Seminars in Arthritis and Rheumatism 44 (2015) 586–591
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Coronary artery disease in giant cell arteritis: A systematic review and meta-analysis Patompong Ungprasert, MDn, Matthew J. Koster, MD, Kenneth J. Warrington, MD Division of Rheumatology, Department of Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905
a r t i c l e in f o
Keywords: Giant cell arteritis Coronary artery disease Epidemiology Meta-analysis
abstract Objective: To investigate the association between giant cell arteritis (GCA) and risk of coronary artery disease (CAD). Methods: We conducted a systematic review and meta-analysis of observational studies that reported relative risks, hazard ratios, or standardized incidence ratios with 95% confidence interval comparing CAD risk in patients with GCA versus non-GCA controls. Pooled risk ratios and 95% confidence intervals were calculated using a random-effect, generic inverse variance of DerSimonian and Laird. Result: Six studies with 10,868 patients with GCA and 245,323 controls were identified and included in our data analysis. The pooled risk ratio of CAD in patients with GCA was 1.51 and did not achieve statistical significance (95% CI: 0.88–2.61). The statistical heterogeneity was high with an I2 of 97%. Conclusion: In contrast to other chronic systemic inflammatory disorders, our meta-analysis did not show any statistically significant increased risk of CAD among patients with GCA. & 2014 Elsevier Inc. All rights reserved.
Introduction The association between chronic inflammation and premature atherosclerosis is well recognized [1,2]. Several studies have demonstrated the detrimental effect of inflammatory cytokines, oxidative stress, and activated leukocytes on endothelial function, resulting in the acceleration of atherosclerosis [3–6]. Chronic inflammation has also been shown to promote the coagulation cascade, impair the anti-coagulation pathway, and inhibit fibrinolysis resulting in a hypercoagulable state [7,8]. These factors may serve as the fundamental pathophysiology of the development of premature coronary artery disease (CAD). Moreover, an increased incidence of CAD has been observed in several chronic inflammatory disorders, such as rheumatoid arthritis, idiopathic inflammatory myopathy, systemic sclerosis, and systemic lupus erythematosus [9–12]. Giant cell arteritis (GCA) is a chronic inflammatory condition characterized by medium- and large-vessel granulomatous vasculitis, typically affecting adults older than 50 years of age [13]. Vascular complications of GCA include ischemic optic neuropathy,
Authors' contributions: Patompong Ungprasert: study design, data search and collection, statistical analysis, and writing manuscript. Matthew J. Koster: data search and collection and revising manuscript. Kenneth J. Warrington: study design and revising manuscript. n Corresponding author. E-mail addresses: [email protected], [email protected] (P. Ungprasert). http://dx.doi.org/10.1016/j.semarthrit.2014.10.010 0049-0172/& 2014 Elsevier Inc. All rights reserved.
stroke, large-vessel stenosis, and aneurysm [14]. Patients with GCA may be at an increased risk of CAD as well. However, the data on CAD risk in these patients remain unclear due to conflicting epidemiological studies [15–17]. Thus, to further investigate this association, we conducted a systematic review and meta-analysis of case–control and cohort studies that compared the risk of CAD in patients with GCA versus non-GCA participants.
Methods Search strategy Two investigators (P.U. and M.J.K.) independently searched published studies indexed in MEDLINE and EMBASE database from inception to August 2014 as well as the American College of Rheumatology annual conference abstract database from 2006 to 2013 using the search strategy described in Appendix 1. A manual search of references of selected retrieved articles was also performed. Inclusion criteria The inclusion criteria were as follows: (1) cohort or case– control study (either prospective or retrospective) published as original study or abstract reporting CAD incidence in patients with GCA; (2) relative risk (RRs), odds ratio (ORs), hazard ratio (HRs) or
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
587
Fig. 1. Outline of our search methodology.
standardized incidence ratio (SIRs), with 95% confidence intervals (CIs) were provided; and (3) non-GCA participants and participants without CAD were used as a reference group for cohort study and case–control study, respectively. Study eligibility was independently determined by each investigator noted above. Differing decisions were resolved by consensus. The quality of the included studies was independently appraised by each investigator using the Newcastle–Ottawa quality assessment scale. Using this scale, each study is assessed on eight items that are categorized into three groups including (1) the selection of the study groups, (2) the comparability of the groups, and (3) the ascertainment of the exposure or outcome of interest for case–control or cohort studies, respectively [18].
Data extraction A standardized data collection form was used to extract the following information: last name of the first author, title of the article, year of publication, country where the study was
conducted, year of publication, study size, study population, criteria used for the diagnosis of GCA, definition and method of verification of coronary artery disease, mean duration of follow-up, and adjusted effect estimates with 95% CI. This data extraction was independently performed by the two investigators. Any differences in data extraction were resolved by consensus.
Statistical analysis Data analysis was performed using Review Manager 5.3 software from the Cochrane Collaboration. Adjusted point estimates and standard errors were extracted from individual studies and were combined by the generic inverse variance method of DerSimonian and Laird [19]. Given the high likelihood of between-study variance with the different study designs and populations, we used a random-effect model rather than a fixedeffect model. Cochran's Q test was used to determine the statistical heterogeneity of this study. This test was complemented with the I2 statistic, which quantifies the proportion of total variation across
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
588
Table Main characteristics of studies included in the meta-analysis Ray et al. [15]
Country Study design Cases
Le Page et al. [16]
Canada Retrospective cohort
France Prospective cohort Newly diagnosed All patients who were cases of GCA diagnosed with GCA were recruited between January 1, 1995 from the and March 31, 2002. Cases participating were identified by using hospitals from The Ontario Health January 1991 to Insurance Plan database April 2004. (which covers 1.5 million senior residents).
Molloy et al. [17]
Amiri et al. [32]
Tomasson et al. [33]
USA Retrospective cohort
Canada Retrospective cohort
UK Retrospective cohort
Udayakumar et al. [34]
Subjects over 65 years of age Sex- and agematched randomly selected from subjects the same database. randomly selected from general population. MI or angina. Definition of MI, angina, or coronary artery revascularization. coronary artery disease Follow-up Until occurrence of any 24 Months cardiovascular event, death, or March 31, 2002.
