Objectives To evaluate coronary atherosclerosis in patients with psoriatic arthritis (PsA) and control subjects using coronary CT angiography (CCTA).
Methods Ninety consecutive patients with PsA (male: 56(62.2%); 50.3±11.1 years) were recruited. 240 controls (male: 137(57.1%); 49.6±10.7 years) without known cardiovascular (CV) diseases who underwent CCTA due to chest pain and/or multiple CV risk factors were recruited for comparison.
Results Patients with PsA and controls were matched in age, gender and traditional CV risk factors (all p>0.2). The prevalence of overall plaque (54(60%)/84(35%), p<0.001), calcified plaque (CP) (29(32%)/40(17%), p=0.002), mixed plaque (MP) (20(22%)/18(8%), p<0.001), non-calcified plaque (NCP) (39(43%)/53(22%), p<0.001) and combined MP/NCP (46(51%)/62(26%), p<0.001) were all significantly higher in patients with PsA. Three-vessel disease was diagnosed in 12(13%) patients with PsA and 7(3%) controls (p<0.001), while obstructive plaques (>50% stenosis) were observed in 8(9%) patients with PsA and 7(3%) controls (p=0.033). After adjusting for traditional CV risk factors, PsA remained an independent explanatory variable for all types of coronary plaques (OR: 2.730 to 4.064, all p<0.001). PsA was also an independent explanatory variable for three-vessel disease (OR: 10.798, p<0.001) and obstructive plaque (3.939, p=0.024). In patients with PsA, disease duration was the only disease-specific characteristic associated with more vulnerable plaques (MP/NCP) in multivariate analysis (1.063, p=0.031). The other independent explanatory variables were age ≥55 years (5.636, p=0.005) and male gender (8.197, p=0.001).
Conclusions Patients with PsA have increased prevalence, burden and severity of coronary atherosclerosis as documented by CCTA. Longer disease duration was independently associated with the presence of vulnerable MP/NCP plaques in patients with PsA.
Trial registration number NCT02232321.
- Psoriatic Arthritis
- Cardiovascular Disease
Statistics from Altmetric.com
Psoriatic arthritis (PsA) is a chronic inflammatory arthritis associated with increased prevalence of cardiovascular (CV) diseases and related mortality.1–3 A recent meta-analysis revealed that the CV morbidity was increased by 43% in patients with PsA compared with the general population.4 The increased risk may be due to an increased prevalence of traditional CV risk factors3 ,5 as well as inflammation. Uncontrolled low-grade inflammation was associated with a 24% increased risk of major adverse CV events (MACE) when a disease-modifying antirheumatic drug (DMARD) was not prescribed.1 Medications such as non-steroidal anti-inflammatory drugs and glucocorticoid may also contribute towards the increased CV risk,6 although conventional synthetic and biological DMARDs may have protective effect.7 ,8 Early identification of patients with high CV risk is important as early aggressive interventions may be beneficial.
Subclinical atherosclerosis is a good surrogate marker for CV disease (CVD) in the general population9 ,10 and patients with rheumatoid arthritis (RA).11–13 There is also an increased prevalence of subclinical atherosclerosis in patients with PsA.14 ,15 A recent meta-analysis confirmed an increased carotid intima-media thickness (IMT) (mean difference: 0.07 mm, p<0.001) and a higher frequency of carotid plaques (OR: 3.12, p=0.04).15 Carotid atherosclerosis has been shown to improve risk prediction for incident CVD over what was achieved by a model with Framingham risk factors alone in the general population,16 as well as improving CV risk stratification in patients with PsA.14 ,17 Nevertheless, previous knowledge on atherosclerosis in patients with PsA was limited to the carotid arteries. Detailed assessment of coronary atherosclerosis may further improve the risk stratification, and may identify patients who need immediate aggressive intervention, such as percutaneous coronary intervention.
