Effects of Hydroxychloroquine on endOthelial function in eLDerly with sleep apnea (HOLD): study protocol for a randomized clinical trial

In a meta-analysis, the prevalence of sleep apnea in the general adult population ranges from 6% to 17%, reaching 49% at advanced ages [1]. In a Brazilian epidemiological study, OSA affected 80 to 95% of the elderly [2]. The most damaging consequence of OSA is an increase in cardiovascular morbidity and mortality [3]. Cardiovascular diseases are responsible for 20% of deaths of Brazilians over 30 years old and lead to increased morbidity and mortality in over 30% of the elderly [4]. Prospective population studies have shown that individuals with severe untreated sleep apnea have a higher risk of general and cardiovascular mortality, independent of traditional cardiovascular risk factors [5]. Apnea patients, compared to the general population, are twice as likely to have hypertension [6], ischemic heart disease [7], and cerebrovascular disease [8]. In 2004, the National Heart, Lung, and Blood Institute committee proposed lines of research to enlighten the consequences of sleep disorders in heart disease [9]. In 2008, the American Heart Association and the American College of Cardiology published a consensus indicating the need for researches like ours that addresses the major cardiovascular consequences of sleep apnea [10]. The main reasons why sleep apnea is associated with cardiovascular diseases are the periods of intermittent hypoxia and the repeated nocturnal awakenings precipitated by apneas. Apnea occurs when anatomical and/or functional changes in the airway, associated with loss of tone in the pharyngeal dilator muscles during sleep, lead to airway collapse and interruption of airflow. Nighttime awakenings cause chronic sympathetic hyperactivity and intermittent hypoxia causes oxidative stress, which causes inflammation, the onset of endothelial dysfunction and, later, atherosclerosis. The association between coronary artery disease and sleep apnea appears to be consistent. In a cross-sectional analysis of the Sleep Heart Health Study cohort, individuals in the highest quartile of apnea-hypopnea index present a 27% risk increase for coronary artery disease. Sleep apnea increases the chance of having ischemic heart disease by 65% [11] and having a heart attack by 71% [12]. Our group demonstrated that sleep apnea is a more robust risk factor for coronary artery disease than classic factors such as cholesterol, in a sample excluding morbidly obese, diabetic, and smokers [13].

The endothelium, by isolated area, is one of the largest organs in the body, composed of trillions of cells, weighing more than 1 kg and covering almost 3 m² in a 70-kg adult male [14]. It interacts with multiple systems and has been implicated in the pathogenesis of neurological, renal, liver, vascular, dermatological, immunological, and cardiovascular diseases. It is a highly specialized tissue, responsible for vascular homeostasis through the regulation of arteriolar tone, platelet aggregation, smooth muscle cell growth, and leukocyte adhesion. The endothelium regulates the vascular tone through the secretion vasodilator, such as nitric oxide (NO), and vasoconstrictors, such as endothelin. Damage to the endothelium and dysfunction of this tissue are associated with the development of arterial hypertension and atherosclerosis. There are many molecular and cellular mechanisms involved with endothelial damage and vascular aging. Since inflammation and oxidative stress are part of these mechanisms, tackling these two processes can be considered as future therapeutic targets for the prevention of cardiovascular disease [15]. A review article on the topic states that a greater understanding of endothelial function provides not only a grasp of the pathophysiology of cardiovascular disease but also an opportunity for clinical treatment, early detection of diseases, stratification of cardiovascular risk, and evaluation of therapeutic response [16].

The assessment of flow-mediated dilation (FMD) was first introduced in the 1990s as a non-invasive approach to examining vasodilator function in vivo. The FMD result quantifies the endothelium-dependent arterial function mediated by nitric oxide, being used as an indirect marker of vascular health [17]. A meta-analysis that included 32 studies and 15,000 individuals concluded that the FMD result is an independent predictor of cardiovascular events and death. Each 1% dilation increase in the FMD result was associated with a 10% lower risk of cardiovascular event or death. In that study, the predictive effect of brachial FMD was more substantial for cardiovascular mortality than for general mortality, suggesting that impaired endothelial function is predominantly a cardiovascular risk factor [18].

The method has limitations that must be considered. Small changes in the methodological approach can critically influence results and decrease exam reproducibility. Therefore, the compliance of guidelines with updated and standardized methodology is indicated to reduce measurement errors and improve FMD reliability in clinical studies [19].