Subjects over 50 years of age randomly selected from the same database.
Mean age of cases (yr) Women (%) Number of cases Number of controls Average range of follow-up (yr) Confounder adjusted
75.2
75.1
NA
USA Retrospective cohort All patients in All patients covered by the All patients who were Olmsted diagnosed with GCA comprehensive British county, between January 1, 1990 Columbia provincial Minnesota, and June 1, 2010. Cases medical service who USA, who were identified by using were diagnosed with were The Health Improvement GCA between 1990 and diagnosed Network database (which 2010. with GCA covers 7.3 million between patients of general 1950 and practitioner in the UK). 2009. Must fulfill the following Diagnostic code from the ACR 1990 GCA criteria (1) 40 years of registry plus at least two classification age or older, prescriptions for oral criteria. (2) diagnosis of GCA at glucocorticoid after the least two visits, and diagnosis. (3) use of steroid within 6 months after the diagnosis. Sex- and ageSex-, age-, and time of Sex- and age-matched matched entry-matched subjects subjects randomly subjects randomly selected from selected from the same randomly the same database. database. selected from the same database. MI MI Unstable angina, MI, NSTEMI, or STEMI. Until death, NA Until occurrence of any emigration cardiovascular event, from the death, emigration from system or the system, or June 1, April 30, 2010. 2013. 75.0 73.1 76.2
59.0 1142
61.0 432
NA 4807
74.0 834
73.2 3408
79.2 245
200,000
483
19,228
8340
17,027
245
2.7
2.0
NA
NA
3.9
9.7
Age, sex, hypertension, DM, cancer, dyslipidemia, and current medication.
Age, sex, history of angina, Age and sex. Age, sex, race, admission COPD, obesity, use of type, mean income, hormonal replacement hypertension, DM, therapy, DM, obesity, dyslipidemia, and dyslipidemia, use of peripheral vascular NSAIDs, and number of disease. hospitalizations.
Age, sex, and calendar year of index date.
Quality assessment (Newcastle– Ottawa scale)
Selection: 4 stars
Age, sex, hypertension, DM, cancer dyslipidemia, smoking, and peripheral vascular disease. Selection: 3 stars
Selection: 3 stars
Selection: 4 stars
Selection: 4 stars
Comparability: 1 star Outcome: 2 stars
Comparability: 1 star
Comparability: 2 stars
Comparability: 2 stars
Outcome: 2 stars
Outcome: 2 stars
Outcome: 2 stars
Selection: 4 stars Comparability: 2 stars Outcome: 3 stars
Diagnosis of Diagnostic code from the GCA registry.
Must fulfill their pre-defined criteria.
Controls
Comparability: 2 stars Outcome: 2 stars
All patients who were hospitalized with a diagnosis of GCA in the year 2004. Cases were identified by using the Nationwide Inpatient Sample database (which covers eight million hospital discharges from 1004 hospitals across 37 states). Diagnostic code from the registry.
NA
NA
USA, United States of America; UK, United Kingdom; GCA, giant cell arteritis; ACR, American College of Rheumatology; NA, Not available; MI, myocardial infarction; STEMI, ST elevation myocardial infarction; NSTEMI, non-ST elevation myocardial infarction; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus.
studies that is due to heterogeneity rather than chance. A value of I2 of 0–25% indicates insignificant heterogeneity, 26–50% low heterogeneity, 51–75% moderate heterogeneity, and 76–100% high heterogeneity [20].
Results Our search strategy yielded 314 potentially relevant studies. Of them, 293 studies were excluded, as they were not cohort/
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
589
Fig. 2. Forest plot of all included studies.
case–control studies or were not conducted in patients with GCA. Overall, 17 studies underwent full-length article review. Eight of them were excluded since they were descriptive studies without a control group [21–28] and two studies were excluded because they reported CAD mortality ratios among patients with GCA but did not report incidence ratios [29,30]. Seven studies (one case–control study [31] and six retrospective cohort studies [15–17,32–34]) met our eligibility criteria. However, two studies utilized the same database [31,34]. Thus, to avoid potential patient duplication, we excluded the study by Machado et al. [31] as the data from the study by Udayakumar et al. [34] is more comprehensive, leaving six studies with 10,868 patients with GCA and 245,323 controls for the data analyses. Figure 1 outlines our search methodology and review process. The detailed characteristics and Newcastle–Ottawa quality assessment scale of the included studies are described in the Table. Our meta-analysis failed to show a statistically significant increased CAD risk among patients with GCA with the pooled risk ratio of 1.51 (95% CI: 0.88–2.61). The statistical heterogeneity was high with an I2 of 97%. All six studies were relatively equally weighted. Figure 2 demonstrates the forest plots of our findings. To further explore the high statistical heterogeneity, we performed two sensitivity analyses. The first sensitivity analysis was conducted by excluding the study by Le Page et al. [16], as it was the only prospective cohort study, in contrast to medical registry-based retrospective cohort design in all other studies. However, the value of I2 after the exclusion of this study remained the same (I2 ¼ 97%, pooled risk ratio ¼ 1.43, 95% CI: 0.83–2.69). The second sensitivity
analysis was performed by excluding the studies by Amiri et al. [32] and Tomasson et al. [33], as these two studies strictly defined CAD as myocardial infarction while the rest of the studies used a broader definition of CAD. This sensitivity analysis decreased the I2 to 83% (pooled risk ratio ¼ 1.13, 95% CI: 0.70–1.84). We further excluded the study by Molloy et al. [17], as this study did not provide its definition of CAD. The value of I2 decreased further to 75% (pooled risk ratio ¼ 1.32, 95% CI: 0.69–2.53). These sensitivity analyses may serve as indirect evidence that statistical heterogeneity was driven, in part, by the difference in CAD definition. We also performed another sensitivity analysis by adding two studies that reported CAD-related mortality ratios by using the mortality ratios as surrogate for incidence ratios [29,30]. Addition of these two studies slightly increased pooled risk ratio to 1.53 but still could not achieve a statistical significance (95% CI: 0.97–2.42). Evaluation for publication bias Funnel plot to evaluate publication bias is shown in Figure 3. The graph is asymmetric, providing a suggestion that publication bias may be present.