Coronary CT angiography (CCTA) accurately evaluates coronary atherosclerosis. The accuracy approaches 100% against conventional angiography and intravascular ultrasound in prospective assessments,18–20 with high reproducibility.21 The severity and burden of coronary plaques can also be evaluated. CCTA independently predicts short and long-term cardiac events in patients with suspected or established coronary arterial disease (CAD).22–24 Moreover, CCTA detected plaque composition can further discriminate the CV risk. Non-calcified plaque (NCP) and mixed plaque (MP) are considered as more vulnerable and carry higher risk compared with calcified plaque (CP).25 ,26 Karpouzas et al 27 found that all types of coronary plaque were more common in patients with RA compared with controls; RA disease activity was associated with high-risk NCP and MP. Since patients with PsA has an increased CV risk which is comparable to patients with RA,28 we hypothesised that coronary plaques will be more common in patients with PsA compared with controls as a result of chronic inflammation.
In this study, we compared the presence, burden and severity of coronary plaques in patients with PsA with age/gender-matched control subjects. We also explored the correlation between PsA disease-specific features and the high-risk NCP and MP in these patients.
Patients and methods
One hundred and thirteen consecutive patients with PsA attending the outpatient clinic of the Prince of Wales Hospital (PWH) who fulfilled the Classification of Psoriatic Arthritis criteria29 were invited for CCTA from June 2014 to June 2016. Patients were ineligible if they (1) had a history of overt CVD (myocardial infarction, percutaneous transluminal coronary angioplasty, surgery for ischaemic heart disease, stroke, transient ischaemic attack, carotid endarterectomy, peripheral arterial reconstructive surgery or limb amputation); (2) were on statins; (3) had contradiction of CCTA such as pregnancy, known iodine allergy, impaired renal function. Twenty patients refused to participate. CCTA scan was not able to perform in three patients due to persistent high heart rate. Ninety patients with PsA were finally included.
Two hundred and forty age-matched and gender-matched control subjects were retrospectively selected from the database of patients who received CCTA assessment from the Department of Radiology at the PWH from January 2010 to March 2016. These subjects were referred for CCTA due to chest pain or discomfort, presence of ≥1 CV risk factors and/or ECG abnormalities. Exclusion criteria comprised (1) any concomitant autoimmune syndromes, (2) established CVD prior to the CCTA and (3) use of statins. The selection of control subjects was performed blinded to the CCTA results.
The study was approved by the Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee (the Joint CUHK-NTEC CREC). Written informed consent was obtained from all patients with PsA according to the Declaration of Helsinki and International Conference on Harmonisation Guideline for good clinical practice (ICH-GCP) guidelines. Due to the retrospective nature, waiver of informed consent from control subjects was approved by the Joint CUHK-NTEC CREC.
For patients with PsA, pain, physicians' and patients' global assessments were evaluated using a 100-point visual analogue scale, where 0 indicated excellent well-being and 100 indicated feeling extremely unwell. Physical examination included the number of tender and swollen joints using the 68 tender/66 swollen joint count, and the number of permanently deformed joints. The Health Assessment Questionnaire was used to evaluate physical function, and the Psoriasis Area and Severity Index was used to assess the extent of skin involvement.30 Overall disease activity was assessed using minimal disease activity (MDA)31 and Disease Activity in Psoriatic Arthritis (DAPSA).32 Anthropomorphic measurements including heights, weight and two consecutive blood pressure readings in sitting position were recorded. Body mass index (BMI) was calculated. Other data obtained from patients with PsA through the interview and chart review included smoking habits, history of diabetes, hypertension and hyperlipidaemia. Drug history was retrieved from case notes or elicited during the clinical assessment. All patients were interviewed and examined using standardised data collection instruments. Complete blood count, liver and renal function tests, erythrocyte sedimentation rate, C reactive protein, fasting blood glucose and lipid profile (total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides) were checked before CCTA. The mean time between lab test and CCTA was 150.7±133.5 days.
For control subjects, traditional CV risk factors, including age, gender, history of diabetes, hypertension, hyperlipidaemia and smoking habit were retrieved through chart review via the Clinical Management System of the Hong Kong Hospital Authority. Results of laboratory tests which dated closest to the CCTA assessment and within 6 months were recorded. All results included in the analysis were retrieved in at least 90% (216) of the control subjects.
