An international audit of the management of dyslipidaemia and hypertension in patients with rheumatoid arthritis: results from 19 countries

Abstract

Introduction

Patients with rheumatoid arthritis (RA) have an increased risk of cardiovascular disease (CVD), associated with a substantial inflammatory burden and increased prevalence of traditional risk factors of CVD.¹ In line with what is observed in the general population, recent studies indicate a modest decline in CVD event rate for RA patients.^2–4 Moreover, a few studies have reported a more pronounced reduction in CVD-related mortality in RA patients relative to the general population.^5,6 However, epidemiological data from Norway in the time period 1995–2008 suggest that all-cause mortality in RA patients is still increased compared with the general population,⁷ indicating a persistent need for optimization of CVD preventive measures in these patients.

Over the last decade, a growing number of studies reported elevated risk of CVD in RA, yet it remains uncertain whether this knowledge has influenced the management of dyslipidaemia and hypertension. In the general population, regular audits on CVD risk factors allow for periodic evaluations of guideline implementation as well as changes over time. A global survey of CVD risk factor management (SURF) has been established with the mission of creating an international audit standard to record and monitor CVD risk factors and improve the prevention of CVD (www.surfriskfactor-audit.com). SURF has previously been performed in patients with coronary heart disease (CHD),⁸ and is currently ongoing among patients with stroke, chronic obstructive pulmonary disease, systemic lupus erythematosus, and antiphospholipid syndrome. To date, no surveys on CVD risk management across several world regions have previously been performed in patients with RA.

The objectives of the SURF in RA (SURF-RA) audit were to describe differences in estimated risk of atherosclerotic CVD among patients with RA from different world regions and to evaluate management of lipid-lowering treatment (LLT) and antihypertensive treatment.

Methods

The data in SURF-RA were derived from already established clinical cohorts, as well as from prospective recording in cardiology and rheumatology clinics, between 2014 and 2019. Patients aged >18 years with a clinically diagnosed RA were eligible for inclusion. The participating centres were divided into Western Europe, Central and Eastern Europe, Mexico, North America (USA and Canada), and Asia. The SURF-RA study has been approved by the Data Protection Officer (DPO) at the Oslo University Hospital, Ullevål (2017/7243), and a general data protection regulation evaluation was performed by the DPO at Diakonhjemmet Hospital (10/10-2018), Oslo, Norway. Each participating centre was responsible for obtaining the correct regulatory approval for participating in SURF-RA.

All data were recorded using a one-page questionnaire. Demographic data included year of birth and sex. The following RA disease-related variables were recorded: rheumatoid factor (RF) and anti-citrullinated protein antibody (ACPA) positivity, C-reactive protein and erythrocyte sedimentation rate (ESR), use of anti-rheumatic medication, and disease activity score using 28 joints with either C-reactive protein or ESR (DAS28CRP and DAS28ESR).

CVD risk factors registered were smoking status, physical activity, hypertension, hyperlipidaemia, obesity, diabetes mellitus, and the most recent CVD risk factor measurements. Moreover, use of lipid-lowering agents and antihypertensive treatment, as well as presence of established CVD, was recorded.

C-reactive protein, ESR, and lipid values were analysed according to each centre’s laboratory standards. For general CVD risk screening, fasting status has been shown not to influence the prognostic value of the blood sample,⁹ and we therefore included both fasting and non-fasting lipid values in the SURF-RA.

CVD risk estimation and treatment definitions

The 10-year risk of a fatal atherosclerotic CVD event was calculated using the European CVD risk calculator, the systematic coronary risk evaluation (SCORE), which includes information on age, sex, smoking status, total cholesterol (TC), HDL cholesterol (HDL-c), and systolic blood pressure (BP) (www.heartscore.org). Atherosclerotic CVD was defined as presence of CHD, stroke, and/or peripheral vascular disease.

