The Complex Sleep Apnea Resolution Study: A Prospective Randomized Controlled Trial of Continuous Positive Airway Pressure Versus Adaptive Servoventilation Therapy

Introduction: Prior studies show that adaptive servoventilation (ASV) is initially more effective than continuous positive airway pressure (CPAP) for patients with complex sleep apnea syndrome (CompSAS), but choosing therapies has been controversial because residual central breathing events may resolve over time in many patients receiving chronic CPAP therapy. We conducted a multicenter, randomized, prospective trial comparing clinical and polysomnographic outcomes over prolonged treatment of patients with CompSAS, with CPAP versus ASV. Methods: Qualifying participants meeting criteria for CompSAS were randomized to optimized CPAP or ASV treatment. Clinical and polysomnographic data were obtained at baseline and after 90 days of therapy. Results: We randomized 66 participants (33 to each treatment). At baseline, the diagnostic apnea-hypopnea index (AHI) was 37.7 +/- 27.8 (central apnea index [CAI] = 3.2 +/- 5.8) and best CPAP AHI was 37.0 +/- 24.9 (CAI 29.7 +/- 25.0). After second-night treatment titration, the AHI was 4.7 +/- 8.1 (CAI = 1.1 +/- 3.7) on ASV and 14.1 +/- 20.7 (CAI = 8.8 +/- 16.3) on CPAP (P <= 0.0003). At 90 days, the ASV versus CPAP AHI was 4.4 +/- 9.6 versus 9.9 +/- 11.1 (P = 0.0024) and CAI was 0.7 +/- 3.4 versus 4.8 +/- 6.4 (P < 0.0001), respectively. In the intention-to-treat analysis, success (AHI < 10) at 90 days of therapy was achieved in 89.7% versus 64.5% of participants treated with ASV and CPAP, respectively (P = 0.0214). Compliance and changes in Epworth Sleepiness Scale and Sleep Apnea Quality of Life Index were not significantly different between treatment groups. Conclusion: Adaptive servoventilation (ASV) was more reliably effective than CPAP in relieving complex sleep apnea syndrome. While two thirds of participants experienced success with CPAP, approximately 90% experienced success with ASV. Because both methods produced similar symptomatic changes, it is unclear if this polysomnographic effectiveness may translate into other desired outcomes.