• epidemiology
• respiratory
• paediatric practice
• neonatology

The use of low-dose inhaled corticosteroids (ICS) to treat children with asthma can be life-transforming, and this is confirmed repeatedly when they are made widely available for the first time in a low and middle income setting.1 However, reading the National Report on Asthma Deaths (NRAD)2 among other documents makes it very clear that progress has stalled. In response to this perception, a Lancet commission has recently been published.3 This annotation reviews some of the implications of that document for paediatrics.

The first proposal is that ‘asthma’ is no more than a clinical description of symptoms such as wheeze, breathlessness and cough, as arthritis describes red, painful joints; neither is a 21st century diagnosis. The proper response of a family whose child has been given a diagnosis of asthma is, ‘what sort of asthma does my child have?’ So the Commission proposes that as far as possible, the airway disease should be deconstructed into identifiable aspects such as airflow limitation, eosinophilic airway inflammation, airway infection and impaired airway defences, and altered cough reflex sensitivity and efficacy. The emphasis is on defining in particular ‘treatable traits’ of airway disease,4 which include such extrapulmonary comorbidities as obesity and allergic rhinoconjunctivitis, and social and environmental factors, in particular adherence. There are important implications of this approach in preschool children, school-age children with an associated airway or systemic disease and also in the era of the new biologicals.

The wheezing preschool child has traditionally been ‘phenotyped’ on the history as having episodic viral wheeze or multiple trigger wheeze,5 which is limited in many ways,6 and can hardly be said to be worthy of 21st century practice. Some but not all likely have eosinophilic airway inflammation, and thus be responsive to ICS. The Lancet approach is to urge characterisation of the airway disease; it is perfectly possible with patience and a good technician to determine if there is fixed or variable airflow obstruction using spirometry in preschool children.7 Induced sputum can be used even in very young children to determine infection status,8 9 although induced sputum cytology in the preschool child does not predict airway eosinophilia9 unlike in older children.10 There is increasing evidence that the peripheral blood eosinophil count predicts airway eosinophilia,9 and response to ICS, as does atopic sensitisation,11 neither of which are difficult to determine even in primary care. So rather than ‘not diagnosing asthma’, we should be asking ‘does this preschool child have eosinophilic airway inflammation and thus is ICS responsive?’ We need to get out of the stagnant rut of making treatment decisions based merely on a conversation with a carer who by definition is reporting the child’s state second hand.

Frequently, the question is asked, do children with, for example, cystic fibrosis (CF), bronchopulmonary dysplasia (BPD) and sickle cell disease (SCD) have asthma?12–14 Rather the question should be asked, what sort of airway disease do they have? Do they have variable airflow obstruction that may be responsive to inhaled β-2 agonist, or eosinophilic inflammation responsive to ICS? So BPD survivors have fixed and variable airflow obstruction, but no evidence of airway inflammation.15–19 We have shown that children with SCD have fixed but not variable airflow obstruction, and no airway inflammation, yet another sort of ‘asthma’.20 Depending on the ‘asthma’, the treatment should be tailored accordingly.

Although differentiating eosinophilic from non-eosinophilic disease is an essential first step, we are learning that there are more than just TH2 pathways that can cause airway eosinophilia.21 22 This is important, as many different biologicals are being used to target the TH2 pathway,23–25 and in the future we will need to find ways of matching the biological to the child’s asthma, which probably means defining endotypes. The alternative is a sterile series of N of 1 trials hoping to hit on what is best for the child.

