Cardiovascular risk markers in patients with psoriatic arthritis: A meta-analysis of literature studies.

INTRODUCTION
Several studies reported an increased cardiovascular (CV) morbidity and mortality in patients with psoriatic arthritis (PsA). We performed a meta-analysis on the impact of PsA on major markers of CV risk.


METHODS
Studies on the relationship between PsA and common carotid artery intima-media thickness (CCA-IMT), prevalence of carotid plaques, flow-mediated dilation (FMD), nitrate-mediated dilation (NMD), pulse wave velocity (PWV), augmentation index (AIx), and ankle-brachial index (ABI) were systematically searched in the PubMed, Web of Science, Scopus, and EMBASE databases.


RESULTS
Sixteen case-control studies (898 cases, 1,140 controls) were included. Compared to controls, PsA patients showed a higher CCA-IMT (MD 0.07 mm; 95% CI 0.04, 0.11; P < 0.0001), and a higher frequency of carotid plaques (OR 3.12; 95% CI 1.03, 9.39; P = 0.04). Moreover, a lower FMD was found in PsA subjects than in controls (MD -2.56%; 95% CI -4.17, -0.94; P = 0.002), with no differences in NMD (MD -0.40%; 95% CI -1.19, 0.39; P = 0.32). Because of the low number of studies, no meta- analytical evaluation was performed for PWV, AIx, and ABI. Despite heterogeneity among studies, PsA appears significantly associated with markers of subclinical atherosclerosis and CV risk.


DISCUSSION
These findings could help to establish more specific CV prevention strategies in this clinical setting.


Introduction
During recent years, a series of case-control studies reported accelerated atherosclerosis (18,19), impaired endothelial function (20,21), and increased arterial stiff ness (22,23) in patients with PsA. However, these data have been challenged in other studies (24,25), and no meta-analytical data providing an overall information about this issue are currently available.
In order to provide a comprehensive overview of the relationship between PsA and subclinical atherosclerosis, we performed a systematic review with meta-analysis of literature studies evaluating the impact of PsA on the major markers of CV risk.

Methods
A protocol for this review was prospectively developed, detailing the specifi c objectives, the criteria for study selection, the approach to assess study quality, the outcomes, and the statistical methods.

Search strategy
To identify all available studies, a detailed search pertaining to PsA and the markers of CV risk (i.e. IMT, FMD, NMD, PWV, AIx, and ABI) was conducted according to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines (26). A systematic search was performed in the electronic databases (PubMed, Web of Science, Scopus, EMBASE), using the following search terms in all possible combinations: psoriatic arthritis, intima-media thickness, carotid plaques, atherosclerosis, fl ow-mediated dilation, nitrate-mediated dilation, endotheliumdependent dilation, endothelium-independent dilation, endothelial dysfunction, pulse wave velocity, augmentation index, arterial stiff ness, ankle-brachial index. Th e last search was performed on 10 November 2014. Th e search strategy was developed without any language or publication year restriction.
In addition, the reference lists of all retrieved articles were manually reviewed. In case of missing data, study authors were contacted by e-mail to try to retrieve original data. Two of the authors (M.N.D.D.M. and P.A.) analyzed each article and performed the data extraction independently. In case of disagreement, a third investigator was consulted (R.L.). Discrepancies were resolved by consensus. Selection results showed a high inter-reader agreement ( κ ϭ 0.97) and have been reported according to the PRISMA fl ow chart (Supplementary Appendix 1, to be found online at http://informahealthcare.com/doi/abs/10.3109/07853890. 2015.1031822).

Data extraction and quality assessment
According to the pre-specifi ed protocol, all studies evaluating the impact of PsA on the markers of CV risk were included. Casereports, case-series without a control group, reviews, and animal studies were excluded. To be included in the analysis, a study had to provide values (means with standard deviation) of at least one variable among the following: common carotid artery IMT (CCA-IMT), brachial artery FMD or NMD, carotid-femoral or carotid-radial PWV, aortic AIx, and ABI. Studies reporting the prevalence of carotid plaques were also included.
In each study, data regarding sample size, major clinical and demographic variables, values of IMT, FMD, NMD, PWV, AIx, and ABI, and prevalence of carotid plaques in PsA patients and healthy controls were extracted.
Given the characteristics of the included studies, the evaluation of methodological quality of each study was performed with the Newcastle -Ottawa Scale (NOS), which is specifi cally developed to assess the quality of non-randomized observational studies (27). Th e scoring system encompasses three major domains (selection, comparability, exposure) and the resulting score ranges between 0 and 8, a higher score representing a better methodological quality. Results of the NOS quality assessment are reported in Supplementary Appendix 2 (to be found online at http://informahealthcare. com/doi/abs/10.3109/07853890.2015.1031822).

Statistical analysis and risk of bias assessment
Statistical analysis was carried out using Review Manager (Version 5.2, The Cochrane Collaboration, Copenhagen, Denmark) provided by Th e Cochrane Collaboration.
Diff erences among cases and controls were expressed as mean diff erence (MD) with pertinent 95% confi dence intervals (95% CI) for continuous variables, and as odds ratio (OR) with pertinent 95% CI for dichotomous variables.
IMT has been expressed in millimeters (mm), FMD, NMD, and AIx as percentage (%), PWV as mm per second (mm/s), and ABI as an absolute number.
Th e overall eff ect was tested using Z scores, and signifi cance was set at P Ͻ 0.05. Statistical heterogeneity between studies was assessed with chi-square, Cochran ' s Q test, and with I 2 statistic, which measures the inconsistency across study results and describes the proportion of total variation in study estimates that is due to heterogeneity rather than sampling error. In detail, an I 2 value of 0% indicates no heterogeneity, 25% low, 25% -50% moderate, and 50% high heterogeneity (28).
Publication bias was assessed by Egger ' s test and represented graphically by funnel plots of the standard diff erence in means versus the standard error. Visual inspection of funnel plot asymmetry was performed to address for possible small-study eff ect, and Egger ' s test was used to assess publication bias, over and above any subjective evaluation. P Ͻ 0.10 was considered statistically signifi cant (29). In order to be as conservative as possible, the random eff ect method was used for all analyses to take into account the variability among included studies.

Sensitivity analyses
We repeated analyses by including only the studies judged as ' high quality ' according to NOS (i.e. NOS Ն the median value found among included studies).
A further analysis was performed aft er excluding studies providing a composite IMT, calculated by averaging the CCA with the internal carotid artery (ICA) and/or the carotid bifurcation (BIF) measurements.

Meta-regression analyses
We hypothesized that diff erences among included studies may be aff ected by demographic variables (mean age, male gender) and clinical data related to disease activity (disease activity score in 28 joints (DAS28), Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Bath Ankylosing Spondylitis Functional Index (BASFI), C-reactive protein (CRP) levels, erythrocyte sedimentation rate (ESR), disease duration), anti-rheumatic treatment (therapy with non-steroidal anti-infl ammatory drugs (NSAIDs), sulfasalazine (SSZ), corticosteroids (CS), methotrexate (MTX), or TNF α -blockers), and coexistence of traditional CV risk factors (hypertension, smoking habit, diabetes mellitus, obesity, hyperlipidemia). To assess the possible eff ect of such variables in explaining diff erent results observed across studies, we planned to perform meta-regression analyses aft er implementing a regression model with changes in IMT, FMD, NMD, PWV, AIx, and ABI values, or presence of carotid plaques as dependent variables ( y ) and the above-mentioned covariates as independent variables ( x ). Th is analysis was performed with Comprehensive Meta-analysis (Version 2, Biostat, Englewood NJ, USA; 2005).

Results
Aft er excluding duplicate results, the search retrieved 786 articles. Of these studies, 414 were excluded because they were off the topic aft er scanning the title and/or the abstract, 352 because they were reviews/comments/case reports or they lacked of data of interest. For 1 study the on-line full-length version was not available, and another 3 studies were excluded aft er full-length paper evaluation.
Th us, 16 articles (on 898 PsA patients and 1,140 healthy controls) were included in the fi nal analysis (Supplementary Appendix 1, to be found online at http://informahealthcare.com/ doi/abs/10.3109/07853890.2015.1031822). In detail, 12 studies with data on CCA-IMT (13 data sets on 759 cases and 937 controls), 9 studies reporting on the prevalence of carotid plaques (10 data sets on 648 cases and 787 controls), 6 studies on FMD (7 data sets on 229 patients and 279 controls), and 4 on NMD (5 data sets on 173 cases and 192 controls) were included.
PWV and AIx have been evaluated in one study each (22,23), while no study tested ABI in PsA patients and healthy controls. Th us, these outcomes were not included in the meta-analytical evaluation.

Study characteristics
All included studies had a case-control design. Major characteristics of populations are shown in Table I, and further data about PsA 41  PsA PsA Th e NOS for quality assessment of included studies showed a median value of 7.
disease activity and ongoing anti-rheumatic treatment of PsA patients are reported in Table II.
Th e number of patients varied from 19 to 224, the mean age from 37.6 to 57.8 years, and the prevalence of male gender from 30% to 82.6%.
Mean values of DAS28 varied from 2.4 to 5.9, BASDAI from 4.6 to 35, CRP from 0.46 to 2.73 mg/dL, ESR from 15.1 to 38 mm/h, and disease duration from 3.4 to 12.1 years. An ongoing treatment with MTX was reported by 27.7% -86.4%, and with TNF α -blockers by 0% -69%.
One study (25) provided separate data for PsA patients and healthy controls stratifi ed for the presence of hypertension. Th e two groups were evaluated as two diff erent data sets.

Publication bias
Because it is recognized that publication bias can aff ect the results of meta-analyses, we attempted to assess this potential bias using funnel plots analysis.
Funnel plots of eff ect size versus standard error for studies evaluating CCA-IMT, FMD, and NMD were rather symmetrical, suggesting the absence of publication bias and of small-study eff ect (Supplementary Appendix 3, to be found online at http:// informahealthcare.com/doi/abs/10.3109/07853890.2015.10318 22), confi rmed by Egger ' s test ( P ϭ 0.45, P ϭ 0.11, and P ϭ 0.86, respectively).
In contrast, studies on the prevalence of carotid plaques showed an asymmetric distribution and were aff ected by publication bias (Egger ' s P ϭ 0.0009).
Similar results were confi rmed also aft er excluding 5 studies (19,23,25,31,36) providing a composite IMT, calculated by averaging the CCA-IMT with the internal carotid artery-IMT and/or the IMT at the level of the carotid bifurcation (Table IV).

Meta-regression analyses
Since meta-regression is eff ective when a covariate is reported by at least 10 studies (39) and several items of information were missing from each included study, a meta-regression approach could be performed only to test the impact of age and male gender on CCA-IMT, and these demographic covariates did not impact on the evaluated outcome ( P ϭ 0.39 and P ϭ 0.92, respectively).

Discussion
Results of the present meta-analysis consistently show that PsA is associated with subclinical atherosclerosis and endothelial dysfunction. In particular, we reported an increased carotid IMT with a high prevalence of carotid plaques and impaired FMD in patients with PsA. Our fi ndings are strengthened by sensitivity analyses. Moreover, regression models were able to refi ne the results further, providing the evidence that demographic variables (age and gender) did not impact on carotid IMT.
Overall, these data clearly show an increased CV risk in patients with PsA and suggest the need for a strict monitoring of CV risk factors and of signs of subclinical atherosclerosis in PsA patients. Accordingly, previous published studies reported an increased risk of major CV events and CV death in patients with PsA (7 -9) and other autoimmune (40) or rheumatic diseases (41).  no meta-analytical evaluation was possible for arterial stiff ness parameters and ABI, it is relevant that higher values of PWV and AIx have also been documented in PsA patients (22,23).
To strengthen our results, it would have been useful to assess also the prevalence of CV events among PsA subjects and healthy controls, but most included studies excluded patients with clinically proven coronary artery disease or history of myocardial infarction or cerebrovascular accidents. Th e clinical relevance of our results can be better understood when we consider that the risk of myocardial infarction increases by 43% every 0.163 mm increase in carotid IMT (51), and that the prevalence of carotid plaques is an even more reliable predictor of CV events than IMT (15,52).
In addition, our results on FMD further confi rm the presence of accelerated atherosclerosis in patients with PsA. Th e rationale for the association between FMD and CV prognosis is the assumption that it refl ects endothelium-dependent dilation and, in turn, NO bioavailability (53). Endothelium-derived NO possesses several anti-atherogenic and plaque-stabilizing properties, including inhibition of cell growth and proliferation, regulation of vascular tone and arterial wall stress, inhibition of leukocyte and platelet adhesion, and anti-thrombotic and fi brinolytic properties (54). In keeping with this, it has been documented that each 1% decrease in FMD is associated with a 12% increase of CV events (55).
In line with these fi ndings, the European League Against Rheumatism (EULAR) recently proposed the application of a 1.5 multiplier to the CV risk calculated in rheumatic patients through the scores currently used for the general population (e.g. the Framingham score) (56). However, this approach still requires a long-term validation.
Our results further support the need for large long-term interventional trials with CV end-points to investigate whether benefi ts in articular disease achieved by aggressive suppression of infl ammation may translate into reduced CV risk in PsA.
Some potential limitations of our study need to be discussed. First, studies included in our meta-analysis have diff erent inclusion and exclusion criteria, and most of the patients included in the analysis had concomitant CV risk factors (hypertension, smoking, obesity, diabetes mellitus, hyperlipidemia), diff erent disease activity status (DAS28, BASDAI, BASFI, CRP levels, ESR, disease duration), and diff erent anti-rheumatic treatment (NSAIDs, SSZ, CS, Many CV risk factors are thought to have a causal role in the atherosclerotic process (42). Although PsA patients exhibit an increased prevalence of these CV risk factors (6), the relationship between subclinical atherosclerosis and PsA seems to be more complex, and the presence of traditional risk factors might not entirely explain the accelerated atherosclerotic process in this clinical setting (43). Th us, other mechanisms (i.e. infl ammatory and immunological) have been proposed to explain the relationship between PsA and atherosclerosis (6).
Immune-mediated infl ammation seems to play a pivotal role in the pathogenesis of atherosclerosis, being involved in endothelial dysfunction, plaque rupture, and thrombosis (44). In keeping with this, some common markers of infl ammation (i.e. CRP, fi brinogen) are emerging predictors of CV disease (45,46), and patients with PsA exhibit elevated levels of these acute-phase proteins (47). ESR, another marker of infl ammation, is commonly found increased in PsA (33), and high ESR has been associated with increased overall mortality in this clinical setting (48). Th e strong correlation between these markers of infl ammation and those of platelet activation (CD62P, CD63) suggests that disease activity is involved in platelet hyperreactivity in rheumatic diseases (49). In line with these data, a direct correlation between infl ammatory status (as expressed by CRP levels) and increasing quartiles of maximal platelet aggregation has also been documented in PsA (50).
Overall, our fi ndings are in line with several experimental and clinical evidences, supporting the hypothesis that premature atherosclerosis may be one of the main features of PsA and that chronic infl ammation plays an important role in its pathogenesis, acting independently and/or synergistically with traditional CV risk factors.
In order to provide a comprehensive overview of the relationship between PsA and subclinical atherosclerosis, all the major recognized markers of CV risk were taken into account in the current meta-analysis. Th us, we reported an increased carotid IMT with a high prevalence of carotid plaques and impaired FMD in patients with PsA as compared with controls, but we did not fi nd a signifi cant diff erence in the NMD. However, when interpreting the latter result, we should consider the limited number of studies ( n ϭ 4) evaluating this outcome. Although  MTX, or TNF α -blockers). However, based on Cochrane ' s collaboration guidelines, a meta-regression approach is eff ective when a covariate is reported by at least 10 studies (39), and several items of information were missing from each included study. Th us, a metaregression approach could be performed only to test the impact of some demographic variables (age and gender) on CCA-IMT, and caution is necessary in overall results interpretation. Second, heterogeneity among the studies was generally signifi cant. However, all results have been confi rmed without heterogeneity aft er the exclusion of one study (24) which has been specifi cally designed to compare the eff ect of diff erent treatments (TNF α -blockers versus traditional disease modifying anti-rheumatic drugs) on the CCA-IMT. In addition, results were also confi rmed by sensitivity analyses. Th us, although we were not able conclusively to ascertain all potential sources of heterogeneity, we are confi dent that the impact on results can be considered minimal.
Th ird, we should consider that fi ve disease patterns are documented in PsA patients (distal joint disease, oligoarthritis, spondylarthropathy, polyarthritis, arthritis mutilans) but none of the included studies reported data on IMT, FMD, and NMD stratifi ed according to PsA clinical subset. Similarly, no information on the type of psoriasis could be extracted from the included studies. Th us, we could not evaluate how the clinical presentation of the disease may impact on our results.
Finally, caution is necessary in the interpretation of the results on FMD and on the prevalence of carotid plaques. While IMT is a somehow reproducible parameter, FMD measurement may be infl uenced by many confounding factors (57), signifi cantly limiting reproducibility of FMD assessment and, in turn, the relevance of results. In particular, some studies indicate that cuff placement (58) or duration of cuff occlusion (59) may aff ect results. Moreover, we have to consider that the defi nition of carotid plaques varied widely among studies. However, it is interesting to highlight that the eff ect of PsA was consistently confi rmed for all evaluated outcomes, strongly suggesting an increased CV risk in these patients.
In conclusion, in our meta-analysis PsA appeared signifi cantly associated with subclinical atherosclerosis and endothelial dysfunction and, in turn, with an increased CV risk. Th us, patients with PsA may benefi t from a periodic assessment of surrogate markers of CV risk. Th is could help to establish more specifi c CV prevention strategies for these patients. P ϭ 0.03 I 2 ϭ 82%; P Ͻ 0.0001 95% CI ϭ 95% confi dence intervals; CCA-IMT ϭ common carotid artery intima-media thickness; MD ϭ mean diff erence; n ϭ number.