Abstract
Previous publications have highlighted the disparity between research trial populations and those in clinical practice, but it has not been established how this relates to randomised controlled trials (RCTs) of phenotype-targeted biological therapies in severe asthma.
Detailed characterisation data for 342 severe asthma patients within the Wessex Severe Asthma Cohort (WSAC) was compared against comprehensive trial eligibility criteria for published phase IIB and phase III RCTs evaluating biological therapies in severe asthma since 2000.
37 RCTs evaluating 20 biological therapies were identified. Only a median of 9.8% (range 3.5–17.5%) of severe asthma patients were found to be eligible for enrolment in the phase III trials. Stipulations for airflow obstruction, bronchodilator reversibility and smoking history excluded significant numbers of patients. A median of 78.9% (range 73.2–86.6%) of patients with severe eosinophilic asthma would have been excluded from participation in the phase III licensing trials of interleukin (IL)-5/IL-5R targeted therapies.
Despite including only well characterised and optimally treated severe asthmatics under specialist care within the WSAC study, the vast majority were excluded from trial participation by criteria designed to re-confirm diagnostic labels rather than by biomarker criteria that predict the characteristic addressed by the treatment.
Abstract
RCTs of biological therapies in severe asthma are poorly generalisable with most patients excluded by outmoded disease concepts despite possessing the targetable trait addressed by the treatment http://ow.ly/iog930md0J3
Introduction
Asthma affects over 300 million people worldwide with an estimated total annual cost in the UK approaching £5 billion [1, 2]. Whilst the majority of people with asthma can be treated effectively with inhaled corticosteroids (ICS) and bronchodilators, 5–10% with severe asthma suffer persistent symptoms, frequent exacerbations and an accelerated decline in lung function despite treatment [3]. Severe asthma is significantly more expensive with increased healthcare resource usage and high-cost medications accounting for much of this additional cost [4].
Cohort studies and cluster analyses have advanced our understanding of severe asthma, establishing it as a heterogeneous condition encompassing multiple phenotypes with underlying endotypes [5–9] defined by specific pathobiological pathways that underpin the manifestations of the disease. To address the significant unmet clinical need in severe asthma, there has been a focus since the turn of the millennium on developing biological therapies to target specific components of these inflammatory pathways (predominantly in those with Type-2 inflammation) [10]. These targeted interventions are recognised as an important step towards personalised medicine for patients with severe asthma. However, such treatments are expensive mandating that their use is rationalised by high-quality clinical evidence of efficacy and effectiveness, and the use of biomarkers to stratify patients and determine those most likely to benefit from treatment.
Recent expert commentary has proposed that less emphasis be placed on historical definitions of asthma, highlighting that most randomised controlled trials (RCTs) study populations that are poorly generalisable to clinical practice [11, 12], with a focus instead on “treatable traits” to identify clinical trial populations who are most likely to benefit from an intervention. Previous studies have shown that only 3.3–6% of patients with asthma fulfilled the eligibility criteria for the clinical trials upon which asthma guidelines are based [13, 14]. However, these studies did not focus on patients with severe asthma under specialist care or biological therapies targeting specific asthma phenotypes and thus it is unclear whether a similar impact on the generalisability of trial data exists.
To investigate this we have therefore aligned the data from a large well characterised cohort of severe asthma patients, the Wessex Severe Asthma Cohort (WSAC), with the clinical trial eligibility criteria of published RCTs assessing biological therapies in severe asthma. Additionally, we have compared the profile of the WSAC with published data from other severe asthma cohorts to evaluate how representative this cohort is of the broader severe asthma population, so that the implications of these findings can be fully appreciated.
Methods
Study design
The WSAC is an observational, cross-sectional study providing detailed characterisation of severe asthma patients recruited from specialist severe asthma clinics at Portsmouth and Southampton Hospitals between April 2009 and January 2014. Study participants had asthma, confirmed by a specialist in accordance with the British Thoracic Society (BTS)/Scottish Intercollegiate Guidelines Network (SIGN) guidelines 2009, which remained poorly controlled with persisting symptoms and exacerbations despite treatment with high-dose ICS (or maintenance systemic corticosteroids), a long-acting β2-agonist (LABA) and/or alternative controller medications (equivalent to step 4 or step 5 of the Global Initiative for Asthma (GINA) management guidelines for asthma 2017 [15]), as well as focused management of any co-morbid conditions (full inclusion criteria are included in the supplementary material).
