Key Points
- •
Problematic, severe asthma comprises wrong diagnosis (‘not asthma at all’), asthma with co-morbidities (‘asthma plus’), difficult asthma and true severe, therapy-resistant asthma. These patients should undergo a detailed, protocolized series of investigations.
- •
More than 50% of children referred to tertiary centers with problematic severe asthma in fact will be well controlled if basic management is optimal; they do not require therapy with ‘beyond guidelines’ medications.
- •
Invasive investigation should be considered in children with apparent severe, therapy-resistant asthma to determine whether there is discordance between symptoms and inflammation; whether there is an unusual pattern of inflammation; whether the child is steroid responsive and to what extent; and whether there is fixed airflow limitation.
- •
Unlike adult severe asthma, pediatric severe, therapy-resistant asthma is characterized by marked atopy and airway eosinophilia not neutrophilia, but without classic Th2 signature cytokine expression; the optimal monitoring strategy is unclear; sputum cellularity is much more variable over time and titrating treatment against sputum eosinophil count is not useful.
- •
There is little or no evidence base for ‘beyond guidelines’ therapy in children who fail standard therapies including omalizumab.
Introduction
Refractory asthma is rare in childhood and probably accounts for less than 5% of all pediatric asthma. It may be becoming less common over time, possibly because of more effective modern treatments. However, this group accounts for an enormous amount of morbidity and healthcare costs, and even mortality, and so although rare, it is a very important topic. This chapter describes the assessment of children aged 6 years and older using the Royal Brompton Hospital approach to apparent refractory asthma. This personal practice is described as no evidence-driven approach exists. The reader is invited to judge the value of these protocols, as well as referring to a recent ERS/ATS guideline. Finally, recent research advances are briefly summarized. Severe preschool wheeze is not discussed here.
Refractory Asthma: Basic Principles
The cardinal sin in asthma management is to continue to escalate and intensify asthma therapy in a child who is not responding to treatment, without posing the question, why is the prescribed treatment not working? We know that most children with asthma will respond very well to low-dose inhaled corticosteroids (ICSs) when properly administered, sometimes in combination with a second controller such as long-acting β 2 agonists (LABAs) or leukotriene receptor antagonists (LTRAs). So, when faced with a child with apparently refractory asthma, the pediatrician should not reach again for the prescription pad, but go through a rigorous protocol to determine what it is about the child and his/her asthma that means the anticipated response is not happening.
Nomenclature
We use the term ‘ problematic severe asthma ’ to describe children who, despite apparently optimal asthma management, have ongoing symptoms ( Table 37-1 ). It should be noted that this definition is arbitrary and not evidence based. Problematic severe asthma itself comprises four categories, the management of each of which is entirely different ( Table 37-2 ).
|
|
The overall aim of these protocols is to determine if the individual child is truly a candidate for ‘beyond guidelines’ therapy or can be better managed by standard approaches.
An overlapping approach which is also a useful conceptual framework is to define risk by considering domains of asthma severity. These are:
- 1.
Level of prescribed treatment
- 2.
Level of baseline asthma control over the previous month
- 3.
Burden and nature of exacerbations over the previous 6 to 12 months
- 4.
Risk of future complications, including failure of normal airway growth (for which there is increasing evidence), risk of future loss of control and/or exacerbations; risk of medication side-effects.
Approach to the Child with Problematic Severe Asthma
The first step is as always a full history and physical examination. Close attention should be paid to the nature of the symptoms. In particular, the word ‘wheeze’ is used very imprecisely, including being applied to crackling noises, upper airway noises and even stridor. The possibility that the child does not in fact wheeze should be considered unless and until a physician has heard wheeze with a stethoscope. Cough-variant asthma is a popular diagnosis, but most children who cough do not have asthma or indeed any other disease. The decision as to whether to investigate further, and what tests to perform, is driven by findings on history and examination. The two classical errors are failure to document physician-heard wheeze and failure to document airflow obstruction which changes over time or with treatment. There has been an increased interest in the diagnostic use of measurements of airway inflammation such as exhaled nitric oxide (FeNO), usually measured at a flow rate of 50 mL/s (FeNO 50 ) and induced sputum cytospin. Although evidence of airway inflammation is not a mandatory diagnostic test for asthma, and in particular a child already on treatment with ICS may be inflammation free, absence of any evidence of airway inflammation in a child apparently symptomatic for asthma should prompt a diagnostic review. Finally, the diagnostic process should not end at this stage; the possibility of a wrong diagnosis should be at the forefront throughout, and further testing considered in the presence of a surprising finding such as airway neutrophilia (see below).