Discussion In contrast to other chronic systemic inflammatory diseases, an increased CAD risk among patients with GCA was not observed in
Fig. 3. Funnel plot.
590
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
this meta-analysis [9–12]. There are few possible explanations for this negative result. First, interestingly, several studies have found a lower prevalence of cardiovascular risk factors, particularly diabetes mellitus (DM) and dyslipidemia among patients with GCA compared with non-GCA controls [16,34–38]. The reason why patients with this form of vasculitis might have a lower prevalence of DM is unclear. One hypothesis is that patients with DM may have a decreased responsiveness of T cells to the inciting antigens being presented by the dendritic cells of the arterial adventitia, a process that is crucial in the development of GCA [39,40]. This lower baseline risk might offset the deleterious effect of chronic inflammation on endothelial cells of coronary artery [35]. Second, GCA is a disease of the elderly which is characteristically seen in the sixth to seventh decade of life [13,14]. In contrast to rheumatoid arthritis and systemic lupus erythematous in which patients are diagnosed on an average at a younger age and experience a longer disease course, patients with GCA might not live long enough after disease diagnosis to observe the detrimental effect of chronic inflammation on their coronary arteries. In fact, the average follow-up duration of most of the included studies was less than 4 years [15,16,33]. Moreover, this large-vessel vasculitis is universally treated with corticosteroids, a very potent anti-inflammatory agent, which can decrease the inflammatory burden and, thus, might further slowdown the progression of associated atherosclerosis [41]. Even though most of the included studies are of high quality, there are some limitations and, thus, these results should be interpreted with caution. First, most of the included studies were conducted using medical registry-based database, raising a concern of coding inaccuracy. The studies by Ray et al. [15] and Tomasson et al. [33] were particularly of concern as the first study had only 25% temporal artery biopsy billing codes and only 47% patients on prednisone, while the latter had high amounts of missing data (only complete data in 43.3% of GCA and 37.3% in reference group). To confirm the robustness of the result, we conducted another sensitivity analysis excluding these two studies. The exclusion of these two studies with a particular coding inaccuracy concern did not significantly alter our result as the pooled risk ratio only slightly decreased to 1.36 without achieving a statistical significance (95% CI: 0.60–2.89; I2 ¼ 96%). Second, as mentioned before, publication bias might be present. Third, the statistical heterogeneity was high in this meta-analysis. We suspect that the difference in CAD definition was the main source of this heterogeneity as suggested by sensitivity analysis described above.
Conclusion In contrast to other rheumatic inflammatory disorders, a significant increased CAD risk among patients with GCA was not observed in this meta-analysis. The reason for this negative finding remains unclear but could possibly be related to the lower baseline cardiovascular risk and the older age at diagnosis.
Appendix A. Supporting Information Supplementary material cited in this article is available online at http://dx.doi.org/10.1016/j.semarthrit.2014.10.010.
Reference [1] Frostegard J. Atherosclerosis in patients with autoimmune disorders. Arterioscler Thromb Vasc Biol 2005;25:1776–85. [2] Libby P. Inflammation in atherosclerosis. Nature 2002;420:868–74.
[3] Montecucco F, Mach F. Common inflammatory mediators orchestrate pathophysiological processes in rheumatoid arthritis and atherosclerosis. Rheumatology (Oxford) 2009;48:11–22. [4] Rho YH, Chung CP, Oeser A, Solus JF, Gebretsadik T, Shintani A, et al. Interaction between oxidative stress and high-density lipoprotein cholesterol is associated with severity of coronary artery calcification in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62:1473–80. [5] Niessner A, Sato K, Chaikof EL, Colmegna I, Goronzy JJ, Weyand CM. Pathogensensing plasmacytoid dendritic cells stimulate cytotoxic T-cell function in atherosclerotic plaque through interferon-alpha. Circulation 2006;114: 2482–9. [6] Hansen S, Lassig G, Pistrosch E, Passauer J. Endothelial dysfunction in young patients with long-term rheumatoid arthritis and low disease activity. Atherosclerosis 2003;170:177–80. [7] Xu J, Lupu F, Esmon CT. Inflammation, innate immunity and blood coagulation. Hamostaseologie 2010;30:8–9. [8] Esmon CT. The interactions between inflammation and coagulation. Br J Haematol 2005;131:417–30. [9] Maradit-Kremers H, Nicola PJ, Crowson CS, Ballman KV, Gabreil SE. Cardiovascular death in rheumatoid arthritis: a population-based study. Arthritis Rheum 2005;52:722–32. [10] Bessant R, Hingorani, Patel L, MacGregor A, Isenberg DA, Rahman A. Risk of coronary heart disease and stroke in a large British cohort of patients with systemic lupus erythematosus. Rheumatology (Oxford) 2004;43:924–9. [11] Ungprasert P, Charoenpong P, Ratanasrimetha P, Thongprayoon C, Cheungpasitporn W, Suksaranjit P. Risk of coronary artery disease in patients with systemic sclerosis: a systematic review and meta-analysis. Clin Rheumatol 2014;33:1099–104. [12] Ungprasert P, Suksaranjit P, Spanuchart I, Leeaphorn N, Permpalung N. Risk of coronary artery disease in patients with idiopathic inflammatory myopathies: a systematic review and