Utility of Coronary Calcium Scoring (CCS) in Connective Tissue Disorders (CTDs) for the Evaluation of Subclinical Coronary Atherosclerosis - A Systematic Review.

OBJECTIVE: To assess the current state of knowledge for the utility of coronary calcium scoring (CCS) in connective tissue disorders (CTDs) as it relates to the presence and quantification of coronary atherosclerosis.
METHODS: Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, a literature search via PubMed, Embase, Scopus, Web of Science Core Collection, CINAHL, and Cochrane Database of Systematic Review retrieved 1019 studies (since database inception on May 7, 2018) from which 121 manuscripts were eligible for review. Inclusion criteria consisted of studies that investigated CCS in adults with respective CTDs. Studies were excluded if a complete manuscript was not written in English or was a case report.
RESULTS: Thirty-one studies were included (27 with healthy age-/gender-matched control group for comparison and 4 without). CTDs analyzed in articles with control group: 11 rheumatoid arthritis (RA), 14 systemic lupus erythematosus (SLE), 4 systemic sclerosis (SSc), 1 idiopathic inflammatory myopathies (IIM), 1 Takayasu arteritis, and 1 psoriasis. Nine out of 11 RA studies, 12 out of 14 SLE studies, and 2 out of 4 SSc studies showed statistically significant increased CCS when compared with the control group. CTDs analyzed in studies without control group: two Kawasaki disease, one juvenile idiopathic arthritis (JIA), and one antiphospholipid syndrome (APS) article, which demonstrated increased coronary arterial calcium burden, however, without statistically significant data.
CONCLUSION: CTDs, especially SLE and RA, are associated with higher CCS compared with the control group, indicating increased risk of coronary atherosclerosis. Our search did not elicit sufficient publications or statistically significant results in many other CTDs.

Introduction

The association between atherosclerosis and inflammation has been well established 1. Atherosclerosis has been attributed to oxidative injury to the endothelial walls from inflammatory cells 2, 3. Most connective tissue disorders (CTDs) are known chronic inflammatory disorders, of which some have been associated with cardiovascular disease (CVD). This CVD association has been mostly shown in systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA). One study demonstrated that the risk of CVD in those with SLE was reported to be more than two times compared with their control group 4. Another study that investigated the risk of CVD in a multitude of inflammatory disorders found the highest risk in those with RA and other CTDs 5. The pathophysiologic mechanisms of why this occurs in diseases such as RA and SLE is still not fully understood, though inflammation seems to play a central role. Given the established association of atherosclerosis with inflammation, all CTDs should be investigated for their possible contributions to CVD.

Traditional cardiovascular risk factors, such as age, race, systolic blood pressure, cholesterol, diabetes, smoking, and others, have been previously validated for their increased risk of CVD and have been used to predict the risk of future cardiovascular events 6. A growing number of studies suggest that inflammatory and autoimmune disorders have both a higher burden of atherosclerosis and a higher number of cardiovascular hard events 4, 5. Furthermore, there is a disproportionate rate of CVD seen in the younger population with CTDs who lack traditional cardiovascular risk factors that accrue with age 7. This could be explained by the ebb and flow of systemic inflammation, which has been well described as an underlying mechanism of atherosclerotic plaque 1. The traditional CVD risk stratification model might underestimate the true cardiovascular risks in CTDs, and new risk calculation models are needed.

Currently, there is no cardiac risk stratification model that includes proinflammatory CTDs as a risk factor for CVD. Coronary event risk calculators such as the American College of Cardiology/American Heart Association Atherosclerotic Cardiovascular Disease (ASCVD) risk calculator 6 do not include CTDs as a contributor to cardiovascular events.

The impact of CTDs on one's cardiovascular health is an important relationship to understand to be able to aid in its prevention. Surrogate markers for CVD, such as coronary calcium score (CCS), have been shown to be an effective tool to predict increased risk of coronary heart disease events 8, such as myocardial infarction. The utility of CCS in identifying the cardiovascular risk in patients with CTDs has not been studied over the spectrum of CTDs (such as vasculitis, myositis, and mixed CTDs).