Coronary atherosclerosis assessment
CCTA scans were performed with a 64-row multidetector CT (LightSpeed VCT XT, GE Healthcare, Milwaukee, Wisconsin, USA) in accordance with the protocol employed in the ACCURACY trial,33 and were analysed by an experienced radiologist (K-TW). Coronary artery calcium score (CAC) was quantified by the Agatston method.34 The presence, site and stenosis level of plaques were recorded. Coronary arteries were standardised to American Heart Association 15-segment model.35 Segment involvement score (SIS) represented the total number of segments harbouring plaque. Lesions rendering over 50% stenosis of the lumen were considered as obstructive. For multiple plaques, the most stenotic one was recorded. The types of plaque were determined to be non-calcified (density lower than the contrast enhanced lumen), calcified (density greater than the contrast enhanced lumen) and mixed (presence of both calcified and non-calcified component in the same plaque).27
Results are expressed as count (percentage), mean±SD or median (IQR) as appropriate. Comparisons between two groups were assessed using Student's t-test or Mann-Whitney U test for continuous variables and χ2 test for categorical variables. Univariate and multivariate logistic regression analysis was used to identify the independent explanatory variables for the presence of various kinds of coronary plaques in patients with PsA and control subjects, or the presence of MP/NCP in patients with PsA. Age ≥55 years, gender, history of hypertension, diabetes, hyperlipidaemia and smoking were included in the multivariate analysis as potential confounding factors.27 All statistical analyses were conducted using IBM SPSS Statistics V.22 (IBM, Armonk, New York, New York, USA). A minimal level of significance of p<0.05 is used.
Clinical characteristics of patients with PsA and control subjects
Patients with PsA and control subjects were well matched in age (50.3±11.1 vs 49.6±10.7 years, p=0.558) and gender (male: 56 (62.2%) vs 137 (57.1%), p=0.399). Traditional CV risk factors, including history of diabetes, hypertension, hyperlipidaemia and smoking habit were also matched (all p>0.2; table 1). The prevalence of subjects having three concomitant CV risk factors were higher in patients with PsA compared with controls, although the prevalence was low in both groups (7 (7.8%) vs 6 (2.5%), p=0.028). None of the patients with PsA and controls was on antiplatelet agents. In the patients with PsA, the overall disease activity was low to moderate (DAPSA: 13 (9–21)), with 24 (26.7%) patients achieving MDA. Other clinical characteristics of the patients with PsA were shown in table 1.
Two hundred and five (85.4%) control subjects received CCTA assessment due to chest pain or discomfort. Eight (3.3%) were due to other symptoms such as shortness of breath, syncope, palpitation or dizziness. Fifteen (6.3%) and 8 (3.3%) patients were due to ECG abnormalities and multiple CV risk factors, respectively. In contrast, only 8 (8.9%) patients with PsA reported history of occasional chest discomfort.
Coronary atherosclerosis in patients with PsA and control subjects
Coronary plaques were identified in 54 (60%) patients with PsA, which were significantly more common than in control subjects (84 (35.0%), p<0.001; figure 1A). The prevalence of all types of plaques was increased from twofold to threefold (figure 1B–E). High-risk MP/NCPs were identified in 46 (51.1%) patients with PsA and 62 (25.8%) control subjects (p<0.001). Thirty-six (40.0%) patients with PsA and 50 (20.8%) control subjects had a CAC >0 (p<0.001; figure 2A); in this subgroup of patients, CAC was significantly higher in patients with PsA (50 (11–121) vs 25 (5–47), p=0.029; figure 2B). Patients with PsA also had significantly higher plaque burden and severity in terms of more three-vessel diseases (12 (13.3%) vs 7 (2.9%), p<0.001) and obstructive lesions (8 (8.9%) vs 7 (2.9%), p=0.033) compared with control subjects (figure 3). SIS was 1.83±2.33 in patients with PsA and 0.66±1.21 in control subjects (p<0.001).
In the multivariate analysis, after adjusting for the traditional CV risk factors, patients with PsA still had increased risk of overall plaque and all types of plaques by 2.7–4.1-folds (table 2). The OR for patients with PsA to have high-risk MP/NCP was 3.244 (95% CI 1.909 to 5.511, p<0.001). PsA was also an independent explanatory variable for CAC >0, the presence of three-vessel diseases and obstructive lesions (table 2).
Disease-specific features and high-risk MP/NCP in patients with PsA
In the univariate analysis, none of the PsA-specific features was significantly associated with increased risk of MP/NCP (table 3). However, after adjusting for the traditional CV risk factors, longer PsA disease duration was independently associated with increased risk of MP/NCP (OR for per 1 year increase: 1.063 (1.006 to 1.123), p=0.031, table 3). Other independent risk factors were as follows: age ≥55 years (OR: 5.636 (1.691 to 18.785), p=0.005) and male gender (OR: 8.197 (2.356 to 28.514), p=0.001). Other PsA features were not associated with MP/NCP.
