The many faces (various clinical phenotypes) of asthma

I was planning to make this blog post very pretty with some graphs and tables from a recent journal alrticle. Imagine my dismay to discover that my tardiness to renew my ATS membership had resulted in the suspension of my access to the online blue journal. All good now; subs paid, but access not yet restored. Still, the blog must be posted. Ugly though it may be….

Around 8% of the adult Australian population suffers from ‘asthma’. This means that asthma is a big public health problem. As such, asthma has been given the ‘pathway’ treatment; that is, rules and guidelines have been generated to direct asthma management. These rules and guidelines are also brought into play when asthma management is evaluated.

However, all of this ‘proceduralising’ of asthma mangement assumes that we know and are agreed on what asthma actually is, and how we diagnose it and measure its severity. Unfortunately, this is more of a problem than it should be!

Classically, asthma has been said to be a disease of allergic inflammation in the airways. Eosinophils and mast cells are supposed to be the main mediators of this inflammation; fibrosis and scarring develops early; corticosteroids (such as prednisolone) settle the
inflammation down. If we are agreed on that definition – and that it applies to all ‘asthmatics’ – we then measure severity of asthma by frequency of symptoms and use of bronchodilator medications (such as salbutamol / Ventolin), measurements of airway ‘obstruction’ (such as spirometry and peak flow variability) and ‘exacerbations’ requiring attendances at the doctor’s clinic or hospital and/or use of oral corticosteroids (such as prednisolone).

Those of us who see lots of patients with asthma know how deeply unsatisfactory this approach to classification can be. Some patients present with profound symptoms which settle quickly on classic treatment with inhaled steroids, while others have persistent symptoms and lung function abnormality in spite of mulitple treatments. Some never take their puffers but struggle on with persistent symptoms, never getting very sick. Others have no symptoms for months and months, take their medication then present desperately unwell after an inadvertent exposure.

We need fresh eyes on asthma classificatoins.

I have been encouraged to read a paper, with accompanying editorial, in the Blue Journal from the middle of February which sought to reevaluate how we ‘categorise’ asthma sufferers. Over seven-hundred subjects from the ‘Severe asthma research program’ (SARP) in the USA completed questionnaires as well as physiological tests of lung function. Biological markers of disease were also measured in some patients (exhaled nitric oxide, induced sputum eosinophils). Not all of the subjects had ‘severe’ asthma.

Ultimately five ‘clusters’ of patients were identified:

Cluster 1: 15% of subjects. Younger, predominantly women (80%), childhood onset / atopic asthma and normal lung function. 40% on no controller medication; those on controllers generally on two or fewer. 70% reported no significant exacerbations in the previous year. However 30 to 40 % had almost daily symptoms (perhaps predominantly exercise related?)

Cluster 2: 44% of subjects. Slightly older, 67% women. Mainly childhood onset / atopic asthma. Baseline pre-bronchodilator lung function normal or can be reversed to normal in 94% of subjects. More prevalent medication use (only 26% on no medication, more on 3 puffers). Higher doses of inhaled steroids.

Cluster 3: only 8% of subjects. Significantly different. older women, with older age at onset of asthma and more overweight (58% BMI > 30). Less likely to be atopic. Lower baseline FEV1 with only 64% whose lung function ‘normalised’ after bronchodilator. Higher doses
of medication use and, in spite of this, health care utilisation. (Despite the fact that they are a small portion of the patient population, they are disproprotionately represented in hospitals I think). They have symptoms / quality of life impairments that seem out of proportion to their physiological impairment.

Cluster 4 and 5: the remaining 33% of subjects. 70 or 80% fulfill the ATS criteria for severe asthma. These are the patients with legitimate bad asthma. Cluster 4 was the only cluster in which both genders were equally represented, and atopy predominated (83%). Cluster 5 was later onset, mainly women with less atopy. Each cluster has long duration of disease, and significantly impaired pre- bronchodilator FEV1 commonly. More subjects in cluster 4 had significant bronchodilator responsiveness. Health care utilisation and medication doses were high in both clusters.

These clusters ring true. They fit with my impression of the population of patients with asthma; they differ significantly in important markers of severity, and in particular health care and medication utilisation; ultimately they may provide us with a step towards identifying groups of patients with asthma who sholud be offered significantly different management programs.

Ultimately, the authors found that discrimination between these clusters was possible 80%of the time simply using FEV1 pre- and post- bronchodilator and gender (this is where the missing diagrams would’ve been helpful).

We should trawl through our patient data and assign patients to ‘clusters’ based on lung function and gender. I suspect that we would be weighted towards cluster 3-5. What would be really interesting would be to see if there were medications or management strategies that worked, or failed, for each cluster. Another audit project? Must discuss it with the clinical staff!

Andrew