Characterisation protocol
All participants underwent a detailed characterisation protocol including clinical, physiological and biological assessments (full details are included in the supplementary material). The study was funded by the UK Medical Research Council (MRC)/National Institute for Health Research (NIHR) Patient Research Cohorts Initiative and was conducted in accordance with the International Conference on Harmonisation and Good Clinical Practice standards and the ethical principles outlined in the Declaration of Helsinki. Independent ethics committee approval was obtained (MREC No. 09/H0502/37) and all participants provided written, informed consent prior to participation in the study.
Identification of trials and eligibility analysis
A systematic search was used to identify all phase IIB and phase III RCTs studying novel treatments in severe asthma between January 2000 and January 2018. Abstracts were reviewed and primary publications sought for relevant RCTs (the reference lists of these publications were also reviewed). Key eligibility criteria were extracted from primary publications, published trial protocols and clinical trial databases where available. Each patient from the WSAC was assessed against the eligibility criteria for each trial to determine the numbers that would have been deemed suitable for enrolment and key criteria which excluded patients from trial participation were reviewed. Criteria were divided into diagnostic criteria (e.g. airflow obstruction and reversibility) and biomarker criteria (e.g. peripheral blood eosinophil count), with the latter used to identify a specific disease phenotype. Where relevant data was unavailable it was assumed patients remained eligible, with the exception of studies mandating sputum eosinophilia where failure of sputum production precluded enrolment. In addition, each patient was assessed against the National Institute for Health and Care Excellence (NICE) treatment recommendations for biological therapies currently licenced for use in asthma in the UK. The primary outcome was the proportion of patients eligible for each of the RCTs identified.
Role of the funding source
The UK MRC and NIHR provided joint funding for the study but did not contribute to its design or to collection, analysis or interpretation of the data.
Results
The WSAC enrolled 342 severe asthmatics, all of whom fulfilled the American Thoracic Society (ATS)/European Respiratory Society (ERS) 2014 definition of severe asthma. A summary of the characteristics of this group is shown in table 1 and further details are included in the supplementary material. The severe asthma patients within the WSAC are demographically comparable to previous cohorts/registries, as detailed in the supplementary material.
Wessex Severe Asthma Cohort (WSAC) characteristics summary (n=342)
37 RCTs, comprising 23 phase II and 14 phase III trials, assessing 20 novel therapies in over 15 000 patients with severe asthma were identified. Of these RCTs, 29 (78%) assessed treatments targeting the Type-2 “high” inflammatory pathway. The most frequent primary endpoint was a reduction in exacerbation frequency (71% of phase III trials).
Only a median of 9.8% (range 3.5–17.5%) of severe asthma patients within the WSAC would have been eligible for enrolment in the phase III trials of biological therapies in severe asthma. The proportion of patients within the WSAC who would have been suitable for enrolment in each RCT is shown in table 2. Whilst there is an increment in eligibility between phase II and phase III RCTs, overall suitability remains low.
Summary of Wessex Severe Asthma Cohort (WSAC) trial exclusions#
Commonly used eligibility criteria and their impact on trial enrolment are highlighted in table 3. The requirements for persistent airflow obstruction and/or significant bronchodilator reversibility were key reasons for trial exclusion and when both criteria were required only 33.6% of severe asthma patients in the WSAC remained eligible. The cumulative effect of multiple eligibility criteria dramatically restricts eligibility for trial participation. However, the use of composite inclusion criteria, allowing bronchodilator reversibility, bronchial hyper-responsiveness and/or measures of variable airflow obstruction, modestly increases median eligibility from 7.6% (range 2.1–30.7%) to 15.8% (range 4.1–39.5%).