meta-analysis of observational studies. Semin Arthritis Rheum 2014;44:63–7. [13] Salvarani C, Cantini F, Hunder GG. Polymyalgia rheumatica and giant-cell arteritis. Lancet 2008;372:234–45. [14] Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz MJ, Miranda-Filloy JA, Gonzalez-Juanatey C, Martin J, et al. Epidemiology of giant cell arteritis and polymyalgia rheumatica. Arthritis Rheum 2009;61:1454–61. [15] Ray JG, Mamdani MM, Geerts WH. Giant cell arteritis and cardiovascular disease in older adults. Heart 2005;91:324–8. [16] Le Page L, Duhaut P, Seydoux D, Bosshard S, Ecochard R, Abbas F, et al. Incidence of cardiovascular events in giant cell arteritis: preliminary results of a prospective double cohort study (GRACG). Rev Med Interne 2006;27: 98–105. [17] Molloy ES, Kirchner HL, Murray PK, Hoffman GS, Langford CA. Cardio-vascular outcomes in giant cell arteritis. Presented at the American College of Rheumatology Annual Scientific Meeting, 6–11 November 2007, Boston, Massachusetts. [Abstract no. 1235]. [18] Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603–5. [19] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trial 1986;7:177–88. [20] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J 2003;327:557–60. [21] Dunstan E, Lester SL, Rischmueller M, Dodd T, Black R, Ahern M, et al. Epidemiology of biopsy-proven giant cell arteritis in South Australia. Intern Med J 2014;44:32–9. [22] García-Martínez A, Arguis P, Prieto-González S, Espígol-Frigolé G, Alba MA, Butjosa M, et al. Prospective long term follow-up of a cohort of patients with giant cell arteritis screened for aortic structural damage (aneurysm or dilatation). Ann Rheum Dis 2014;73:1826–32. [23] Souza AW, Okamoto KY, Abrantes F, Schau B, Bacchiega AB, et al. Giant cell arteritis: a multicenter observational study in Brazil. Clinics 2013;68: 317–22. [24] Kisza K, Murchison AP, Dai Y, Bilyk JR, Eagle RC Jr, Sergott R, et al. Giant cell arteritis incidence: analysis by season and year in mid-Atlantic United States. Clin Experiment Ophthalmol 2013;41:577–81. [25] Espitia O, Néel A, Leux C, Connault J, Espitia-Thibault A, Ponge T, et al. Giant cell arteritis with or without aortitis at diagnosis. A retrospective study of 22 patients with longterm followup. J Rheumatol 2012;39:2157–62. [26] Abdul-Rahman AM, Molteno AC, Bevin TH. The epidemiology of giant cell arteritis in Otago, New Zealand: a 9-year analysis. N Z Med J 2011;124:44–52. [27] Singh S, Balakrishnan C, Mangat G, Samant R, Bambani M, Kalke S, et al. Giant cell arteritis in Mumbai. J Assoc Physicians India 2010;58:372–4. [28] Aziz S, Al-Ansari A, Bankart J, McFadzean R. Clinical manifestations and laboratory tests in biopsy proven giant cell arteritis in Glasgow. Eur J Intern Med 2009;20:e146. [29] Gonzalez-Gay MA, Rubiera G, Pineiro A, Garcia-Porrua C, Pego-Reigosa R, Gonzalez-Juanatey C, et al. Ischemic heart disease in patients from northwest Spain with biopsy proven giant cell arteritis. A population based study. J Rheum 2005;32:502–6. [30] Uddhammar A, Eriksson AL, Nystrom L, Stenling R, Rantapaa-Dahlqvist S. Increased mortality due to cardiovascular disease in patients with giant cell arteritis in northern Sweden. J Rheum 2002;29:737–42. [31] Machado EBV, Gabriel SE, Beard CM, Michet CJ, O'Fallon WM, Ballard DJ. A population-based case-control study of temporal arteritis: evidence for an
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
[32]
[33]
[34]
[35]
association between temporal arteritis and degenerative vascular disease. Int J Epidemiol 1989;18:836–41. Amiri N, Choi HK, Sayre EC, Shojania K, Avina-Zubieta JA. Risk of myocardial infarction in patients with giant cell arteritis: a populationbased study. Presented at the American College of Rheumatology Annual Scientific Meeting, 25–30 October 2013, San Diego, California. [Abstract no. 1673]. Tomasson G, Peloquin C, Mohammad A, Love TJ, Zhang Y, Choi HK. Risk for cardiovascular disease early and late after a diagnosis of giant-cell arteritis. A cohort study. Ann Intern Med 2014;160:73–80. Udayakumar PD, Chandran AK, Crowson CS, Warrington KJ, Matteson EL. Cardiovascular risk and acute coronary syndrome in giant cell arteritis: a population based retrospective cohort study. Arthritis Care Res (Hoboken) 2014. http://dx.doi.org/10.1002/acr.22416 [Epub ahead of print]. Matthews JL, Gilbert DN, Farris BK, Siatkowski RM. Prevalence of diabetes mellitus in biopsy-positive giant cell arteritis. J Neuroophthalmol 2012;32: 202–6.
591
[36] Schmidt J, Kermani TA, Muratore F, Crowson CS, Matteson EL, Warrington KJ. Statin use in giant cell arteritis: a retrospective study. J Rheumatol 2013;40:910–5. [37] Gonzalez-Gay MA, Pineiro A, Gomez-Gigirey A, Garcia-Porrua C, Pego-Reigosa R, Dierssen-Sotos T, et al. Influence of traditional risk factors of atherosclerosis in the development of severe ischemic complications in giant cell arteritis. Medicine 2004;83:342–7. [38] Gonzalez-Juanatey C, Lopez-Diaz MJ, Martin J, Llorca J, Gonzalez-Gay MA. Atherosclerosis in patients with biopsy-proven giant cell arteritis. Arthritis Rheum 2007;57:1481–6. [39] Ly K-H, Regent A, Tamby MC, Mouthon L. Pathogenesis of giant cell arteritis: more than just an inflammatory condition? Autoimmun Rev 2010;9:635–45. [40] Casey J, Sturm C Jr. Impaired response of lymphocytes from non-insulin dependent diabetics to staphage lysate and tetanus antigen. J Clin Microbiol 1982;15:109–14. [41] Roman MJ, Shanker BA, Davis A, Lockshin MD, Sammaritano L, Simantov R, et al. Prevalence and correlates of accelerated atherosclerosis in systemic lupus erythematosus. N Engl J Med 2003;349:2399–406.