Peripheral arterial tonometry (PAT) is a non-invasive method of assessing endothelial function in clinical practice. The method was incorporated into several population and multicenter studies, such as the Framingham Heart Study. The results are based on digital pulse amplitude variation during reactive hyperemia induced by a 5-min forearm cuff occlusion. It is operator/interpreter independent and provides immediate results, which are advantages compared to FMD. The reactive hyperemia index (RHI) obtained by PAT is considered to correlate significantly with coronary endothelial function [20]. PAT is a predictor of cardiovascular risk in high-risk patients, according to several authors [21‐24]. The method also has good reproducibility, according to studies by Brant et al and Reisner et al. [25] In two large population studies, totaling more than 10,000 participants, there is only a modest correlation between FMD and PAT [26, 27]. Therefore, both methods will be used in this study, in order to quantify aspects of both macro- and microvascular circulation and compare the two methods.

Sleep apnea is related to systemic inflammation, oxidative stress, and endothelial dysfunction [28, 29]. These factors are protagonists in the process that leads patients with sleep apnea to develop cardiovascular disease. The association between endothelial dysfunction and sleep apnea was initially considered ambiguous because there are numerous common features among patients with endothelial dysfunction and sleep apnea. Age, obesity, smoking, and alcohol consumption were factors cited as potential causes of the two conditions, confusing a possible cause-effect relationship. A meta-analysis by Wang et al. concluded that moderate/severe sleep apnea was significantly associated with endothelial dysfunction, increased arterial stiffness and increased serum levels of inflammatory markers. The meta-regression data suggest that the adverse effect of moderate-severe sleep apnea on endothelial function is not modified by potential confounders, such as body mass index [30]. Cross-sectional analyses of population studies and case-control studies consistently demonstrated an association between obstructive sleep apnea and impaired endothelium-dependent vasodilation [31]. In vitro experiments demonstrate that the serum of patients with sleep apnea impairs the migration of coronary endothelial cells [32]. Our group published studies demonstrating that intermittent hypoxia causes oxidative stress [33], inflammation, and damage to lipids and proteins [34].

Hydroxychloroquine was initially an antimalarial drug. Due to its anti-inflammatory properties, it is commonly used to treat rheumatic diseases, in particular rheumatoid arthritis and connective tissue diseases such as systemic lupus erythematous. HCQ reduces the activation of the innate immunity system by inhibiting the stimulation of Toll-like receptors [35], which can play an important role in the activation of inflammatory cells in atherosclerotic patients [36]. Some studies suggest that hydroxychloroquine also reduces the production of cytokines important in the pathogenesis of atherosclerosis, such as interleukin-1 and 6 and tumor necrosis factor alpha (TNF-α) [37, 38]. TNF-α block therapy was associated to a risk reduction of coronary artery disease among patients with rheumatoid arthritis [39].

In addition to its anti-inflammatory effects, HCQ has other properties that may be beneficial in the treatment of coronary artery disease [40]. In patients with lupus, chloroquine inhibits the synthesis of several members of the matrix metalloproteinase family, especially MMP-9 [41], enzymes capable of degrading interstitial collagen in the fibrotic layer of the atherosclerotic plaque [42].

Cohort studies have shown that HCQ lowers cholesterol in patients with lupus and rheumatoid arthritis [43‐46]. In addition, in rheumatoid arthritis, the use of HCQ has been associated with a lower incidence of type 2 diabetes [47‐49], lower levels of glycosylated hemoglobin (HbA1c) [50], and better sensitivity to insulin and beta cell function [51]. In two randomized studies of diabetics with inadequate glycemic control, HCQ significantly reduced glucose levels [52, 53].

In addition, HCQ may have antithrombotic properties [54, 55]. In mice, HCQ reduced the size and duration of the thrombus and caused a decrease in the thickness of the vascular wall and the progression of atherosclerosis [56, 57]. In a case-control study, among patients with lupus, the use of HCQ was associated with a reduced risk of thromboembolic complications [58].

Sharma et al. reported in a retrospective study involving 1266 patients with rheumatoid arthritis an association between HCQ use and a 72% reduced risk of an outcome composed of acute coronary syndrome, cardiac revascularization, stroke, transient ischemic accident, peripheral arterial disease, and sudden death [59].

Evidence from studies in patients with rheumatic diseases converges to a role for HCQ in reducing cardiovascular risk. However, its cardiovascular effects in patients at higher cardiovascular risk but without rheumatic diseases are unknown.

Some common side effects of hydroxychloroquine, occurring in 1-10% of users, include anorexia, emotional lability, headache, blurred vision, abdominal pain, nausea, rash, and itching. Rarely, in 0.1–1% of high-dose users, Stevens-Johnson syndrome, cardiomyopathy, and retinopathy may occur.

Considering that sleep apnea is a chronic disease with an inflammatory component, as well as rheumatic conditions, HCQ, due to its favorable effects in the reduction of several cardiovascular risk factors, may improve endothelial dysfunction associated with sleep apnea.

Considering the low cost of HCQ, the high prevalence of sleep apnea in the elderly, and the high morbidity and mortality of cardiovascular diseases, the search for feasible approaches in primary care that can reduce these outcomes is justified. Thus, the use of HCQ in patients with sleep apnea and at high risk for coronary artery disease represents an entirely new approach to this disease.

To test in a randomized clinical trial the effect of hydroxychloroquine on endothelial function and its correlates, in elderly people with sleep apnea.

Biochemical assessment will be performed on all participants at the beginning of the study and after 8 weeks of treatment. The exams will be held early in the morning, with participants fasting. Blood samples will be collected in five different sampling tubes with a capacity of 9 ml. The samples will be taken to the laboratory to measure plasma levels of ultra-sensitive C-reactive protein, total cholesterol, HDL-cholesterol, triglycerides, glucose, and glycated hemoglobin.

The microvascular endothelial function will be determined by an automatic device (EndoPAT2000, Itamar Medical, Cesarea, Israel), with the technique already described. The cuff will be placed on the non-dominant arm, 2 cm above the cubital fossa, and RH-PAT tests on the pulp of each index finger will be performed. The basal pulse amplitude will be measured for 5 min. The arterial blood flow will be interrupted on one side for 5 min by insufflation of the cuff, with an occlusion pressure of 200 mmHg, or 60 mmHg above systolic blood pressure. After 5 min of occlusion, the cuff is deflated and reactive hyperemia, the RH-PAT sign, is registered in both hands for an additional 5 min. The contralateral finger is a control for changes in systemic vasodilation. The reactive hyperemia index (RHI) is automatically calculated by the RH-PAT equipment according to the manufacturer’s formula. RHI is defined as the ratio between post-deflation pulse amplitude, 90 to 150 s after the cuff is released, and the mean baseline pulse amplitude. This result is divided by the corresponding proportion of the control finger, and multiplied by a correction factor in relation to the baseline. Low RHI values are related to an inadequate endothelial vasodilator response.

High-resolution ultrasound equipment (Esaote) will be used to assess arterial endothelial function. A high frequency transducer will be used to obtain longitudinal images of the brachial artery walls. The images of diameter and arterial flow, at each moment of the protocol, will be recorded simultaneously on a computer with the aid of a capture card (Easycap). To minimize operational errors, both the transducer and the subject's arm will be positioned and maintained in the same position during the examination. Baseline images will be recorded for 1 min and then the pressure cuff on the forearm will be inflated to 200 mmHg and maintained for 5 min, characterizing reactive hyperemia. After deflating the cuff, 3 min of endothelium-dependent brachial artery dilation will be recorded. After the exam, the images recorded in MPEG will be converted to MP4 and then analyzed using specific software (Cardiovascular Suite) for the diameter and arterial flow before and after reactive hyperemia.

The portable sleep monitoring device Somnocheck Micro (Weinmann Medical Technology, Hamburg, Germany) will be attached to the patient’s wrist. The patient will also use a nasal cannula to register airflow. The device will be programmed to turn on at 22 pm and turn off at 10 am. A combination of photoplethysmography-derived pulse wave analysis and respiratory flow signals may enable differentiation between obstructive and central apnea and provide information regarding the extent of sleep fragmentation. In detail, respiratory effort derives by analyzing fluctuations of the pulse wave analysis signal caused by intrathoracic pressure changes during spontaneous breathing cycles. Central apneas, RERAs, mean and minimum oxygen saturation, and the apnea-hypopnea index will be reported according to the rules of the American Academy of Sleep Medicine. The portable sleep monitoring exam will be performed before and after 8 weeks of treatment.