Published ahead of print on August 6, 2003, doi:10.1164/rccm.200304-566OC
© 2003 American Thoracic Society Obstructive Sleep Apnea and Thoracic Aorta DissectionRespiratory Department, Cardiology Department, Neurophysiology Department, and Nephrology Department, Hospital Universitari Vall d'Hebron, Barcelona, Spain Correspondence and requests for reprints should be addressed to Gabriel Sampol, M.D., Servei de Pneumologia, Hospital Universitari Vall d'Hebron, Passeig Vall d'Hebron, 119-129, 08035 Barcelona, Spain. E-mail: gsampol{at}vhebron.net
Obstructive sleep apnea syndrome (OSAS) is a process that is associated with the development of arterial hypertension, the main risk factor for aortic dissection and during obstructive episodes of the upper airways with marked increases in transmural pressure of the aorta wall. The aim of this work was to study the association between aortic dissection and OSAS. Nineteen consecutive patients with thoracic aorta dissection and 19 hypertensive patients of similar age, sex, and body mass index were studied by clinical questionnaire and polysomnography. Snoring and nonrefreshing sleep were common in both groups. Thirteen patients (68%) from each group showed an apnea–hypopnea index of more than 5 per hour. However, patients with aortic dissection presented a higher apnea–hypopnea index (28 [30.3] versus 11.1 [10.4], p = 0.032). Seven patients with dissection presented an apnea–hypopnea index of more than 30 versus 1 patient in the control group (p = 0.042). Patients with thoracic aorta dissection presented a high prevalence of previously undiagnosed and frequently severe OSAS. Further studies, including this diagnosis as a prognostic variable in the follow-up of patients with aortic dissection, are required. Our results suggest that in patients with aortic dissection and symptoms consistent with OSAS, a sleep study should be considered in their clinical management.
Key Words: obstructive sleep apnea • sleep • thoracic aorta Aortic dissection constitutes a life-threatening medical emergency that is associated with high morbidity and mortality, with the most frequent complications being recurrent dissection and aortic dilation and rupture (1). It is important, therefore, to define factors that may influence its onset and/or evolution with the aim of incorporating new therapeutic strategies to improve the prognosis. Obstructive sleep apnea syndrome (OSAS) is a frequent process (2) that is characterized by episodes of occlusion or near occlusion of the pharynx during sleep with persistence of inspiratory effort during the interruption of airflow. These episodes are repetitive, and in severe cases, a large number occur every night. They produce a sympathetic activation that, as demonstrated in animal (3) and human (4–6) studies, can lead to the development of arterial hypertension, the main risk factor for aortic dissection (7). On the other hand, apneic episodes imply repeated inspiratory effort against an occluded airway, thereby determining strongly negative intrathoracic pressures. These negative pressures affect all intrathoracic structures and increase progressively during apneas (8). Furthermore, they are associated with marked cyclical fluctuations in sympathetic activity and blood pressure, which shows a progressive increase during apnea (9). Thus, in each apnea, OSAS patients present marked increases in transmural pressure of the aorta wall, determined by the difference between arterial pressure and intrathoracic extravascular pressure. These mechanisms were previously suggested by Cistulli and colleagues as being implicated in the frequent development of aortic dilation in patients with Marfan's syndrome after a high OSAS prevalence was detected in these patients (10) and an attenuation of this dilation with OSAS treatment using nasal continuous positive airway pressure (11). The aim of this study was to evaluate the presence of OSAS in a group of patients with thoracic aorta dissection. Because arterial hypertension is a risk factor for aortic dissection and, on the other hand, several studies have shown OSAS to be more prevalent in hypertensive patients (12–14), a control group of hypertensive patients was included.
Patients Dissection group. As a participating center in the International Registry of Acute Aortic Dissection, data on patients with this disease seen at our hospital are prospectively collected (7). Thirteen of the 41 consecutive patients with thoracic aortic dissection treated at our center between January 2000 and April 2001 died during admission. These 13 patients included 9 men and 4 women, mean age 63.5 (9) years, with ascending aorta dissection (Stanford type A) in 11 and a dissection distally affecting the aorta from the origin of the brachiocephalic trunk (Stanford type B) in 2. Of the 28 survivors, 3 were excluded for an age of more than 80, 2 for poor health status (one paraplegia and one advanced neoplasia), 2 for residing more than 150 km from the hospital, and 2 who could not be located. The remaining 19 composed the study group. None had had a previous aortic dissection or had a family history or clinical characteristics of Marfan's syndrome, and in no case was the aortic dissection secondary to chest trauma. None of the patients had been diagnosed previously of OSAS or had consulted a physician for suggestive symptoms. Ten patients presented a Stanford type A dissection, which extended to the descending aorta in six; nine patients had a Stanford type B dissection. All dissections were acute, except in three patients with type A dissection of more than 2-weeks' evolution from the onset of symptoms, which was considered chronic. The 10 patients with ascending aorta dissection underwent surgery; the remaining nine patients were treated medically. After discharge, patients were followed in the outpatient clinic at intervals of approximately 4 months for control of their blood pressure and aortic disease.
