Pericarditis among giant cell arteritis patients: From myth to reality

Abstract

Background

Giant cell arteritis (GCA) is an inflammatory disease of unknown etiology affecting adults age > 50 years. GCA (also known as temporal arteritis) is a vasculitis of large and medium‐size vessels that involves the extracranial branches of the carotid artery. Common manifestations include constitutional symptoms, headache, jaw claudication, scalp tenderness, and vision loss. Cardiac involvement in GCA is considered to be as low as 5%, and < 30 cases of pericarditis among GCA patients have been reported in the literature. The aim of this study was to evaluate the association between GCA and pericarditis by conducting a cross‐sectional study utilizing the database of the largest healthcare provider in Israel.

Hypothesis

GCA is associated with pericarditis.

Methods

The proportion of past documentation of pericarditis among patients diagnosed with GCA was compared with that of their age‐ and sex‐matched controls. Univariate analysis was performed using the χ² and t tests; multivariate analysis was performed using logistic regression.

Results

The study included 4329 GCA patients and 21 611 controls. GCA patients had higher rates of cardiovascular risk factors. Pericarditis was observed in 53 GCA patients and 72 controls (1.22% vs 0.33%, respectively; P < 0.001), significantly higher among GCA patients in comparison with controls. A significant interaction was found between GCA, pericarditis, and young age (<70 years).

Conclusions

The study showed an independent association between GCA and pericarditis, especially among young patients. Proper screening should be applied whenever a suspicion arises as to the existence of comorbidity in patients with either disease.

1. INTRODUCTION

Giant cell arteritis (GCA), also known as temporal arteritis, is a vasculitis of large and medium‐size vessels that commonly involves the extracranial branches of the carotid artery. The onset of GCA tends to be subacute; its manifestations are related to the ischemia that might result. GCA has been associated with constitutional symptoms (fever, malaise, fatigue), as well as headache, jaw claudication, scalp tenderness, and loss of vision.1 The incidence peaks between the ages of 70 to 79 and appearance before the age of 50 is rare.2

An article written in 1980 by How et al3 argued that although GCA has been associated with various cardiological conditions, such as myocardial infarction and congestive heart failure, the disease is often overlooked by cardiologists. One of the cardiological manifestations that was mentioned in their study was pericarditis.

Pericarditis, in common with GCA, is also an inflammatory process, but one that mainly involves the pericardial sac. Its etiology and underlying mechanism are mostly unknown; infections and autoimmune and autoinflammatory processes have all been suggested playing a role in the development of the idiopathic form of the disease.4, 5 It is characterized by a sharp and severe pleuritic chest pain and can also present with constitutional symptoms. Other manifestations include specific electrocardiographic findings and a pericardial friction rub heard on physical examination, which is pathognomonic for the disease.

Although evidence supporting the association between GCA and pericarditis has been accumulating for more than half a century, it is still often overlooked today.6 Only several case studies have reported pericardial effusions in patients diagnosed with GCA,7, 8, 9, 10, 11 and many of these patients were presented due to unusual manifestations of 1 of the 2 diseases. Overall, there are about 2 dozen case reports to date,12 and the interest in this association seems to have subsided in recent years.

In this study, we wished to rekindle the investigation into the association between these 2 conditions and to examine it in a large, population‐based database. The grand scale of this study will help shed light on the true nature of this association.

2. METHODS

This study is one of a series of explorative and analytic studies based on the chronic disease registry of Clalit Health Services (CHS), the largest healthcare provider in Israel. We performed a “big‐data” analysis by utilizing the medical database of CHS. CHS has a comprehensive computerized database and a chronic disease registry, both of which receive real‐time input from pharmaceutical, medical, and administrative computerized operating systems in a continuous manner. The CHS chronic disease registry is composed of diagnoses of different medical conditions derived from records of community physicians or from hospital discharge letters.

The study was designed as a cross‐sectional study, as depicted in Figure 1. The proportion of documented history of pericarditis was compared between patients with GCA and their age‐ and sex‐matched controls. GCA patients were defined as patients who had documented diagnosis of GCA at least once in their medical records as outpatients by a community physician, either a primary‐care physician or a specialist, or who were diagnosed with GCA in their hospital discharge papers as inpatients (or both as inpatients and outpatients, when appropriate). Pericarditis was defined similarly, but specifically using the International Classification of Diseases codes 393, 420, and 423, and their subdivisions. All GCA patients detected in the CHS database were included in this study. Controls were randomly selected from the CHS database, with the exclusion of GCA patients. Five controls were matched by age and sex for each GCA patient (Figure 1). Age matching was based on year of birth. We decided to limit the age of subjects in the study to 85 (born in 1933 or later) due to a significant age disparity between GCA patients and controls. This inequality was caused by the fact that many controls who developed comorbidities were deceased; hence, their age was lower at the time of study conduction as compared with their GCA counterparts. The problematic issue of recruiting elderly controls was described before by Salthouse.13

Figure 1.

Study flowchart. Abbreviations: CHS, Clalit Health Services; GCA, giant cell arteritis

Data available from the CHS database included age, sex, body mass index (BMI), socioeconomic status (SES), as well as diagnoses of chronic diseases, such as hypertension, hyperlipidemia, and smoking status (ever smoked). Due to the limitations of the database, it was not possible to determine the etiology of each disorder. However, the validity of the diagnoses in the registry was found to be high in previous studies.14, 15, 16

2.1. Statistical analysis

The proportion of pericarditis and other covariates was compared between GCA patients and controls in the study sample group, and it was also analyzed independently to seek for interactions across all strata of categorical and continuous variables. For the matter of interactions, age variable was treated as dichotomous (age < 70 and ≥ 70). The cutoff was based on a review by Gonzalez‐Gay et al. that stated that incidence of GCA is rising after age 70.2 The χ² test was used to assess the distribution of categorical variables between GCA patients and controls, and the t test was applied for continuous variables. The association between pericarditis and GCA was evaluated by a multivariate logistic regression model. Odds ratios (ORs), as well as 95% confidence intervals (CIs), are presented. Statistical analysis was performed using R Statistical Software version 3.3.0 (R Foundation for Statistical Computing, Vienna, Austria).

The study was approved by the ethics committee of CHS, located in Soroka Medical Center, Be'er Sheva, Israel.

3. RESULTS

The study included 4329 GCA patients and 21 611 controls (Table 1). The mean ages of the GCA patients and controls were 65.9 and 64.1 years, respectively. Both groups had a clear female predominance and were overweight according to their mean BMI values. GCA patients presented significantly more cardiovascular risk factors, including smoking. Pericarditis was observed in 53 GCA patients vs 72 controls, producing a significantly higher rate of pericarditis among GCA patients in comparison with controls (1.22% vs 0.33%, respectively; P < 0.001). GCA patients also exhibited a greater proportion of high SES when compared with controls.

Table 1.

Basic characteristics of GCA patients and matched controls

Characteristic	Controls (No GCA), N = 21 611	GCA, N = 4329	P Value
Age, y	64.1 ±14.2	65.9 ± 14.1	<0.001
Female sex	14 817 (68.6)	2967 (68.5)	0.98
BMI, kg/m2	28.4 ± 5.91	28.4 ± 5.70	0.08
SES
Low	4051 (37.8)	1616 (37.4)	—
Medium	4361 (40.7)	1743 (40.4)	0.31
High	2302 (21.5)	958 (22.2)	<0.001
HTN	5648 (26.1)	2432 (56.2)	<0.001
Hyperlipidemia	7382 (34.2)	3223 (74.5)	<0.001
DM	3290 (15.2)	1396 (32.2)	<0.001
Smoking	3496 (16.2)	1578 (36.5)	<0.001
Pericarditis	72 (0.33)	53 (1.22)	<0.001

Abbreviations: BMI, body mass index; DM, diabetes mellitus; GCA, giant cell arteritis; HTN, hypertension; SD, standard deviation; SES, socioeconomic status.

Data are presented as n (%) or mean ± SD.

