Mechanisms underlying the supine related phenotype of obstructive sleep apnoea
2017-02-27T05:16:01Z (GMT) by
Obstructive sleep apnoea (OSA) is a common medical disorder that affects up to 24% of men and 9% of women (Young, Palta et al. 1993). Between 50 - 60% of patients with OSA who present to sleep clinics for overnight polysomnography experience twice as many respiratory events in supine sleep compared to lateral sleep (Pevernagie and Shepard 1992, Oksenberg, Silverberg et al. 1997, Richard, Kox et al. 2006, Joosten, Hamza et al. 2012), while approximately 25-30% of the same population experience respiratory events exclusively in the supine position (Mador, Kufel et al. 2005, Gillman, Roebuck et al. 2012, Joosten, Hamza et al. 2012). The cause of the observed preponderance of respiratory events in the supine sleeping position remains elusive. Certainly several studies have demonstrated a less collapsible airway in the lateral compared to the supine position in male OSA patients (Boudewyns, Punjabi et al. 2000, Penzel, Moller et al. 2001, Ong, Touyz et al. 2011); that is, a more negative pressure needs to be applied to the airway lumen to cause the airway to collapse when lateral – with the collapsing pressure referred to as the pharyngeal critical closing pressure (Pcrit). However, the reason for the fall in Pcrit is unclear with a postulated role for upper airway size and shape changes with body position. Recent advances in the understanding of OSA pathogenesis have revealed that non-anatomical factors such as ventilatory control instability and arousal threshold play an important role in airway obstruction in a significant proportion of OSA patients (Eckert, White et al. 2013). Importantly, there are no studies that examine the role of body position on ventilatory control instability, arousal threshold or active Pcrit (the pharyngeal critical closing pressure when airway muscles are activated) and there are no studies that examine upper airway size and shape, and lung volume in supine related OSA patients compared to matched controls. For patients with supine related OSA, the use of positional therapy can be an effective treatment. A number of positional modification devices have been demonstrated to be efficacious in preventing supine sleep and thus reducing the overall apnoea and hypopnoea index (AHI) (Jokic, Klimaszewski et al. 1999, Loord and Hultcrantz 2007, Permut, Diaz-Abad et al. 2010, Bignold, Mercer et al. 2011). Indeed, the use of positional modification devices for supine related OSA constitutes the most common form of individualised treatment for OSA sufferers. Despite the common occurrence of supine related OSA, and a seemingly cheap and readily available treatment, there remain a number of important unanswered questions that this thesis addresses. SELECTING PATIENTS FOR THE USE OF POSITIONAL MODIFICATION DEVICES The use of positional modification devices in the treatment of supine related OSA relies on three assumptions. Firstly, that the supine AHI is high from night-to-night (i.e. that there is a problem that warrants intervention), secondly, and partly related, that supine sleep time remains sufficient from night-to-night to warrant treatment, and thirdly that the lateral AHI remains low from night-to-night (i.e. that the intervention will be effective ongoing). Because these assumptions have never been tested, and because there is a recognised variability in the AHI from night-to-night (Le Bon, Hoffmann et al. 2000, Bittencourt, Suchecki et al. 2001) coupled with poor treatment compliance with traditional discomfort based treatments the take-up of positional modification by clinicians is poor. Chapter 2 delivers important information that will help guide clinicians’ use of positional modification devices. We demonstrate for the first time that male subjects with four times as many respiratory events in supine sleep compared to lateral sleep have a repeatable phenotype of supine related OSA. In addition, we show that a low lateral AHI is repeatable from night-to-night, thus giving clinicians confidence when using positional modification in selected patients. UPPER AIRWAY SHAPE AND LUNG VOLUME The observed fall in passive Pcrit (the Pcrit when upper airway muscles are relaxed) when subjects move from the supine to lateral position is frequently attributed to less favourable airway shape and size in the supine position. However, many studies show no significant change in upper airway cross-sectional area when subjects move from supine to lateral (Jan, Marshall et al. 1994, Martin, Marshall et al. 1995, Pevernagie, Stanson et al. 1995, Walsh, Leigh et al. 2008), whilst there are conflicting results on the effect of body position on airway shape. Importantly, no studies compare patients with supine related OSA to controls matched for key potential confounders such as age, gender, degree of obesity and body mass distribution. Another potential mechanism by which passive Pcrit may be reduced in lateral sleep is a change in lung volume. Previous studies have demonstrated that a reduced lung volume raises the Pcrit (Stanchina, Malhotra et al. 2003, Jordan, White et al. 2009) and that body position can affect lung volume with a reduced lung volume in the supine position in normal subjects. Chapter 3 clarifies some of the confusion present in the upper airway imaging literature. We have demonstrated that there are no significant differences between supine OSA patients and matched controls with regard to airway cross sectional area and shape when moving from supine to lateral position. However, there was a significant improvement in lung volume in the lateral position. This suggests that lung volume is likely to, at least in part, explain the consistently observed fall in passive Pcrit in the lateral position in OSA patients. THE EFFECT OF BODY POSITION ON PATHOPHYSIOLOGICAL FACTORS THAT CONTRIBUTE TO AIRWAY OBSTRUCTION The effect of body position on active Pcrit, ventilatory control stability and arousal threshold Recent advances in the understanding of OSA pathogenesis have highlighted a number of important contributors including: the ability of the upper airway muscles to stiffen and dilate the airway (McGinley, Schwartz et al. 2008, Edwards and White 2011), ventilatory control instability (i.e. a high loop gain) (Wellman, Jordan et al. 2004) and arousal threshold (Younes 2004). Despite the clear contribution of these factors to OSA, how they are affected by body position is not known. By applying a recently developed method for measuring these parameters in a non-invasive manner (Wellman, Edwards et al. 2013) we have been able to demonstrate that lateral positioning improves the ability of the airway to stiffen and dilate with no significant effect on arousal threshold or ventilatory control instability. The important findings in Chapter 4 highlight the enticing prospects for combination treatments for OSA that could include positional modification to improve passive Pcrit and the ability of the upper airway to stiffen and dilate, together with treatments that improve ventilatory control instability and arousal threshold, such as acetazolamide (Edwards, Sands et al. 2012) and non-muscle relaxant sedatives (Eckert, Owens et al. 2011). CONCLUSIONS AND FUTURE DIRECTIONS The work in this doctoral thesis demonstrates that there is a sub-population of OSA patients with a repeatable form of supine related OSA and a reliably low lateral AHI who are likely to benefit from positional modification. We have shed light on the mechanism by which lateral positioning reduces Pcrit by showing that supine related OSA patients have an improvement in lung volume in the lateral position compared to matched controls, without any significant differences in upper airway size and shape. By applying a recently developed method for measuring the traits contributing to OSA we have demonstrated important improvements in the passive and active airway anatomy in the lateral position, with no significant improvement in ventilatory control instability or arousal threshold. This raises the possibility of future treatments that combine positional treatment with treatments known to reduce loop gain and raise the arousal threshold.