This is the first study to thoroughly evaluate coronary atherosclerosis in patients with PsA compared with control subjects. Patients with PsA without known CAD had a threefold to fourfold increased prevalence of all types of coronary plaque, which can probably account for the 68% increased risk for myocardial infarction.4 The plaque burden and severity were also higher in patients with PsA. These differences were independent of traditional CV risk factors, supporting the hypothesis that PsA itself is an independent risk factor. In patients with PsA, besides age and gender, cumulative inflammatory burden as reflected by the PsA disease duration was independently associated with the presence of high-risk MP/NCP plaque.
Both non-obstructive and obstructive coronary plaques were associated with increased mortality.24 In our study, 60% of patients with PsA had at least one coronary plaque, comparing with only 35% of control subjects. The prevalence of CP (32% vs 17%), NCP (43% vs 22%) and MP/NCP (51% vs 26%) was doubled in PsA, while the prevalence of MP was almost tripled (22% vs 8%). CAC was also higher in patients with PsA. These results were consistent with our previous finding of a higher prevalence of carotid atherosclerosis in patients with PsA,14 as well as coronary atherosclerosis in other chronic inflammatory diseases such as RA.27 In the RA study, the prevalence of overall plaque, CP, MP and NCP were 71%, 22%, 27% and 57% for patients with RA and 45%, 11%, 11% and 38% for controls, respectively.27 The prevalence of various types of plaques was lower in both the patients with PsA and control subjects in the current study, probably due to ethnic differences.36 ,37 Nevertheless, the adjusted OR of coronary atherosclerosis for PsA was even higher than that for RA. These data supported the previous notion that the increased CV risk in PsA was at least similar to that of RA.28
Similar to the findings in patients with RA, more patients with PsA compared with controls had three-vessel disease (13% vs 3%), obstructive lesions (9% vs 3%) and a higher SIS, which indicated higher coronary plaque burden and severity. In the CONFIRM study which followed over 20 000 subjects for 2.3 years, obstructive plaques were associated with an even higher mortality (CV risk factors adjusted HR compared with subjects without plaque: 2.60) than non-obstructive plaques (HR compared with subjects without plaque: 1.60), and each unit increase in SIS was associated with an increased adjusted risk of mortality by 10%.24 In our study, PsA is the only independent explanatory variable for obstructive lesions, and one of the only two independent explanatory variables for three-vessel disease (table 2). These data highlighted the fact that most of these PsA subjects with plaques are having silent ischaemia; therefore, more aggressive CV evaluation strategy should be considered in patients with PsA.
NCP and MP were vulnerable plaques which carried a higher risk than CP,25 ,26 with HRs for NCP and MP in predicting MACE being 5.30 and 9.54, respectively, compared with CP.25 In another study which included 1102 patients with non-obstructive plaques only, the HRs for all-cause mortality were 7.4 for NCP and 3.2 for MP, compared with CP.26 Patients with PsA had a threefold increased risk of MP/NCP compared with control subjects (table 2), highlighting the role of inflammation in accelerating coronary atherosclerosis. To further explore this issue, we investigated the correlation between disease-specific characteristics and MP/NCP in patients with PsA, and found that the PsA disease duration was an independent explanatory variable for MP/NCP. Increase in disease duration for 1 year was associated with a 6% increased risk in having MP/NCP, supporting the hypothesis that higher cumulative inflammatory burden (as reflected by the longer disease duration) promotes the development of vulnerable plaques. Inflammation initiates the formation of the atherosclerotic plaque and involves in the thinning of the fibrous cap, which may lead to plaque rupture and acute infarction of coronary arteries.38 This may explain the correlation between PsA disease duration and the vulnerable plaque type (MP/NCP). The precise mechanism of how chronic inflammation promotes atherosclerosis remains unresolved. It is possible that chronic inflammation induces the expression of adhesion molecules (intercellular adhesion molecule-1, vascular cell adhesion molecule-1, E-selectin, etc]); promotes endothelial dysfunction; stimulates smooth cell survival, migration and proliferation in the arterial intima; and subsequently leads to atherosclerotic plaque formation.8 ,38 Proinflammatory cytokines may also promote the release of matrix metalloproteinases, which alters the balance of elastin/collagen and increases the plaque vulnerability by plaque remodelling and thinning of the fibrous cap.8 ,38 Our results further highlighted the importance of cumulative inflammatory burden and coronary atherosclerosis. In patients with PsA with longer disease duration, a more comprehensive CV risk surveillance strategy may be needed. Further large-scale studies should be conducted to determine the most appropriate and cost-effective strategy for these patients.