Control group. No patient of either group presented ischemic heart disease, heart failure, or renal failure. Only one patient in the aortic dissection group, with chronic obstructive pulmonary disease and mild obstruction on respiratory function tests, had lung disease.
Procedures No patient took sedatives or alcohol before the study. All were in a stable phase at the time of the sleep study, with no changes in their symptoms or medication in the 4 previous weeks. In the aortic dissection group, the mean time elapsed between discharge and the sleep study was 12 (4.7) months, and during this period, no patient presented significant weight change (± 3 kg). All patients underwent a full overnight polysomnographic study, which included recording of oronasal flow (thermistor), thoracoabdominal movements (strain gauges), electrocardiography, submental electromyography, bilateral electro-oculography, electroencephalography (C4-A1, C3-A2), oxyhemoglobin saturation (finger pulse oxymeter), body position, leg movements, and snoring. All sleep studies were reviewed manually by an expert scorer who was unaware of the group to which the patient belonged. Sleep was staged, and arousals were identified according to standard criteria (17, 18). An apnea was defined as a cessation of airflow with a duration of at least 10 seconds. Differentiation was made between obstructive and central apneas according to the respiratory effort channels (presence or absence of thoracoabdominal movement). Hypopnea was defined as a more than 50% reduction in thermistor tracing with a duration of at least 10 seconds associated with a cyclical dip in SaO2 of 4% or more. The apnea–hypopnea index (AHI) was defined as the sum of the number of apneas plus hypopneas divided by total sleep time.
Statistical Analysis
Clinical characteristics and results of the sleep questionnaire are shown in Tables 1 and 2 and in Table E1 in the online supplement. The majority of the patients were mild to moderately overweight men who frequently presented cardiovascular risk factors besides arterial hypertension. No patient had morbid obesity (body mass index of more than 35), and no differences were observed between groups in the prevalence of smoking, diabetes, or dyslipemia. Both groups had a high prevalence of snoring on most nights for many years. Nonrefreshing sleep was frequent in both groups; however, in general, the patients did not report daytime somnolence, with mean values of the Epworth scale being less than 10, a usual cut-off value for defining the presence of excessive daytime sleepiness. Only one patient of the dissection group presented sleepiness in active situations.
Evolution time of hypertension from its detection was similar in both groups, 7.3 (7.1) years in the dissection group and 8.3 (9.1) years in the control group (p = NS), as were the number of antihypertensive drugs used in their treatment (2.2 [0.9] vs. 2.1 [1.1], respectively, p = NS) and blood pressure on the day of the sleep study. Nine patients in the dissection group and 10 in the control group presented nocturnal or diurnal systolic or diastolic blood pressure of more than 140 mm Hg or more than 90 mm Hg, respectively. The results of the sleep study are shown in Table 3 . Thirteen patients of each group (68%) had an AHI of more than 5. However, apnea and hypopnea were more frequent in patients with aortic dissection. Seven patients with dissection (37%) (three type A and four type B) presented with an AHI of more than 30, a value usually accepted to define severe OSAS; in contrast, only one patient (5%) in the control group had an AHI of more than 30 (p = 0.042). Compared with patients with dissection and an AHI of less than 30, those with aortic dissection and severe OSAS had greater body mass index (29.3 [1.7] kg/m2 vs. 25.9 [3.1] kg/m2, p = 0.015] and during sleep showed a lower mean SaO2 (90.1% [2.0] vs. 93.6% [2.6], p = 0.008) and minimum SaO2 (71.6% [10.2] vs. 86.3% [4.2], p = 0.008); in contrast, they were of similar age (56.4 [15.0] vs. 59.6 [14], p = NS), years of arterial hypertension evolution (6.2 [7.6] vs. 6.1 [6.4], p = NS), number of hypertensive drugs (2.4 [1] vs. 2.1 [0.9], p = NS), and years of snoring (18.8 [6.7] vs. 20 [14.1], p = NS).
In the last imaging test performed before the sleep study, persistence of the aortic dissection was observed in 6 of the 10 patients treated surgically and in all of those in whom surgery was not performed. Four patients were considered to show progression of their aortic disease: one with type A and three with type B dissections. The latter three presented severe OSAS: Two showed progressive aortic dilation, and one presented evolution to dissection of an intramural hematoma.
The results of this study show an association between thoracic aortic dissection and OSAS. In particular, a higher AHI was found in patients with aortic dissection compared with a control group of hypertensive patients. Arterial hypertension is the main known risk factor for aortic dissection, and previous studies showed a high prevalence of OSAS in hypertensive patients referred to a hypertension unit (12–14), which could justify our findings in a group of patients with aortic dissection. However, the AHI found in our dissection patients, with seven suffering more than 30 apneas–hypopneas per hour of sleep, is higher than that previously reported in hypertensive patients. Furthermore, it was significantly higher than that of a control group of hypertensive patients of similar sex, age, body mass index, and upper body obesity, all known risk factors for the development of OSAS (19–21). Isaksson and Svanborg (22) affirmed that OSAS is more common in patients with poorly controlled hypertension, although this was not encountered by other authors (14). However, apart from an AHI higher than that reported by these authors, the aortic dissection patients in our study presented arterial hypertension of a number of years of known evolution, number of drugs required for its control, and blood pressure values on the day of the sleep study similar to those of the control group. Antihypertensive treatment differed slightly between groups, and the frequency of ß-blockers, a type of medication that some authors have suggested might worsen OSAS (23), was somewhat higher in the group of patients with dissection. However, this adverse effect has not been confirmed in studies comparing the effects of ß-blockers and placebo (24) or other antihypertensive drugs (25). These facts raise the hypothesis that OSAS, besides favoring the presence of arterial hypertension, could be a contributing factor to dissection in some patients through the mechanical stress on the aorta wall caused by repeated episodes of apnea and hypopnea. Inspiratory efforts against an occluded upper airway determine progressively negative intrathoracic pressures, which reach final mean peak values of approximately -60 cm H2O (8, 26). These negative pressures are transmitted to all intrathoracic structures and have been related to worsening of left ventricle function (27) and gastroesophageal reflux (28). In an animal model, Peters and colleagues (29, 30) observed an increase in systolic and diastolic aortic diameters during obstructive apnea episodes. In parallel to this development of progressively negative intrathoracic pressures, a marked increase in sympathetic activity and blood pressure is produced during apneas, which at the end of the obstructive event may double the basal systolic values (9, 31, 32). Upper airway obstructive episodes during sleep are known to be frequently asymptomatic (2) and may have been evolving for years before being clinically detected. Thus, it could be speculated that in our patients, the sudden rises in the transmural pressure of the aortic wall, repeated hundreds of times nightly over years, could have contributed to dissection of the aortic wall, already weakened by factors such as diabetes, dyslipemia, or smoking. Apart from by this increase in the sheer forces, OSAS could contribute to this weakening of the aorta wall because various mechanisms have been suggested relating it to arteriosclerosis development (33–35) and an increase in intima-media thickness of great arteries has been demonstrated in OSAS patients (36). Our study has several limitations. The sleep study was conducted several months after the aortic dissection had been diagnosed; however, we believe that such a short period of time, together with the absence of significant changes in weight, suggests that the detected sleep-disordered breathing was present at the time of the dissection. Furthermore, a case-control study does not permit us to elucidate whether OSAS is a risk factor for aortic dissection. Although the main known variables for the development of both entities had been controlled, other confounding factors may have existed to influence our results. A complementary alternative approach to our study would be to include OSAS as a prognostic variable in follow-up studies, including a greater number of patients with aortic dissection. Despite these limitations, we believe that our findings may have repercussions on the management of these patients. Treatment of OSAS with nasal continuous positive airway pressure prevents obstructive episodes of the upper airway and, consequently, the development of intrathoracic negative pressures (26), sympathetic discharges, and their associated rise in blood pressure (9, 37, 38), all of which are desirable in the aortic dissection patient. Furthermore, it has been demonstrated that antihypertensive medication does not achieve optimum control of blood pressure in hypertensive OSAS patients (25, 39), which is achieved when nasal continuous positive airway pressure is added (40). Although this study was not oriented toward assessing the evolution of patients with dissection, despite the shortness of follow-up, three of the four patients who showed aortic disease progression presented severe OSAS. Thoracic aorta dissection frequently requires surgical treatment. OSAS is known to be associated with an increase in perioperative morbidity and mortality (41, 42) because of a rise in the frequency and duration of upper airway obstructive episodes caused by the use of analgesics, sedatives, and anesthetics (43). This may be particularly important in patients with aortic dissection and OSAS undergoing surgery in whom, in addition to complications secondary to OSAS per se, rises in transmural pressure during upper airway obstructive episodes may have a particularly adverse effect on the recently surgically repaired thoracic aorta. Because the anesthetic and postoperative management of these patients with OSAS benefits from specific measures (44, 45), we believe that the early detection of OSAS could contribute to better perioperative management of these patients. Our results indicate the need to assess the presence of symptoms suggestive of OSAS in patients with thoracic aorta dissection. Given the relative absence of sleepiness detected and the lack of specificity of other symptoms such as snoring, we believe that their presence should be additionally studied by simple screening tests such as nocturnal pulsioxymetry or a limited sleep study and, when the patient is stable, with full polysomnography. In summary, in this study, a high mean AHI was found in patients with thoracic aorta dissection. We speculate that the coexistence of OSAS may impose an additional risk of aortic dissection in predisposed patients or determine worse evolution because of the increase in aortic transmural pressure implied. Because effective treatment for OSAS is available, we believe its diagnosis should be considered in the overall assessment of patients with aortic dissection.
The authors thank Christine O'Hara for help with the English version of the paper and Rosa Llòria for editorial assistance.
Supported in part by Red Respira (Instituto Carlos III. FIS. RTYC-C03/11)–SEPAR. This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Conflict of Interest Statement: G.S. has no declared conflict of interest; O.R. has no declared conflict of interest; A.S. has no declared conflict of interest; J.L.T. has no declared conflict of interest; P.L. has no declared conflict of interest; T.S. has no declared conflict of interest; A.E. has no declared conflict of interest. Received in original form April 24, 2003; accepted in final form August 5, 2003
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