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Coronary artery disease in giant cell arteritis: A systematic review and meta-analysis Patompong Ungprasert, MDn, Matthew J. Koster, MD, Kenneth J. Warrington, MD Division of Rheumatology, Department of Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905
a r t i c l e in f o
Keywords: Giant cell arteritis Coronary artery disease Epidemiology Meta-analysis
abstract Objective: To investigate the association between giant cell arteritis (GCA) and risk of coronary artery disease (CAD). Methods: We conducted a systematic review and meta-analysis of observational studies that reported relative risks, hazard ratios, or standardized incidence ratios with 95% confidence interval comparing CAD risk in patients with GCA versus non-GCA controls. Pooled risk ratios and 95% confidence intervals were calculated using a random-effect, generic inverse variance of DerSimonian and Laird. Result: Six studies with 10,868 patients with GCA and 245,323 controls were identified and included in our data analysis. The pooled risk ratio of CAD in patients with GCA was 1.51 and did not achieve statistical significance (95% CI: 0.88–2.61). The statistical heterogeneity was high with an I2 of 97%. Conclusion: In contrast to other chronic systemic inflammatory disorders, our meta-analysis did not show any statistically significant increased risk of CAD among patients with GCA. & 2014 Elsevier Inc. All rights reserved.
Introduction The association between chronic inflammation and premature atherosclerosis is well recognized [1,2]. Several studies have demonstrated the detrimental effect of inflammatory cytokines, oxidative stress, and activated leukocytes on endothelial function, resulting in the acceleration of atherosclerosis [3–6]. Chronic inflammation has also been shown to promote the coagulation cascade, impair the anti-coagulation pathway, and inhibit fibrinolysis resulting in a hypercoagulable state [7,8]. These factors may serve as the fundamental pathophysiology of the development of premature coronary artery disease (CAD). Moreover, an increased incidence of CAD has been observed in several chronic inflammatory disorders, such as rheumatoid arthritis, idiopathic inflammatory myopathy, systemic sclerosis, and systemic lupus erythematosus [9–12]. Giant cell arteritis (GCA) is a chronic inflammatory condition characterized by medium- and large-vessel granulomatous vasculitis, typically affecting adults older than 50 years of age [13]. Vascular complications of GCA include ischemic optic neuropathy,
Authors' contributions: Patompong Ungprasert: study design, data search and collection, statistical analysis, and writing manuscript. Matthew J. Koster: data search and collection and revising manuscript. Kenneth J. Warrington: study design and revising manuscript. n Corresponding author. E-mail addresses: [email protected], [email protected] (P. Ungprasert). http://dx.doi.org/10.1016/j.semarthrit.2014.10.010 0049-0172/& 2014 Elsevier Inc. All rights reserved.
stroke, large-vessel stenosis, and aneurysm [14]. Patients with GCA may be at an increased risk of CAD as well. However, the data on CAD risk in these patients remain unclear due to conflicting epidemiological studies [15–17]. Thus, to further investigate this association, we conducted a systematic review and meta-analysis of case–control and cohort studies that compared the risk of CAD in patients with GCA versus non-GCA participants.
Methods Search strategy Two investigators (P.U. and M.J.K.) independently searched published studies indexed in MEDLINE and EMBASE database from inception to August 2014 as well as the American College of Rheumatology annual conference abstract database from 2006 to 2013 using the search strategy described in Appendix 1. A manual search of references of selected retrieved articles was also performed. Inclusion criteria The inclusion criteria were as follows: (1) cohort or case– control study (either prospective or retrospective) published as original study or abstract reporting CAD incidence in patients with GCA; (2) relative risk (RRs), odds ratio (ORs), hazard ratio (HRs) or
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
587
Fig. 1. Outline of our search methodology.
standardized incidence ratio (SIRs), with 95% confidence intervals (CIs) were provided; and (3) non-GCA participants and participants without CAD were used as a reference group for cohort study and case–control study, respectively. Study eligibility was independently determined by each investigator noted above. Differing decisions were resolved by consensus. The quality of the included studies was independently appraised by each investigator using the Newcastle–Ottawa quality assessment scale. Using this scale, each study is assessed on eight items that are categorized into three groups including (1) the selection of the study groups, (2) the comparability of the groups, and (3) the ascertainment of the exposure or outcome of interest for case–control or cohort studies, respectively [18].
Data extraction A standardized data collection form was used to extract the following information: last name of the first author, title of the article, year of publication, country where the study was
conducted, year of publication, study size, study population, criteria used for the diagnosis of GCA, definition and method of verification of coronary artery disease, mean duration of follow-up, and adjusted effect estimates with 95% CI. This data extraction was independently performed by the two investigators. Any differences in data extraction were resolved by consensus.
Statistical analysis Data analysis was performed using Review Manager 5.3 software from the Cochrane Collaboration. Adjusted point estimates and standard errors were extracted from individual studies and were combined by the generic inverse variance method of DerSimonian and Laird [19]. Given the high likelihood of between-study variance with the different study designs and populations, we used a random-effect model rather than a fixedeffect model. Cochran's Q test was used to determine the statistical heterogeneity of this study. This test was complemented with the I2 statistic, which quantifies the proportion of total variation across
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
588
Table Main characteristics of studies included in the meta-analysis Ray et al. [15]
Country Study design Cases
Le Page et al. [16]
Canada Retrospective cohort
France Prospective cohort Newly diagnosed All patients who were cases of GCA diagnosed with GCA were recruited between January 1, 1995 from the and March 31, 2002. Cases participating were identified by using hospitals from The Ontario Health January 1991 to Insurance Plan database April 2004. (which covers 1.5 million senior residents).
Molloy et al. [17]
Amiri et al. [32]
Tomasson et al. [33]
USA Retrospective cohort
Canada Retrospective cohort
UK Retrospective cohort
Udayakumar et al. [34]
Subjects over 65 years of age Sex- and agematched randomly selected from subjects the same database. randomly selected from general population. MI or angina. Definition of MI, angina, or coronary artery revascularization. coronary artery disease Follow-up Until occurrence of any 24 Months cardiovascular event, death, or March 31, 2002.