Does the Child Actually have Asthma?
Numerous conditions may mimic asthma ( Table 37-3 ). The likelihood of particular diagnoses will vary across the world. Important clues to an alternative diagnosis include neonatal onset of symptoms; chronic productive cough for many sequential weeks; and evidence of systemic disease. The nonatopic child with apparently severe asthma should always be carefully assessed, since most such children have multiple positive skin prick tests and specific IgE, as well as a raised total IgE. A review of diagnostic testing for all the conditions listed in the Table 37-3 is beyond the scope of this chapter, but as a principle, testing should be focussed rather than take a scattergun approach.
| Class of Diagnosis | Examples |
|---|---|
| Local immunodeficiency | Cystic fibrosis, primary ciliary dyskinesia, persistent bacterial bronchitis |
| Systemic immunodeficiency | Any, including B cell and T cell dysfunction |
| Intraluminal bronchial obstruction | Foreign body, carcinoid, other tumor |
| Intramural bronchial obstruction | Bronchomalacia, complete cartilage rings, intramural tumor |
| Extraluminal bronchial obstruction | Vascular ring, pulmonary artery sling, congenital lung cyst, enlarged lymph nodes due to tumor or tuberculosis, other mediastinal masses |
| Direct aspiration | Bulbar or pseudobulbar palsy; laryngeal cleft |
| Aspiration by direct contamination | H-type fistula |
| Aspiration secondary to reflux | Any cause of gastroesophageal reflux, including hiatus hernia and esophageal dysmotility |
| Complications of prematurity | Bronchomalacia, structuring secondary to intubation, vocal cord palsy secondary to surgery for patent arterial duct |
| Congenital heart disease | Bronchial compression from enlarged cardiac chambers or great vessels; pulmonary edema |
| Interstitial lung disease | Any not presenting with neonatal respiratory failure |
| Dysfunctional breathing | Vocal cord dysfunction, hyperventilation syndromes |
Are There Significant Co-morbidities Which Need Addressing?
A co-morbidity may not be easy to treat (e.g. obesity), but it is difficult to argue that potentially toxic biologicals are appropriate for a child who is not being supported by his/her family to lose weight. Co-morbidities have recently been reviewed in detail.
Gastroesophageal Reflux
The possible relationships between asthma and respiratory symptoms, and gastroesophageal reflux are complex. Reflux can cause symptoms either by direct contamination of the lower airway or indirectly by an esophagobronchial reflex; respiratory disease can cause reflux, including via mechanisms secondary to abnormal pleural pressure swings or configuration of the diaphragm; and reflux may be an asymptomatic fellow-traveler. The best evidence is that, irrespective of any symptoms suggestive of reflux, therapy with, for example, a proton pump inhibitor does not improve severe asthma; and that has certainly been our experience, despite hitherto performing pH-metry as part of our work-up for severe asthma.
Rhinosinusitis
Upper airway disease worsens quality of life, and should be treated on its own merits in any context. There is increasing evidence that treating rhinosinusitis may be beneficial at least in mild to moderate asthma. The mechanisms of any benefit remain conjectural. In our series, significant rhinosinusitis is unusual in severe asthma.
Obesity
There are complex potential interactions between asthma and obesity. Obesity may cause breathlessness and ‘wheeze’ without evidence of asthma, leading to inappropriate treatment. Obesity may be associated with a pauci-inflammatory form of asthma, at least in adults, although it is sometimes unclear whether this is true asthma ; asthma with an eosinophilic phenotype on airway wall biopsy with raised sputum interleukin 5 (IL-5) ; and steroid resistance. Obesity is a proinflammatory state, as is obstructive sleep apnea (OSA) (see below). Asthma (via reduced exercise performance) and its treatment (prednisolone bursts or long-term therapy) may cause or contribute to obesity. Hence, particular care is necessary before escalating therapy for ‘asthma’ in the obese child with respiratory symptoms. Weight reduction is always beneficial in the obese child, but is difficult to achieve.
