Clodagh M. Ryan, M.D., Director of the Sleep Fellowship Program; Owen D. Lyons, Clinical Supervisor, Sleep Fellowship Program

Reviewed by Elisa Perger

Obstructive sleep apnea (OSA) is associated with increased cardiovascular risk, mediated through pathophysiological mechanisms that include intermittent hypoxia, excessive sympathetic nervous activation, and exaggerated swings in negative intrathoracic pressure (2, 3). Although it has previously been established from randomized controlled trials that treatment of OSA with continuous positive airway pressure (CPAP) reduces blood pressure, with the most marked effects seen in drug-resistant hypertension (4), data to support a role for CPAP therapy to reduce cardiovascular mortality comes largely from observational studies (5, 6). To address this, McEvoy and colleagues conducted a multicenter, randomized, parallel-group trial to evaluate the efficacy of CPAP in reducing cardiovascular mortality in patients with moderate to severe OSA (oxygen desaturation index ≥ 12) and a history of coronary artery disease or cerebrovascular disease who were mildly or non-sleepy (Epworth Sleepiness Scale ≤ 15) (1). Patients were excluded if they had severe hypoxia (oxygen saturation < 80%) or if they had a Cheyne-Stokes respiration pattern. The primary endpoint included a composite of death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for unstable angina, heart failure, or transient ischemic attack. A total of 2,687 subjects were randomized to either “usual care” (n = 1,341) or “usual care” plus CPAP (n = 1,346). After a mean follow-up of 3.7 years, there was no significant difference in the occurrence of the primary endpoint between the groups (hazard ratio [HR] with CPAP added, 1.10; 95% confidence interval [CI], 0.91–1.32; P = 0.34). Mean duration of adherence to CPAP therapy was 3.3 hours per night. A one-to-one propensity score analysis performed to compare 561 adherent patients (CPAP used for more than 4 h/night) and 561 patients in the usual care group showed no significant difference in the primary endpoint (HR, 0.80; 95% CI, 0.60–1.07; P = 0.13) but a lower risk of cerebrovascular events among the CPAP group (HR, 0.52; 95% CI, 0.30–0.90; P = 0.02).

The results of this relatively large randomized controlled trial are clearly an important addition to the current knowledge base and certainly, on the basis of this one study, CPAP cannot be recommended as a therapy in patients with moderate to severe OSA with established cardiovascular disease if the sole purpose is to reduce cardiovascular complications. This trial affirms the results of other studies in highlighting the uncertain efficacy of CPAP therapy in the reduction of cardiovascular risk in nonsymptomatic patients with OSA over the short to medium term and also highlights the challenge of CPAP adherence (7, 8). However, it is important that these results are not extrapolated to those patients with OSA who do have excessive daytime sleepiness or significant hypoxia, given these patients were excluded from the study.

Reviewed by Carolina Gonzaga-Carvalho

Acute decompensation of heart failure (ADHF) and subsequent hospitalization has been shown to be associated with high mortality (10). Untreated sleep-disordered breathing (SDB) has been shown to be independently associated with increased mortality in patients with chronic heart failure (11) and, as such, may represent a potentially modifiable risk factor in patients admitted to the hospital with ADHF.

Khayat and colleagues conducted a prospective cohort study of patients hospitalized with ADHF in an attempt to determine the effect of SDB on postdischarge mortality (9). Admitted stable patients who had a reduced left ventricular ejection fraction less than or equal to 45% with no prior diagnosis of SDB underwent attended, inpatient, overnight polygraphy. SDB was defined by an apnea–hypopnea index (AHI) greater than or equal to 15 events/h, and patients were classified as having OSA or central sleep apnea (CSA) on the basis of established criteria (12); minimal or no SDB (nmSDB) was defined by AHI less than 15 events/h. Subjects with SDB were classified as “treated” if confirmed by either sleep physician report or PAP download to have adherence of more than 4 h/night between 6 and 12 months after discharge. “Untreated” patients were those who refused, had not started PAP within 6 to 12 months after discharge, or for whom PAP usage was not confirmed following records review.