Subjects over 50 years of age randomly selected from the same database.
Mean age of cases (yr) Women (%) Number of cases Number of controls Average range of follow-up (yr) Confounder adjusted
75.2
75.1
NA
USA Retrospective cohort All patients in All patients covered by the All patients who were Olmsted diagnosed with GCA comprehensive British county, between January 1, 1990 Columbia provincial Minnesota, and June 1, 2010. Cases medical service who USA, who were identified by using were diagnosed with were The Health Improvement GCA between 1990 and diagnosed Network database (which 2010. with GCA covers 7.3 million between patients of general 1950 and practitioner in the UK). 2009. Must fulfill the following Diagnostic code from the ACR 1990 GCA criteria (1) 40 years of registry plus at least two classification age or older, prescriptions for oral criteria. (2) diagnosis of GCA at glucocorticoid after the least two visits, and diagnosis. (3) use of steroid within 6 months after the diagnosis. Sex- and ageSex-, age-, and time of Sex- and age-matched matched entry-matched subjects subjects randomly subjects randomly selected from selected from the same randomly the same database. database. selected from the same database. MI MI Unstable angina, MI, NSTEMI, or STEMI. Until death, NA Until occurrence of any emigration cardiovascular event, from the death, emigration from system or the system, or June 1, April 30, 2010. 2013. 75.0 73.1 76.2
59.0 1142
61.0 432
NA 4807
74.0 834
73.2 3408
79.2 245
200,000
483
19,228
8340
17,027
245
2.7
2.0
NA
NA
3.9
9.7
Age, sex, hypertension, DM, cancer, dyslipidemia, and current medication.
Age, sex, history of angina, Age and sex. Age, sex, race, admission COPD, obesity, use of type, mean income, hormonal replacement hypertension, DM, therapy, DM, obesity, dyslipidemia, and dyslipidemia, use of peripheral vascular NSAIDs, and number of disease. hospitalizations.
Age, sex, and calendar year of index date.
Quality assessment (Newcastle– Ottawa scale)
Selection: 4 stars
Age, sex, hypertension, DM, cancer dyslipidemia, smoking, and peripheral vascular disease. Selection: 3 stars
Selection: 3 stars
Selection: 4 stars
Selection: 4 stars
Comparability: 1 star Outcome: 2 stars
Comparability: 1 star
Comparability: 2 stars
Comparability: 2 stars
Outcome: 2 stars
Outcome: 2 stars
Outcome: 2 stars
Selection: 4 stars Comparability: 2 stars Outcome: 3 stars
Diagnosis of Diagnostic code from the GCA registry.
Must fulfill their pre-defined criteria.
Controls
Comparability: 2 stars Outcome: 2 stars
All patients who were hospitalized with a diagnosis of GCA in the year 2004. Cases were identified by using the Nationwide Inpatient Sample database (which covers eight million hospital discharges from 1004 hospitals across 37 states). Diagnostic code from the registry.
NA
NA
USA, United States of America; UK, United Kingdom; GCA, giant cell arteritis; ACR, American College of Rheumatology; NA, Not available; MI, myocardial infarction; STEMI, ST elevation myocardial infarction; NSTEMI, non-ST elevation myocardial infarction; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus.
studies that is due to heterogeneity rather than chance. A value of I2 of 0–25% indicates insignificant heterogeneity, 26–50% low heterogeneity, 51–75% moderate heterogeneity, and 76–100% high heterogeneity [20].
Results Our search strategy yielded 314 potentially relevant studies. Of them, 293 studies were excluded, as they were not cohort/
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
589
Fig. 2. Forest plot of all included studies.
case–control studies or were not conducted in patients with GCA. Overall, 17 studies underwent full-length article review. Eight of them were excluded since they were descriptive studies without a control group [21–28] and two studies were excluded because they reported CAD mortality ratios among patients with GCA but did not report incidence ratios [29,30]. Seven studies (one case–control study [31] and six retrospective cohort studies [15–17,32–34]) met our eligibility criteria. However, two studies utilized the same database [31,34]. Thus, to avoid potential patient duplication, we excluded the study by Machado et al. [31] as the data from the study by Udayakumar et al. [34] is more comprehensive, leaving six studies with 10,868 patients with GCA and 245,323 controls for the data analyses. Figure 1 outlines our search methodology and review process. The detailed characteristics and Newcastle–Ottawa quality assessment scale of the included studies are described in the Table. Our meta-analysis failed to show a statistically significant increased CAD risk among patients with GCA with the pooled risk ratio of 1.51 (95% CI: 0.88–2.61). The statistical heterogeneity was high with an I2 of 97%. All six studies were relatively equally weighted. Figure 2 demonstrates the forest plots of our findings. To further explore the high statistical heterogeneity, we performed two sensitivity analyses. The first sensitivity analysis was conducted by excluding the study by Le Page et al. [16], as it was the only prospective cohort study, in contrast to medical registry-based retrospective cohort design in all other studies. However, the value of I2 after the exclusion of this study remained the same (I2 ¼ 97%, pooled risk ratio ¼ 1.43, 95% CI: 0.83–2.69). The second sensitivity
analysis was performed by excluding the studies by Amiri et al. [32] and Tomasson et al. [33], as these two studies strictly defined CAD as myocardial infarction while the rest of the studies used a broader definition of CAD. This sensitivity analysis decreased the I2 to 83% (pooled risk ratio ¼ 1.13, 95% CI: 0.70–1.84). We further excluded the study by Molloy et al. [17], as this study did not provide its definition of CAD. The value of I2 decreased further to 75% (pooled risk ratio ¼ 1.32, 95% CI: 0.69–2.53). These sensitivity analyses may serve as indirect evidence that statistical heterogeneity was driven, in part, by the difference in CAD definition. We also performed another sensitivity analysis by adding two studies that reported CAD-related mortality ratios by using the mortality ratios as surrogate for incidence ratios [29,30]. Addition of these two studies slightly increased pooled risk ratio to 1.53 but still could not achieve a statistical significance (95% CI: 0.97–2.42). Evaluation for publication bias Funnel plot to evaluate publication bias is shown in Figure 3. The graph is asymmetric, providing a suggestion that publication bias may be present.
Discussion In contrast to other chronic systemic inflammatory diseases, an increased CAD risk among patients with GCA was not observed in
Fig. 3. Funnel plot.
590
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
this meta-analysis [9–12]. There are few possible explanations for this negative result. First, interestingly, several studies have found a lower prevalence of cardiovascular risk factors, particularly diabetes mellitus (DM) and dyslipidemia among patients with GCA compared with non-GCA controls [16,34–38]. The reason why patients with this form of vasculitis might have a lower prevalence of DM is unclear. One hypothesis is that patients with DM may have a decreased responsiveness of T cells to the inciting antigens being presented by the dendritic cells of the arterial adventitia, a process that is crucial in the development of GCA [39,40]. This lower baseline risk might offset the deleterious effect of chronic inflammation on endothelial cells of coronary artery [35]. Second, GCA is a disease of the elderly which is characteristically seen in the sixth to seventh decade of life [13,14]. In contrast to rheumatoid arthritis and systemic lupus erythematous in which patients are diagnosed on an average at a younger age and experience a longer disease course, patients with GCA might not live long enough after disease diagnosis to observe the detrimental effect of chronic inflammation on their coronary arteries. In fact, the average follow-up duration of most of the included studies was less than 4 years [15,16,33]. Moreover, this large-vessel vasculitis is universally treated with corticosteroids, a very potent anti-inflammatory agent, which can decrease the inflammatory burden and, thus, might further slowdown the progression of associated atherosclerosis [41]. Even though most of the included studies are of high quality, there are some limitations and, thus, these results should be interpreted with caution. First, most of the included studies were conducted using medical registry-based database, raising a concern of coding inaccuracy. The studies by Ray et al. [15] and Tomasson et al. [33] were particularly of concern as the first study had only 25% temporal artery biopsy billing codes and only 47% patients on prednisone, while the latter had high amounts of missing data (only complete data in 43.3% of GCA and 37.3% in reference group). To confirm the robustness of the result, we conducted another sensitivity analysis excluding these two studies. The exclusion of these two studies with a particular coding inaccuracy concern did not significantly alter our result as the pooled risk ratio only slightly decreased to 1.36 without achieving a statistical significance (95% CI: 0.60–2.89; I2 ¼ 96%). Second, as mentioned before, publication bias might be present. Third, the statistical heterogeneity was high in this meta-analysis. We suspect that the difference in CAD definition was the main source of this heterogeneity as suggested by sensitivity analysis described above.
Conclusion In contrast to other rheumatic inflammatory disorders, a significant increased CAD risk among patients with GCA was not observed in this meta-analysis. The reason for this negative finding remains unclear but could possibly be related to the lower baseline cardiovascular risk and the older age at diagnosis.
Appendix A. Supporting Information Supplementary material cited in this article is available online at http://dx.doi.org/10.1016/j.semarthrit.2014.10.010.
Reference [1] Frostegard J. Atherosclerosis in patients with autoimmune disorders. Arterioscler Thromb Vasc Biol 2005;25:1776–85. [2] Libby P. Inflammation in atherosclerosis. Nature 2002;420:868–74.
[3] Montecucco F, Mach F. Common inflammatory mediators orchestrate pathophysiological processes in rheumatoid arthritis and atherosclerosis. Rheumatology (Oxford) 2009;48:11–22. [4] Rho YH, Chung CP, Oeser A, Solus JF, Gebretsadik T, Shintani A, et al. Interaction between oxidative stress and high-density lipoprotein cholesterol is associated with severity of coronary artery calcification in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62:1473–80. [5] Niessner A, Sato K, Chaikof EL, Colmegna I, Goronzy JJ, Weyand CM. Pathogensensing plasmacytoid dendritic cells stimulate cytotoxic T-cell function in atherosclerotic plaque through interferon-alpha. Circulation 2006;114: 2482–9. [6] Hansen S, Lassig G, Pistrosch E, Passauer J. Endothelial dysfunction in young patients with long-term rheumatoid arthritis and low disease activity. Atherosclerosis 2003;170:177–80. [7] Xu J, Lupu F, Esmon CT. Inflammation, innate immunity and blood coagulation. Hamostaseologie 2010;30:8–9. [8] Esmon CT. The interactions between inflammation and coagulation. Br J Haematol 2005;131:417–30. [9] Maradit-Kremers H, Nicola PJ, Crowson CS, Ballman KV, Gabreil SE. Cardiovascular death in rheumatoid arthritis: a population-based study. Arthritis Rheum 2005;52:722–32. [10] Bessant R, Hingorani, Patel L, MacGregor A, Isenberg DA, Rahman A. Risk of coronary heart disease and stroke in a large British cohort of patients with systemic lupus erythematosus. Rheumatology (Oxford) 2004;43:924–9. [11] Ungprasert P, Charoenpong P, Ratanasrimetha P, Thongprayoon C, Cheungpasitporn W, Suksaranjit P. Risk of coronary artery disease in patients with systemic sclerosis: a systematic review and meta-analysis. Clin Rheumatol 2014;33:1099–104. [12] Ungprasert P, Suksaranjit P, Spanuchart I, Leeaphorn N, Permpalung N. Risk of coronary artery disease in patients with idiopathic inflammatory myopathies: a systematic review and meta-analysis of