Upper Airway Obstruction/Sleep Disordered Breathing
There is an increasing literature on asthma and OSA. Although the literature reports associations in at least mild to moderate asthma, in our patients OSA is very rare in severe asthma, except in the presence of concomitant obesity (see above). We do not routinely perform polysomnography on children with severe asthma who are not obese. OSA was reported to cause sputum neutrophilia rather than eosinophilia in one study; this inflammatory pattern rarely if ever occurs in true severe, therapy-resistant asthma, and if seen, should prompt a diagnostic re-evaluation.
Dysfunctional Breathing
Vocal cord dysfunction and other forms of dysfunctional breathing are common in asthma, although not all groups report this and the symptoms are frequently misattributed to asthma, with inappropriate escalation of therapy. There is much less work in children than in adults. Fifteen percent of children investigated following our protocol had evidence of dysfunctional breathing, including hyperventilation and vocal cord dysfunction. Clues include the absence of symptoms at night, and often stridor rather than expiratory noises. An experienced respiratory physiotherapist should be asked to assess breathing pattern, and to treat an abnormal breathing pattern with a training program, although currently there is no randomized controlled trial evidence of benefit. Parental video-recording of an attack may be illuminating. In older children, direct laryngoscopy during an exercise test enables direct demonstration of the problem. Dyspnea perception has been little studied in severe pediatric asthma, but adults with severe asthma do not become as dyspneic as those with mild asthma during bronchoconstriction. The possibility of poor symptom perception as a cause of an apparent sudden catastrophic deterioration should be borne in mind.
Food Allergy
Atopy is almost inevitable in severe pediatric asthma, but patients with asthma and food allergy are over-represented in such cohorts. Whether food allergy is causative of the problem or a marker is unclear; certainly anaphylaxis at rest and on exercise enters the differential diagnosis of acute severe asthma, and should always be considered because it is treatable. Food allergy should always be properly documented if exclusion diets are proposed; blind dietary exclusions are frequently tried and in our experience are of no benefit.
Next Steps: Is True Severe, Therapy-Resistant Asthma Likely?
Once it is clear that the diagnosis is almost certainly asthma, and co-morbidities have been excluded or identified, we proceed with a more detailed evaluation, led by the specialist asthma nurse and often involving the clinical psychologist and respiratory physiotherapist. The assessment will include both a hospital outpatient visit and a nurse-led community assessment, with the nurse arranging a visit to the family home and making contact with the school.
Hospital Visit
Details of the hospital visit are given in Table 37-4 . We evaluate allergic sensitization by both skin prick tests (SPTs) and sIgE tests, because there is imperfect concordance (76–83%) between them. This includes fungal sensitization ; because severe asthma with fungal sensitization [SAFS] ( Table 37-5 ) has different treatment options (see below), we do not routinely perform double-blind food challenges. FeNO is measured ideally at multiple flow rates as an indirect assessment of proximal (J NO ) and distal (C ALV ) airway inflammation. Although there is evidence that distal inflammation may be important in severe asthma (below), variable-flow FeNO has not been evaluated as a clinical tool to guide therapy in an individual child, so this is largely a research technique.
| Issue to be Addressed | Tests Performed |
|---|---|
| Symptom pattern | Asthma control test, prednisolone bursts, unscheduled visits |
| Psychosocial factors | Questionnaires |
| Lung function | Spirometry before and after bronchodilator |
| Allergic sensitization | Skin prick tests, specific IgE |
| Aeroallergens | Grass and tree pollen, house dust mite, cockroach, cat and dog, and any others suggested by the clinical history |
| Food allergens | Peanut, milk, egg and any others suggested by the clinical history |
| Fungi (see Table 37-5 ) | Fungi Table 37-3 |
| Airway inflammation | FeNO 50 , multiple flow rates Induced sputum if FEV 1 is >70% predicted |
| Tobacco exposure | Urine or salivary cotinine |
| Medication adherence | Serum prednisolone and theophylline levels if prescribed; serum inhaled corticosteroid levels if available |
| Adult Criteria | Proposed Pediatric Criteria * |
|---|---|
| Treatment with 500 µg fluticasone/day or continuous oral corticosteroids, or four prednisolone bursts in the previous 12 months or 12 in the previous 24 months, and | Meets criteria for problematic severe asthma (above) |
| IgE <1000 (exclude ABPA) | No IgE exclusion |
| Negative IgG precipitins to Aspergillus fumigatus | No IgG exclusion |
| Sensitization (SPT, sIgE) to at least one of Aspergillus fumigatus, Alternaria alternata, Cladosporium herbarum, Penicillium chrysogenum, Candida albicans, Trichophyton mentagrophytes and Botrytis cinerea | As adult criteria |
* There is no agreed definition in children, but given the rarity of allergic bronchopulmonary aspergillosis (ABPA) in children with asthma, we have proposed to eliminate the total IgE and IgG criteria, from the diagnostic criteria.
It should be remembered that children are not mini-adults. In particular, it should be noted that the interpretation of spirometry and imaging has a developmental perspective.
Spirometry, Bronchodilator Responsiveness and Bronchial Challenge Testing.
Unlike in adults, spirometry is poorly discriminatory between asthma of different severities in children. Spirometry is of course useful as part of the definition of an exacerbation, and to monitor progression of lung growth over time in epidemiological studies. Epidemiological evidence is that, for groups , spirometry in severe asthma tracks over decades, but the pattern of lung growth may be abnormal in really severe asthma; for example, the Melbourne study showed a failure of the adolescent airway growth spurt. There is other worrying evidence from studies showing that individuals with apparently good control of symptoms with ICSs may have abnormal airway growth over time. One post hoc analysis suggested that this was related to the exacerbating phenotype, but only in patients not treated with ICSs. This study requires prospective confirmation, but is in line with studies of other airway diseases showing so-called ‘exacerbations’ lead to a worse overall long-term outlook. Bronchodilator responsiveness may be used diagnostically and to define persistent airflow limitation (PAL). Little has been written about bronchial challenge testing as a clinical tool in children with severe asthma. In many, it will be too hazardous because of poor baseline spirometry and/or extreme bronchial hyperreactivity. There is one situation in which direct airway challenge usually is useful in the diagnostic work-up, namely in a child with normal spirometry but reported severe symptoms, and in whom a negative challenge would make uncontrolled asthma unlikely. In such children, a combined protocol of hypertonic saline challenge and sputum induction may be useful; in genuine severe asthma, the need for albuterol pretreatment before sputum induction as a safety measure precludes this approach. The role of challenge testing in children with PAL or obliterative bronchiolitis (OB), as part of confirmation that further escalation of therapy is not useful, is not clear; in all probability, a systemic steroid trial would be preferred.
High-Resolution Computed Tomographic (HRCT) Scanning.
HRCT is not a routine part of our protocol, and there is no pediatric evidence to suggest the need for it. HRCT may be performed as part of the diagnostic work-up if, for example, the patient is nonatopic or bronchiectasis is suspected. However, in adult studies, bronchial wall dilatation is common in severe asthma, and it is important not to over-diagnose bronchiectasis in children, in whom minor degrees of airway dilatation may be reversible, even when related to an immunodeficiency. It is essential to note that HRCT scans may be unable to distinguish severe asthma from OB. In adults, but not in children, there is evidence that HRCT scans may be a useful biomarker of asthma severity. In children, HRCT changes consistent with asthma are less apparent (and also less well defined) than in adults, and bronchial wall thickening has no or only the weakest correlation with reticular basement membrane (RBM) thickening and forced expiratory volume in 1 second (FEV 1 ). Air trapping on HRCT may allow an estimate of distal airway disease, but in severe asthma has not been compared with sophisticated tests of distal airway function such as lung clearance index. There are no studies on the usefulness or otherwise of HRCT scans as a monitoring mechanism for severe pediatric asthma, and the radiation risk of even low-dose HRCT, especially in young children, is not a trivial consideration.
