Endoscopically Active Ulcerative Colitis Is Associated With Asymptomatic Atherosclerotic Vascular Disease: A Case-Control Study

Abstract

Background

Chronic inflammation in immune-mediated conditions has been associated with an increased risk in atherosclerotic disease. There is paucity of evidence regarding the prevalence of asymptomatic atherosclerosis in patients with ulcerative colitis (UC) and its association with disease activity. We sought to compare the prevalence of asymptomatic atherosclerotic disease between young patients with UC with and without mucosal healing (MH) and healthy control individuals.

Methods

An observational study was conducted in 2 hospitals in Buenos Aires, Argentina. Patients with UC 18 to 50 years of age with at least 1 previous colonoscopy in the last year were enrolled, along with age- and sex-matched healthy control individuals. Carotid and femoral ultrasound assessments were performed to determine the prevalence of atherosclerotic lesions and abnormal intima-media thickness (IMT). We compared the prevalence of atherosclerotic disease and the prevalence of abnormally increased IMT in at least 1 vascular territory.

Results

Sixty patients with UC and 60 healthy control individuals were enrolled. Mean age was 38 years and 53.33% were men. Although the prevalence of atherosclerotic lesions was similar in patients with UC without MH when compared with both patients with UC with MH and control individuals (3.7% vs 0% vs 6.67%; P = .1), we found a significant increase in abnormal IMT in at least 1 vascular territory in UC patients without MH when compared with healthy control individuals (48.15% vs 26.67%; P = .05).

Conclusions

Patients with UC with active mucosal inflammation showed a significantly increased odds of asymptomatic femoral or carotid vascular disease when compared with control individuals.

Introduction

Chronic inflammation has been associated with the development of atherosclerosis as well as thrombotic events.^1,2 Patients with immune-mediated conditions like psoriasis or rheumatoid arthritis are at a higher risk of accelerated atherosclerosis.^3,4 There is increasing evidence showing that inflammatory bowel disease (IBD), which comprises both ulcerative colitis (UC) and Crohn’s disease (CD), is also associated with an increased risk of atherosclerotic cardiovascular disease (ACVD) even when traditional risk factors are contemplated.^5,6 The reasons for this association may include an increased risk of thrombosis, endothelial dysfunction caused by systemic inflammation, changes in the gut microbiome, lipid profile abnormalities due to IBD medications, and chronic steroid exposure.⁷ There is also evidence suggesting that disease activity can further increase the risk of symptomatic cardiovascular events.^8,9

The evidence becomes scarce when assessing the prevalence of asymptomatic ACVD among patients with IBD.^10-12 IBD is usually diagnosed at an early age, in which ACVD is not prevalent; hence, there is a lack of systematic diagnostic approach to assess ACVD among young individuals. Whether asymptomatic ACVD is more prevalent among young patients with IBD, or has an association with disease activity, remains to be determined. Endoscopic inflammatory activity has become in recent years a key prognostic element in IBD, which is more accurately and easily ascertained in patients with UC.¹³ There is paucity of evidence regarding the association between endoscopic inflammation and the prevalence of asymptomatic ACVD. Hence, the present study aimed to compare the prevalence of asymptomatic ACVD between young UC patients with and without endoscopic healing (MH) vs healthy control (HC) individuals. We hypothesized that ACVD is more prevalent among young patients with UC when compared with young, healthy, age- and sex-matched HC individuals; we further hypothesized that it becomes more evident among patients with UC who have persistent inflammatory activity.

Methods

Study Design and Population

From April to December 2022, an observational prospective study was undertaken in 2 academic hospitals from Buenos Aires, Argentina. The study was approved by a local ethics committee (approval number #857). Consecutive patients with a diagnosis of UC between 18 and 50 years of age were invited to participate in the study. UC patients were enrolled regardless of their clinical activity and should have at least a colonoscopy performed within a year before enrolment to determine if MH had been achieved. Patients with UC with a recent diagnosis (<6 months) were excluded from the study. In addition, age- and sex-matched individuals without a diagnosis of UC and/or any other gastrointestinal condition from primary care outpatient clinic were included in a 1:1 ratio (HC individuals).

