Abstract
Moderate–severe obstructive sleep apnoea (OSA) has been associated with several pro-atherogenic mechanisms and increased cardiovascular risk, but it is not known if minimally symptomatic OSA has similar effects. Circulating cell-derived microparticles have been shown to have pro-inflammatory, pro-coagulant and endothelial function-impairing effects, as well as to predict subclinical atherosclerosis and cardiovascular risk.
In 57 patients with minimally symptomatic OSA, and 15 closely matched control subjects without OSA, AnnexinV-positive, platelet-, leukocyte- and endothelial cell-derived microparticles were measured by flow cytometry.
In patients with OSA, median (interquartile range) levels of AnnexinV-positive microparticles were significantly elevated compared with control subjects: 2,586 (1,566–3,964) μL−1 versus 1,206 (474–2,501) μL−1, respectively. Levels of platelet-derived and leukocyte-derived microparticles were also significantly higher in patients with OSA (2,267 (1,102–3,592) μL−1 and 20 (14–31) μL−1, respectively) compared with control subjects (925 (328–2,068) μL−1 and 15 (5–23) μL−1, respectively). Endothelial cell-derived microparticle levels were similar in patients with OSA compared with control subjects (13 (8–25) μL−1 versus 11 (6–17) μL−1).
In patients with minimally symptomatic obstructive sleep apnoea, levels of AnnexinV-positive, platelet- and leukocyte-derived microparticles are elevated when compared with closely matched control subjects without obstructive sleep apnoea. These findings suggest that these patients may be at increased cardiovascular risk, despite being minimally symptomatic.
Obstructive sleep apnoea (OSA) is characterised by repetitive apnoeas/hypopnoeas during sleep, which are associated with oxygen desaturations and arousals from sleep. It has been estimated that 2–4% of the adult population in Western countries suffer from symptomatic OSA, and it is becoming more prevalent as the average population body weight rises 1. The prevalence of minimally symptomatic OSA among middle-aged adults has been shown to be as high as 26% 1, making OSA one of the most frequent disorders, and of major epidemiologic interest.
Cross-sectional and prospective studies have implicated OSA as an important potential causal factor in the development of cardiovascular disease 2. The mechanisms underlying the association between OSA and cardiovascular disease are currently not fully understood. Multiple causal factors leading to vessel wall damage and atherosclerotic plaques have been proposed, including reflex sympathetic activation, consequent increases in blood pressure, endothelial dysfunction, activation of platelets, systemic inflammation, reactive oxygen species and hypoxia-induced apoptosis of endothelial cells 3–6.
Microparticles (MPs) circulating in peripheral blood originate from the plasma membrane of diverse activated or apoptotic cells of platelet, leukocyte and endothelial origin, and are involved in the pathogenesis of vascular inflammation, thrombosis, endothelial dysfunction and atherosclerosis 7, 8. Plasma levels of MPs have been shown to be elevated in patients with cardiovascular risk factors 9, 10, have been shown to predict subclinical atherosclerosis burden in asymptomatic subjects 11, and appear to be a predictor of the recurrence of myocardial infarction or death in patients with acute coronary syndromes 12. As a result, there is growing interest in early detection of these novel measures of cardiovascular risk with the anticipation that an intervention could reduce their levels and, thus, benefit patients who are at increased risk for future cardiovascular events.
However, there are no studies measuring circulating MPs in patients with highly prevalent minimally symptomatic OSA, and none investigating diverse subgroups of MPs originating from platelets, leukocytes and endothelial cells in patients with OSA.
To explore this novel hypothesis regarding the mechanism of vascular damage in OSA, a cross-sectional study measuring circulating cell-derived MPs was performed in patients with minimally symptomatic OSA, and compared with a closely matched control group.
