1480

Obstructive sleep apnoea in patients with dilated cardiomyopathy: effects of continuous positive

airway pressure

The combined effects of negative intrathoracic pressure swings during obstructive sleep apnoeas (OSAs) and increased sympathetic nervous system tone associated with hypoxia and sleep arousal may lead to pulmonary oedema or left-ventricular hypertrophy. Therefore, we have done a study of patients with congestive heart failure secondary to idiopathic dilated cardiomyopathy to see whether OSA could contribute to impaired left-ventricular function and to assess nasal continuous positive airway pressure (NCPAP) for treatment. Eight men (aged 29-69 years) took part in the study; all were obese. Left-ventricular ejection fraction (LVEF) was measured while on stable medication and then 4 weeks after the start of nocturnal NCPAP. NCPAP was associated with abolition of OSA (mean [SE] number of apnoeas and hypopnoeas per hour of sleep 54·1 [7·2] and 1·0 [0·4] for pretreatment and NCPAP nights, LVEF Mean (SE) respectively, p 70% of forced vital capacity. Arterial blood gas analyses were essentially normal, apart from a slight reduction in PaOz (mean [SE] 80 [5] mm Hg), which would be expected in obese patients or in those with mild pulmonary venous congestion. One patient had a reduced Pa02 (54 mm Hg), which developed after the onset of congestive heart failure and was secondary to chronic pulmonary oedema and toxicity from chronic amiodarone therapy for control of ventricular arrhythmias. This resulted in a reduction in his lung diffusing capacity to 25% of that predicted. However, he had no evidence of obstructive or restrictive lung disease. The patient was therefore included in the study because congestive heart failure preceded

hypoxia. Before the start of the study, all patients were kept on stable medication for at least 2 months. All patients were being treated with digoxin, diuretics, and angiotensin-coverting-enzyme

inhibitors, and ADDRESSES

one

patient,

as

noted

above,

was

receiving

Sleep Research Laboratory, Queen Elizabeth Hospital, Toronto, Ontario, Canada (R Rutherford, BSc, A. Xie, MD); and Departments of Medicine, Toronto Hospital (S. Malone, MD, P P Liu, FRCPC, T Douglas Bradley, FRCPC) and Mount Sinai Hospital (R Holloway, FRCPC), University of Toronto, Toronto. Correspondence to Dr T Douglas Bradley, 212-10 EN Toronto General Hospital, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4.

1481

TABLE I-PATIENT CHARACTERISTICS

chest-wall motion .5 Hypopnoea was defined as a 50% or greater decrease in tidal volume lasting longer than 10 S.S,21 To measure intrathoracic negative pressure swings during obstructive apnoea, oesophageal pressure was recorded by a balloon catheter.22 The apnoea-hypopnoea index was defined as the number of apnoeas and hypopnoeas per hour of sleep. All patients were found to have OSA and underwent a trial of nocturnal NCPAP. NCPAP was adjusted so that apnoeas and hypopnoeas were abolished (10-12-5 cm H2O). Sleep studies were repeated while patients were receiving appropriate NCPAP.

Cardiac function All patients were male. Arterial blood gas tensions were obtained on room air while patients were awake. *This patient had amiodarone pulmonary toxicity and chronic pulmonary oedema BMI = body mass mdex, FEV1 =forced expiratory volume in 1 s, FVC =forced vital

capacity

amiodarone for

treatment of ventricular premature beats. Despite therapy, cardiomegaly and exertional dyspnoea (Canadian

Cardiovascular

Society

class III in three

patients

and class II in

five18) persisted.

