-grade inflammatory activity despite an apparently appropriate response to steroids arise from the repeatedly reported finding of inflammatory infiltrates in surgical or necropsy aortic specimens (4,19,20). However, a critical analysis of the reported cases reveals that the characteristics and extent of inflammatory infiltrates as well as the dose and duration of the corticosteroid treatment received until surgery or necropsy are not described in detail in most reports. Therefore it is not clear whether active or residual inflammation is observed and whether specimens are obtained during active disease or in patients in remission under the current treatment strategy in terms of dose and duration of corticosteroids (13, 20, 23). Dense granulomatous lesions are usually described in specimens obtained from patients dying from aortic complications during active disease or in patients treated with low corticosteroid doses or treated for short periods (4, 13). This important point was addressed by Lie who examined 35 aortic specimens from patients with GCA (19). Lie remarked that active granulomatous lesions were found in patients in whom the samples were obtained shortly after diagnosis, whereas the intensity and extent of inflammatory infiltrates were lower in treated patients. Our data, obtained from a cohort of prospectively treated patients according to the current standard of care, do not support that patients with smoldering or relapsing disease or patients with persistently elevated acute-phase reactants or proinflammatory cytokines are more prone to develop aneurysm/dilatation. The aortic specimens obtained during elective surgery from 2 of our patients showed scattered inflammatory foci in the media. Infiltrating leukocytes immunostained positive for MMP-2 and MMP-9. MMPs, particularly MMP-9, have been considered to be involved in elastin degradation and generation of aortic aneurysm in several models (24-27). However, gelatinolytic signal of MMP-9, which is mainly produced by activated inflammatory cells, was faint in aortic aneurysm compared with active GCA lesions in a temporal artery obtained at diagnosis. In contrast, active MMP-2, which can also be expressed by vascular smooth muscle cells and is involved in vascular reparative mechanisms (28, 29), was detected equally in both active GCA lesions and aortic aneurysm. We cannot exclude that remaining infiltrates or MMP-2 produced during vascular remodeling increases vessel wall damage over the years. However, persisting inflammatory infiltrates were very scarce, whereas elastic lamellae disruption, which is an early finding in experimental aneurysm formation (24), was extensive, possibly as a consequence of the initial inflammatory injury.
Characteristics of the target tissue may play a significant role in the extent of the initial injury. Some patients may have unique substrate characteristics in their aortic tissue, making it more susceptible to aortic inflammation, whereas in others the aorta may remain relatively spared. Once inflammation and injury are established, characteristics of vascular remodeling may vary in different aortic segments. Necropsy studies have indeed demonstrated that inflammatory lesions in GCA equally target the thoracic and the abdominal aorta (4). This is in accordance with recent studies showing a similar proportion of thoracic and abdominal FDG uptake in individuals with active disease (11). However, in all series, thoracic aneurysms are much more frequent than abdominal aneurysms in patients with GCA (12-15). Thoracic and abdominal aortas differ greatly in lumen diameter, wall thickness, vasa vasorum density, content of elastic and collagen fibers, propensity to atherosclerosis, and susceptibility to infection-induced vasculitis (30, 31). Thoracic and abdominal aortas may then respond differently to inflammatory injury. In addition, the thoracic aorta is subjected to a higher pressure, which might favor progressive dilatation of a weakened wall. Sex may also influence the development of aortic damage. In our series, aortic structural abnormalities were more frequently observed in men. Male predominance in susceptibility to experimental aortic aneurysms has also been demonstrated in experimental settings (32, 33).
Contrary to data gathered from retrospective studies (14, 15), we did not observe a higher prevalence of aortic structural damage in patients with traditional cardiovascular risk factors. This may be due to the prospective nature of this study, in which tight control of vascular risk factors was part of the therapeutic approach. The higher prevalence of hypercholesterolemia, and consequently statin therapy, among patients with a preserved aortic wall raises the hypothesis of statins as protective agents against aortic wall structural damage.
In summary, prospective screening shows that a remarkable proportion of patients with GCA develop aneurysm/dilatation, in some instances severe enough to warrant surgical repair. The life-threatening nature of the potential complications derived from aortic structural damage indicates that patients with GCA should be subjected to a continuous surveillance by clinical examination and imaging. Our data do not support that in patients treated according to the current standard of care, aortic aneurysm formation results mainly from persistent activity; our data suggest interplay of heterogeneous factors. Investigating mechanisms involved in the development of aortic structural damage and its progression is of major relevance for patients with GCA.
AUTHOR CONTRIBUTIONS
Dr. Cid had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. García-Martínez, Arguis, Segarra, Lozano, Cid.
Acquisition of data. García-Martínez, Hernández-Rodríguez, Arguis, Paredes, Segarra, Lozano, Nicolau, Ramírez, Lomeña, Josa, Pons, Cid.
Analysis and interpretation of data. García-Martínez, Arguis, Paredes, Segarra, Lozano, Ramírez, Lomeña, Cid.
Manuscript preparation. García-Martínez, Cid.
Statistical analysis. García-Martínez, Cid.
