Sleep Strategies

CPAP vs noninvasive ventilation for obesity hypoventilation syndrome


The conventional approach to treat hypoventilation has been to use noninvasive ventilation (NIV), while continuous positive airway pressure (CPAP) that does not augment alveolar ventilation improves gas exchange by maintaining upper airway patency and increasing functional residual capacity. Why, then, are we debating the use of CPAP vs NIV in the treatment of obesity hypoventilation syndrome (OHS)? To understand this rationale, it is important to first review the pathophysiology of OHS.

The hallmark of OHS is resting daytime awake arterial PaCO2of 45 mm Hg or greater in an obese patient (BMI > 30 kg/m2) in absence of any other identifiable cause. To recognize why only some but not all obese subjects develop OHS, it is important to understand the different components of pathophysiology that contribute to hypoventilation: (1) obesity-related reduction in functional residual capacity and lung compliance with resultant increase in work of breathing; (2) central hypoventilation related to leptin resistance and reduction in respiratory drive with REM hypoventilation; and (3) upper airway obstruction caused by upper airway fat deposition along with low FRC contributing to pharyngeal airway narrowing and increased airway collapsibility (Masa JF, et al. Eur Respir Rev. 2019; 28:180097).


Let us examine some of the studies that have compared the short-term efficacy of CPAP vs NIV in patients with OHS. In a small randomized controlled trial (RCT), the effectiveness of CPAP and NIV was compared in 36 patients with OHS (Piper AJ, et al. Thorax. 2008;63:395). Reduction in PaCO2 at 3 months was similar between the two groups. However, patients with persistent nocturnal desaturation despite optimal CPAP were excluded from the study. In another RCT of 60 patients with OHS who were either in stable condition or after an episode of acute on chronic hypercapnic respiratory failure, the use of CPAP or NIV showed similar improvements at 3 months in daytime PaCO2, quality of life, and sleep parameters (Howard ME, et al. Thorax. 2017;72:437).

In one of the largest randomized control trials, the Spanish Pickwick study randomized 221 patients with OHS and AHI >30/h to NIV, CPAP, and lifestyle modification (Masa JF, et al. Am J Respir Crit Care Med. 2015:192:86). PAP therapy included NIV that consisted of in-lab titration with bilevel PAP therapy targeted to tidal volume 5-6 mL/kg of actual body weight or CPAP. Life style modification served as the control group. Primary outcome was the change in PaCO2 at 2 months. Secondary outcomes were symptoms, HRQOL, polysomnographic parameters, spirometry, and 6-min walk distance (6 MWD). Mean AHI was 69/h, and mean PAP settings for NIV and CPAP were 20/7.7 cm and 11 cm H2O, respectively. NIV provided the greatest improvement in PaCO2 and serum HCO3 as compared with control group but not relative to CPAP group. CPAP improved PaCO2 as compared with control group only after adjustment of PAP use. Spirometry and 6 MWD and some HRQOL measures improved slightly more with NIV as compared to CPAP. Improvement in symptoms and polysomnographic parameters was similar between the two groups.

In another related study by the same group (Masa JF, et al. Thorax. 2016;71:899), 86 patients with OHS and mild OSA (AHI <30/h), were randomized to NIV and lifestyle modification. Mean AHI was 14/h and mean baseline PaCO2 was 49 +/-4 mm Hg. The NIV group with mean PAP adherence at 6 hours showed greater improvement in PaCO2 as compared with lifestyle modification (6 mm vs 2.8 mm Hg). They concluded that NIV was better than lifestyle modification in patients with OHS and mild OSA.