Home Visit
The nurse-led home visit is a key part of the work-up of problematic, severe asthma. Professors sitting in the clinic know little to nothing of what is really happening at home. Five areas are explored: adherence, tobacco smoke, allergens, psychosocial issues and asthma education. If the patient has been referred from a distant center, the home visit may be performed by a local specialist nurse after discussion with our own team. This approach may not be feasible everywhere, but in our hands, allows the identification of significant and potentially reversible factors in more than half of those referred with problematic severe asthma. It is certain that relying purely on a hospital-based assessment by a pediatrician will lead to many mistakes.
Adherence.
This is discussed in detail in chapter 38 . It was found that less than half of patients had picked up more than 80% of the required prescriptions, and nearly one third had picked up less than 50%. We also enumerated the collection of excessive prescriptions of short-acting β2-agonists (SABAs); collecting ≥6 prescriptions/year is associated with a poor outcome.
It was common to find that medication was past its expiry date; in 25% of cases a complete set of in-date medications could not be produced when the nurse visited. Other issues were total inaccessibility of any medications, and medications unopened in their original wrapping, neither of which inspired confidence that medications were actually being taken.
Parental Supervision.
In one study, even young children (20% of 7-year olds, 50% of 11-year olds) were left to take asthma medications unsupervised. Often parents think they are supervising treatment, but do not in fact actually witness their child take the therapy, instead merely calling out reminders to their child who is in a different room in the home. This situation requires sensitive exploration.
Use of Inhaler Devices.
These are often wrongly used and regular instruction in their use may lead to improvements. However, even multiple teaching sessions are not enough to ensure good inhaler technique; all the children in our series had had repeated instruction in specialized centers, and yet still had a poor technique. A common issue is using pressurized metered dose inhalers without spacers, because spacers are considered babyish; this omission guarantees minimal airway deposition of medications.
Environmental Tobacco Smoke and Other Irritants.
Active smoking by adults with asthma causes steroid resistance, and it seems likely that passive smoke exposure has the same effect. Passive smoke exposure is common in children with asthma; the prevalence of active smoking is unknown, but unlikely to be low. We found that 25% of children with problematic, severe asthma were exposed to tobacco smoke.
The mechanisms of tobacco smoke-induced steroid resistance have been researched mainly in adults; the phenotype is neutrophilic. In a study of children with severe, therapy-resistant asthma, we found that parental smoking reduced histone deacetylase protein expression and activity, and reduced the in vitro inhibitory effects of dexamethasone on tumor necrosis factor-α-induced IL-8 release from alveolar macrophages. Bronchoalveolar lavage (BAL) showed higher IL-8 concentrations and neutrophil counts and the children had lower asthma control test (ACT) scores compared with non-passive smoke-exposed children; these findings are supported by adult data. Additionally, it is also likely that symptoms are exacerbated by a direct irritant effect of smoke.
Other environmental irritants sometimes encountered include incense or joss sticks, and the extensive use of air fresheners and other aerosol sprays. Environmental pollution is also important, but difficult to modulate except at the level of public health.
Allergen Exposure.
This is discussed in detail elsewhere. Allergen exposure in the home combined with evidence of sensitization has been clearly implicated in the etiology of viral-induced asthma attacks.
In our study, 17 of 30 children who owned furry pets were sensitized to that pet on skin prick testing; only two had any allergen avoidance precautions in place. Thirty-one children were thought to have clinically significant house dust mite (HDM) exposure; five were using comprehensive allergen avoidance measures, 15 partial and 11 none. Reduction of mold exposure may be particularly important if SAFS (see above) is suspected. Allergen exposure in school may also be important, but this is an even more difficult area in which to intervene.
Psychosocial Issues.
Acute and chronic stress may trigger asthma exacerbations; there may be quite complex time relationships between the two. Stress has been shown to amplify the airway eosinophilic response to allergen challenge. In our study, psychosocial issues were common, especially anxiety and depression, a finding reported by others. Most issues only emerged during discussions in the home. Altogether, about half the families were referred to clinical psychology for a more detailed assessment. It is not productive to try to determine whether anxiety and depression are the cause or result of severe asthma; both are treated on their individual merits.
Asthma Education.