The purpose of this systematic review is to evaluate the coronary atherosclerotic disease risk of all CTDs by computed tomography (CT) scan with coronary arterial calcification (CAC) or CCS.

Protocol and registration

This descriptive systematic review was registered with PROSPERO (registration number: CRD42019128607) and was performed following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) 9.

Search strategy

A structured search of published studies relating to coronary artery calcium scoring and CTDs was conducted in Medline via PubMed, Embase, Scopus, Web of Science Core Collection, CINAHL, and Cochrane Database of Systematic Review from database inception through May 7, 2018.

Search strategies were customized for each database and included appropriate controlled vocabulary terms and keywords related to CAC and CTDs. Full details of the strategies for each database are available (Appendix S1). The reference lists of all included studies were hand searched to identify any additional relevant publications.

The Medical Subject Headings (MeSH) used to identify CAC were the terms that were used to illustrate the CTDs and CT scan findings in the search concept (Table 1).

Table 1

Search terms for connective tissue disorders

The Four Domains of Connective Tissue Disorders

Rheumatoid arthritis Seronegative spondyloarthropathies Ankylosing spondylitis Reactive arthritis Enteropathic arthropathy OR spondylitis associated with inflammatory bowel disease Psoriatic arthritis Undifferentiated spondyloarthropathy Connective tissue diseases Systemic sclerosis Primary Sjogren's syndrome Systemic lupus erythematosus Antiphospholipid syndrome Relapsing polychondritis Idiopathic inflammatory myopathies: polymyositis, dermatomyositis, antisynthetase syndrome, inclusion‐body myositis, necrotizing autoimmune myopathy Mixed connective tissue disorder Undifferentiated connective tissue disease Vasculitis: Large vessel vasculitis: Takayasu arteritis Giant cell arteritis Medium vessel vasculitis: Polyarteritis nodosa Kawasaki disease Small‐vessel vasculitis Microscopic polyangiitis Granulomatosis with polyangiitis Eosinophilic granulomatosis with polyangiitis Variable‐vessel vasculitis Behcet's disease Cogan's syndrome Immune complex small‐vessel vasculitis Anti–glomerular basement membrane disease Cryoglobulinemic vasculitis IgA vasculitis (Henoch‐Schonlein) Hypocomplementemic urticarial vasculitis (anti‐C1q vasculitis)

Abbreviation: IgA, immunoglobulin A.

Eligibility criteria

Inclusion criteria consisted of studies that illustrated the CCS/CAC of the respective CTD in adults. Studies were excluded if they were a case report, not written in English, or were an abstract article without a full manuscript to evaluate. Each manuscript was evaluated to look for CAC detected from chest CT scans in each respective CTD (Figure 1).

Figure 1

Flow diagram ‐ search design.

Study selection, data extraction, and data items

Two reviewers evaluated the data and studies independently. A total of 1907 abstracts were initially obtained from our search, which was reduced to 1019 after duplicates were removed. The 1019 abstracts were analyzed for relevancy, which resulted in 898 abstracts being excluded because the title and/or abstract did not evaluate CCS/CAC of CTDs. In the end, 121 full manuscripts remained, which were analyzed for mean, median, and/or prevalence/incidence of CCS/CAC in each respective CTD as either a primary outcome of the study or as a secondary outcome. CCS/CAC was measured by CT scan of the chest without intravascular contrast and interpreted by a reviewer for CCS or CAC incidence/prevalence. All CT image modalities, including multidetector row CT (MDCT) or electron beam tomography (EBT) were included. The units used to describe the CCS was described by Agatston et al 10 and are reported in this review as Agatston units. Studies were also investigated for variables such as age, gender, ethnicity, and cardiovascular risk factors, such as hypertension, diabetes, smoking, and hyperlipidemia.