observational studies. Semin Arthritis Rheum 2014;44:63–7. [13] Salvarani C, Cantini F, Hunder GG. Polymyalgia rheumatica and giant-cell arteritis. Lancet 2008;372:234–45. [14] Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz MJ, Miranda-Filloy JA, Gonzalez-Juanatey C, Martin J, et al. Epidemiology of giant cell arteritis and polymyalgia rheumatica. Arthritis Rheum 2009;61:1454–61. [15] Ray JG, Mamdani MM, Geerts WH. Giant cell arteritis and cardiovascular disease in older adults. Heart 2005;91:324–8. [16] Le Page L, Duhaut P, Seydoux D, Bosshard S, Ecochard R, Abbas F, et al. Incidence of cardiovascular events in giant cell arteritis: preliminary results of a prospective double cohort study (GRACG). Rev Med Interne 2006;27: 98–105. [17] Molloy ES, Kirchner HL, Murray PK, Hoffman GS, Langford CA. Cardio-vascular outcomes in giant cell arteritis. Presented at the American College of Rheumatology Annual Scientific Meeting, 6–11 November 2007, Boston, Massachusetts. [Abstract no. 1235]. [18] Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603–5. [19] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trial 1986;7:177–88. [20] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J 2003;327:557–60. [21] Dunstan E, Lester SL, Rischmueller M, Dodd T, Black R, Ahern M, et al. Epidemiology of biopsy-proven giant cell arteritis in South Australia. Intern Med J 2014;44:32–9. [22] García-Martínez A, Arguis P, Prieto-González S, Espígol-Frigolé G, Alba MA, Butjosa M, et al. Prospective long term follow-up of a cohort of patients with giant cell arteritis screened for aortic structural damage (aneurysm or dilatation). Ann Rheum Dis 2014;73:1826–32. [23] Souza AW, Okamoto KY, Abrantes F, Schau B, Bacchiega AB, et al. Giant cell arteritis: a multicenter observational study in Brazil. Clinics 2013;68: 317–22. [24] Kisza K, Murchison AP, Dai Y, Bilyk JR, Eagle RC Jr, Sergott R, et al. Giant cell arteritis incidence: analysis by season and year in mid-Atlantic United States. Clin Experiment Ophthalmol 2013;41:577–81. [25] Espitia O, Néel A, Leux C, Connault J, Espitia-Thibault A, Ponge T, et al. Giant cell arteritis with or without aortitis at diagnosis. A retrospective study of 22 patients with longterm followup. J Rheumatol 2012;39:2157–62. [26] Abdul-Rahman AM, Molteno AC, Bevin TH. The epidemiology of giant cell arteritis in Otago, New Zealand: a 9-year analysis. N Z Med J 2011;124:44–52. [27] Singh S, Balakrishnan C, Mangat G, Samant R, Bambani M, Kalke S, et al. Giant cell arteritis in Mumbai. J Assoc Physicians India 2010;58:372–4. [28] Aziz S, Al-Ansari A, Bankart J, McFadzean R. Clinical manifestations and laboratory tests in biopsy proven giant cell arteritis in Glasgow. Eur J Intern Med 2009;20:e146. [29] Gonzalez-Gay MA, Rubiera G, Pineiro A, Garcia-Porrua C, Pego-Reigosa R, Gonzalez-Juanatey C, et al. Ischemic heart disease in patients from northwest Spain with biopsy proven giant cell arteritis. A population based study. J Rheum 2005;32:502–6. [30] Uddhammar A, Eriksson AL, Nystrom L, Stenling R, Rantapaa-Dahlqvist S. Increased mortality due to cardiovascular disease in patients with giant cell arteritis in northern Sweden. J Rheum 2002;29:737–42. [31] Machado EBV, Gabriel SE, Beard CM, Michet CJ, O'Fallon WM, Ballard DJ. A population-based case-control study of temporal arteritis: evidence for an
P. Ungprasert et al. / Seminars in Arthritis and Rheumatism 44 (2015) 586–591
[32]
[33]
[34]
[35]
association between temporal arteritis and degenerative vascular disease. Int J Epidemiol 1989;18:836–41. Amiri N, Choi HK, Sayre EC, Shojania K, Avina-Zubieta JA. Risk of myocardial infarction in patients with giant cell arteritis: a populationbased study. Presented at the American College of Rheumatology Annual Scientific Meeting, 25–30 October 2013, San Diego, California. [Abstract no. 1673]. Tomasson G, Peloquin C, Mohammad A, Love TJ, Zhang Y, Choi HK. Risk for cardiovascular disease early and late after a diagnosis of giant-cell arteritis. A cohort study. Ann Intern Med 2014;160:73–80. Udayakumar PD, Chandran AK, Crowson CS, Warrington KJ, Matteson EL. Cardiovascular risk and acute coronary syndrome in giant cell arteritis: a population based retrospective cohort study. Arthritis Care Res (Hoboken) 2014. http://dx.doi.org/10.1002/acr.22416 [Epub ahead of print]. Matthews JL, Gilbert DN, Farris BK, Siatkowski RM. Prevalence of diabetes mellitus in biopsy-positive giant cell arteritis. J Neuroophthalmol 2012;32: 202–6.
591
[36] Schmidt J, Kermani TA, Muratore F, Crowson CS, Matteson EL, Warrington KJ. Statin use in giant cell arteritis: a retrospective study. J Rheumatol 2013;40:910–5. [37] Gonzalez-Gay MA, Pineiro A, Gomez-Gigirey A, Garcia-Porrua C, Pego-Reigosa R, Dierssen-Sotos T, et al. Influence of traditional risk factors of atherosclerosis in the development of severe ischemic complications in giant cell arteritis. Medicine 2004;83:342–7. [38] Gonzalez-Juanatey C, Lopez-Diaz MJ, Martin J, Llorca J, Gonzalez-Gay MA. Atherosclerosis in patients with biopsy-proven giant cell arteritis. Arthritis Rheum 2007;57:1481–6. [39] Ly K-H, Regent A, Tamby MC, Mouthon L. Pathogenesis of giant cell arteritis: more than just an inflammatory condition? Autoimmun Rev 2010;9:635–45. [40] Casey J, Sturm C Jr. Impaired response of lymphocytes from non-insulin dependent diabetics to staphage lysate and tetanus antigen. J Clin Microbiol 1982;15:109–14. [41] Roman MJ, Shanker BA, Davis A, Lockshin MD, Sammaritano L, Simantov R, et al. Prevalence and correlates of accelerated atherosclerosis in systemic lupus erythematosus. N Engl J Med 2003;349:2399–406.