Sleep studies The eight patients underwent overnight sleep studies. Sleep stage determined by standard criteria.19 A respiratory inductance plethysmograph (’Respitrace’, Ambulatory Monitoring, White Plains, USA)2O was used to measure rib-cage and abdominal movements. Electrocardiogram recordings were made with a single precordial lead. Arterial oxyhaemoglobin saturation (Sa02) was measured with an ear oximeter (’BIOX II’, Omeda, Boulder, USA). Mean nocturnal Sa02 was determined by averaging the highest and lowest Sa02 every 40 s, and mean low nocturnal Sa02 was determined by averaging the lowest Sa02 every 40 s throughout the night. Obstructive apnoea was defmed as absence of a tidal volume for greater than 10 s in the presence of paradoxical was

Patients were asked to grade their cardiac-function status before and after 4 weeks of NCPAP treatment according to established criteria.’8 As an objective measure of left-ventricular contractile function, we measured LVEF at rest in the supine position by gated equilibrium radionuclide angiography just before the start of nightly NCPAP. LVEF was measured again after patients had been on nocturnal NCPAP for 4 weeks. Radionuclide scans were interpreted by a doctor not involved in the study who had no knowledge of patients’ use of NCPAP. Four patients agreed to a 1-week withdrawal of NCPAP. These patients were admitted to hospital every night for 7 nights during which time they did not receive NCPAP. Overnight ear oximetry was done, and in all cases cyclical dips in Sa02 confirmed recurrence of OSA. At the end of 1 week, resting LVEF was reassessed. The four patients then restarted NCPAP. One patient underwent a progressive stage-one exercise study at baseline and then after 1 month of NCPAP, with increments in power of 16-1W.23 Differences between baseline data and data obtained while patients were on NCPAP were analysed by paired t tests. Among the four patients who did not receive NCPAP for 1 week, one-way analysis of variance with Bonferroni correction for multiple comparisons24 was done on values of LVEF. Differences were regarded as significant when p was less than 0-05.

Results

Sleep study data Table

II shows that all patients had severe OSA with a (SE) apnoea-hypopnoea index of 54.1 (7.2) per hour of sleep, associated with hypoxia and arousals from sleep. Patients’ peak inspiratory oesophageal pressure swings during episodes of apnoea were subatmospheric, ranging mean

from - 25 to - 90 cm H2O before treatment with NCPAP. This would produce a substantial acute increase in leftventricular transmural pressure and afterload by increasing the difference between intracardiac and intrathoracic pressure.4 A tracing of oesophageal pressure during obstructive hypopnoea in one patient is shown in fig la. Fig lb shows reversal of OSA and the return of a normal breathing pattern during treatment with NCPAP in the same patient. Exaggerated negative intrathoracic pressure swings during OSA were abolished by positive airway pressure, as shown by NCPAP mask pressure (fig 1 b). OSA was reversed in all patients while on NCPAP, with an TABLE II-MEAN

(SE) EFFECTS OF NCPAP ON OBSTRUCTIVE SLEEP APNOEA

Fig 1-Tracings made during an obstructive hypopnoea in one patient during stage-2non- rapid -eye-movement sleep (a), and the same patient during NCPAP therapy (b). EEG = electroencephalogram, EKG =electrocardiogram; EMGat= anterior tibial electromyogram; EMGsm=submenta! electromyogram; EOG=electrooculogram; Sa02=oxyhaemoglobin saturation, VT=tidal volume

1482

Fig 2-Effects of NCPAP therapy on LVEF (top) and function (bottom) of all eight patients.

cardiac

"Control" indicates pretreatment LVEF and cardiac function.

associated improvement in nocturnal Sa02 (table II). Sleep patterns became normal during NCPAP treatment, with increased slow-wave sleep and a reduction in arousals. In the one patient who was hypoxic while awake, Pa02 did not improve during the study, which suggests that NCPAP had no primary effect on amiodarone pulmonary toxicity. Cardiac function Cardiac function improved in all patients after 4 weeks’ home NCPAP therapy. Mean (SE) LVEF increased signifiantly from 37 (4)% before therapy to 49 (5)% (p < 0-0001) after 4 weeks’ NCPAP (fig 2). Patients also had significant improvement in cardiac-function class (fig 2). Among the four patients who withdrew from NCPAP for 1 week, mean (SE) LVEF increased from 43 (6)% to 53 (6)% (p < 0-001) while on NCPAP and then decreased to 46 after NCPAP withdrawal. One-way (5)% (p