METHODS
Patients
Patients with possible OSA were referred to the Oxford Sleep Unit, Oxford Centre for Respiratory Medicine (Oxford, UK), by general practitioners, ear, nose and throat surgeons, or other hospital consultants, because of severe snoring or witnessed apnoeas. Patients were eligible for the study if they were aged 45–75 yrs, had proven OSA with a severity defined as >7.5 oxygen desaturations of >4% per hour (oxygen desaturation index (ODI) >7.5 events·h−1), and no history of excessive daytime sleepiness or any other daytime symptom of OSA (such as tiredness or fatigue) which would have justified a trial of continuous positive airway pressure therapy. The level of symptoms was assessed at clinical interview, with no specific upper threshold for the Epworth sleepiness scale (ESS) value to exclude entry. All eligible patients were offered participation in the study.
Control subjects were mainly identified and recruited from a general practitioner's database and were eligible for the study if they were aged 45–75 yrs with no history of OSA, had <5 oxygen desaturations of >4% per hour (ODI <5 events·h−1) and an apnoea/hypopnoea index <5 events·h−1. Control subjects were matched to patients with OSA for sex, age, body mass index (BMI), waist to hip ratio and cardiovascular comorbidities. The study was approved by the Oxford research ethics committee (REC number 05/Q1604/159), and written informed consent was obtained from all participants.
Sleep study and assessment of sleepiness
In patients, OSA was diagnosed from a one-night in-hospital respiratory polygraphic sleep study. Patients' body movements, heart rate and pulse transit time (PTT) changes were recorded as measures of arousal from sleep. Pulse oximetry, snoring and increases in the respiratory swing in PTT were used as markers of breathing pattern and respiratory effort (Win-Visi monitoring system; Stowood Scientific Instruments, Oxford, UK) as previously described and validated 13, 14.
The results of the sleep study were scored automatically, with manual review to ensure accuracy of the data. OSA was diagnosed from a review of all data, and severity was quantified as the number of oxygen desaturations >4% per hour of study (ODI). In control subjects, OSA was excluded by home sleep studies using the apnoeaLinkTM device (MAP medicine technology; ResMed, Martinsried, Germany). This device records the patient's nasal respiratory pressure signal and finger oximetry during sleep; it has been validated as an accurate instrument for the detection of snoring, apnoea/hypopnoea and oxygen desaturations 15. The results of the sleep study were scored automatically with dedicated software (MAP medicine technology; ResMed), with manual review to ensure accuracy of the data. Apnoeas were defined as a cessation of airflow lasting >10 s, and hypopnoeas as a reduction in airflow of ≥50% lasting >10 s, associated with a drop in oxygen saturation of >4%.
Comparisons of measurements of the ODI in seven control subjects on two different nights by both sleep study methods (Win-Visi monitoring system and apnoeaLinkTM) revealed a small and nonsignificant bias (mean difference in ODI was -0.29 desaturations·h−1; p = 0.72); all seven subjects had an ODI of <5 events·h−1 in both sleep studies.
Subjective sleepiness was assessed using the ESS, which assesses the tendency to fall asleep during eight typical daytime situations 16. Objective sleepiness was measured with one sleep resistance challenge (Osler test), which tests the ability to stay awake in a darkened and sound isolated room 17.
Cardiovascular risk score
A cardiovascular risk score (Framingham index) was used to objectively assess an individual’s 5-yr risk of death due to cardiovascular events 18. The risk score is based on 11 factors, including: age, sex, systolic blood pressure, serum total cholesterol concentration, height, serum creatinine concentration, cigarette smoking, diabetes, left ventricular hypertrophy (evaluated by resting ECG; Sokolow-Lyon index; SV1+RV5/V6 >3.5 mV), history of stroke and myocardial infarction. The risk score is an integer, with points added for each factor according to its association with risk. The sum score, and the corresponding risk of a fatal cardiovascular event within 5 yrs, were derived from individual patient data according to Pocock et al. 18.
Office blood pressure was measured in the sitting position after a period of rest for 5 min with a standard digital automatic monitor (Omron Healthcare Company, Kyoto, Japan). The mean value of three readings was used for analysis.
Measurement of MPs
MPs were measured by flow cytometry as previously described by Biro et al. 19 and blind of the patient group.
Briefly, blood was drawn from fasting participants in the morning between 09:00 h and 10:00 h using a 19 G needle. Within 1 min of the blood being taken, the tubes were centrifuged at 1,550×g for 20 min at 20°C, in order to produce platelet-free plasma. 250 μL of plasma was aliquoted, frozen immediately and stored at -80°C.
The 250-μL samples were thawed on melting ice for 1 h and then spun at 18,000×g for 30 min at 20°C. 225 μL of supernatant was removed, and 225 μL of PBS–citrate, 0.32%, was added. The samples were spun again at 18,000×g for 30 min at 20°C. 225 μL of supernatant was again removed and replaced with 75 μL of PBS–citrate 0.32%. The diluted monoclonal antibodies (MAbs) were spun again at 18,000×g for 5 min, to remove MAb–MAb complexes, and the supernatant transferred to a fresh tube.
Annexin-V fluoresceinisothiocyanate (AnnV-FITC; Becton Dickinson, Oxford, UK) was used to stain all phosphatidylserine-positive MPs of endothelial, platelet and lymphocyte origin. AnnV-FITC in the absence of calcium was used as a negative control.
CD31-phycoerythrin (CD31-PE), CD41-PE-Cy5 (Becton Dickinson) were used to differentiate between platelet-derived MPs (PMPs; CD31+CD41+) and endothelial cell-derived MPs (EMPs; CD31+CD41-). CD45-allophycocyanin (CD45-APC; Becton Dickinson) was used as a marker for lymphocyte-derived MPs (LMPs; CD45+). Appropriate PE, PE-Cy5 and APC isotypes were used as negative controls.
Each sample was stained in two tubes, the first containing AnnV-FITC, CD31-PE and CD41-PE-Cy5 to detect PMPs and EMPs, the second containing AnnV-FITC and CD45-APC, to detect LMPs.
Diluted AnnV-FITC or MAbs (5 μL) were added to PBS-calcium 2.5 mmol·L−1 to make up a total volume of 50 μL. 10 μL of sample was incubated with the appropriate MAbs for 30 min at room temperature, protected from light. After the incubation, 900 μL of PBS-calcium was added.
Acquisition of samples
Samples were acquired using FACSCalibur® (Becton Dickinson). Compensations were checked by acquiring single colour stained tubes. The MP gate was checked at each run by acquiring a tube containing 10 μL of MPs and 5 μL of 1/1,000 dilution of 1 μM beads (Sigma-Aldrich L-2778, St Louis, MO, USA) in 950 μL of PBS-calcium.
Data analysis
Data are expressed as mean±sd if data were normally distributed, and as median (interquartile range) if data were not normally distributed. Differences between the OSA and the control group in patient characteristics and MPs were compared by independent t-tests if data were normally distributed, or by the Mann–Whitney U-test if data were not normally distributed. For comparison of frequencies, Chi-squared test of independence was used. Spearman's rank test was used for correlation analysis. A p-value <0.05 was considered statistically significant.
RESULTS
Subject characteristics
A total of 57 patients with minimally symptomatic OSA and 15 control subjects without OSA were recruited. Control subjects were well matched for age, sex, BMI, waist to hip ratio and cardiovascular risk profile. There was no statistically significant difference in prescribed antihypertensive and cholesterol- or glucose-lowering medications between the two groups. As would be expected, patients with OSA had a higher ODI, a greater neck circumference, and tended to have a higher ESS, although objective sleepiness assessed by the Osler test was similar in the two groups (table 1⇓).
AnnexinV-positive MPs
The median level of AnnexinV+ MPs was significantly higher in patients with minimally symptomatic OSA (2,586 (1,566–3,964) μL−1) than in matched control subjects (1,206 (474–2,501) μL−1; p = 0.007), as illustrated in figure 1⇓.
There was no correlation between the level of AnnexinV+ MPs and ODI (r = 0.12, p = 0.31).
CD31+CD41+ PMPs
Patients with OSA had significantly higher median levels of PMPs (2,267 (1,102–3,592) μL−1) than matched control subjects without OSA (925 (328–2,068) μL−1; p = 0.009; fig. 2⇓). There was no correlation between the level of PMPs and ODI (r = 0.09, p = 0.43).
CD31+CD41- EMPs
Levels of peripheral blood EMPs were slightly higher in patients with OSA (13 (8–25) μL−1) compared with control subjects (11 (6–17) μL−1), although this difference did not reach statistical significance (p = 0.21; fig. 3⇓). There was no correlation between the level of EMPs and ODI (r = 0.12, p = 0.32).
CD45+ LMPs
The median level of LMPs was significantly higher in patients with minimally symptomatic OSA (20 (14–31) μL−1) than in matched control subjects (15 (5–23) μL−1; p = 0.02; fig. 4⇓). There was a statistically significant correlation between the level of LMPs and ODI (r = 0.30, p = 0.01).
DISCUSSION
To the best of the current authors’ knowledge, the present study is the first to look at circulating MPs originating from different cell types in patients with OSA. It was found that patients with minimally symptomatic OSA have higher blood levels of AnnexinV+ MPs, PMPs and LMPs when compared to a well-matched control group without OSA. Although levels of EMPs were slightly higher in patients with OSA compared with control subjects, this did not reach statistical significance.
There has been growing interest in circulating cell-derived MPs in recent years, as plasma levels of MPs have been shown to be elevated in patients with cardiovascular risk factors 9, 10, have been shown to predict subclinical atherosclerosis burden in asymptomatic subjects 11, and appear to be a predictor of the recurrence of myocardial infarction or death in patients with acute coronary syndromes 12. Therefore, quantification of MPs seems to be a valuable tool for the exploration of cardiovascular risk in asymptomatic patients.
The finding that patients with minimally symptomatic OSA had twice the plasma level of AnnexinV+ MPs compared with closely matched control subjects without OSA (fig. 1⇑) suggests that OSA could play an independent role in the formation of MPs. There are several possible links between OSA and the creation of MPs: platelets have been shown to release MPs after activation and in response to high vascular shear stress (e.g. during acute rises in blood pressure) 20, and endothelial cells and leukocytes release MP after activation by inflammatory cytokines (e.g. tumour necrosis factor-α), aggregated low-density lipoproteins or reactive oxygen species 21, 22. All of these mechanisms of MP formation have previously been shown to be associated with OSA 3, 23.
The present study found that levels of CD31+CD41+ PMPs were more than twice as high in OSA patients than in control subjects (fig. 2⇑). This is in contrast to the report of Geiser et al. 24, who found no difference in PMP levels between 12 patients with moderate–severe OSAS and six healthy controls. Possible explanations for this negative finding may be the small sample size and the different technique used to detect PMPs in the study by Geiser et al. 24.
PMPs have been shown to have pro-thrombogenic and/or pro-atherogenic effects in experimental vascular models by enhancing monocyte arrest on activated endothelium through P-selectin, glycoprotein IIb/IIIa, RANTES (Regulated upon Activation, Normal T-cell Expressed, and Secreted) and intercellular adhesion molecule (ICAM)-1 25, 26. In addition, PMPs express CD40L, which has been shown to stabilise arterial thrombi 27. As elevated levels of P-selectin, CD40L and ICAM-1 have been associated with OSA in previous reports 28, 29, the elevated PMP levels found in the present study may be a novel important link between these pro-atherogenic factors and OSA.
The two-fold elevation in PMP levels found in patients with OSA compared with the control group is comparable to the difference in PMP levels reported between patients with severe arterial hypertension and normotensive control subjects 9. Therefore, it appears that the difference found in the present study may be of clinical importance.