The impact of commonly used eligibility criteria on Wessex Severe Asthma Cohort (WSAC) eligibility
In table 2, eligibility is subdivided into asthma and biomarker criteria, which demonstrates that the majority of patients are excluded by non-phenotypic criteria. The percentage of patients with blood eosinophil counts of ≥300 cells·µL−1 who would have been eligible for enrolment in published phase III trials of IL-5 and IL-5R targeted therapies is illustrated in figure 1. Again this demonstrates that most are excluded by non-phenotypic criteria. A similar effect is seen with an eosinophilic population defined as ≥2% or ≥3% of sputum eosinophils, or by blood eosinophil counts of ≥150 cells·µL−1 (supplementary material). The median eligibility for RCTs assessing biological agents targeting Type-2 asthma (8.8%, range 2.1–26.9%) was lower than for non-Type-2 asthma (14%, range 5.3–39.5%) and comparative RCTs of novel non-biological therapies for severe asthma (33.9%, range 20.8–76.9%).
Trial eligibility for phase III interleukin (IL)-5 targeted treatments in severe asthmatics with a blood eosinophil count of ≥300 cells·µL−1
Within the WSAC, 26% of severe asthmatics were current smokers or ex-smokers with a smoking history of ≥10 pack-years. Of those patients who successfully produced sputum (59 out of 90; 62%) or had a peripheral blood count (81 out of 90; 90%), 56% had a sputum eosinophil count of ≥2% and/or a peripheral blood eosinophil count of ≥300 cells·µL−1 but were not eligible for enrolment in most trials targeting Type-2 “high” disease due to their smoking status. The impact of smoking on Type-2 biomarker status in the WSAC is shown in table 4.
The impact of smoking on Type-2 biomarker status in the Wessex Severe Asthma Cohort (WSAC)
That less than 50% of the severe asthmatics in the WSAC who fulfilled the NICE recommendations for treatment with mepolizumab and reslizumab would have been eligible for inclusion in the phase III trials of these therapies (mepolizumab 45.3%; reslizumab 33.9%) is demonstrated in figure 2.
A comparison of National Institute for Health and Care Excellence (NICE) treatment recommendations and phase III trial eligibility. All phase III trials were included using eligibility criteria from NICE technology appraisal guidance TA431 (January 2017) and TA479 (October 2017). For full criteria see the supplementary material.
Discussion
The WSAC was established with the aim of evaluating real-world severe asthma patients and is comparable to the severe asthma populations described in previously published cohorts and registries [7, 16–19], from which cluster analyses have identified the currently recognised asthma phenotypes. Despite including only well characterised and optimally treated severe asthmatics under specialist care, the vast majority of patients in the WSAC (90.2%; range 82.5–96.5%) would have been excluded from the landmark phase III trials of biological therapies published to date.
Severe asthma is a heterogeneous condition and biological therapies are only likely to benefit subsets of the population. It would seem reasonable to assume that the majority of patients with poorly-controlled severe eosinophilic asthma (despite high-dose inhaled steroids) should have been eligible for inclusion in RCTs of therapies targeting inflammatory mediators of the Type-2 asthma pathway. However, a median of only 21.1% (range 13.4–26.8%) of patients with severe eosinophilic asthma in the WSAC (as defined by a blood eosinophil count of ≥300 cells·µL−1) would have been eligible for the phase III trials of these therapies.
There is significant heterogeneity in the eligibility criteria used between trials despite the aim of reconfirming a diagnosis of asthma in a similar target population. For example, whilst 30 of the 37 RCTs required demonstrable airflow obstruction, 12 unique criteria were used to define this. Whilst some studies adopted pragmatic composite eligibility criteria to broaden inclusion, many still required specific evidence of bronchodilator reversibility and/or persistent airflow limitation, which dramatically reduced patient eligibility. In the WSAC, of those patients with a sputum eosinophil count of ≥3%, an asthma control questionnaire (ACQ) score greater than 1.5 and at least one severe exacerbation in the past year, 23% had no evidence of persistent airflow limitation, 56% did not have 12% bronchodilator reversibility and 61% were excluded when both features were required.
In a similar fashion to other cohorts, 26% of the severe asthmatics in the WSAC were current smokers (5.8%) or ex-smokers with a smoking history of ≥10 pack-years (20.5%). Whilst it is reported that asthma in smokers is generally associated with non-eosinophilic inflammation [20], a significant proportion in the WSAC did have demonstrable airway eosinophilia. It is recognised that asthmatics who smoke have impaired responsiveness to corticosteroids and suffer more frequent exacerbations and an accelerated rate of lung function decline [21, 22]. Arbitrary exclusion of these patients from trials has led to a paucity of evidence upon which to base treatment decisions for them. It is also important to reflect that other factors impacting upon airway biology, including obesity, persisting allergen exposure and bacterial dysbiosis, are not commonly included within trial eligibility criteria, raising a question of equity.