It was considered unethical to subject healthy controls to CCTA due to the risk of radiation exposure and therefore subjects without prior CVD who were indicated for CCTA were recruited as controls instead. We would need to emphasise that the control subjects in our study were not healthy controls, but symptomatic patients without known CAD who underwent clinically indicated CCTA. Therefore, the risk of coronary atherosclerosis in patients with PsA could be even higher when compared with healthy controls. Subjects (for both patients with PsA and controls) on statins were excluded as statins have been proven to be effective in preventing atherosclerosis.39
The strength of our study was that we compared the prevalence, burden and severity of coronary atherosclerosis in patients with PsA and control subjects for the first time and in a large cohort. The patients and controls were well matched in age, gender and CV risk profiles. Based on our case number, the statistical power was estimated to be 0.94 for comparing overall plaque and 0.96 for comparing MP/NCP, with α=0.01. Our study also has a few limitations. First, the control subjects were retrospectively selected; this may affect the repeatability of the results. In addition, some of the clinical information, such as BMI, was missing. Ideally, blood samples from the controls should be checked for the inflammatory markers for more comprehensive analysis. Nevertheless, to ensure that the controls were appropriated recruited, CCTA results were blinded to the investigators during case selection. Second, due to the cross-sectional design, we were not able to evaluate the correlation between the CCTA identified atherosclerosis and CV events. Single determination of inflammatory markers and disease activity assessment may not accurately represent concentration over time and cumulative burden of exposure. In addition, we were unable to precisely quantify lifetime dosages of medications to examine more fully the effect of pharmacologic therapy on the development of atherosclerosis. Third, as patients with PsA suffer from more severe subclinical carotid atherosclerosis compared with patients with psoriasis (PsO),40 it would be interesting to include patients with PsO to address whether this is also true for coronary atherosclerosis. Fourth, disease activity in these patients with PsA was low to moderate. Therefore, overall plaque occurrence and burden in our population may not be generalisable to patients with PsA with higher disease activity. Last but not least, the prevalence of coronary plaque was much higher than the carotid plaque which we previously reported (18%).14 In a substudy (manuscript under preparation), results of the carotid ultrasound were compared with CCTA in all these patients. No significant association between carotid plaque and coronary plaque (p=0.191) were found. However, mean and maximum IMT were significantly increased in patients with coronary plaque (p=0.013 and p=0.026, respectively). Mean IMT was also significantly increased in patients with multivessel disease (p=0.02) and those with obstructive stenosis (p=0.017). These findings support the concept that atherosclerosis is a systemic process. However, the lack of association between carotid and coronary plaque indicates that atherosclerosis may have a heterogeneous distribution, which may be affected by the risk factor profile.
In conclusion, patients with PsA have increased prevalence, burden and severity of coronary atherosclerosis documented by CCTA. Longer disease duration is independently correlated with the presence of vulnerable MP/NCP plaque in patients with PsA.
This study was supported by the Health and Medical Research Fund (No 01120496). We acknowledge the brilliant medical students for their great assistance in this project: Kit Li, Ben M Cheung; Diana U Ming; Eric T Lui; Vanessa W Tao; Zoe S Tsang.
Handling editor Tore K Kvien
Contributors L-ST 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. LST, KTW and JS: study design. L-ST, K-TW, ITC, QS, EKL, PW, EWK, MYL, YR, IY, SY, ML, TKL, C-KW, TYZ, MC and AP-WL: acquisition of data. L-ST, JS, ITC and JJ-WL: analysis and interpretation of data. JS and L-ST: manuscript preparation.
Funding Health and Medical Research Fund (01120496).
Competing interests None declared.
Ethics approval Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee (the Joint CUHK-NTEC CREC).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.