Antihypertensive treatment

Hypertension was defined as measured BP ≥ 140/90 mmHg and/or use of antihypertensive treatment. The treatment goal of antihypertensive treatment was set at <140/90 mmHg according to guidelines at the time of the survey, which is a modest approach to the most recent European guidelines where defined targets depend upon age and side effects.¹⁰

Lipid-lowering therapy

The calculated CVD risk by SCORE was multiplied by 1.5 as recommended by the European League Against Rheumatism.¹¹ Patients included in the high CVD risk group were (i) diabetes mellitus, (ii) TC > 8.1 mmol/L, (iii) SCORE ≥ 5% and <10%, and/or (iv) LLT use. The European guideline-recommended LDL cholesterol (LDL-c) goal was <2.6 mmol/L for patients in the high CVD risk group at the time of the survey.¹²

The following patients were classified in the very high CVD risk group: (i) established CVD and/or (ii) SCORE ≥ 10%. For patients in the very high CVD risk group, the LDL-c goal was <1.8 mmol/L.¹²

Statistical analyses

Descriptive summaries of continuous variables are given by mean (SD) or median (IQR) as appropriate, while categorical variables are summarized by frequencies and percentages. Between-group comparisons of continuous variables were carried out using the Kruskal–Wallis test, while the χ² test was used for categorical outcomes. Missing data were not imputed in any case, each variable being summarized using all available data. All analyses were performed using the statistical software R, version 3.4.4.

Results

In total, 14 503 RA patients from 19 countries and 53 centres throughout three world regions were included (Supplementary material online, Table S1). Patient characteristics are shown in Table 1. Mean age was 59.8 ± 13.6 years. There was a strong female preponderance (75%). On average, the RA disease duration was 10.8 ± 9.5 years.

More than half of the patients were RF and/or ACPA positive. The median levels of C-reactive protein and ESR were low, and mean RA disease activity, as measured by the composite scores DAS28CRP and DAS28ESR, also reflected a modest disease activity (Supplementary material online, Table S2).

Mean body mass index was 27.4 kg/m². Current smoking was most common in Central and Eastern Europe (29%), in contrast to America and Asia (∼8–10%). Mean TC was 5.0 mmol/L, while LDL-c ranged from 2.5 mmol/L in America to ∼3.0 mmol/L in Europe. Average BP was in the normal range (mean 128/77 mmHg), particularly, with lowest measurements in Mexico (mean 118/74 mmHg). Hypertension was reported as present in nearly two-thirds of all the patients (Table 1).

The prevalence of known atherosclerotic CVD was 13% (Table 1). Premature familial CVD was not a prominent risk factor, except in patients from Central and Eastern Europe (15%). Thirteen per cent of the patients had diabetes mellitus, mainly type 2 diabetes (12%) (data not shown).

Overall, the distribution of patients in the different CVD risk categories by SCORE was comparable. The majority of the patients in the highest CVD risk classes were from Central and Eastern Europe and North America (Table 1).

Use of CVD preventive medication is presented in Table 2. Around one-fourth of all RA patients were on statin therapy, except those in the Asian and Mexican cohorts, where the reported statin use was only 6% and 16%, respectively. The use of other LLT was low in all investigated regions (1–7%). The proportion of patients at LDL-c goal was higher among patients using statins in combination with other LLT (41.8%), than in patients using statins in monotherapy (31.2%).

The use of antihypertensive treatment also differed substantially between the cohorts, with a mean prevalence of 29%, and lowest use reported in Western Europe and Mexico. A minority of the total cohort received a two- or three-drug antihypertensive treatment combination. The proportion of patients at BP goal receiving antihypertensive treatment in monotherapy or as a two- or three-drug combination was comparable (62.5%, 63.8%, and 64.5%, respectively).

BP goal attainment is presented in Figure 1: ∼40% of the hypertensive patients did not receive antihypertensive treatment, and among the patients who were using antihypertensive treatment, 63% were at the defined BP goal. In comparison, the BP goal attainment declined to 50% when all patients with an indication for antihypertensive treatment (both patients using and not using antihypertensive treatment) were included, with lowest numbers in Central and Eastern Europe (41%) and highest in North America (64%).

Figure 1

The proportion of patients achieving blood pressure goals across five geographical regions. Abbreviation: BP, blood pressure.

Open in new tabDownload slide

Lipid goal attainment is shown in Figure 2. Fifty-two per cent had an indication for LLT, of which 44% used LLT. Among the patients treated with LLT, recommended LDL-c goals were obtained by 58% in the high and 37% in the very high CVD risk group. Out of those with indication for LLT (both patients using and not using LLT), only 45% and 18% were at LDL-c targets in the high and very high CVD risk classes, respectively. There were considerable differences in LDL-c goal attainment between different geographical areas, with the highest goal attainment rate in North America and lowest in Central and Eastern Europe, Mexico, and Asia. Further information on BP and lipid goal attainment within each world region is illustrated in Figure 3A–C and Supplementary material online, Figure S1.

Figure 2

The proportion of patients reaching lipid goals across five geographical areas. Abbreviations: CVD, cardiovascular disease; LLT, lipid-lowering treatment; LDL-c, LDL cholesterol.

Open in new tabDownload slide

Figure 3

The proportion of patients achieving guideline-recommended goals for lipids and blood pressure across 19 countries. Abbreviations: BP, blood pressure; LLT, lipid-lowering treatment.

Open in new tabDownload slide

Table 3 shows the BP and lipid goal attainment for patients with indication for CVD preventive treatment. Attainment of recommended BP and LDL-c goals was comparable across patients with different RA disease activity levels. Although females had a higher degree of BP target achievement than men, the opposite was observed regarding LDL-c goal attainment in patients at very high risk of CVD, where a lower frequency at 16% was observed in women compared with 21% in men (P = 0.001). The proportion of patients who achieved target BP was significantly higher in those >70 years (P < 0.001), while no differences in lipid goal achievement in the various age categories were observed.

Discussion

SURF-RA is the first international audit of CVD risk management in patients with RA including data on risk factor levels and management from three world regions. We report considerable geographical differences in estimated CVD risk, initiated preventive treatment, and attainment of guideline-recommended goals for BP and lipids in RA patients, in line with what has previously been reported for patients with CHD.¹³ In our data, a large proportion of Central and Eastern European and North American patients had high or very high estimated risk of CVD. Interestingly, the highest fraction with goal attainments for LLT and antihypertensive treatment was found in North America, in contrast to poor treatment results in Central and Eastern Europe. In fact, data from a recent large nationwide polish study showed that there was a high prevalence of CVD and CVD risk factors also in non-RA persons from this region.¹⁴ There are several possible explanations for the geographical variations in estimated CVD risk and treatment of risk factors, including differences in health care resources and systems, drug availability, national guidance on preventive therapy, and finally differences between adopted guidelines.

Approximately half of the patients in the SURF-RA had an indication for LLT based on estimated CVD risk, and indeed, the actual proportion may be even higher considering previous reports on inaccurate CVD risk prediction in patients with RA.^15,16 Unfortunately, attempts to develop an RA-specific CVD risk calculator have failed to improve risk prediction compared with already established CVD algorithms.^17,18 Hence, further efforts to more accurately predict CVD risk in patients with RA are needed.

The benefit of LLT in reducing CVD morbidity and mortality is well documented in the general population.^19,20 One large statin trial has been performed in patients with RA, the Trial of Atorvastatin for the Primary Prevention of Cardiovascular Events in Patients with Rheumatoid Arthritis (TRACE RA). The study was unfortunately terminated prematurely due to lower CVD event rates than anticipated.²¹ Although not statistically significant, TRACE RA did demonstrate a 34% relative risk reduction of major CVD events, which is comparable to reports from other patient populations.¹⁹ Moreover, a post-hoc analysis of two large statin trials confirmed a comparable effect of statins in terms of reduction of LDL-c and CVD events in patients with and without RA.²² The LDL-c hypothesis has been confirmed by several studies over the last years; the lower the LDL-c, the better CVD outcome. This may be achieved by adding ezetimibe, a cholesterol absorption inhibitor, to the maximally tolerated statin dose.²³ Taking into consideration the evidence regarding efficacy of LDL-c lowering on CVD outcomes in patients with RA, optimal statin dosing and increasing the modest use of other LLT than statins observed in the SURF-RA is a potential area of improvement. In many countries, CVD risk evaluation is performed by primary care physicians. Patients with RA do not see their primary care physician as often as other individuals, possibly due to regular consultations at the rheumatology department.²⁴ One may also speculate that lack of communication between rheumatologists and cardiologists/lipidologists/hypertensiologists contributes to the poor handling of CVD risk in patients with RA. Other possible reasons for why only about half of the SUF-RA patients with indication for LLT actually received such treatment may be the fair of drug–drug interactions and statin intolerance that may be more prevalent in patients with rheumatic diseases. However, in a position paper on statin intolerance, it is stated that statins should always be considered in patients with rheumatic diseases and high CVD risk.²⁵

The evidence regarding CVD risk management in RA is inconsistent. In a Swedish cohort study, the risk of recurrent acute coronary syndromes and mortality was increased among patients with RA compared with non-RA subjects, yet the authors found no discrepancies in prescribed secondary preventive medication between the groups.²⁶ Similarly, a UK general practice study reported no differences in CVD risk management between RA and non-RA persons.²⁷ Data from the USA confirmed that traditional CVD risk factors are not less aggressively managed in patients with RA compared with controls.²⁸ In contrast, significant underuse of statins prescribed as primary prevention has been reported in both the USA and the UK.^29,30 Moreover, results from a Danish nationwide study showed a lower rate of preventive therapy initiation after myocardial infarction in RA patients compared with non-RA subjects.³¹ In our study, the degree of lipid goal achievement (18%) was lower than what has been reported for CHD patients without RA (32% in men and 23% in women).³² Another recent study confirmed that there are still gaps between guideline recommendations and clinical practice for lipid management across Europe, which actually will be exacerbated due to more aggressive treatment goals in the 2019 guidelines.^33,34 There are many possible reasons for the inconsistent evidence regarding CVD management in RA compared with non-RA individuals, including different practices in various geographical regions, in general vs. specialist health care, and variations in CVD risk prediction and treatment target recommendations. However, a lipid goal attainment of only 18%, as seen in the RA patients at the highest risk of CVD in SURF-RA, is undoubtedly an area that needs improvement, especially since we have previously shown that LDL-c goal attainment for secondary prevention purposes is feasible in almost three-fourths of RA patients.³⁵ Further, we have demonstrated that lipid goal attainment was not influenced by inflammation or use of anti-rheumatic medications.³⁶ This is in line with the SURF-RA data in that the level of RA disease activity was not associated with either BP or lipid goal attainment.

Bartels et al. found that RA patients had ∼30% lower hazard ratio of being diagnosed with hypertension than non-RA persons.³⁷ In the present study, we found a prevalence of hypertension at around 60% in RA patients, slightly lower than the estimate reported from the UK in 2007 (70%).³⁸ While this difference may be explained by geographical variations, it may also be a result of more attention on the beneficial effects of BP lowering over the last decade. The UK study showed further that only 22% of hypertensive RA patients who were on antihypertensive treatment had reached BP target,³⁸ which is substantially lower than the BP goal attainment of 63% in SURF-RA. Taking into account all patients with an indication for antihypertensive treatment (both those using and not using antihypertensive treatment), the rate of BP goal attainment was ∼50%. This latter estimate is comparable to data from the general population.³⁹ In the latest European guidelines on hypertension, poor adherence to prescribed antihypertensive treatment is pointed out as a key hindrance to BP goal attainment.¹⁰ To our knowledge, no such studies discussing compliance of antihypertensive treatment in patients with RA have been published. However, increased focus on drug adherence is likely to raise the proportion of RA patients who reach BP target. European guidelines recommend a dual combination of antihypertensive medication as first-line treatment, except in fragile elderly and those at an overall low risk of CVD.¹⁰ Interestingly, our data demonstrate that only a minor proportion of RA patients received such combination therapy. We argue that increased use of dual or triple combinations of antihypertensive drugs may improve BP goal attainment among patients with RA.

The prevalence of CVD in the SURF-RA was relatively low at 13%. This may be due to a global trend towards reduced CVD mortality that has been reported during the last few years for both the general population and patients with RA.^5,6 The mean age of the study population was ∼60 years and almost 75% of the patients were females, which may also possibly explain the relatively low CVD prevalence. As for non-RA persons, it has previously been shown that there is a sex difference in CVD event rates in patients with RA independent of traditional CVD risk factors and markers of RA disease activity.⁴⁰

The SURF-RA audit has limitations; the analyses included data from both established cohorts and consecutively recruited patients, which may represent differences in RA disease activity, treatment, etc. This non-systematic data collection and inclusion difference may have represented a bias regarding estimated risk of CVD due to age and inflammation level, but it is less likely that cohort differences have affected adherence to treatment guidelines on hypertension and dyslipidaemia. Differences in data collection including missing data may also represent a potential source of bias. Furthermore, we employed the same definitions of CVD risk calculation and treatment targets (i.e. European guidelines) to enable uniform comparison of data from different regions. However, comparisons were hampered since several centres were non-European and thus have had other guidelines with diverse treatment goals to adhere to. Variations in the number of patients recruited from different countries, lack of data from several countries (e.g. from Poland, which is the largest country in Central and Eastern Europe), and also discrepancies in CVD risk between countries representing each region are limitations to the SURF-RA data. As for all studies including patient reported variables, there was a possibility of responder bias. Finally, the low reported prevalence of CVD at 2.5% in the Mexican population may partly be explained by lower age, higher proportion of females (92%), and low smoking prevalence compared with the other investigated regions, but there is also a possibility of a selection bias of included patients from Mexico. The SURF-RA also has several strengths, including a large sample size and data from different world regions, which are lacking in other audits on CVD risk management performed in patients with RA.

Conclusions

This large international survey on RA patients, the SURF-RA, revealed considerable geographical differences in estimated CVD risk and preventive treatment among different world regions. Special attention should be paid to enhance preventive treatment for patient with RA in Central and Eastern Europe. We observed lower goal attainment for LLT than what has previously been reported for individuals with CHD, and only half of the patients had obtained BP goal. Despite a high focus on the increased CVD risk in patients with RA over the last decade, there is still a substantial potential for improvement in CVD preventive measures. We recommend to establish protocols and standard operating procedures to assist health care professionals looking after RA patients.

Conflict of interest: S.R. and A.G.S. disclose collaborative agreement for independent research support from Eli Lilly who had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. C.A.H. has received grants from Pfizer Canada (unrelated research work), USB Canada (unrelated research work), and Covid Immunity Task Force (unrelated research work). A.B. is Chair of Government Formulary List Advisory Committee (Malta). P.D. has received honoraria from Pfizer, Lilly, and Galapagos. M.T. has received a grant from Ministry of Health Czech Republic (023723, NV18-01-00161A). A.G.S. has received honoraria for lectures from AbbVie, Novartis, Bayer, Eli Lilly, and Sanofi. E.I., G.W., J.S., C.S.C., P.v.R., G.D.K., I.G., S.R.D., G.K., E.M., M.A.G.-G., P.P.S., M.G.T., A.L., D.V., B.K., M.S.S., V.P.-R., D.A.G.-D., P.F., D.P.M., R.M., E.M.M., D.G., S.M., L.K., T.P., A.T., M.V., J.L., P.H., and H.M. have nothing to disclose.

Data availability statement: The data underlying this article will be shared on reasonable request to the corresponding author.

References