Some adherence issues relate to basic misunderstandings of the purpose of treatment, and these are also addressed. If the child does not have a detailed asthma plan, this is put in place and communicated to the school. It is to be hoped that all this will have been done previously, but any gaps are sought and closed.
Lesson Learned the Hard Way
School Visit.
Despite the best efforts of my specialist respiratory nurses, I have made many mistakes. The importance of contact with the school cannot be overemphasized. Teachers are an important resource; they are experienced in assessing children and spend many hours each day with them. If there is any suspicion that symptoms are being overcalled by the parents, talk to the teacher. In one particularly egregious case (details modified to preserve patient confidentiality) a girl with so-called severe asthma was in fact the captain of field hockey, had the nickname of ‘the Greyhound’ and her teachers did not know that she even had asthma, let alone that a reserve supply of SABA was kept for her!
Management of Exacerbations.
In theory, the treatment of an asthma attack is based on objective evidence of present symptoms. In practice, the previous history is influential. When assessing the child with apparent multiple severe exacerbations, it is essential to determine what (if any) objective assessments were carried out before instituting treatment, and whether or not the level of treatment was objectively justified. One child under our care was actually started on intravenous albuterol despite being fully saturated on room air!
Parental Doublespeak.
The basic tenet of pediatrics is to believe the parents. On occasion, parents not telling the truth may be the underlying problem, and this is very difficult to detect. This is distinct from the effect of a very understandable parental anxiety leading to exaggeration of symptoms. Motivation for the former may include access to financial and other benefits as a result of having a ‘sick’ child, right up to deliberate fabrication of symptoms (Munchausen by proxy). Unfortunately the result is often a long delay in diagnosis.
Multidisciplinary Team Discussion
The assessment generates a mass of detailed information, and this is next assessed in a dedicated, multidisciplinary team meeting. We aim to determine whether the child has difficult asthma for which the basic management steps need to be addressed, or potentially severe, therapy-resistant asthma, which justifies further invasive investigations.
Does the Child Have ‘Difficult Asthma’ or Severe, Therapy-Resistant Asthma?
In more than half of the children, no further investigations are undertaken, or at least, invasive testing is deferred pending other interventions. This unfortunately does not mean the problem is necessarily solved; for example, it is one thing to identify poor adherence as an issue, but quite another to address it. Often, the child’s poor adherence to medication has not been appreciated by the parents, and identification of the problem leads to them finding a solution. Adherence may be addressed by giving the child an inhaler with a microchip to record when medication is taken. We never attempt deception, and always tell the child and family that a recording is being made. If the inhaler is used and the child improves, this is gratifying. However, it is still illuminating if the inhaler is not used and asthma symptoms continue unabated. We have sometimes used directly observed therapy at school, on the basis that 5 days of treatment a week is better than none; if this is put in place, it is essential to check that the seemingly obvious, namely that therapy is directly observed, is actually happening. One child told us that the school nurse was always working at a desk with her back to the child when supposedly directly observing therapy!
Deferring investigations for psychological intervention are common. Psychological interventions may be effective. In general, individualized plans work best. Of course, these may be combined with the other interventions outlined above, and also with invasive investigations.
Addressing allergen exposure in the home is notoriously difficult, especially if the allergen concerned originates from a much-loved family pet (the English disease). Frequently, the cat will be sent away for a short time and then allowed to return because the child’s asthma has not improved; however, at least a year’s separation is necessary substantially to reduce the allergen load. We do not prescribe omalizumab to any child unless every effort to reduce the environmental allergen burden has been made.
The fact that many children with apparent severe asthma just need to get basic management right has been reported by others, and has been the bugbear of intervention study design; those with apparently severe asthma tend to melt away when proper, protocol-driven management is put in place.
The remainder of the children continue in the protocol as outlined in the next section.
Is the Distinction Between ‘Difficult Asthma’ and ‘Severe, Therapy-Resistant Asthma’ Meaningful?
