Nocturnal hypoxemic burden during positive airway pressure treatment across different central sleep apnea etiologies

INTRODUCTION

Nocturnal hypoxemia is associated with increased cardiovascular mortality. Here, we assess whether positive airway pressure by adaptive servo-ventilation (ASV) reduces nocturnal hypoxemic burden in patients with primary central sleep apnea (primary CSA), or heart failure related central sleep apnea (CSA-HF) and treatment emergent central sleep apnea (TECSA).

METHODS

Overnight oximetry data from 328 consecutive patients who underwent ASV initiation between March 2010 and May 2018 were retrospectively analyzed. Patients were stratified into three groups: primary CSA (n = 14), CSA-HF (n = 31), TECSA (n = 129). Apnea hypopnea index (AHI) and time spent below 90% SpO₂ (T90) was measured. Additionally, T90 due to acute episodic desaturations (T90_Desaturation) and due to non-specific and non-cyclic drifts of SpO₂ (T90_Non-specific) were assessed.

RESULTS

ASV reduced the AHI below 15/h in all groups. ASV treatment significantly shortened T90 in all three etiologies to a similar extent. T90_Desaturation, but not T90_Non-specific, was reduced by ASV across all three patient groups. AHI was identified as an independent modulator for ΔT90_Desaturation upon ASV treatment (B (95% CI: -1.32 (-1.73; -0.91), p < 0.001), but not for ΔT90 or ΔT90_Non-specific. Body mass index was one independent predictor of T90.

CONCLUSIONS

Across different central sleep apnea etiologies, ASV reduced AHI, but nocturnal hypoxemic burden remained high due to a non-specific component of T90 not related to episodic desaturation. Whether adjunct risk factor management such as weight-loss can further reduce T90 warrants further study.

Long-term outcome with focus on pulmonary hypertension in Obesity Hypoventilation Syndrome

INTRODUCTION

Pulmonary Hypertension (PH) is a frequent comorbidity in Obesity Hypoventilation Syndrome (OHS).

OBJECTIVE

We investigated long-term outcome of OHS with a particular emphasis on PH.

METHODS

In a prospective design, 64 patients with OHS and established noninvasive positive pressure ventilation (NPPV), were assessed by serum biomarkers, right heart catheterization, blood gases analysis, lung function, Epworth-Sleepiness Scale (ESS), Pittsburgh Sleep Quality Index (PSQI), World Health Organization-functional class (WHO-FC) and health-related quality of life (HRQL) via the Severe Respiratory Insufficiency (SRI) questionnaire. After a planned follow-up of 5 years patients were reassessed regarding vital status, WHO-FC, ESS, SRI, PSQI, body mass index (BMI) and NPPV use. Prognostic markers were explored using univariate and multivariate Cox regression analyses.

RESULTS

At the 5-year follow-up, BMI tended to decrease (P = 0.05), while WHO-FC, ESS and PSQI remained unchanged. HRQL deteriorated in terms of SRI summary score and most subdomains (P < .05 each). NPPV adherence still was high (89%), while daily NPPV use increased from 6.7 (5.1; 8.0) h/d to 8.2 (7.4; 9.0) h/d (P < .05). After a 5-year follow-up, mortality was 25.8%. In univariate regression analyses only age > 69.5 years (HR = 4.145, 95%-CI = 1.180-14.565, P = 0.016), NT-proBNP > 1256 pg/mL (HR = 5.162, 95%-CI = 1.136-23.467, P = 0.018), diffusion capacity for carbon monoxide (DLCO, %pred) (HR = 0.341, 95%-CI = 0.114-1.019, P = 0.043) and higher oxygen use during daytime (HR = 5.236, 95%-CI = 1.489-18.406, P = 0.004) predicted mortality. No independent factor predicting mortality was detected in multivariate analysis.

CONCLUSION

Despite a high long-term NPPV use HRQL worsened. Age, oxygen use at baseline, DLCO (%pred) and NT-proBNP, as a surrogate parameter for PH, were related to long-term survival.

[Studies with the fully automated EEG sleep analysis system QUISI]

To judge sleep reliably would be interesting not only for outpatient sleep diagnostics, but also in the inpatient sphere. A step in this direction is the easy-to-apply fully automated sleep-EEG-analysis system QUISI. Nine males patients with obstructive sleep apnoea syndrome were examined simultaneously by polysomnography and the QUISI system. A comparison of the sleep profiles evaluated by polysomnographic recordings using Rechtschaffen and Kales' rules with the automated classification using QUISI recording shows that the difference of the sleep parameters sleep period time (SPT), total sleep time (TST), REM, stage 1, stage 2, slow-wave sleep and the awake stage of maximum five patients ranged between +/- 10 minutes, respectively +/- 5 per cent. Only two of nine sleep profiles could be compared as far as the sequence of the sleep stages was concerned. According to our results it is important to edit the original data and to treat the fully automated sleep profile. We cannot recommend QUISI to establish sleep profiles for the obstructive sleep apnoea syndrome.