Personalised treatment is nearly here, but there’s already much that can be done beyond the one-size-fits-all approach.
CPAP is not the be all and end all for treating obstructive sleep apnoea, and a better understanding of causes and patient characteristics is bringing us closer to personalised care, according to a recent review from the Oxford Centre for Respiratory Medicine in the UK.
Although personalised treatment is still in the research phase, it’s already possible for Australian patients to utilise a variety of treatments, guided by their GP, the president of the Australasian Sleep Association Associate Professor Sutapa Mukherjee told The Medical Republic.
But we’re not quite there yet when it comes to pinpointing individual causes and, furthermore, there were systemic barriers to connecting patients with treatment, she said.
When a patient comes in complaining of snoring and sleep disturbance, either obesity or upper airway crowding was usually assumed the culprit and continuous positive airway pressure (CPAP) the solution to addressing the patient’s susceptibility to upper airway collapse.
The reviewers emphasised that CPAP was a very effective treatment – when used.
In practice, many find it invasive and intolerable, with 25% of people stopping within a year and over half only using it irregularly or for less than four hours a night.
But recent advances in identifying distinct causes and patient subgroups allow better targeted treatments, the study authors wrote. Along with obesity-induced upper airway collapsibility, endotypes (pathophysiological mechanisms) include enlarged tonsils, sensitive arousal threshold (more easily disturbed sleep), low dilator muscle responsiveness and elevated loop gain (where the brain doesn’t get the signal to breathe due to unstable levels of CO2).
OSA is thought to be caused by a combination of these, although the authors said it was “unclear whether each of these traits are truly an endotype or arise as a result of OSA itself”. And phenotypes include excessive sleepiness, disturbed sleep and minimal symptoms.
But alternatives to CPAP are available.
Weight loss, which may involve bariatric surgery, was an option for obese patients. Mandibular advancement devices (MADS, commonly referred to as splints) were sometimes recommended to patients who are thinner, have milder disease, and retrognathia (colloquially referred to as overbite), and were made by specialist dentists.
Professor Mukherjee noted that dentists were also good at picking up sleep apnoea, because they couldn’t help but notice retrognathia and oropharyngeal crowding.
Tonsillectomy was an option for those with enlarged tonsils. Other potential treatments included supplemental oxygen, hypoglossal nerve stimulation and sedatives. These treatments, especially in combination, could bring down the severity of sleep apnoea.
The efficacy of CPAP had, in some ways, held back the development of these other treatments to their full potential, the authors pointed out. But while the results of these alternative treatments weren’t as good as those from correctly used CPAP, the outcomes could certainly be better than inadequately used CPAP, they noted.
“CPAP is really the best therapy for severe sleep apnoea, but when you get to the moderate or mild levels, other options pop up,” Professor Mukherjee told TMR.
What stands in the way of specifically targeting treatment to the individual is the lack of ability to easily diagnose endotypes.
“A detailed attended in-lab PSG with additional oesophageal/epiglottic balloon catheters and genioglossus electrodes is the current gold standard to measure endotypes. This clearly limits our ability to measure endotypes at scale, or to measure endotypes over the longer term using repeated sleep studies, and simpler methods are needed,” the authors wrote.
Professor Mukherjee said that with the current availability of options, and lack of ability to easily categorise patients, it was difficult to determine which therapy would suit which patient.
“But with better ways of phenotyping and endotyping people, we can hopefully come up with an algorithm in the future, which will make it a lot easier to say that this person might do best with a mouthguard as a first line, and then we might add in this and then we might add in that,” she said.
While the currently available tools were somewhat blunt, they were still useful.
Generally, patients were referred for a home or in-lab sleep study, which measured the number of obstructions per hour (the Apnoea Hypopnoea Index). But the AHI was “not a great measure of sleep apnoea”, said Professor Mukherjee because the scores could vary widely from one night to the next. “Particularly in the mild and moderate groups, you might over-call it or under-call it.”
Sleep study reports could provide important information about oxygen saturation levels, in order to ascertain an individual’s level of hypoxic burden, she said.
“There are some people that are having obstructive events, but they don’t last very long – maybe 30 seconds where the levels just drop down to, say, 80. But then when they wake up, it goes straight back up to 99 very quickly. So they might have short events that don’t occur often. Others have events that last one and a half or two minutes with oxygen levels going down to 60 and only get back up to 80 before the next event.”
Better phenotyping and ways of targeting treatment accordingly are in the works and becoming more available, said Professor Mukherjee. This includes, unsurprisingly, artificial intelligence devices for both collecting data, which is sometimes referred to as “neurobles” – like “wearables for the brain”. These used big data to inform a more accurate diagnosis.
Rather than getting a possibly unrepresentative single night data set, a monitor under the mattress, for example, could provide a fuller picture over days or longer.
At Flinders Health and Medical Research Institute: Sleep Health, Professor Mukherjee and her colleagues have been conducting research to improve phenotyping and provide patients with targeted therapies like overnight oxygen therapy, surgery and drugs such as acetazolamide. “There’s a lot coming up on the horizon in terms of medication,” she said.
Meanwhile, therapies besides CPAP could be used to good effect, she said. Prescribing weight loss, for instance, was likely to be more effective now with the proliferation of injectables like semaglutide. Positional therapy (helping patients change their sleep position) and mandibular splints could be successful on their own or as an adjunct to CPAP.
Physiological causes, like the function of genioglossus and the tensor palatini muscles that control the tongue, could be determined at the sleep study level, said Professor Mukherjee.
“If there are abnormalities in the way the muscles work, the loop gain and the arousal threshold, they might be these other markers that are involved in actually being the reason why people who are not overweight can still get sleep apnoea,” she said.
The barriers to diagnosis and treatment of OSA are also systemic, according to Professor Mukherjee.
“A lot of it is being driven by industry, and the GP and the sleep physician are really kind of minor players,” she said.
GP guidelines on obstructive sleep apnoea were developed by the National Centre for Sleep Health Services Research and released in March this year.
GPs might refer a complex case to a sleep or respiratory physician, but it can be a long wait. “In the meantime, the patient has already bought the [CPAP] machine, and they sell them the most expensive machine as well, which they often don’t need,” she said.
They might even send them to a company who conducts the sleep study who may be “very invested in selling the CPAP and they often don’t mention the other therapies”, said Professor Mukherjee.
“It’s a very complicated area that we’re in at the moment in terms of sleep apnoea, that’s the that’s what I would say.
“The whole idea that it’s a one size fits all, whether you’re talking about how to diagnose it or how to treat it, is being knocked on its head.”
BMJ 2023, online 22 May