Of 4,874 patients screened, 1,117 were included. Of these, 47% had OSA, 31% CSA, and 22% nmSDB. The Cox proportional hazard models for time to all-cause death over the 3-year follow-up, which included left ventricular ejection fraction, age, sex, creatinine, and diabetes as covariates, showed that the HRs for CSA and OSA versus nmSDB were 1.61 (95% CI, 1.1–2.4; P = 0.02) and 1.53 (95% CI, 1.1–2.2; P = 0.02), respectively. By the end of the first year after discharge, 58 (17%) patients with CSA and 103 (20%) patients with OSA were confirmed to have been on PAP therapy. There was no difference in mortality between those with SDB on PAP therapy and patients with nmSDB. However, those patients with CSA and OSA who were not on PAP by the end of the first year after discharge had higher mortality compared with nmSDB (HR, 1.9; 95% CI, 1.3–2.9; P = 0.001; vs. HR, 1.8; 95% CI, 1.3–2.7; P = 0.001).

The study had several limitations, including a relatively large number of polygraphy recording failures (18.7%), an absence of information regarding PAP usage or adherence after the first year, and no postdischarge polygraphy after treatment of the ADHF. Furthermore, the difference in the primary outcome seen between treated and untreated patients with SDB must be interpreted in the context of the inherent flaws of a nonrandomized protocol.

Nonetheless, this is the first study to demonstrate that newly diagnosed OSA or CSA are both independently associated with postdischarge mortality in patients with HF. Whether screening for and subsequent treatment of SDB can reduce mortality in this high-risk population remains to be determined.

Reviewed by Toru Inami

OSA is prevalent in patients with coronary artery disease (CAD) and has been associated with progression of coronary plaque burden (14, 15). Previously, it has been shown that after percutaneous coronary intervention (PCI) for acute coronary syndrome with bare metal stents, the presence of OSA was associated with poorer quantitative angiographic outcomes, including late luminal loss and subsequent restenosis (16). However, in the modern era of drug-eluting stents, the rates of restenosis and target vessel revascularization have decreased to 10%. Therefore, in this multicenter international prospective cohort study, Lee and colleagues aimed to assess the impact of OSA on major adverse cardiac and cerebrovascular event (MACCE) after PCI (13). MACCE were defined as a composite of cardiovascular mortality, nonfatal myocardial infarction, nonfatal stroke, and planned revascularization.

A total of 1,311 enrolled patients who underwent a successful PCI had polysomnography performed within 7 days of their PCI. Drug-eluting stents were inserted in 80.1%. OSA defined as an AHI greater than or equal to 15 events/h was present in 45.3% of the cohort. Compared with the non-OSA group, the OSA group had a higher mean age and body mass index, a higher proportion of men, and a higher prevalence of both hypertension and diabetes. During the median follow-up of 1.9 years, the OSA group had a higher incidence of an MACCE compared with the non-OSA group (3-year estimate, 18.9% vs. 14.0%; P = 0.001). The incidence of cardiovascular mortality was also higher in the OSA group (3-year estimate, 4.2% vs. 1.7%; P = 0.035). In Cox regression analysis, OSA remained a predictor of MACCEs, with a hazard ratio of 1.57 (95% CI, 1.1–2.2; P = 0.013) after adjustment for age, sex, ethnicity, body mass index, hypertension, and diabetes mellitus. There were no significant differences in stent-related adverse events between groups with and without OSA, suggesting the influence of OSA on clinical outcomes may not have been directly related to the PCI procedure or underlying CAD per se. However, in the absence of routine angiographic follow-up it is difficult to determine if there were different rates of restenosis between the groups that were not clinically apparent. In addition, no analysis was performed to evaluate the association between OSA severity and the oxygen desaturation index on MACCE.

However, this study clearly demonstrates an independent association between OSA and major cardiac or cerebrovascular events in patients with CAD after PCI. These findings provide the rationale for randomized trials to test the hypothesis that the treatment of OSA in patients with CAD undergoing PCI can improve cardiovascular outcomes.