A significant interaction between GCA and pericarditis was observed among all strata of study covariates, namely younger age (<70 years), except the following groups: low SES, BMI <25 kg/m², diabetes, and with and without smoking (Figure 2). In a multivariate logistic regression model (Table 2), GCA was found to be independently associated with pericarditis (OR: 1.69, 95% CI: 1.16–2.144). Smoking was also found to be independently associated with pericarditis. However, other covariates, such as age and sex, failed to demonstrate a similar independent association.

Figure 2.

Interaction between GCA and pericarditis by strata of study covariates. Abbreviations: BMI, body mass index; GCA, giant cell arteritis; SES, socioeconomic status

Table 2.

Multivariate logistic regression: covariates associated with pericarditis

Characteristic	OR	95% CI	P Value
Age	1.01	1.00–1.03	0.15
Female sex	0.78	0.52–1.17	0.23
BMI, 1‐kg/m2 increment	0.98	0.94–1.01	0.22
SES
Medium vs low	0.89	0.59–1.35	0.58
High vs low	0.80	0.47–1.31	0.38
HTN	1.36	0.88–2.13	0.18
Hyperlipidemia	1.41	0.85–2.46	0.20
DM	0.85	0.56–1.29	0.46
Smoking	1.55	1.05–2.27	0.03
GCA	1.69	1.16–2.44	0.01

Abbreviations: BMI, body mass index; CI, confidence interval; DM, diabetes mellitus; GCA, giant cell arteritis; HTN, hypertension; OR, odds ratio; SD, standard deviation; SES, socioeconomic status.

4. DISCUSSION

In the current study, we explored the association between GCA and pericarditis; we demonstrated that even after adjustment for age, sex, and other cardiovascular risk factors was performed, the 2 diseases remained significantly associated. The association was especially prominent among younger (age < 70 years) GCA patients. For years, the existence of the association between GCA and pericarditis was supported only by anecdotal case reports. This population‐based study is the first to quantify the proportion of pericarditis among GCA patients, as well as the first to establish an independent connection between the 2 diseases.

Overall, cardiac disease (endocardial, myocardial, or pericardial) is estimated to occur in about 5% of GCA patients.17 However, regarding pericardial involvement, the rates are uncertain. Our study has demonstrated that the rate of pericarditis among GCA patients was 1.22%, as opposed to a French study by Zenone and Puget,18 which described pericardial effusion in 3.5% of patients. This discrepancy could be explained by the nature of the 2 disorders, which are uncommon and therefore are challenging to diagnose, even for experienced clinicians. The presentation of GCA may precede pericarditis or coincide with it; indeed, in some cases, pericarditis and even pericardial tamponade were the first overt manifestations of GCA.19

Pericarditis is a term used to describe a group of diseases arising from different etiologies. Autoimmune pericarditis, which may occur in rheumatoid arthritis or systemic lupus erythematosus, is mediated by an antibody‐provoked interferon response.20 On the other hand, an autoinflammatory systemic process such as GCA may induce pericarditis via the inflammasome pathway, in which the activation of sensor molecules triggers a predominant T helper 17 cell response that results in excessive interleukin‐1 (IL‐1) production.

Currently, high‐dose corticosteroid therapy is still considered the mainstay of pharmacotherapy in GCA, which holds a great number of side effects. However, in recent years, evidence is accumulating toward the role of IL‐1 inhibitors, such as anakinra and canakinumab, in the treatment of GCA. Presently, IL‐1 inhibitors are labeled for rheumatoid arthritis, juvenile idiopathic arthritis, and autoinflammatory conditions such as familial Mediterranean fever. Several small studies that have attempted to treat refractory GCA with IL‐1 inhibitors have enjoyed relative success.21, 22 Although Imazio et al23 demonstrated the efficacy of using colchicine and anti‐inflammatory drugs for the treatment of acute pericarditis, IL‐1 inhibitors were also tried as a novel agent for treating idiopathic recurrent pericarditis with a high rate of remission, but only in small‐scale studies as well.24 It is only reasonable to believe that in the future, IL‐1 inhibitors will become key players in the treatment of pericarditis.25 Therefore, IL‐1 seems to be involved in both diseases, and might, at least partly, explain their common pathways. Interestingly, the IL‐6 receptor α inhibitor tocilizumab showed promising results in treating GCA, with longer to glucocorticoid‐free remission compared with placebo.26 This might open a door for future IL‐6 pathway inhibition in also treating pericarditis.

