Presented in part at the 70th Annual Scientific Meeting of the American College of Rheumatology, Washington, DC, November 2006, and at the 13th International Vasculitis Meeting and ANCA Workshop, Cancun, Mexico, April 2007.
Giant Cell Arteritis
Clinical relevance of persistently elevated circulating cytokines (tumor necrosis factor α and interleukin-6) in the long-term followup of patients with giant cell arteritis†
Version of Record online: 28 MAY 2010
Copyright © 2010 by the American College of Rheumatology
Arthritis Care & Research
Volume 62, Issue 6, pages 835–841, June 2010
How to Cite
García-Martínez, A., Hernández-Rodríguez, J., Espígol-Frigolé, G., Prieto-González, S., Butjosa, M., Segarra, M., Lozano, E. and Cid, M. C. (2010), Clinical relevance of persistently elevated circulating cytokines (tumor necrosis factor α and interleukin-6) in the long-term followup of patients with giant cell arteritis. Arthritis Care Res, 62: 835–841. doi: 10.1002/acr.20043
- Issue online: 28 MAY 2010
- Version of Record online: 28 MAY 2010
- Manuscript Accepted: 22 JAN 2010
- Manuscript Received: 20 AUG 2009
- Ministerio de Ciencia y Tecnología and Fondo Europeo de Desarrollo Regional. Grant Number: SAF 08/04328
- Research award from the Hospital Clínic
- Fundació Pedro Pons
- Instituto de Salud Carlos III
To assess the clinical relevance of increased circulating cytokines in patients with giant cell arteritis (GCA) after long-term followup.
We performed a cross-sectional evaluation of 54 patients with biopsy-proven GCA prospectively followed for a median of 5.4 years (range 4–10.5 years). GCA-related complications, vascular events, relapses, current prednisone dose, time required to achieve a maintenance prednisone dosage <10 mg/day, cumulated prednisone at that point, and adverse effects during followup were recorded. Serum interleukin-6 (IL-6) and tumor necrosis factor α (TNFα) were determined by immunoassay.
All patients were in clinical remission. Both cytokines were significantly higher in patients than in controls (mean ± SD 21 ± 35 versus 5 ± 11 pg/ml; P < 0.001 for IL-6 and mean ± SD 32 ± 14 versus 16 ± 9 pg/ml; P < 0.001 for TNFα). No differences were found in patients with or without GCA-related complications or vascular events during followup. Circulating cytokines were significantly higher in patients who had experienced relapses (mean ± SD 25 ± 39 versus 10 ± 11 pg/ml; P = 0.04 for IL-6 and mean ± SD 34 ± 15 versus 25 ± 11 pg/ml; P = 0.042 for TNFα). IL-6 was significantly higher in patients still requiring prednisone (mean ± SD 29 ± 45 versus 13 ± 17 pg/ml; P = 0.008), and TNFα correlated with cumulated prednisone dose (r = 0.292, P = 0.04). No significant relationship was found between elevated cytokines and prednisone adverse effects or patients' quality of life.
Circulating TNFα and IL-6 may persist elevated in GCA patients after long-term followup and remain higher in patients who have experienced more relapsing disease. However, in this patient cohort, elevated circulating cytokines were not associated with increased frequency of GCA complications, vascular events, or treatment-related side effects.
Giant cell arteritis (GCA) is the most common systemic vasculitis among people age >50 years. GCA inflammatory lesions preferentially target large and medium-sized vessels. Typical symptoms of the disease (headache, jaw claudication, scalp tenderness, facial aches, and visual loss) derive from inflammatory involvement of the carotid artery branches. Involvement of other arteries such as the aorta and its major tributaries remains asymptomatic unless complications such as dilatation or stenosis occur (1–4).
Vascular inflammatory infiltrates are mainly composed of T lymphocytes and macrophages, which are the source of a variety of inflammatory mediators, including proinflammatory cytokines interleukin-1β (IL-1β), tumor necrosis factor α (TNFα), and IL-6 (5–9). These cytokines are mainly involved in local autocrine/paracrine responses, but TNFα and IL-6 may be released into the bloodstream and trigger systemic effects, including fever, malaise, weight loss, anemia of chronic disease type, and elevation of acute-phase proteins, all common in patients with GCA. Accordingly, local production of proinflammatory cytokines TNFα, IL-1β, and IL-6 in involved temporal arteries and circulating TNFα and IL-6 correlate with the intensity of the acute-phase response at diagnosis (8, 10). Interestingly, a strong systemic inflammatory response and markedly increased expression of IL-6 are negatively associated with the development of disease-related cranial ischemic events at diagnosis (11, 12).
Patients with GCA experience a rapid relief of their symptoms with high-dose corticosteroids. However, disease activity may not be completely abrogated, and 40–60% of patients relapse when corticosteroids are tapered (13, 14). In addition, corticosteroid-treated patients may develop GCA-related vascular complications during followup: approximately 10–15% of patients with visual symptoms continue to have deteriorating vision during the first weeks of treatment (13, 15), 22.2% of patients develop significant aortic dilatation (3), and 5–15% develop extremity artery stenosis (2, 16). It is not clear at present whether these vascular complications arise from subclinical vascular inflammation or result from maladaptive remodeling driven by the initial inflammatory injury.
High-dose corticosteroid treatment results in sharp decreases in erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level (17–19). However, when corticosteroids are tapered, acute-phase proteins may remain slightly or moderately elevated above the normal levels in some patients, even maintaining clinical remission (17–19). In keeping with this observation, it has been shown that circulating IL-6 persists elevated for several months, but it is not known whether IL-6 persists elevated after long-term followup (20). Elevated IL-6 in asymptomatic patients with GCA has been considered to reflect remaining subclinical vascular inflammation. Persistent vascular inflammation raises concerns about its long-term clinical consequences, including late GCA-related complications, accelerated atherosclerotic disease, or IL-6–induced osteopenia (20).
The aim of our study was to determine whether circulating proinflammatory cytokines persist elevated in patients with GCA after long-term followup, and whether persistent increase in circulating cytokines is associated with a higher frequency of GCA-related complications or other vascular events, clinically apparent disease activity, corticosteroid requirements, and corticosteroid-derived side effects.
PATIENTS AND METHODS
The study group consisted of 54 patients (14 men and 40 women) with biopsy-proven GCA and a median age of 79 years (range 63–91 years) who were prospectively treated and followed by the authors for a median of 5.4 years (range 4–10.5 years). This patient cohort was subjected to a cross-sectional screening for aortic dilatation in a previous study, and details regarding patient selection have been previously published (3). Briefly, patients were consecutively selected among those who had regular followup visits every 4–6 months for at least 4 years. All of the patients were treated according to a defined protocol (3). A relapse was defined as reappearance of disease-related symptoms (cranial, polymyalgic, systemic symptoms, or anemia of chronic disease type not attributable to other causes) that resolved with an increase in prednisone dose 10 mg above the previous dose able to maintain remission.
At the time of the screening for aortic aneurysm, patients were subjected to a clinical evaluation in search of disease-related symptoms. Patients' quality of life was self-estimated with a visual analog scale (VAS; 0–100 mm) for pain or other physical limitations, psychological well-being, independence for self-care, and the ability to perform work, social, or recreational activities. The average of these 4 assessments was considered.
Followup data were categorized into disease-related complications, other vascular events, disease activity, corticosteroid requirements, and corticosteroid-related adverse events. Disease-related complications included aortic dilatation and visual deterioration due to anterior ischemic optic neuritis during followup (confirmed by an ophthalmologist). Other vascular events included clinically symptomatic cardiovascular (angina or myocardial infarction), cerebrovascular (transient ischemic attack or stroke), or lower extremity arteriopathy (intermittent claudication or ischemia). Disease activity data comprised the number of relapses and the corticosteroid requirement assessed as time (in weeks) necessary to achieve a maintenance prednisone dosage <10 mg of prednisone/day, cumulated prednisone dose at that point, and prednisone treatment (any dose) at the time of evaluation. Corticosteroid-related side effects included new or worsening hypertension, diabetes mellitus and hypercholesterolemia, symptomatic fractures, gastrointestinal bleeding, mild or serious (requiring hospitalization) infection, and symptomatic cataracts requiring intervention.
At the time of the evaluation, general laboratory analysis, including hemoglobin and acute-phase reactants ESR, CRP, and haptoglobin, was performed. Circulating levels of IL-6 and TNFα were measured by immunoassay using Quantikine kits from R&D Systems, according to the manufacturer's instructions. Other cytokines thought to be relevant in the pathogenesis of GCA such as interferon-γ and IL-1β were not determined because concentrations of these cytokines in human serum are usually around the detection threshold. IL-6 and TNFα were also measured in 15 healthy donors with similar age and sex distribution. Mann-Whitney test or Student's t-test, when applicable, was employed for quantitative variables, and Spearman's or Pearson's test was employed for correlations.
The study was approved by the ethics committee of our institution, and all of the patients signed informed consent.
IL-6 and TNFα concentrations in sera from patients with GCA after long-term followup.
The median IL-6 concentration in the patient cohort was 13 pg/ml (range 0–237) and the median TNFα concentration was 30 pg/ml (range 6–77). Seventy-eight percent of patients with GCA had IL-6 levels and 79% had TNFα concentrations above the reference values for the general population (0–5 pg/ml for IL-6 and 0–20 pg/ml for TNFα). As shown in Figure 1, circulating levels of IL-6 and TNFα were significantly higher in GCA patients than in healthy controls with similar age and sex distribution.
At the time of the evaluation, the median ESR was 32 mm/hour (range 8–66), the median CRP level was 0.7 mg/dl (range 0.2–5.5), the median haptoglobin level was 1.63 gm/liter (range 0.08–2.86), and the median hemoglobin level was 129 gm/liter (range 108–167). Circulating IL-6 significantly correlated with TNFα concentrations (r = 0.378, P = 0.005) and with CRP plasma levels (r = 0.296, P = 0.03). No significant correlations were found between IL-6 or TNFα and the rest of the laboratory parameters determined (Table 1).
|IL-6, pg/ml||TNFα, pg/ml|
|CRP level, mg/dl||0.296||0.03†||0.19||ns|
IL-6 and TNFα concentrations and the development of GCA-related complications during followup.
As previously published, 12 (22.2%) of the 54 patients developed aortic aneurysm or dilatation during the followup period (3). Two patients (3.7%) experienced GCA-related worsening of vision after the initiation of corticosteroid treatment. As shown in Table 2, no differences in cytokine concentrations were found between patients who had or had not developed disease-related vascular complications during followup.
|IL-6, pg/ml||TNFα, pg/ml|
|Present, mean ± SD||Absent, mean ± SD||P||Present, mean ± SD||Absent, mean ± SD||P|
|Aortic aneurysm/dilatation (n = 12)†||13 ± 8||24 ± 39||ns||30 ± 19||32 ± 13||ns|
|Worsening of vision (n = 2)||15 ± 3||22 ± 36||ns||17 ± 6||32 ± 14||ns|
|Other vascular events (n = 7)||19 ± 29||22 ± 35||ns||32 ± 13||31 ± 15||ns|
IL-6 and TNFα concentrations and the development of vascular events.
Seven patients (13%) experienced symptomatic vascular complications in other territories. The relative contribution of GCA versus traditional vascular risk factors in the development of vascular disease could not be fully ascertained. Four patients presented lower extremity ischemia that required percutaneous revascularization and stenting (1 patient), bypass surgery (1 patient), and extremity amputation (1 patient). Three patients developed transient cerebral ischemic attacks and a Doppler sonography disclosed significant carotid stenosis. One patient experienced a stroke 1 month after the diagnosis of GCA, and the magnetic resonance angiography exhibited thrombosis of the right carotid artery. Finally, 1 patient developed myocardial infarction that required percutaneous angioplasty and stenting. Overall, 2 of the 7 patients exhibited symptomatic vascular involvement in more than one territory (extremity and cerebrovascular ischemia). As shown in Table 2, there were no significant differences in circulating levels of proinflammatory cytokines between patients with or without symptomatic vascular events during followup.
Correlation between IL-6 and TNFα concentrations and GCA activity and corticosteroid requirements.
At the time of the evaluation, all of the patients were in stable clinical remission with no evidence of relapse, infection, or other chronic inflammatory diseases within the previous 4 months. Four patients had an ESR ≥50 mm/hour and 5 had a CRP concentration ≥2 mg/dl (normal value <1). These patients had persistent mild elevation of acute-phase reactants with no development of disease-related symptoms during the following 6 months. Thirteen patients (24.1%) had not presented disease flares during the entire followup, 15 (27.8%) had experienced one relapse, and 26 (48.1%) had presented more than one. At the time of the evaluation, 27 patients had successfully discontinued prednisone, whereas the remaining 27 patients were still receiving low-dose prednisone treatment (median 3.75 mg/day, range 1.25–12.5).
Patients who had experienced at least one relapse during followup had higher levels of TNFα and IL-6 than patients with no relapsing disease (Figures 2A and B). IL-6 concentrations were significantly higher in patients who still required corticosteroid treatment at the time of the evaluation (Figure 3A). No significant differences were found in TNFα levels between patients still receiving prednisone compared with those who had successfully discontinued corticosteroid treatment (Figure 3B). Circulating levels of both IL-6 and TNFα remained significantly higher in patients who had been able to discontinue therapy than in healthy controls (mean ± SD 13 ± 17 versus 5 ± 11 pg/ml; P < 0.001 for IL-6 and mean ± SD 32 ± 12 versus 16 ± 9 pg/ml; P = 0.005 for TNFα). TNFα concentrations tended to correlate with the time required to reach a maintenance daily prednisone dosage <10 mg (r = 0.235, P = 0.09) and significantly correlated with the cumulative prednisone dose at that point (r = 0.292, P = 0.04) (Figure 3C). As shown in Table 3, the longer duration of treatment observed in patients with elevated TNFα or IL-6 levels did not result in more corticosteroid-related side effects.
|IL-6, pg/ml||TNFα, pg/ml|
|Present, mean ± SD||Absent, mean ± SD||P||Present, mean ± SD||Absent, mean ± SD||P|
|Hypertension (n = 25)||12 ± 10||30 ± 45||0.041||31 ± 16||32 ± 13||ns|
|Diabetes mellitus (n = 7)||17 ± 11||22 ± 37||ns||30 ± 23||32 ± 13||ns|
|Hypercholesterolemia (n = 20)||16 ± 17||25 ± 42||ns||31 ± 17||32 ± 13||ns|
|Osteoporotic fractures (n = 7)||22 ± 28||21 ± 36||ns||27 ± 10||32 ± 15||ns|
|Mild infection (n = 6)||56 ± 90||17 ± 18||ns||34 ± 14||31 ± 14||ns|
|Serious infection (n = 6)||24 ± 26||21 ± 36||ns||31 ± 14||32 ± 14||ns|
|Cataracts (n = 10)||36 ± 72||18 ± 19||ns||31 ± 9||32 ± 15||ns|
The median VAS score in the entire series was 90 mm (range 37–100). No significant correlation was found between IL-6 or TNFα levels and VAS scores (r = −0.228, not significant for IL-6 and r = −0.048, not significant for TNFα).
Previous studies have shown that IL-6 may persist elevated for several months after the beginning of corticosteroid treatment in patients in remission, but longer followup studies have not been performed (18, 20). To our knowledge, the present study is the first attempt to evaluate circulating cytokine levels in patients in remission after long-term followup and indicates that circulating proinflammatory cytokines IL-6 and TNFα remain elevated in a substantial proportion of patients with GCA.
The source of elevated cytokines in patients in remission is not clear. Surgical or necropsy specimens from long-term treated patients with GCA have disclosed extensive vascular remodeling with persistent, small foci of inflammatory cells (3, 21, 22). However, it is important to remark that inflammatory cells may not be the only source of cytokines. We have previously shown that cultured myointimal cells derived from temporal arteries are able to produce substantial amounts of IL-6 (23, 24). Therefore, both remaining inflammatory cells and regenerating smooth muscle cells may contribute to proinflammatory cytokine production, particularly IL-6. Increased circulating cytokines in patients with GCA may then represent long-lasting persistence of low-grade inflammatory activity and/or continuous vascular remodeling. Differences observed among patients may reflect the extent of persistent subclinical vascular inflammation/remodeling or may be due to functional polymorphisms in cytokine genes. Polymorphisms in the TNFα or IL-6 gene promoters resulting in higher cytokine production have been identified (25, 26).
Several studies have shown that chronic inflammatory diseases such as rheumatoid arthritis or systemic lupus erythematosus are associated with accelerated atherosclerosis and a higher risk of cardiovascular disease, which has been attributed to persisting low-grade inflammatory activity (27–30). Moreover, moderately increased serum levels of IL-6 and CRP in the general population are associated with a higher risk of cardiovascular events (31, 32). In our patients, persistent increase in circulating cytokines was not associated with higher frequency of vascular complications, either GCA or atherosclerosis related, during followup. However, these results must be interpreted with caution, due to the relatively small number of patients included and the relatively low frequency of disease- or atherosclerosis-related complications in GCA patients in remission (15, 33). We cannot exclude that a much larger series or a longer followup could evidence an association between elevated circulating cytokines and a higher frequency of vascular complications. However, the physiologically limited lifespan of patients with GCA reduces the significance of studies with a much longer followup and restricts its relevance to the youngest subset of patients.
Previous studies have shown that increased TNFα expression in lesions at diagnosis is associated with persistent disease activity. This observation is in accordance with results generated by several groups showing an association between a strong acute-phase response at diagnosis and more relapsing disease, both in patients with GCA and in patients with polymyalgia rheumatica (34, 35). The present study shows that patients who have experienced more relapses or have required more corticosteroid doses still maintain significantly increased circulating TNFα and IL-6 levels after long-term followup. Although increased serum TNFα and IL-6 concentrations are associated with more refractory disease, TNFα blockade failed to reduce relapses and spare corticosteroids, indicating that elevated TNFα, even being a marker of disease persistence, may not be crucial in maintaining disease activity or may be compensated by redundant cytokines (36). Blocking IL-6 has not been attempted in GCA and may or may not face similar limitations. Taken together, these findings suggest caution in attributing functional roles to these or other elevated biomarkers and underline the need for functional studies before they can be considered candidate therapeutic targets (37).
Of interest, although patients with increased TNFα or IL-6 levels had higher corticosteroid requirements, this was not associated with an increase in corticosteroid-related adverse events. This observation may again be limited by the relatively small size of the patient cohort. Moreover, regarding corticosteroid-induced osteoporosis, only symptomatic fractures were taken into account and the asymptomatic collapse of dorsal vertebrae was not systematically assessed. In addition, the development of corticosteroid-related side effects does not only depend on the cumulated doses because some patients are particularly prone to develop these complications. Patients with osteopenia at diagnosis or patients with underlying metabolic syndrome are especially susceptible to developing related complications when receiving corticosteroid therapy (38, 39).
Although not associated with major clinical consequences, subclinical inflammation may potentially produce malaise, fatigue, or reduction in well-being, impairing quality of life. Self-estimated quality of life was surprisingly high in our elderly patient cohort. This may not be representative of the overall GCA population since patients included had been able to maintain a regular long-term followup. Although increased cytokines were associated with more relapsing disease, no correlation was found between cytokine levels and patients' quality of life scores at the time of the evaluation. A limitation of this conclusion may be that quality of life was evaluated with a VAS and not with a validated instrument. An instrument to specifically measure quality of life in GCA patients is in development but is still awaiting validation (40). Illiteracy, sight problems, and lack of training in answering questionnaires were the main reasons for choosing a simple VAS in this particular patient cohort. It may be possible that a more sensitive instrument could have detected differences between patients with or without persistent subclinical inflammatory activity.
In summary, our study shows long-term persistence of elevated circulating cytokines in patients with GCA. Although patients with higher cytokine levels had experienced a more relapsing disease, persistent elevation of circulating cytokines was not associated with clinically relevant complications related to GCA, atherosclerotic disease, or corticosteroid treatment.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. 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 conception and design. García-Martínez, Cid.
Acquisition of data. García-Martínez, Hernández-Rodríguez, Espígol-Frigolé, Prieto-González, Butjosa, Segarra, Lozano, Cid.
Analysis and interpretation of data. García-Martínez, Cid.
- 2Incidence and predictors of large-artery complications (aortic aneurysm, aortic dissection, and/or large artery stenosis) in patients with giant cell arteritis: a population-based study over 50 years. Arthritis Rheum 2003; 48: 3522–31., , , , .
- 3Development of aortic aneurysm/dilatation during the followup of patients with giant cell arteritis: a cross-sectional screening of fifty-four prospectively followed patients. Arthritis Rheum 2008; 59: 422–30., , , , , , et al.
- 9Elevated production of interleukin-6 is associated with a lower incidence of disease-related ischemic events in patients with giant-cell arteritis: angiogenic activity of interleukin-6 as a potential protective mechanism. Circulation 2003; 107: 2428–34., , , , , , et al.
- 10A strong initial systemic inflammatory response is associated with higher corticosteroid requirements and longer duration of therapy in patients with giant cell arteritis. Arthritis Rheum 2002; 47: 29–35., , , , , , et al.
- 24Proinflammatory functions of vascular smooth muscle cells (VSMC) in giant cell arteritis (GCA) [abstract]. Arthritis Rheum 2008; 58 Suppl: S928–9., , , , , , et al.
- 38Baseline osteopenia is associated with subsequent bone loss and fractures in glucocorticosteroid-treated patients with giant-cell arteritis (GCA) receiving calcium and vitamin D supplements [abstract]. Arthritis Rheum 2000; 43 Suppl: S1793., , , , , , et al.