We have published follow-up data for 78 children. Those assigned to the difficult asthma group were able to reduce their prescribed dose of ICS (although perhaps in reality they were actually taking it for the first time) while increasing their FEV 1 and having fewer prednisolone bursts than the children with severe, therapy-resistant asthma. This is in accord with the multifaceted intervention studies reported in those with less severe asthma. The severe, therapy-resistant group had smaller increases in FEV 1 but were unable to reduce treatment. We concluded that the groups behaved differently, and the distinction was useful. We also assessed whether or not the two groups could be distinguished from basic measurements; however, although children with severe, therapy-resistant asthma had a lower FEV 1 % predicted, more bronchodilator reversibility and a higher FeNO 50 than those with difficult asthma, the overlap between the groups made placing all individuals into the correct category an impossibility. A similar result was reported by the Severe Asthma Research Program (SARP) group, namely statistically significant but clinically unreliable differences between severe and moderate asthma in terms of FEV 1 and FeNO 50 .
Research Implications.
The categorization into difficult and severe, therapy-resistant asthma has important implications for the interpretation of research studies, such as genetic association studies. Cohorts of children with ‘severe asthma’ who have not gone through a detailed filtering process will be diluted by at least 50% with children for whom the basics have simply not been addressed, and therefore may not be comparable to those with truly therapy-resistant asthma.
Invasive Testing
Typically children with severe, therapy-resistant asthma have very major problems with asthma ( Table 37-6 ), thus justifying an invasive approach. There are a number of differences between the typical child with severe, therapy-resistant asthma and the adult with this diagnosis. In children, there is no female predominance, but a strong atopic history was common (85% atopic with a total median IgE of 386 [115–1286]). The children are also not obese. Although we have published a number of research papers based on our bronchoscopic studies, the aim of testing is primarily clinical to answer four key questions ( Table 37-7 ). An individualized treatment plan is devised on the basis of the answers to these questions.
| Parameter | Mean Result (range) |
|---|---|
| Symptoms (ACT) | 13/25 (9–17) |
| Symptom duration (years) | 10.1 (9.3–12.7) |
| Number intubated | 11/53 or 21% |
| FEV 1 %predicted | 69 (55–87) |
| Acute FEV 1 response to 1 mg of albuterol (%) | 15.6 (5.5–23.4) |
| FeNO 50 (ppb) | 50 (29–70) |
| Mean sputum eosinophils (%) | 7.5 (3.2–30.4) |
|
Protocol: Invasive Investigation of Severe, Therapy-Resistant Asthma
The child is assessed invasively and noninvasively on the same day. Noninvasive investigations are symptom assessment, spirometry and acute response to bronchodilator, sputum induction and FeNO measurement, as performed in the initial assessment. The child then undergoes a bronchoscopy under a general anesthetic, with performance of BAL and endobronchial biopsy, and often bronchial brushings as a research procedure, as described in detail elsewhere. At the end of the procedure, we administer a single dose of intramuscular triamcinolone (40–80 mg depending on the size of the child). The child is kept in hospital overnight, even if a pH study is not to be performed. The child is then seen again 4 weeks later, and all the noninvasive measurements are repeated.
The key requirement for the safe performance of bronchoscopy in these children is a skilled anesthetist. The child may be premedicated with SABAs. Adverse events are very rare. We have seen acute severe bronchospasm on one occasion, which rapidly responded to treatment. Close post-procedure monitoring of the child is essential.
Assessment of Airway Inflammation.
We do not see neutrophilic inflammation in children with severe asthma, which is in marked contrast to what is seen in adults. Instead, induced sputum, BAL and endobronchial biopsy are eosinophil dominated, although there are marked variations between individuals in the extent of eosinophilia. Hence, a neutrophilic phenotype should prompt reconsideration of the diagnosis. There tends to be agreement between induced sputum and BAL, but discordance between these luminal compartments and the airway wall histology. It is not clear which is most relevant to disease pathophysiology. The relationship between FeNO and airway eosinophilia is discussed below.