To determine the long-term clinical effectiveness of CPAP vs NIV, patients in the Pickwick study, who were initially assigned to either CPAP or NIV treatment group, were continued on their respective treatments, while subjects in the control group were again randomized at 2 months to either CPAP or NIV (Masa JF, et al. Lancet. 2019;393:1721). All subjects (CPAP n=107; NIV n=97) were followed for a minimum of 3 years. CPAP and NIV settings (pressure-targeted to desired tidal volume) were determined by in-lab titration without transcutaneous CO2 monitor, and daytime adjustment of PAP to improve oxygen saturation. Primary outcome was the number of hospitalization days per year. Mean CPAP was 10.7 cm H2O pressure and NIV 19.7/8.18 cm H2O pressure with an average respiratory rate of 14/min. Median PAP use and adherence > 4 h, respectively, were similar between the two groups (CPAP 6.0 h, adherence > 4 h 67% vs NIV 6.0/h, adherence >4 h 61%). Median duration of follow-up was 5.44 years (IOR 4.45-6.37 years) for both groups. Mean hospitalization days per patient-year were similar between the two groups (CPAP 1.63 vs NIV 1.44 days; adj RR 0.78, 95% CI 0.34-1.77; p=0.561). Overall mortality, adverse cardiovascular events, and arterial blood gas parameters were similar between the two groups, suggesting equal efficacy of CPAP and NIV in this group of stable patients with OHS with an AHI >30/h. Given the low complexity and cost of CPAP vs NIV, the authors concluded that CPAP may be the preferred PAP treatment modality until more studies are available.

An accompanying editorial (Murphy PB, et al. Lancet. 2019; 393:1674), discussed that since this study was powered for superiority as opposed to noninferiority of NIV (20% reduction in hospitalization with NIV when compared with CPAP), superiority could not be shown, due to the low event rate for hospitalization (NIV 1.44 days vs CPAP 1.63 days). It is also possible optimum NIV titration may not have been determined since TCO2 was not used. Furthermore, since this study was done only in patients with OHS and AHI >30/h, these results may not be applicable to patients with OHS and low AHI < 30/h that are more likely to have central hypoventilation and comorbidities, and this group may benefit from NIV as compared with CPAP.

Novel modes of bi-level PAP therapy

There are limited data on the use of the new bi-level PAP modalities, such as volume-targeted pressure support ventilation (PS) with fixed or auto-EPAP. The use of intelligent volume-assured pressure support ventilation (iVAPS) vs standard fixed pressure support ventilation in select OHS patients (n=18) showed equivalent control of chronic respiratory failure with no worsening of sleep quality and better PAP adherence (Kelly JL, et al. Respirology. 2014;19:596). In another small randomized, double-blind, crossover study, done on two consecutive nights in 11 patients with OHS, the use of auto-adjusting EPAP was noninferior to fixed EPAP (10.8 cm vs 11.8 cm H2O pressure), with no differences in sleep quality and patient preference (McArdle N. Sleep. 2017;40:1). Although the data are limited, these small studies suggest the use of new PAP modalities, such as variable PS to deliver target volumes and auto EPAP could offer the potential to initiate bi-level PAP therapy in outpatients without the in-lab titration. More studies are needed before bi-level PAP therapy can be safely initiated in outpatients with OHS.


In summary, how can we utilize the most effective PAP therapy for patients with OHS? Can we use a phenotype-dependent approach to PAP treatment options? The answer is probably yes. Recently published ATS Clinical Practice Guideline (Am J Respir Crit Care Med. 2019;200:e6-e24) suggests the use of PAP therapy for stable ambulatory patients with OHS as compared with no PAP therapy, and patients with OHS with AHI >30/h (approximately 70% of the OHS patients) can be initially started on CPAP instead of NIV. Patients who have persistent nocturnal desaturation despite optimum CPAP can be switched to NIV. On the other hand, data are limited on the use of CPAP in patients with OHS with AHI <30/h, and these patients can be started on NIV. PAP adherence >5-6 h, and weight loss using a multidisciplinary approach should be encouraged for all patients with OHS.

Dr. Dewan is Professor and Program Director, Sleep Medicine; Division of Pulmonary, Critical Care and Sleep Medicine; Chief, Pulmonary Section VA Medical Center; Creighton University, Omaha, Nebraska.

Next Article: