Effects of auto-servo ventilation on patients with sleep-disordered breathing, stable systolic heart failure and concomitant diastolic dysfunction: subanalysis of a randomized controlled trial.

Birner, Christoph; Series, Frederic; Lewis, Keir; Benjamin, Amit; Wunderlich, Silke; Escourrou, Pierre; Zeman, Florian; Luigart, Ruth; Pfeifer, Michael; Arzt, Michael

BACKGROUND Systolic heart failure (HF) is frequently accompanied by diastolic dysfunction and sleep-disordered breathing (SDB). OBJECTIVES The objective of this subset analysis was to determine effect sizes of auto-servo ventilation (ASV and biphasic positive airway pressure ASV) on echocardiographic measures of diastolic function in patients with systolic HF and SDB. METHODS Thirty-two patients with stable systolic HF, concomitant diastolic dysfunction [age 66 ± 9 years old, left ventricular (LV) ejection fraction: 30 ± 7% and New York Heart Association class II: 72%] and SDB (apnea-hypopnea index, AHI: 48 ± 19/h; 53% had predominantly obstructive sleep apnea) receiving either ASV (n = 19) or optimal medical treatment (control, n = 13) were analyzed in a randomized controlled clinical trial. Polysomnographic and echocardiographic measurements were obtained at baseline and after 12 weeks. RESULTS AHI significantly improved in the ASV group compared to the control group (-39 ± 18 vs. -0.2 ± 13.2/h, p < 0.001). At baseline, 24 (75%) patients had impaired LV relaxation, and 8 (25%) had a pseudo-normalized filling pattern. At the 12-week control visit, diastolic function assessed by the isovolumetric relaxation time (-10.3 ± 26.1 vs. 9.3 ± 49.1, p = 0.48) and deceleration time (-43.9 ± 88.8 vs. 12.4 ± 68.8, p = 0.40) tended to improve after ASV treatment, but did not reach statistical significance. Likewise, the proportion of patients whose diastolic dysfunction improved was nonsignificantly higher in the ASV than in the control group, respectively (37 vs. 15%, p = 0.25). CONCLUSIONS ASV treatment efficiently abolishes SDB in patients with stable systolic HF and concomitant diastolic dysfunction, and was associated with a statistically nonsignificant improvement in measures of diastolic dysfunction. Thus, these data provide estimates of effect size and justify the evaluation of the effects of ASV on diastolic function in larger randomized controlled trials.