Patients would be excluded if they had any known history of cardiovascular disease: stroke, coronary heart disease, renal vascular disease, chronic kidney disease, peripheral vascular disease, or diabetes mellitus.

Endpoints and Variables of Study

After signing informed consent, all patients underwent physical examination, a blood test and both carotid and femoral ultrasound assessment. Ultrasound assessment was performed following the protocol explained further subsequently. The primary endpoint was the prevalence of at least 1 atherosclerotic lesion in the territories explored between patients with UC and HC individuals. We also compared the prevalence of an abnormal intima-media thickness (IMT) in at least 1 vascular territory. In addition, we compared the prevalence of at least 1 atherosclerotic lesion between patients with UC with and without evidence of MH and HC individuals. We compared the distribution of these vascular lesions between groups.

Abnormal IMT was defined as an IMT above 0.8 mm. An atherosclerotic lesion was defined as a focal protrusion in the lumen measuring at least IMT >1.5 mm, a protrusion at least 50% greater than the surrounding IMT, or an arterial lumen encroaching >0.5 mm.¹⁴ This definition was used for carotid territories as well for femoral territories, as done in previous studies such as the PESA (Prevalence, Vascular Distribution and Multiterritorial Extent of Subclinical Atherosclerosis in a Middle-Aged Cohort) study.¹⁵

The following variables were registered for each patient: age, sex, smoking, history of arterial hypertension, history of dyslipidemia, body mass index, use of antihypertensive medication, use of lipid-lowering medication, and use of aspirin; blood tests were performed to determine total cholesterol, high-density lipoprotein (HDL), low-density lipoprotein, triglycerides, and C-reactive protein (CRP). In patients with UC, the following variables were also registered: disease extent according to Montreal classification and disease duration. Severity of UC was assessed as follows: MH (defined as Mayo endoscopic subscore of 0 or 1 in the last colonoscopy) or the use of UC-specific medication (mesalamine, thiopurines, systemic steroids, tofacitinib, ozanimod, or any biologic agent [anti-tumor necrosis factor α [TNFα], vedolizumab, or ustekinumab).

Vascular Ultrasound Assessment

Carotid and femoral IMT were determined using a high-resolution duplex ultrasonography device (Phillips HD-11 and Philips Affinity 50) equipped with a 7-MHz linear transducer. Briefly, longitudinal and transverse section images were obtained from the carotid bulb and common femoral systems. According to the leading-edge method, the distance between the first (lumen-intima border) and second (media-adventitia border) echogenic lines was recorded as the IMT. An average of 10 measurements was performed on both sides and the mean of these 10 measurements was considered as the baseline value. All operators were experts in vascular ultrasonography and had a high interrater agreement, as evidenced by a kappa coefficient of 0.77.

Statistical Analysis

Stata software (version 13; StataCorp) was used for statistical analysis. The prevalence of subclinical atherosclerotic lesions among young to middle-aged individuals has been underexplored in the literature. However, the PESA study¹⁵ evaluated middle-aged asymptomatic individuals and reported an overall prevalence was 63% for such lesions. In our study, we hypothesized that there would be a 20% increase in subclinical atherosclerotic lesions patients with UC. To achieve a statistical power of 80% and maintain an alpha error of <5%, we determined that a sample size of 65 patients per group would be required.

Categorial variables were presented as percentages, while numerical variables were described as mean ± SD or as median and interquartile range for non-normally distributed variables. To assess univariable differences between patients and HC individuals, Student’s t test, Mann-Whitney U test, chi-square test, or Fisher exact test were used based on the distribution of the variables or the number of individuals. Variables associated with the prevalence of subclinical atherosclerotic lesions and having a P value of <.1 were included in a multivariate analysis, which employed a multinomial logistic regression model, taking the variable ACVD as the dependent variable.

Ethical Considerations

We complied with the ethical principles for clinical research described in the Declaration of Helsinki. The study protocol was approved by the institution’s ethics committees. All patients we invited to participate and signed an informed consent form.