This search strategy resulted in 33 studies that were eligible for full manuscript review (Figure 1). Of these, 27 studies compared the CCS/CAC of the respective CTD to a control group, whereas 6 studies did not have a control group for comparison. Of these six studies, two were excluded (one RA, one SLE) because of the sufficient number of higher‐quality RA and SLE studies that had a control group for comparison. The remaining four studies (two Kawasaki disease, one juvenile idiopathic arthritis [JIA], and one antiphospholipid syndrome [APS]) were included in this review despite the absence of a control group for comparison because of the lack of better‐quality studies that investigated this association. This resulted in a total of 31 articles that were evaluated in this systematic review. Of these, the quantity and type of CTDs analyzed with a control group for comparison were as follows: 11 RA, 14 SLE, 4 systemic sclerosis (SSc), 1 idiopathic inflammatory myopathies (IIM), 1 Takayasu arteritis, and 1 psoriasis. The quantity and type of CTDs analyzed without a control group for comparison were as follows: 2 Kawasaki, 1 JIA, and 1 APS.

Risk of bias in individual studies

The quality of articles included in this review were assessed using the Newcastle‐Ottawa Scale (NOS) 11 assessment for cohort studies and the Agency for Healthcare Research and Quality (AHRQ) 12 criteria for case‐control and cross‐sectional studies. The NOS assessment assigns a maximum of nine points to each study. The assessment scale analyzes three broad perspectives of each study: the selection of the study groups, the comparability of the groups, and the ascertainment of either the exposure or outcome of interest for case‐control or cohort studies 11. When a study receives more than the median number of stars, it is considered to be of good quality (or at low risk of bias); otherwise, it is deemed to be of low quality (or at high risk of bias). All NOS assessments for cohort studies in this review had a score greater than 4, indicating good quality (Appendix S2). The individual assessments using AHRQ did not have a score value but did assess important study qualities through a series of investigational questions to assess for possible bias. The AHRQ assessment of the included case‐control and cross‐sectional studies subjectively demonstrated that they are also of good or fair quality (Appendix S3).

Results

CCS/CAC of CTDs compared with healthy matched control group (Table 2)

Connective tissue diseases summary

A collective total 11 RA, 14 SLE, 4 SSc, 1 IIM, 1 Takayasu arteritis, and 1 psoriasis article was analyzed. Not all studies investigated CCS in the respective CTD as a primary outcome yet were able to be included as they met inclusion/exclusion criteria. Some studies investigated the CCS/CAC of multiple CTDs compared with a control group. MDCT or EBT was used in these studies to detect CAC/CCS. Some articles (see Refs. 16, 18, and 30) used participants from the Multi‐Ethnic Study of Atherosclerosis (MESA) 40 study, which evaluated CAC in different ages, sex, ethnicities, and cardiovascular risk factors to match for their control group. Table 2 outlines the details of each study, including the type of CTD, CCS/CAC findings, statistical significance, and sample size. Descriptive characteristics of each article are outlined in Appendix S4.

