The Positive and Negative about Positive Airway Pressure Therapy

Cardiovascular disease is the leading cause of death in the world, with an estimated 17.3 million deaths per year globally and more than 375,000 deaths annually in the United States alone (1). In the United States, the 2020 goal is to improve cardiovascular health by 20% and reduce deaths from cardiovascular diseases and stroke by 20%, but the list of targeted risk factors has generally not included improving sleep health or screening and treating sleep-disordered breathing (2). Specifically, the Life’s Simple 7 are to not smoke, physical activity, healthy diet, body weight, and control of cholesterol, blood pressure, and blood sugar. The absence of a recommendation for widespread screening and treatment of sleep-disordered breathing, despite many observational data, stems from the lack of prospective randomized controlled trials (RCTs) of effective interventions aimed at improving clinical events, as opposed to surrogate end-points (3, 4). It is for this reason that the long-term trial published in this issue of the Journal by Peker and colleagues (pp. 613–620) is most welcome (5).

Peker and colleagues report the effect of continuous positive airway pressure (CPAP) on long-term clinical events in adult patients with angiography-verified coronary artery disease who had recently (<6 mo) undergone a percutaneous coronary intervention or coronary artery bypass grafting with evidence for moderate obstructive sleep apnea (OSA) without significant sleepiness (5). The lack of sleepiness was operationalized as an Epworth sleepiness score less than 10, and participants were randomized to either automatic CPAP (autoCPAP) or no CPAP therapy. Over the course of 57 months of follow-up, the researchers measured tangible clinical events; specifically, a primary composite end-point of repeat revascularization, myocardial infarction, stroke, and cardiovascular mortality. Intention-to-treat analysis did not reveal any difference in the primary end-point, but on-treatment analysis showed a significant reduction in cardiovascular events in those who were adherent to CPAP (nightly usage, ≥4 h) when compared with those who were nonadherent to CPAP or did not receive CPAP treatment. The researchers found that in such an enriched sample of patients with coronary artery disease who were not sleepy, treatment of OSA with CPAP did not lead to reduction of the primary composite endpoint in intention-to-treat analysis. However, on-treatment analysis revealed a greater than threefold reduction in the chance for the primary end-point in patients with OSA who were treated with CPAP when compared with patients who were not treated with such CPAP therapy.

There are many important takeaways from this study, and the authors should be commended for this long and arduous study. First, the cardiovascular benefit of CPAP therapy, based on on-treatment analysis, is unerringly similar in magnitude to prior observational studies, suggesting that if we were to improve CPAP adherence in our ongoing and future trials, we could conclusively answer the question as to whether treatment of OSA with PAP therapy can improve cardiovascular outcomes (6). Second, this study underscores that our quest to address the aforementioned knowledge gap hinges on our ability to promote CPAP adherence in research participants. Previous efforts to reduce attrition resulting from poor long-term CPAP adherence by instituting a placebo wash-in period for 1 week before randomization have met with less than desired levels of CPAP adherence (7). Such less-than-optimal long-term CPAP adherence brings into sharp focus the lack of high-quality evidence of educational, supportive, or behavioral interventions aimed at promoting CPAP adherence (8). Moreover, at this time, the optimal timing and duration and long-term effectiveness of interventions aimed at promoting CPAP adherence remain uncertain (8). More recently, a large trial of patients with heart failure and central sleep apnea revealed that 28% of participants in the intervention groups did not use a form of PAP therapy termed adaptive servo-ventilation (9). This led to concerns about the interpretation of the lack of benefit and the potential harms of the intervention when such a large proportion of participants were nonadherent to adaptive servo-ventilation (10, 11). As a community of scientists, patients, clinicians, and other stakeholders, we need to develop and successfully test and implement effective long-term interventions that can promote and sustain CPAP adherence. However, a search in ClinicalTrials.gov site reveals only six ongoing trials with a primary focus of promoting CPAP adherence. More needs to be done with regard to improving CPAP adherence both in trials and in the real world for research to translate into benefits to patients (12, 13).

The study by Peker and colleagues focused on nonsleepy apneics to avoid the ethical constraints of performing long-term RCTs against no treatment in symptomatic patients with OSA (5). Such a combination of circumstances could continue to plague our efforts to answer important questions unless we embark in a different direction. Specifically, we need to perform comparative effectiveness research between CPAP therapy and other treatment options, such as mandibular advancement devices in symptomatic patients, considering an accumulating body of evidence of cardiovascular benefits of such alternative treatments (14). Other promising treatment options would conceivably allow us to ethically undertake long-term trials in symptomatic patients (15). However, we may have to contend with differential drop-outs in comparator arms and the high cost and consequent feasibility of such trials. A third and very important aspect to the trial by Peker and colleagues is that the on-treatment analysis should be interpreted with caution, considering that such patients may be more adherent to other therapies, such as cardiovascular medications (16). We await the results of larger, ongoing trials, which could potentially provide conclusive answers of the effect of PAP therapies on cardiovascular outcomes: SAVE (Sleep Apnea Cardiovascular Endpoints Study), ADVENT-HF (Effects of Adaptive Servo Ventilation on Survival and Frequency of Cardiovascular Hospital Admissions in Patients with Heart Failure and Sleep Apnea), and ISAAC (Impact of CPAP on Patients with Acute Coronary Syndrome and Nonsleepy OSA) (7, 17). The study by Peker and colleagues sets the stage nicely for such studies. Whether these trials yield a conclusive positive or negative result with regard to the effect of CPAP therapy on cardiovascular outcomes hangs precariously on positive news regarding CPAP adherence in these trials.