Abstract
Severity of oxygen desaturation is predictive of early atherosclerosis in obstructive sleep apnoea (OSA). Leukotriene (LT)B4 is a lipid mediator involved in atherogenesis.
In 40 non-obese OSA patients, free of a cardiovascular history, and 20 healthy volunteers, the following were evaluated: 1) LTB4 production by polymorphonuclear leukocytes (PMNs) stimulated with A23187; and 2) the relationships between LTB4 production and both OSA severity and infraclinical atherosclerosis markers. The effect of continuous positive airway pressure (CPAP) on LTB4 production was also studied. An overnight sleep study was followed by first-morning blood sampling. Isolated PMNs were stimulated with A23187 in order to induce LTB4 production, which was measured by liquid chromatography–tandem mass spectrometry. Carotid intima-media thickness (IMT) and luminal diameter were measured in subset groups of 28 OSA patients and 11 controls.
LTB4 production was increased in OSA patients compared with controls. LTB4 levels correlated with the mean and minimal arterial oxygen saturation (Sa,O2). LTB4 production correlated with luminal diameter data in patients with a mean Sa,O2 of ≤94% but not with IMT. Lastly, CPAP significantly reduced LTB4 production by 50%.
Leukotriene B4 production is increased in obstructive sleep apnoea in relation to oxygen desaturation. Leukotriene B4 could promote early vascular remodelling in moderate-to-severe hypoxic obstructive sleep apnoea patients.
Obstructive sleep apnoea (OSA) is characterised by recurrent episodes of partial or complete upper airway obstruction occurring during sleep. These episodes of upper airway obstruction are usually associated with a desaturation–reoxygenation sequence, which is an acknowledged detrimental stimulus for the cardiovascular system. Recent data indicate that OSA is associated with an increased prevalence of fatal and nonfatal cardiovascular events 1, and is an independent risk factor for death from any cause 2. Among the intermediary mechanisms that could explain the link between OSA and cardiovascular morbidity, the role of early atherosclerosis has been proposed. It has now been demonstrated that, even after adjustment for confounding factors, OSA per se may lead to atherosclerosis, and that the intensity of the vascular damage is more specifically related to the amount of nocturnal oxygen desaturation 3–5. Moreover, 4 months of continuous positive airway pressure (CPAP) application seems sufficient to partly reverse early atherosclerosis 6.
Leukotriene (LT)B4 is an inflammatory mediator that is derived from the 5-lipoxygenase (5-LO) pathway of arachidonic acid metabolism. LTB4 synthesis is initiated by the activation of 5-LO 7 and its subsequent interaction with the nuclear-membrane-bound 5-LO-activating protein (FLAP) 8 of inflammatory cells. In polymorphonuclear leukocytes (PMNs), the activation of 5-LO depends upon intracellular calcium concentration, which is increased by the addition of calcium ionophores 9. When released from cell membranes by the action of phospholipase A2, arachidonic acid is converted into 5-hydroperoxy-6,8,11,14-eicosatetraenoic acid by 5-LO, which also catalyses its further transformation to LTA4. In PMNs, LTA4 is then converted to LTB4 by LTA4 hydrolase. LTB4 then binds to specific LTB4 receptors (BLTs), namely BLT1 and BLT2, to elicit its biological effects 10, including stimulation of leukocyte chemotaxis, adhesion to vascular endothelium, and degranulation.
A recent growing body of evidence suggests a major role of the 5-LO pathway in the pathogenesis and progression of atherosclerosis. First, stimulated PMNs from individuals with a past history of myocardial infarction produce more LTB4 than PMNs from controls 11. In addition, expression of the 5-LO pathway (5-LO, FLAP, LTC4 synthase and cysteinyl LT receptors) is increased in atherosclerotic lesions at various stages of development in human aorta and coronary and carotid arteries 12, 13. Furthermore, recent human genetic studies have shown that a promoter variant of 5-LO is associated with an increase in carotid intima–media thickness (IMT) in healthy subjects 14, and certain FLAP haplotypes have been linked to an almost two-fold increase in risk of myocardial infarction or stroke 11, 15.
Few studies have assessed the role of local LTB4 production in OSA. These studies have been performed in children and have demonstrated an increased concentration of LTB4 in the upper airway lymphoid tissues of paediatric OSA patients compared with those with recurrent tonsillitis, as well as enhanced levels of LTB4 in the exhaled breath condensate of these children 16. The main objective of the present study was to compare LTB4 production by stimulated PMNs in a group of 40 OSA patients, free of any cardiovascular history and medications, to that of a control group of 20 healthy volunteers. The secondary objectives were: 1) to study the relationship between OSA severity and LTB4 production; 2) to evaluate the relationship between LTB4 production and validated markers of early atherosclerosis (carotid luminal diameter and IMT); and 3) to determine the effect of CPAP on LTB4 production.
METHODS
Population
Patients with newly diagnosed OSA (n = 47) were prospectively included in the present study, as well as 20 control subjects. Patients were referred to the sleep laboratory of Grenoble University Hospital (A. Michallon Hospital, Grenoble, France) for symptoms suggesting OSA. The controls were healthy volunteers who received no compensation for their participation in the present study. All patients and controls underwent full polysomnography. They were nonsmokers and were free of symptoms or a history of medical or surgical treatment for cardiovascular diseases. Exclusion criteria were as follows: known hypertension, disease potentially affecting blood pressure regulation (Parkinson’s disease, renal or cardiac transplantation, and severe cardiac heart failure), atrial fibrillation or frequent premature beats (≥10 beats·min−1), smoking, shift work, diabetes mellitus, asthma, chronic obstructive pulmonary disease, atopy, rhinitis, arthritis, oral appliances or maxillofacial surgery, or pharmacological treatment that could affect LT level. In order to minimise confounding risk factors for atherosclerosis, subjects aged >60 yrs and those with a body mass index (BMI) of >30 kg·m−2 were excluded.
The control group was free of any acute or chronic cardiovascular, inflammatory or sleep disorders, and of any medication.
Of the 47 OSA patients, 40 were matched for age, BMI and sex with the 20 control subjects for comparison of LTB4 production, and 15 were treated with CPAP for ≥3 months.
The present study was approved by the local ethics committee in accordance with the Declaration of Helsinki. All of the participants gave their written informed consent.
Polysomnography
The diagnosis of OSA was established by full polysomnography, which included recording of oronasal flow and thoracoabdominal movements, ECG, submental and pretibial electromyography, electro-oculography, electroencephalography (EEG) and transcutaneous measurement of arterial oxygen saturation (Sa,O2). An apnoea was defined as a complete cessation of airflow for ≥10 s and a hypopnoea as a reduction in the nasal pressure signal of ≥50% or a reduction of 30–50% associated with either oxygen desaturation of ≥3% or an EEG arousal (defined according to the Chicago report 17), both lasting for ≥10 s. The apnoea index (AI) was defined as the number of apnoeas (obstructive or mixed) per hour of sleep. The apnoea/hypopnoea index (AHI) was calculated as the total number of apnoeas and hypopnoeas (obstructive or mixed apnoeas plus obstructive hypopnoeas) per hour of sleep. The respiratory disturbance index (RDI) was calculated and defined as the total number of respiratory events (obstructive or mixed apnoeas, obstructive hypopnoeas and inspiratory flow limitation episodes) per hour of sleep (full polysomnography) or per hour of recording (polysomnography without EEG recording). Sleep apnoea was defined as an AHI ≥5 events·h−1 and symptoms or an RDI >15 events·h−1 17, 18.
Venous blood for stimulated PMN experiments was collected at 07:00 h, immediately following nocturnal polysomnographic recordings.
Carotid ultrasonography
Carotid ultrasonography was performed on 28 OSA patients and 11 controls as previously described 3. The right common carotid artery was studied in the long axis with a probe incidence permitting good-quality images. The images were recorded at end-diastole and end-systole and then stored on an optical disc for subsequent analysis by a specific validated program (TIMC laboratory of Grenoble University Hospital). The common carotid IMT and luminal diameter were automatically measured. Carotid ultrasonography was performed by the same sonographer, who was blinded to the other study data. Analysis of the carotid parameters, using the specific software, was performed by the same operator for the duration of the present study.
Isolation of human PMNs
Venous blood was drawn from all OSA patients and control subjects, and collected on citrate as anticoagulant. PMNs were isolated by dextran sedimentation, followed by Ficoll-PaqueTM PLUS centrifugation (GE Healthcare, Stockholm, Sweden) as previously described 19. Contaminating erythrocytes were eliminated by hypotonic lysis, and PMNs were washed in PBS (pH 7.4) containing 0.133 g·L−1 CaCl2 and 0.1 g·L−1 Mg2+ (Sigma, L’Isle d’Abeau, France). PMNs were finally resuspended in the same buffer at a concentration of 2×106 cells·mL−1. Cellular viability was >98% as judged by the trypan blue exclusion method.
Cell stimulation
PMNs (2×106 cells·mL−1) were incubated for 15 min at 37°C in the presence of 10 μM A23187 (Sigma) or vehicle as previously described 11. Incubations were stopped by centrifugation for 5 min at 5,000×g at 4°C, and the supernatants were stored at -80°C until subsequent analysis.