Table 2

Descriptive summary of CCS/CAC of CTDs with control group for comparison

Author, Year, & Reference	CTD/Sample Size	Mean, Median, or Incidence/Prevalence of CAC or CCS Results (Agatston units)	Statistical Significance
Abdel‐Khalek (2011) 13	RA – 60 Control – 20	RA mean CCS: 126 ± 115.23 Control mean CCS: 4.7 ± 4.03	P < 0.001
Asanuma (2007) 14	Early RA – 90 Established RA – 67 Control – 87	Early RA (< 6 years) median CCS: 0 (0‐47) Established RA (> 10 years) median CCS: 63 (0‐368) Control Median CCS: 0 (0‐18)	P < 0.001
Avalos (2007) 15	Early RA – 57 Late RA – 60 Control – 65	Early RA (< 6 years) – median CCS 0 (0‐33.8) Late RA (> 10 years) – median CCS 65.5 (0‐400.5) Controls – Median CCS 0 (0‐16.4)	P < 0.001
Chung (2013) 16	RA – 155 Control – 835a	Median CCS RA: 3.1 (0‐135.1) Median CCS control: 6.4 (0‐119.6)	NS
Chung (2005) 17	Early RA – 70 Established RA – 71 Control – 86	Early RA (< 5 years): median CCS 0 (0‐42.6), CAC in 42.9% Established RA (> 10 years): median CCS 40.2 (0‐358), CAC in 60.6% Control: median CCS 0 (0‐19.2), CAC in 38.4%	P = 0.001
Giles (2009) 18	RA – 195 Control – 1073a	RA mean CCS: 175 ± 31 Control mean CCS: 122 ± 13	P = 0.002
Kakuta (2016) 19	RA – 37 SSc – 24 SLE – 33 Control – 74	Median CCS RA: 0 (0‐136) Median CCS SSc: 0 (0‐111) Median CCS SLE: 0 (0‐138) Median CCS control: 30 (0‐225)	NS
Kao (2008) 20	SLE – 105 RA – 105 Control – 105	Prevalence of CAC: SLE: 47.6%; RA: 47.6%; Control: 35.2%	P = 0.02
Paccou (2014) 21	RA – 75 Control – 75	RA CAC prevalence: 65.3% Control CAC prevalence: 49.3%	P = 0.04
Wang (2009) 22	RA – 85 Control – 85	RA mean CCS: 62.8 ± 197.0 Control mean CCS: 11.3 ± 38.5	P = 0.002
Yiu (2012) 23	RA – 85 SLE – 69 Control – 106	RA and SLE mean CCS: 42.2 ± 154.3 Control mean CCS: 1.4 ± 13.0	P < 0.01
Asanuma (2003) 24	SLE – 65 Control – 69	SLE mean CCS: 68.9 ± 244.2 Control mean CCS: 8.8 ± 41.8	P = 0.002
Chung (2006) 25	SLE – 93 Control – 65	SLE CAC incidence and mean CCS: 19.4% and 39 ± 200 Control CAC incidence and mean CCS: 6.2% and 4 ± 30	P = 0.02
Chung (2008) 26	SLE – 113 Control – 80	SLE mean CCS: 43.4 ± 189.8 Control mean CCS: 3.8 ± 27.9	P = 0.002
Heshmat (2015) 27	SLE – 30 Control – 30	SLE mean CCS: 42 ± 111.09 Control mean CCS: 0, no CAC was detected	P = 0.04
Kiani (2015) 28	SLE – 80 Control – 241a	Age 45‐54 CAC prevalence: SLE = 58%; control = 22/125 (36%) Age 55‐64 CAC prevalence: SLE = 57%; control = 42/116 (36%)	Age 45‐54: P < 0.001 Age 55‐64: NS
Lertratanakul (2014) 29	SLE – 149 Control – 124	CAC was more prevalent in SLE patients and had significantly higher progression	NS
Othman (2013) 30	SLE – 60 Control – 60	SLE mean CCS: 59.2 ± 20.3 Control mean CCS: 2.6 ± 1.85	P < 0.001
Romero‐Diaz (2018) 31	SLE – 95 Control – 100	SLE – CAC incidence: 18% Control – CAC incidence: 7%	P = 0.03
Romero‐Diaz (2012) 32	SLE – 139 Control – 100	SLE – CAC incidence: 7.2% Control – CAC incidence: 1%	P = 0.02
Seyahi (2013) 33	Takayasu – 47 SLE – 43 Control – 70	Takayasu CAC incidence: 11% SLE CAC incidence: 21% Control CAC incidence: 3%	Takayasu: NS SLE: P = 0.010
Yiu (2009) 34	SLE – 50 Control – 50	SLE CAC prevalence: 42% Control CAC prevalence: 8%	P < 0.01
Khurma (2008) 35	SSc – 17 Control – 17	SSc mean CCS: 126.6 ± 251.0 Control mean CCS: 14.7 ± 52.2	P = 0.003
Mok (2011) 36	SSc – 53 Control – 106	SSc: 56.5% had CCS > 101 Control: 29.4% had CCS > 101	P = 0.01
Seung‐Geun (2013) 37	SSc – 41 Control – 123	SSc median CAC: 0 (0‐133.5) Control median CAC: 0 (0‐454.1)	NS
Diederichsen (2015) 38	IIM – 76 Control – 48	IIM: median CCS 18 (0 ‐ >400) Control: median CCS 5 (0 ‐ >400)	NS
Seremet (2014) 39	Psoriasis – 40 Control – 42	Psoriasis mean CCS: 9.9 ± 35.2 Control mean CCS: 2.8 ± 12.0	NS

Abbreviation: CAC, coronary artery calcium; CCS, coronary calcium score; CTD, connective tissue disorder; IIM, idiopathic inflammatory myopathy; NS, not significant; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SSc, systemic sclerosis.