Results

Between April and December 2022, we reviewed the medical records of 389 UC patients who would potentially be included in the study; 60 UC patients who fulfilled inclusion criteria were finally recruited. Table 1 provides an overview of the characteristics of the participants included in the study. The mean age of the enrolled patients was 38 years, with 53.3% of the individuals being male. All but 2 patients had a colonoscopy performed during the previous 4 months before enrollment; the other 2 patients had their colonoscopies 6 and 8 months before enrollment. None of the patients with UC had a history of UC-related surgery. Comparative analysis revealed that the age and sex distributions were similar between the patients with UC and the HC group. However, we did observe a lower prevalence of dyslipidemia among the UC group (6.7% vs 20%; P = .03). Furthermore, the patients with UC exhibited significantly lower levels of HDL (57.61 ± 16.18 mg/dL vs 63.27 ± 12.74 mg/dL; P = .003). In contrast, patients with UC demonstrated elevated levels of triglycerides (105.2 ± 44.24 mg/dL vs 75.09 ± 31.17 mg/dL; P = .002) as well as higher levels of CRP (4.9 ± 5.2 mg/L vs 2.4 ± 2.3 mg/L; P = .008).

Among patients with UC, 43.3% had an extensive colitis; 26.6% were under treatment with a biologic agent, whereas 10% were not receiving any pharmacological treatment at the time of the study; and 13.3% were on systemic steroids at the moment of inclusion to the study. Mean disease duration was 9.05 years, and 55% showed MH in endoscopic assessment. The mean CRP of patients with UC was 4.9 ± 5.2 mg/L.

Table 2 provides a comparison between patients with UC with and without MH. Among those with MH, there was a higher prevalence of extensive colitis (60.6% vs 22.2%; P = .003). Additionally, patients with UC with MH had a higher body mass index compared with those without MH (24.94 ± 3.29 kg/m² vs 22.2 ± 3.2 kg/m²; P = .01), and they exhibited higher levels of HDL cholesterol (63.26 ± 19.1 mg/dL vs 51.43 ± 9.2 mg/dL; P = .006).

We did not find a significant difference in the prevalence of at least 1 atherosclerotic lesion in the territories explored between overall patients with UC and HC individuals (Figure 1). However, patients with UC without MH showed a significant increase in abnormal IMT in at least 1 vascular territory when compared with HC individuals (48.15% vs 26.67%; P = .05) (Figure 2). Table 3 summarizes the vascular ultrasound findings in both patients with UC and HC individuals.

On multivariate analysis, age (odds ratio [OR], 1.1; 95% confidence interval [CI], 0.99-1.25), disease duration (OR, 1.13; 95% CI, 1.01-1.37), and CRP levels (OR, 25.13; 95% CI, 1.6-394.18) were significantly associated with increased odds of asymptomatic vascular lesions finding in at least 1 vascular territory (Table 4). Age and disease duration did not show collinearity in our model; because age could inevitably influence on the association between disease duration and the endpoint, we included these 2 variables (age and disease duration) in a linear regression model: we found a significant association (P = .002), but r² did not show a strong correlation (r² = 0.29), showing that, regardless of age, disease duration is associated with asymptomatic vascular lesions in at least 1 vascular territory.

Discussion

Our findings indicate that the prevalence of ACVD was similar in both patients with UC and age- and sex-matched HC individuals. However, patients with UC with active mucosal inflammation were at a higher risk of asymptomatic femoral or carotid vascular disease, as measured by IMT, when compared with HC individuals.

Previous evidence has shown that patients with IBD may have an increased risk of cardiovascular disease.^5,6 A Taiwanese study by Huang et al¹⁶ showed an increased risk of ischemic stroke among patients with IBD; this was also suggested in a large cohort of patients with CD.¹⁷ A landmark Danish cohort study including 28 833 IBD patients followed up for up to 13 years showed an increased risk of cardiovascular events after adjusting for confounders.¹⁸ However, this study did not address the risk of asymptomatic ACVD among patients with IBD. The available evidence on this matter is relatively scarce and shows heterogeneous results. A small study by Bruzzese et al¹⁹ compared carotid IMT between patients with both CD and UC vs HC individuals. The study found no correlation between increased thickness and IBD status.¹⁹ In a study conducted by Ozturk et al,²⁰ it was observed that carotid IMT was not significantly different between individuals with IBD and HC individuals who were matched by age and sex. However, the study revealed that patients with IBD had an increased pulsed wave velocity, which could indicate endothelial vascular dysfunction. It is important to note that neither study assessed the femoral IMT, which could have led to an underestimation of vascular lesions.²¹ Üstün et al²² found a nonsignificant increase in carotid IMT among patients with IBD when compared with HC individuals. In addition, CRP levels were significantly correlated to IMT. Likewise, we found that there is a significant association between CRP levels and the likelihood of having at least 1 vascular lesion in the carotid or femoral territories.

Our results not only show an association between UC without endoscopic improvement and subclinical vascular disease, but also suggest that UC inflammatory burden can be associated with the odds of increased IMT or ACVD. Age was found to be the only significant factor associated with the combined outcome of increased IMT and ACVD in patients with UC. Despite the presence of other commonly known risk factors, no significant association was observed. However, disease duration and CRP were significantly increased among patients with UC with increased IMT or ACVD. This finding suggests that the chronic and repetitive proinflammatory state could be one of the primary mechanisms involved in risk of developing ACVD in patients with UC. Cytokines such as interleukin (IL)-1β, IL-6, or TNFα can have an impact on endothelial function and thus ultimately could lead to atherosclerosis over time, regardless of the presence of typical risk factors.⁷ Other immune-mediated disorders that share similarities with IBD, such as psoriasis or rheumatoid arthritis, have a similar association with ACVD.²³ Interestingly, a clinical study by Ridker et al²⁴ showed that the use of a monoclonal antibody (canakinumab) against IL-1β decreased the risk of further cardiovascular events among a large cohort of patients without any immune-mediated inflammatory disease and with a history of myocardial infarction.

Another potential risk factor among patients with IBD that could explain an increased risk of cardiovascular disease would be recurrent exposure to systemic steroids. We were unable to find any association between steroid use and ACVD or an increase in IMT. In addition, we did not find any difference in vascular findings between patients under biologic treatment. Although not definite, some evidence among rheumatologic patients suggests that exposure to anti-TNFα can reduce cardiovascular dysfunction. It remains to be seen if such a reversal is due to their anti-inflammatory effect or if there is a direct effect of the suppression of TNFα on cardiovascular function.²⁵

There are some limitations worth mentioning in our study. First, to ensure the accuracy of our study, we included only patients with UC and excluded those with CD. This decision was made because endoscopic assessment in patients with UC is readily available, and including patients with CD could introduce bias to our study. It is worth mentioning that most of the published studies involving asymptomatic ACVD included predominantly patients with UC. Second, we also lacked the adequate power to assess the association between chronic treatment with biologics and the prevalence of asymptomatic ACVD. This association would be better assessed with a prospective cohort study instead of a case-control study like ours. Our sample size was relatively small if smaller differences in the proportion of ACVD between patients with UC and HC were to be identified. We did not contemplate some cardiovascular risk factors, such as serum lipoprotein A measurement or physical activity, a protective factor against cardiovascular disease, as potential factors to be assessed in the comparison of patients.

Conclusions

In summary, our research indicates that patients with UC with long duration of disease and who have not achieved effective control of their condition are more likely to have asymptomatic ACVD. Therefore, cardiovascular risk assessment should be considered in such patients, even if they do not have any typical risk factors.

Funding

A grant from Pfizer (grant number 66581437) was received to cover expenses related to the conduction of this study.

Conflicts of Interest

J.L. has received honoraria for lectures from Pfizer, Takeda Pharmaceuticals, Janssen, and AbbVie; and fees for medical advisory from Pfizer, GSK, and Janssen. I.Z. has received honoraria for lectures from Janssen, AbbVie, Takeda Pharmaceuticals, and Pfizer; and fees for medical advisory from AbbVie, Takeda Pharmaceuticals, Janssen, and Pfizer. All others disclose no conflicts.

References