A review by Cantarini et al27 states that GCA is one of the systemic diseases that may cause idiopathic recurrent pericarditis on top of acute pericarditis. The partially common pathogenic pathways and shared response to pharmacotherapy of GCA and pericarditis may explain the association between the 2 disorders that was demonstrated in our study. Another plausible explanation for the increased rate of pericarditis among GCA patients is their compromised immune state due to immunosuppressive therapy, leading to higher susceptibility to opportunistic infections. The resulting postviral pericarditis may require treatment with nonsteroidal anti‐inflammatory drugs or colchicine.20

Our study showed a significant interaction between GCA, pericarditis, and younger age (age < 70 years). These findings correspond well with most case reports and case series previously published on the subject.8, 9, 19, 28, 29, 30, 31 However, several articles have reported cases of adults age > 70 years with simultaneous presentations of both GCA and pericarditis, which usually regressed after corticosteroid treatment.12, 32, 33, 34, 35, 36 A Spanish case report described a 76‐year‐old female patient with pectoral girdle pain and a pleural effusion on x‐ray who was eventually diagnosed with polymyalgia rheumatica (PMR) accompanied by pleuropericarditis and whose high levels of CA‐125 indicated serosal involvement, rather than malignancy.37 Based on these observations, it has been suggested that while investigating idiopathic pericarditis among elderly patients, a temporal artery biopsy should be obtained even in the absence of classical GCA manifestations.38 Current guidelines39 note that pericarditis, acute or recurrent, reflects the presence of systemic autoimmune disease in 5% to 15% of the cases; in such suspected cases, the authors warrant a targeted search for etiology and recommend gaining rapid control over the primary disease.

The strength of our study is derived from its large scale, and its novelty; it is the first study to date to describe comorbidity of GCA and pericarditis in such a large number of patients. Our study also provides a good characterization of the patients who suffer from both diseases, as the diagnoses were done in the setting of both community and hospitals, and the computerized data registration started approximately at the beginning of the millennium until the year 2016, when data for the current study were collected.

The clinical significance of this study lies in its ability to shed light on GCA as an often‐forgotten etiology for pericarditis, especially among adults age < 70 years. GCA may be a challenging diagnosis; physicians should consider it in some difficult cases of pericardial diseases, including probably unexplained chronic effusions with elevated CRP; in these cases, a computed tomography scan may demonstrate the thickening of large‐vessel walls and positron emission tomography may show some capitation at the level of vessel walls.

4.1. Study limitations

Our study does have several drawbacks. The main limitation is that the diagnoses of both GCA and pericarditis were not based on standardized diagnostic criteria, but rather on diagnostic labels drawn from CHS physician reports. However, previous studies have supported the high validity of the diagnoses in this database.40 In this study, past diagnosis of pericarditis represented either acute pericarditis or chronic pericarditis, or pleural effusion (or a combination), which may arise from different etiologies. Another possible disadvantage of this study is that it was designed as a cross‐sectional study, so it was impossible to determine whether pericarditis occurred before, during, or after the diagnosis of GCA. Hence, this theoretically prevents the discussion of causal relationship between GCA and pericarditis; however, as discussed previously, pericarditis may be the first presentation of an occult GCA. Although PMR is closely related to GCA, it was not included in this study due to the low documentation rate of this comorbidity alongside GCA in the CHS database. In a previous study we discussed this discrepancy, which is due to registration and documentation habits in real life, rather than lower‐than‐expected rates of PMR.40

5. CONCLUSION

The current study has shown that GCA is independently associated with pericarditis, especially among patients age < 70 years. Proper screening methods should be applied whenever suspicion arises to the existence of comorbidity in patients with either disease.

Author contributions

Arnon D. Cohen, MD, and Howard Amital, MD, share equal contribution.

Conflicts of interest

The authors declare no potential conflicts of interest.

Tiosano S, Adler Y, Azrielant S, et al. Pericarditis among giant cell arteritis patients: From myth to reality. Clin Cardiol. 2018;41:623–627. 10.1002/clc.22927