Given the marked eosinophilic phenotype, a Th2 cytokine profile would be expected, but two pediatric severe asthma studies have failed to confirm this. We studied induced sputum supernatant, BAL using both the Luminex and CBA platforms, and performed immunohistochemistry on endobronchial biopsies, and found scant evidence of the Th2 signature cytokines IL-4, IL-5 and IL-13. The SARP group also failed to find evidence of Th2-driven inflammation. They compared severe, therapy-resistant asthma with mild-moderate disease and showed that the cytokines which best discriminated between the two groups were GRO (CXCL1), RANTES (CCL5), IL-12, interferon-gamma (IFN-γ) and IL-10. They concluded that pediatric severe, therapy-resistant asthma was neither a Th1- nor a Th2-driven disease. Gene expression studies of the sort that have generated the Th2 high and low groups in adults have so far not been performed in children. Whether those few children in whom Th2 cytokines are detectable (see above) are in fact Th2 high remains to be seen. In this context it should be noted that periostin, a useful serum biomarker in adults for the Th2 high phenotype, is not useful in children because it is released from growing bone.
The findings of these two large studies described above do not necessarily show that Th2 cytokines are unimportant in the pathophysiology of severe asthma. All children with severe, therapy-resistant asthma are by definition prescribed high-dose ICS, and it may be that the Th2 component of their disease is abrogated by this therapy. What these studies do suggest is that the factors driving ongoing disease are different from those in adults. Attention is shifting to the epithelial-derived cytokines (IL-25, IL-33, thymic stromal lymphopoietin [TSLP]) as possible candidates, perhaps interacting with innate lymphoid cells. We have shown that IL-33 promoted collagen synthesis by fibroblasts from pediatric patients with severe asthma. We also showed that increased cellular expression of IL-33, but not IL-13, was associated with increased reticular basement membrane thickness in endobronchial biopsies. IL-33 also stained strongly in the endobronchial biopsies. The results were supported by animal data. Also, a recent study of a monoclonal antibody (AMG 157), which prevents TSLP interacting with its receptor, led to marked attenuation of both the early and late phase allergen response in adults with mild asthma, a surprising finding given that only a single cytokine was blocked.
The Th17 pathway has been considered another candidate, at least in adults, but it possibly more likely drives a neutrophilic phenotype, and thus is an unlikely candidate in children. However, eosinophil chemoattraction by IL-17 has been reported, at least in animal models. Furthermore, a recent trial of the monoclonal anti-IL-17 receptor antibody brodalumab in adults showed disappointing results. Further work is clearly needed in order to delineate the proinflammatory mechanisms driving pediatric severe, therapy-resistant asthma.
There is an important implication of the above findings. An increasing number of studies of anti-Th2 cytokine strategies, such as mepolizumab (anti-IL-5, lebrikizumab (anti-IL-13) and dopilumab (IL-4 receptor alpha chain) have recruited children aged 12 years and older, although they are largely dominated by adult participants. It is essential that the promising results in these trials are not uncritically extrapolated to children. It may be justified to perform a therapeutic n-of-1 trial in a child doing badly on all therapies, but further trials in children are needed, including measurements of Th2 cytokines, before Th2 cytokine strategies can be widely recommended in pediatric severe, therapy-resistant asthma.
Assessment of Steroid Responsiveness.
It is likely that steroid responsiveness is a continuum; true steroid unresponsiveness is likely confined to rare cases of mutations in the corticosteroid receptor. The widely accepted adult definition of steroid response is ≥15% predicted increase in FEV 1 in patients with bronchodilator reversibility (BDR) of ≥12% from baseline and an abnormal FEV 1 (≤80%) prior to the trial. There is no accepted definition in children, and the adult definition cannot be applied to around half of our pediatric patients with severe, therapy-resistant asthma, mainly because their baseline spirometry is not sufficiently abnormal prior to the trial, and also because spirometry is often a poor reflection of asthma severity in children (see above). Furthermore, there is no consensus on the dose and duration of corticosteroids for the trial. We opted to use triamcinolone intramuscularly to ensure adherence. We use a multimodality definition of steroid responsiveness ( Table 37-8 ). Most children are partial responders, about 10% are total nonresponders and 10% completely responsive in all domains (complete responders). Some children who were nonresponsive according to the adult definition were partial responders in the multidomain assessment. We have also shown that additional doses of triamcinolone do not change the category of responsiveness (unpublished observations). While the clinical utility of this approach has yet to be confirmed, it might serve as a useful template for current research; so, for example, an intervention with an anti-inflammatory medication might be expected primarily to affect the inflammatory domain.