OSA is considered to represent a pro-inflammatory state, as evidenced by increased levels of pro-inflammatory proteins and cytokines, and thereby leading to endothelial dysfunction and atherosclerosis 23. The finding that LMP levels were higher in patients with minimally symptomatic OSA compared with control subjects (fig. 4⇑) suggests that even minimally symptomatic OSA may be associated with systemic inflammation, as it has been shown that LMP formation is enhanced by inflammatory stimuli 30; in turn, LMPs induce interleukin (IL)-6, IL-8 and monocyte chemotactic protein-1 production 31. The clinical significance of this finding is supported by a recently published study which reported that LMP levels predict subclinical atherosclerosis, as measured by plaque burden of carotid, abdominal aorta and femoral arteries in 216 asymptomatic patients without overt cardiovascular disease 11.
The level of EMPs has been shown to be an index of endothelial injury in patients with coronary artery disease and is possibly useful for identifying asymptomatic patients with diabetes mellitus at increased risk of coronary artery disease 32, 33. Although levels of CD31+CD41- EMPs were slightly higher in patients with OSA compared with control subjects in the present study, this difference did not reach statistical significance (fig. 3⇑). This in contrast to a report of El Solh et al. 6, who found increased levels of apoptotic endothelial cells in 14 patients with moderate–severe OSA when compared with healthy control subjects. It must be mentioned that in the latter study the authors aimed to measure apoptotic circulating endothelial cells rather than EMPs (apoptotic endothelial cells are considerably larger than MPs), and the detection of endothelial cells in their study was based on a relatively nonspecific marker (CD146), which is also recognised on other cell lines such as tumour, smooth muscle, dendritic and stroma cells 34.
However, a possible explanation for the negative finding in the present study could be that patients with minimally symptomatic OSA may not yet have structural damage of endothelial cells and, therefore, do not produce high levels of CD31+CD41- EMPs, which have been shown to reflect structural damage of endothelial cells rather than endothelial cell activation 35. Therefore, in future studies looking at EMPs in patients with OSA, markers that are more specific for endothelial cell activation, such as CD62E+ EMP 35, should be analysed as well. Furthermore, the relatively small number of EMPs in the plasma might have reduced the power to detect a significant difference between patients with OSA and the control group.
At this point it must be mentioned that, although the OSA and control group were matched as closely as possible, it still remains possible that there are other unmeasured and important differences between the two groups that might account for the differences in MPs found. Possible evidence of this is that OSA patients had a larger neck circumference, despite identical BMI and waist-to-hip ratio. This may indicate a difference in fat distribution that, in turn, may reflect metabolic differences. Therefore, the present study can only generate the hypothesis that minimally symptomatic OSA could raise the cardiovascular risk and that this is reflected in elevated MP levels. Controlled interventional trials are needed to see if such changes are reversible before cause and effect can be implied.
Furthermore, the question of whether circulating levels of MPs represent a cause or simply a marker of vascular disease in patients with OSA remains unsolved. It is possible that intermittent hypoxia may directly induce cell apoptosis and thereby leads to elevated levels of MPs.
In conclusion, the present study found elevated levels of total, platelet-derived and leukocyte-derived microparticles in patients with minimally symptomatic obstructive sleep apnoea when compared to well-matched control subjects without obstructive sleep apnoea, although this was not associated with a significantly increased level of endothelial cell-derived microparticles. The elevated levels of microparticles that was found may provide an important link between obstructive sleep apnoea and pro-atherogenic mechanisms such as thrombosis, inflammation and endothelial dysfunction. The findings suggest that patients with minimally symptomatic obstructive sleep apnoea might have increased cardiovascular risk, but any cause and effect relationship has to be proven in randomised controlled interventional studies.
Support statement
M. Kohler is a recipient of a European Respiratory Society (No 118) and University of Zurich research fellowship.
Statement of interest
A statement of interest for this study can be found at www.erj.ersjournals.com/misc/statements.shtml
- Received July 17, 2008.
- Accepted November 6, 2008.
- © ERS Journals Ltd