Sample patients from Ref. 40 Multi‐Ethnic Study of Atherosclerosis (MESA) Study.

Rheumatoid arthritis

Nine out of 11 (82%) RA articles showed a statistically significant increase in CAC prevalence and/or CCS in RA patients compared with the control group. Among the nine articles, three of them 14, 15, 17 analyzed the CCS and CAC prevalence in early RA (less than 5 and less than 6 years of disease duration) and late RA (longer than a 10‐year duration). All of these studies found that the CAC prevalence and CCS were higher in the late RA group. One article 20 only analyzed female patients and a control group. Two articles 16, 19 did not demonstrate increased CAC in RA; however, neither of these studies demonstrated statistical significance. One article 16 illustrated that both prevalence and progression of CAC were similar between the RA and control group, and the other 19 demonstrated that the median CAC was not increased in RA patients when compared with the control group (in a small sample size of 37 patients).

Systemic lupus erythematosus

Twelve out of 14 (86%) studies demonstrated with statistical significance that CCS, CAC incidence/prevalence, and/or CAC burden was increased in SLE when compared with the control group. Three articles 28, 29, 30 only analyzed female patients and controls. One study 31 only analyzed male patients and controls. Two of the 14 studies did not comment on disease duration at time of CAC/CCS. The median or mean disease duration of the other 12 studies were at least 5 years or more in duration. Of the two studies that did not demonstrate statistical significance, one study 29 showed that the SLE group had higher prevalence of CAC and higher rates of CAC progression when compared with the control group. The other study 19 showed that the median CAC was not increased in SLE patients, but it had a small sample size of 33 patients with SLE.

Systemic sclerosis

Two out of four studies (50%) showed statistically significant increased CCS or CAC incidence/prevalence compared with the control group, though one study 35 had a very small population size (17 patients for each group). The other two studies 19, 37 did not demonstrate statistical significance and illustrated that CAC/CCS was not increased in SSc when compared with the control group. The mean or median disease duration in all studies was at least 6 years.

Idiopathic inflammatory myopathy

One article 38 did show a slight increase in the median CCS for IIM (also known as dermatomyositis/polymyositis) when compared with the control group (not statistically significant; P = 0.27). The mean disease duration was 9 years. The IIM group had more patients with CCS score 400 or greater (20%) compared with the control group (4%) (P = 0.04). However, multivariate analysis demonstrated that the confounding factors associated with this were age and smoking, and there was no significant association found between the number of patients with higher CCS and IIM.

Takayasu arteritis

One article 33 illustrated increased CAC incidence in Takayasu arteritis when compared with a control group; however, these data were not statistically significant. The mean disease duration was 9.5 years. This same study did show that the incidence of CAC was greater in SLE than in Takayasu arteritis (with statistical significance).

Psoriasis

One article 39 showed that those with psoriasis had slightly higher CCS than the control group; however, this was not statistically significant. The mean disease duration was 16 years. The psoriasis and control group populations had similar prevalence of CAC.

16 patients had CCS 1‐106 patients had CCS > 10

CCS/CAC of CTDs without a control group for comparison (Table 3)

The following articles were included in this study because of the lack of better‐quality studies to evaluate CAC/CCS in these respective CTDs compared with a control group. These consisted of one APS article 43 that analyzed serologically positive APS antibodies from a previous study 45 of patients that had CCS/CAC calculated, two Kawasaki articles 41, 42 that did not have a control group for comparison, and one JIA article 44 that did not have a control group for comparison. Table 3 illustrates the details of each study, including the type of CTD, CCS/CAC findings, statistical significance, and sample size. MDCT or EBT was used in these studies to detect CAC/CCS. The results demonstrate a higher CAC burden in people with these diseases than would be expected in their respective age groups; however, there is no control group for comparison, and data were not statistically significant. Descriptive characteristics of each article are outlined in Appendix S4.