RCTs of Type-2 targeted therapies excluded significantly more participants on the basis of diagnostic criteria than those evaluating non-Type-2 and non-biological therapies. This suggests hyper-selection within this population, further limiting generalisability of the results. Licensing body and healthcare funder recommendations extrapolate from RCTs which we have demonstrated to be poorly representative of real-life severe asthma populations. The disparity between NICE treatment recommendations and the trial populations is highlighted in figure 2. The residual uncertainty this has created as to the benefit of treatment for many patients has led to a reliance on treatment trials. Avoiding eligibility criteria that are not relevant to the biological trait targeted would allow for a more inclusive trial population that is better able to be generalised to the subsequent treatment population.
Regulatory authorities have a major influence over the design of RCTs required for product licencing. The European Medicines Agency guidelines for asthma trials specify “the aim should be to study a homogeneous population of patients with asthma” and recommend evidence of reversible airflow obstruction for a secure diagnosis of asthma, substantially limiting trial eligibility [23]. Phase IV pragmatic and non-randomised trials have proved crucial in demonstrating treatment efficacy in those patients excluded from licencing trials [24–26]; however, this creates a significant delay in the generation of evidence for many severe asthma patients and carries the risk of inflating the impact, as real-world evidence is not placebo-controlled. Pragmatic phase III RCTs which better reflect real-world populations and clinical practice may improve external validity and equity of access [27]; however, this will require engagement between clinicians, licencing authorities, funding bodies and the pharmaceutical industry.
Our findings show that RCTs in severe asthma lack external validity, with the majority of patients excluded by criteria designed to confirm arbitrary diagnostic labels rather than by biomarker criteria that predict the characteristic or trait addressed by the treatment [11, 12]. Failure to adopt an exclusively phenotypic approach to trial inclusion will perpetuate the limited generalisability of evidence on effectiveness and health economics used by regulatory bodies. This risks the missing of opportunities for the application of novel therapies and also the propagation of the vast unmet need in severe asthma.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERJ-01444-2018_Supplement
Acknowledgements
We would like to thank all of the participants who took part in the Wessex Severe Asthma Cohort (WSAC).
Footnotes
Author contributions: T. Brown, T. Jones, A. Chauhan and P. Howarth contributed to the literatures search, study design, data collection, data analysis, data interpretation and writing of this manuscript. K. Gove, C. Barbar and S. Elliott contributed to the data collection, data analysis and the writing of this manuscript. All other members of the Wessex Severe Asthma Cohort (WSAC) team contributed to the data collection.
Members of the Wessex Severe Asthma Cohort (WSAC) team: L. Aitkin, S. Babu, P. Dennison, R. Djukanovic, C. Grainge, L. Hewitt, N. Jayasekera, R. Kurukulaaratchy, S. Kerley, L. Lau, D. Laws, J. Owen, E. Ray, D. Reynish, H. Rupani and O. Scullion-Win.
Conflict of interest: T. Brown reports personal fees from Teva, Chiesi Farmaceutici, Novartis and Napp, and non-financial support from AstraZeneca, outside the submitted work. T. Jones reports non-financial support from Teva, personal fees and non-financial support from Chiesi Farmaceutici, outside the submitted work. K. Gove has nothing to disclose. C. Barber has nothing to disclose. S. Elliott has nothing to disclose.
Conflict of interest: A. Chauhan reports personal fees and non-financial support from Teva, non-financial support from Boehringer Ingelheim, and personal fees from AstraZeneca, outside the submitted work.
Conflict of interest: P. Howarth reports grants from the Medical Research Council, during the conduct of the study; and is a part-time employee of GlaxoSmithKline.
This article has supplementary material available from erj.ersjournals.com
Support statement: This work was funded by Research Councils UK: Medical Research Council (MRC; G0800649). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received July 30, 2018.
- Accepted October 5, 2018.
- Copyright ©ERS 2018
References