The next important area tackled is the understanding of the early evolution of the asthmas. This is particularly important if we are to move asthma treatment from mere palliative care with ICS, to prevention or cure of the disease in the early stages. A large number of birth cohort studies have provided important information about the natural history of early wheezing, but the molecular pathways are not determined, and they are not steroid responsive.26–28 The Commission advocates moving from the reductionist approach, which has been so successful in the past, to a systems biology approach that allows a multidimensional, multilevel approach. New longitudinal studies must incorporate outcomes reflecting lung function and immunological, inflammatory, metabolic, genetic and epigenetic compartments. The profound effects of the environment on the endotypes need to be modelled, including a developmental perspective, because compartments may mature at different rates. There also needs to be an extrapulmonary focus on, for example, cardiovascular, endocrine, metabolic and neurological systems. Work published since the commission has thrown this into sharp focus; by the third decade, those in the lowest spirometry tertile have increased mortality,29 probably reflecting that impaired spirometry is the ‘canary in the mine’ indicating multisystem issues.30 31 A step-change in our developmental perspectives is needed, with the goal being an understanding of the pathways to established eosinophilic asthma, biomarkers for those pathways and a strategy to interrupt them to prevent long-term eosinophilic inflammation32 33 and loss of lung function which last greatly increases the risk of adult chronic obstructive pulmonary disease (COPD).34

Another crucial area is asthma attacks; the feeble word ‘exacerbation’ should not be used because it lulls people into a false sense of security. These should not be seen as a mere inconvenience, readily put right by 3–5 days of prednisolone. Rather, they are a huge sentinel event. NRAD2 has highlighted they are a risk factor for asthma deaths, and they are certainly a risk factor for further attacks. Attacks of severe wheeze in the preschool years are associated with progression to persistent wheeze,35 and attacks in the school-age years with impaired spirometry36 and thus adult COPD.34 There should be zero tolerance for asthma attacks. We have much to learn from cardiology; a heart attack prompts a focused response to understand what went wrong and what future prevention steps should be taken. The same should be for lung attacks, whether due to asthma, CF37 38 or primary ciliary dyskinaesia,39 because they are not transient inconveniences in any of these diseases. The Commissioners also make clear the need to rethink the definition of severe asthma. NRAD highlighted that 60% of asthma deaths were in people who did not meet the current definition of severe asthma, surely a ludicrous non sequitur. The report highlighted the numerous social and environmental traits that were treatable but not addressed, in particular adherence to ICS, overuse of bronchodilators and failure to attend review appointments.2 Adherence is a particularly important aspect that the Commission discusses; we cannot rely on asking the patient—we have to use more objective measures.40 Thus, the Commission proposes a new and broader definition of severe asthma.

Finally, the Commissioners call for better research, a cry that all paediatricians will echo. We need research in children; we cannot simply extrapolate from adults. So if we are studying early life events, we need developmentally appropriate animal models,41 sensitised appropriately (when was the last time you saw a child who had received an intraperitoneal injection of ovalbumin?). Research is an area where better phenotyping is essential. Genome-wide association studies (GWASs) and other genetic investigations are becoming ever more sophisticated, but the clinical phenotypes are often crude in the extreme—‘Dr-diagnosed asthma’—we know how unreliable that is42 43—or someone somewhere gave the subject an inhaler for some reason. We need to insist on better definitions of what we are actually studying. It is interesting to note that the asthma susceptibility gene CDHR3, which turns out to be the receptor for rhinovirus C,44 was not picked up on big GWASs of all comers with ‘asthma’, but only in a smaller, more focused study where the subjects had to have had an attack of acute severe wheeze45; was this because these subjects were actually not diluted by people with non-specific, asthma-unrelated, respiratory symptoms?

There are important implications for the delivery of care to people with asthma. Clearly, if the diagnosis is objective and secure, and the response to low-dose therapy is excellent, further testing would not be needed. However, if the response to low-dose therapy is not good, rather than sleep-walking up the steps in the guidelines, essentially giving more and more of the same therapy, we suggest that the patient ought to be seen in a specialist setting where phenotyping can be carried out, even if subsequent management is devolved back to primary care.

The Commission finishes with seven recommendations (box); it is a deliberately controversial document and meant to stimulate thought. The overall ambitious aim is to initiate a new era of personalised therapy for airway disease. We need to get the right amount of ICS to the right airways, not put steroids in the tap water; we need to make measurements, not just rely on chit-chat; and we need in all sorts of ways to move into a 21st century way of thinking about and researching airways disease. Asthma is dead—long live The Asthmas.