Table 3

Descriptive summary of CCS/CAC of CTDs without control group for comparison

Author, Year, & Reference	CTD/Sample Size	Mean, Median, or Incidence/Prevalence of CAC or CCS Results (Agatston units)	Statistical Significance
Kahn (2012) 41	Kawasaki – 70 No control group	No coronary dilation (44/70) – none had CAC Transient dilation (12/70) – 1/11 patients had CAC With coronary aneurysm (14/70) – all patients had CAC and the highest CAC burden	NS
Kahn et al (2017) 42	Kawasaki – 116 No control group	No coronary dilation (100/160) – 0 CAC Transient/persistent dilation (33/160) – 1 out of 33 patients had CAC Aneurysm (9/33) – all had CAC and the highest CAC burden	NS
Majka et al (2013) 43	APS – 2203a No control group	APS: CAC was prevalent in 9.5% of young adults (age 18‐30) with APS antibodies.	NS
Aulie et al (2014) 44	JIA – 84 No control group	22 of 84 JIA patients (26%) had a CCS above 0 16 patients had CCS 1‐10 6 patients had CCS > 10	NS

Abbreviation: APS, antiphospholipid syndrome; CAC, coronary artery calcium; CCS, coronary calcium score; CTD, connective tissue disorder; JIA, juvenile idiopathic arthritis; NS, not significant.

Sample patients from Ref. 45 Coronary Artery Risk Development In Young Adults (CARDIA) study.

Kawasaki arteritis

Two articles 41, 42 demonstrated that CAC was increased in most patients with Kawasaki who developed a coronary aneurysm, though these data were not statistically significant. The median disease duration was at least 14.8 and 19.7 years. The median ages at the time of CT scan were 20 and 19.7 years in each study, respectively. There was no control group for comparison in either study. Both studies demonstrated that all of the participants without coronary dilation or aneurysm had no CAC detected.

Antiphospholipid syndrome

One article 43 was analyzed from a pool of patients in the Coronary Artery Risk Development in Young Adults (CARDIA) study 45, which consisted of young adults aged 18 to 30 years old who enrolled in the study in 1985. CAC was measured at 15 and 20 years. This study demonstrated that CAC was more prevalent in these patients with serum positivity for APS than would be expected for their age group. Antiphospholipid antibodies (immunoglobulin G [IgG] and IgA anti‐β2‐Glycoprotein I [GPI] antibodies) were associated with CAC level greater than 0 at year 15 after adjustment for traditional cardiovascular factors, gender, and race. Anti‐β2‐GPI antibodies (more so anti‐β2‐GPI IgG) were associated with CAC levels greater than 0 at year 20, but the relationship was not as strong as that for CAC at year 15.

Juvenile idiopathic arthritis

One article 44 discovered that 26% of patients were found to have CAC (not statistically significant), demonstrating an increased CAC burden than would be expected at this young age. The mean disease duration was 29.2 years, with a median age group of 38 years at the time of CT scan with CAC calculation.

Discussion

The goal of this study was to evaluate the cardiovascular risk of CTDs as measured by CAC prevalence/incidence or CCS. Overall, the articles that compared their respective CTD to a control group effectively matched to age and gender and had mostly similar cardiovascular risk factors. Unfortunately, there were not an adequate number of publications that evaluated CAC/CCS in CTDs meeting this study's inclusion/exclusion criteria, with the exception of SLE and RA. Therefore, this systematic review was unable to definitively assess the cardiovascular risk through means of CCS/CAC in the other CTDs, though the suspicion for CVD is still high based on all of the studies evaluated. Nevertheless, this systematic review did confirm that there is a strong association of CAC, and thus coronary atherosclerosis, in RA and SLE. All of the studies with statistical significance included in this systematic review that compared the RA and SLE patient population to a control group demonstrated increased CAC/CCS in the RA and SLE groups. Given these findings, CCS has been shown to be an effective modality in the evaluation of coronary atherosclerosis in these two CTDs. It is important for future studies to address whether the strict control of other risk factors of coronary artery disease (CAD), such as hypertension, smoking, and hyperlipidemia, in addition to the control of disease activity itself, can assist in preventing future cardiovascular events in this patient population.

SSc had conflicting data on its cardiovascular risk, with half of the studies demonstrating increased CAC/CCS compared with their respective control groups. However, two out of four studies with data of statistical significance did demonstrate increased CAC burden in the SSc group. In addition, one systematic review that investigated the incidence of CAD from autopsy findings, CAC, and coronary angiographic findings demonstrated that SSc is associated with increased incidence of CAD 46. These findings are more supportive of an increased cardiovascular burden from SSc, though additional studies with larger patient population sizes are needed to confirm this.

One IIM study did show that patients with IIM had a higher number of individuals with CCS levels at 400 or greater than the healthy control group; however, that was found to be attributed to confounding risk factors of tobacco smoking and patient age and not IIM. More studies are needed to investigate this correlation. The Takayasu arteritis and psoriasis studies also demonstrated increased CAC (though not statistically significant); however, more studies are needed with a larger patient population size for reliable assessment.

Overall, the analysis of articles without a control group for comparison demonstrated higher CAC/CCS in the Kawasaki, APS, and JIA population than would be expected for their younger patient population (not statistically significant), but additional studies are needed for direct comparison to a control group as well as larger sample sizes. Kawasaki disease is unique in that there is a direct coronary artery injury in the pediatric population complicated by posttraumatic effects, such as coronary aneurysm. These studies were consistent in finding that those with coronary aneurysms did have increased CAC, though the data were not statistically significant. The JIA study also demonstrated increased CAC prevalence than would be expected in its young patient population. The APS study showed a CAC burden greater than zero in those with serum positivity for antiphospholipid antibodies, which is not expected to be found in a young population of patients aged 18 to 30 years old; however, a control group would be helpful in validating these findings.

This systematic review has several limitations. First, this is a descriptive review and lacks statistical analysis to summarize the results of these studies and assist in eliminating some bias. The authors believe that publication bias could be present but it cannot be assessed in this review because of the different expected magnitude of effect and different publication bias across these CTDs. Another limitation of this study is the few number of articles that studied CCS in CTDs besides SLE and RA, which prevents proving or disproving the hypothesis that CCS/CAC is higher in all CTDs when compared with healthy control groups. Third, this review included many cross‐sectional observation studies that could not evaluate the temporal relationship of each respective CTD and CAC.

From the evidence of articles evaluated, this systematic review demonstrates that some CTDs (RA and SLE) have higher CCS and/or CAC incidence or prevalence compared with normal controls, and thus, may be an independent risk factor of coronary atherosclerosis, whereas the SSc data are still ambiguous. For SLE and RA, we suggest this risk from studies that demonstrated this association with statistical significance, though the quality of the articles evaluated (cross‐sectional and cohort studies) must be considered. Based on the current published data regarding other CTDs, increased coronary atherosclerosis also seems likely; however, better‐quality studies are needed to prove this association. This review also identified that aside from Takayasu arteritis and Kawasaki disease, there is currently no published data in the spectrum of vasculitides, which have been associated with high inflammatory burden. It is unclear if MDCT‐ or EBT‐measured CCS is a reliable and generalizable tool to assess subclinical atherosclerosis across the spectrum of CTDs, yet we ascertain that it can be useful for cardiovascular risk assessment in patients with SLE and RA, with consideration of its disadvantages including radiation exposure and cost.

Author contributions

Farshad drafted the manuscript; Farshad, Halalau, Schiopu revised the manuscript. Townsend reviewed the manuscript.

Study Conception and Design

Farshad, Schiopu.

Acquisition of Data

Townsend.

Analysis and Interpretation of Data

Farshad, Schiopu.

Appendix S1. CCS in CTD Search Strategies.

Click here for additional data file.

Appendix S2. Newcastle‐Ottawa Quality Assessment Scale for Cohort Studies.

Click here for additional data file.

Appendix S3. Agency for Healthcare Research and Quality (AHRQ) for Cross‐Sectional and Case‐Control Studies.

Click here for additional data file.

Appendix S4. Descriptive characteristics of CTD articles.

Click here for additional data file.