Quantification of LTB4 by liquid chromatography–tandem mass spectrometry
Quantification of LTB4 was performed by liquid chromatography–tandem mass spectrometry (LC-MS/MS) using a method adapted from a method previously described for LTE4 20. The measurement of LTB4 was performed on 400 μL centrifuged supernatant. Deuterated LTB4 (LTB4-d4; 2 ng) was added to each sample as an internal standard. Solid-phase extraction was performed as previously described 20. Methanolic extracts were dried under nitrogen flow at room temperature and reconstituted in 40 μL mobile phase (methanol and 10 mM ammonium formate; 80:20 volume:volume). After centrifugation, 10 μL of reconstituted extract were injected into the LC-MS/MS system previously described 20. The chromatographic separation was obtained on a 5-μm Kromasil C8 column (125×2 mm; Macherey-Nagel, Hoerdt, France) maintained at 30°C. The mobile phase consisted of 10 mM ammonium formate (phase A) and methanol (phase B) delivered at a flow rate of 200 μL·min−1 as follows: initial 50% B maintained for 6 min, then increased in a linear gradient to 80% B in 6 min and maintained at 80% B for 11 min.
MS/MS acquisitions were made in the negative-ion mode using multiple reaction monitoring, and monitoring the m/z transitions from 335.0 to 195.1 for LTB4 and from 339.1 to 196.9 for LTB4-d4. Calibration curves were constructed using weighted (1/x) linear least-square regression. The lower limit of quantification was 60 pg·mL−1 for LTB4.
Statistical analysis
Continuous data are presented as mean±sd, and noncontinuous data as n (%). Normality was evaluated using skewness and kurtosis tests. Comparisons between continuous variables were made using an unpaired t-test or Mann–Whitney U-test. Noncontinuous variables were compared using a Chi-squared test. Comparisons between the three groups (two OSA groups and control subjects) were made using ANOVA or the Kruskal–Wallis test; subsequent pairwise comparisons were performed using the Bonferroni or Kruskal–Wallis multiple comparison test. Correlations were analysed using the Pearson or Spearman rank test. A multiple regression analysis was performed taking into account the variables that correlated with the dependant variable LTB4. The differences between baseline and post-CPAP values were analysed by means of a paired t-test or the Wilcoxon signed-rank test. A p-value of <0.05 was considered significant.
RESULTS
LTB4 production stimulated by A23187
The baseline characteristics of the study population are described in table 1⇓. There were no significant differences between controls and OSA patients in terms of age, BMI, sex and blood pressure. Conversely, triglyceride levels were significantly higher in OSA patients than in control subjects. As expected, sleep respiratory disturbance parameters differed significantly between OSA patients and controls. Early markers of atherosclerosis (carotid diameter and IMT) did not differ significantly between groups.
PMNs stimulated with 10 μM A23187 produced LTB4, whereas unstimulated PMNs did not (data not shown). The production of LTB4 by PMNs stimulated with A23187 was increased in OSA patients compared with control subjects (14.3±4.7 versus 12.0±4.5 ng·mL−1; p<0.05).
As shown in table 2⇓, a significant correlation was found between LTB4 production and mean nocturnal Sa,O2, minimal nocturnal Sa,O2, percentage of time spent with an Sa,O2 of <90% and AHI in OSA patients. No significant correlation was observed between LTB4 production and age or metabolic variables (BMI, cholesterol, triglycerides, insulin, fasting glucose or homeostasis model assessment of insulin resistance index; see table 2⇓).
In an attempt to further define the relationship between either LTB4 production and hypoxia or LTB4 production and early signs of atherosclerosis, post hoc analyses were performed.
First, multiple regression analysis was conducted taking into account the variables correlated with the dependant variable LTB4 production (AHI and mean Sa,O2). Since there was a trend towards a correlation between low-density lipoprotein (LDL) cholesterol and LTB4 production (p = 0.09), this variable was also included in the model. This analysis yielded a model in which mean Sa,O2 was the strongest independent predictor of LTB4 production (p = 0.0006; p = 0.026 for LDL cholesterol and p = 0.75 for AHI). Therefore, OSA patients were stratified on the basis of mean Sa,O2. The median mean Sa,O2 in OSA patients (i.e. 94%) was used to separate the OSA patients into two groups: mild hypoxic OSA (mean Sa,O2 of >94%), and moderate-to-severe hypoxic OSA (mean Sa,O2 of ≤94%). The baseline characteristics of the two groups of OSA patients are detailed in table 1⇑. As shown in figure 1⇓ and table 1⇑, production of LTB4 by PMNs stimulated with A23187 was significantly higher in the moderate-to-severe hypoxic OSA group than in the mild hypoxic OSA group and control group.
Secondly, in severe hypoxic OSA patients, the influence of the increased production of LTB4 on early markers of atherosclerosis was investigated. As shown in figure 2⇓, LTB4 production correlated with the systolic (r = 0.55; p = 0.034) and diastolic (r = 0.54; p = 0.036) carotid diameters in severe hypoxic OSA. A significant correlation was also found between systolic luminal diameter and LTB4 production (r = 0.39; p<0.002) and diastolic diameter (r = 0.41; p = 0.01) in the whole population. Conversely, IMT did not correlated with LTB4 production in moderate-to-severe OSA patients (p = 0.79) or in the whole population (p = 0.96).
Effect of CPAP on LTB4 production in moderate-to-severe hypoxic OSA patients
LTB4 production was evaluated in 15 OSA patients (mean age 55±7 yrs) who were successfully treated with CPAP for ≥3 months (mean duration 178±96 days) and were regularly using their CPAP (5.2±1.3 h·night−1). As shown in table 3⇓, CPAP significantly decreased AI, AHI and RDI, increased both minimal and mean nocturnal Sa,O2, and decreased the percentage of time spent with an Sa,O2 of <90%. No significant change occurred in BMI after treatment with CPAP. Conversely, CPAP significantly decreased A23187-mediated LTB4 production (table 3⇓; fig. 3⇓). During the same period, LTB4 production remained unchanged in control subjects (11.4±1.2 and 12.4±3.2 ng·mL−1 at the beginning and end of the 3-month study period, respectively; n = 8). The production of LTB4 did not differ significantly between CPAP-treated OSA patients and controls (p = 0.26).
DISCUSSION
The present study represents the first demonstration of increased production of LTB4 in OSA in relation to nocturnal oxygen desaturation severity. Moreover, in moderate-to-severe hypoxic OSA patients, the enhanced production of LTB4 was associated with an increased carotid luminal diameter. Finally, 3 months of CPAP treatment significantly reduced LTB4 production. These results suggest that LTB4 could be one of the mediators relating oxygen desaturation severity and early vascular changes in OSA patients.
Previous studies demonstrated activation of the LTB4 pathway in patients with cardiovascular diseases. In particular, enhanced production of LTB4 by stimulated PMNs has been reported in patients with a past history of myocardial infarction or stroke 11. As in this previous study, production of LTB4 was evaluated by challenge with calcium ionophore. A23187 induces a rise in the intracellular calcium level of PMNs and the translocation of 5-LO from the cytosol to the nuclear membrane 9, thereby permitting the direct evaluation of 5-LO pathway activity independently of any receptor-dependent signalling pathway.
A classic issue in clinical research addressing cardiovascular consequences associated with OSA is confounding factors. The inclusion of obese OSA patients with severe desaturation is generally criticised owing to the prominent role of BMI. For example, several studies addressing oxidative stress or inflammation in obese OSA patients have demonstrated that obesity is the main contributor to these biological changes 21, 22. The classic means of avoiding this limitation is thus to match controls and OSA patients, which results in the inclusion of OSA patients exhibiting moderate oxygen desaturation. As already mentioned, it was decided to include only carefully selected middle-aged non-obese OSA patients and also to exclude patients exhibiting any cardiovascular events, including known hypertension, myocardial infarction and stroke. This strict selection permitted the rigorous comparison of LTB4 production in controls and OSA patients, being free of any confounding factor, and thus highlighting the specific role of even moderate intermittent hypoxia in LTB4 production in OSA. Although triglyceride levels were significantly higher in OSA patients, and total cholesterol tends to be increased in OSA patients, these factors did not correlate with LTB4 production, suggesting that they may not contribute to the increased production of LTB4 in OSA.
In the present study, it was demonstrated, on multivariate analysis, that the main determinant of increased LTB4 production was mean Sa,O2, suggesting that intermittent hypoxia, leading to oxygen desaturation, may play a major role in the increased LTB4 production evidenced in OSA patients. The desaturation–reoxygenation sequence is a typical pattern coupled with the majority of respiratory events in OSA patients. This sequence leads to oxidative/nitrosative stress, with production of reactive oxygen species 23 and reactive nitrogen species 24, which are the most important free radicals. The increased levels of reactive oxygen species contribute to the generation of adhesion molecules 25, activation of leukocytes 26 and production of systemic inflammation 27. Since 5-LO activity is regulated by the cellular redox status and reactive oxygen species 28, the increased production of reactive species in leukocytes from OSA patients 29 could contribute to the activation of the LTB4 pathway in OSA. Exposure of isolated PMNs to hypoxia/normoxia sequences is required to provide definite evidence regarding the role of intermittent hypoxia in LTB4 release.
Since increased production of LTB4 in moderate-to-severe hypoxic OSA patients was clearly demonstrated, and since LTB4 is a mediator of atherogenesis 30, 31, the potential relationship between LTB4 production and various markers of early vascular remodelling that have been demonstrated to be associated with infraclinical atherosclerosis was investigated. Carotid imaging was performed in a more limited group of controls and OSA patients but this subgroup did not differ significantly in terms of anthropometric variables and severity of sleep apnoea. Previous studies have reported increased carotid IMT in OSA patients 3, 4, 32. Having excluded other cardiovascular risk factors in the present carefully selected population is the probable explanation for the nonsignificant difference found between OSA patients and controls regarding IMT. Indeed, previous studies showing early signs of atherosclerosis in OSA have generally been performed in overweight patients (a BMI of 28.1±0.6 and 29.3±0.6 kg·m−2 in the studies of Minoguchi et al. 32 and Drager et al. 4, respectively). Similarly, an increased IMT was found in OSA by Silvestrini et al. 33; however, their studied population included smokers (22%), hypertensive subjects (65%) and patients with diabetes (17%). Finally, in the studies of both Drager et al. 4 and Baguet et al. 3, only OSA patients exhibiting the most severe oxygen desaturation showed carotid hypertrophy. IMT is an established predictor of atherosclerosis 34, but luminal diameter has also been recommended for measurement since there is evidence for an association between increased diameter and the early stages of vascular remodelling 35–37. Moreover, Drager et al. 3 have used the same parameter in assessing atherosclerosis in OSA. Interestingly, whereas carotid IMT was increased only in the most severe patients, carotid diameter was significantly higher in both moderate and severe patients. This suggests that carotid luminal diameter is a more sensitive marker of early atherosclerosis in OSA, and might explain why it was the only parameter that correlated with LTB4 production in the present study. A demonstration of a reduction in carotid diameter under CPAP would have strengthened these data, but such measurements were not available in the present study.
The crucial role of LTB4 in the early stages of atherogenesis is now well established 10. LTB4 is a potent chemoattractant that facilitates recruitment and endothelial cell adhesion of neutrophils to the inflammatory site and promotes recruitment of inflammatory cells into tissues. Recruitment of leukocytes and leukocyte invasion of the arterial wall are critical steps in the development of atherogenesis. Consistent with these findings, pharmacological blockade of the 5-LO pathway 38 prevents atherosclerosis development in mice, and genetic experiments have identified 5-LO as a major gene contributing to atherosclerosis susceptibility in mice 39.
If the hypoxic stress of OSA is a causal factor in promoting LTB4 pathway activation, then treatment with CPAP should reduce LTB4 formation. In the present study, it was shown that the production of LTB4 by stimulated PMNs is reduced after a 3-month minimum period of CPAP in compliant patients. During the same time, LTB4 production remained unchanged in control subjects, demonstrating reliability and reproducibility of these measurements. These data are consistent with a recent study showing that 4 months of treatment with CPAP reduces early signs of atherosclerosis 6. Thus further evidence is provided that, under conditions in which confounding factors and comorbid conditions are minimised, CPAP reduces LTB4 production and could thereby limit atherosclerosis development. With regard to the 40% of OSA patients noncompliant with CPAP treatment, targeting the LTB4 pathway could represent a new therapeutic strategy in the prevention of the cardiovascular consequences of OSA. However, this should be further validated in interventional studies.
In conclusion, leukotriene B4 production is increased in obstructive sleep apnoea patients, and correlates with the severity of oxygen desaturation. The present results are the first to suggest that leukotriene B4 could be a new candidate mediator for explaining the relationship between oxygen desaturation severity and early atherosclerosis in obstructive sleep apnoea patients.
Support statement
This study was supported by a grant from the Délégation Régionale à la Recherche Clinique du Centre Hospitalier Universitaire de Grenoble (Grenoble, France), the Conseil Scientifique de l’Association Nationale pour le Traitement À Domicile de l’Insuffisance Respiratoire Chronique (Paris, France) and the Académie Nationale de Médecine (Paris, France).
Statement of interest
None declared.
Acknowledgments
The authors are grateful to C. Nahum and K. Scalabrino for expert technical assistance and C. Deschaux for statistical analysis.
- Received October 17, 2007.
- Accepted February 11, 2008.
- © ERS Journals Ltd