Development and Validation of Personalized Prediction to Estimate Future Risk of Severe Exacerbations and Uncontrolled Asthma in Patients with Asthma, Using Clinical Parameters and Early Treatment Response

Investigating Machine Learning Techniques for Predicting Risk of Asthma Exacerbations: A Systematic Review

Asthma, a common chronic respiratory disease among children and adults, affects more than 200 million people worldwide and causes about 450,000 deaths each year. Machine learning is increasingly applied in healthcare to assist health practitioners in decision-making. In asthma management, machine learning excels in performing well-defined tasks, such as diagnosis, prediction, medication, and management. However, there remain uncertainties about how machine learning can be applied to predict asthma exacerbation. This study aimed to systematically review recent applications of machine learning techniques in predicting the risk of asthma attacks to assist asthma control and management. A total of 860 studies were initially identified from five databases. After the screening and full-text review, 20 studies were selected for inclusion in this review. The review considered recent studies published from January 2010 to February 2023. The 20 studies used machine learning techniques to support future asthma risk prediction by using various data sources such as clinical, medical, biological, and socio-demographic data sources, as well as environmental and meteorological data. While some studies considered prediction as a category, other studies predicted the probability of exacerbation. Only a group of studies applied prediction windows. The paper proposes a conceptual model to summarise how machine learning and available data sources can be leveraged to produce effective models for the early detection of asthma attacks. The review also generated a list of data sources that other researchers may use in similar work. Furthermore, we present opportunities for further research and the limitations of the preceding studies.

Asthma exacerbations during the pandemic: Time to rethink clinical markers

Background

Reductions in asthma exacerbations during the coronavirus disease 2019 (COVID-19) pandemic may have an impact on clinical trial enrollment and outcomes.

Objective

Our aim was to review clinical studies and reports evaluating asthma exacerbations before and during the COVID-19 pandemic.

Methods

We reviewed clinical studies conducted with biologics over the past decade that evaluated asthma exacerbations as the primary end point. We also reviewed recent clinical reports evaluating asthma exacerbations during the COVID-19 pandemic.

Results

We showed that studies requiring at least 2 exacerbations in the prior year resulted in a higher number of exacerbations on study in the placebo arm, and conversely, those studies in which exacerbations were not required for entering the study failed to meet the primary end point. This result confirmed that history of prior exacerbations is a good maker to predict future exacerbations. In addition, a review of the literature confirmed a reduction of asthma exacerbations during the COVID-19 pandemic. The data presented are descriptive; no formal statistics were used.

Conclusion

Because of the COVID-19 pandemic, historical exacerbations may no longer be the best predictor for exacerbations in a clinical trial or clinical practice. Other clinical markers associated with exacerbations, such as blood eosinophil count and fractional exhaled nitric oxide level, should be considered for enrollment in clinical studies assessing asthma exacerbations.

Characteristics, phenotypes, mechanisms and management of severe asthma

ABSTRACT

Severe asthma is "asthma which requires treatment with high dose inhaled corticosteroids (ICS) plus a second controller (and/or systemic corticosteroids) to prevent it from becoming 'uncontrolled' or which remains 'uncontrolled' despite this therapy." The state of control was defined by symptoms, exacerbations and the degree of airflow obstruction. Therefore, for the diagnosis of severe asthma, it is important to have evidence for a diagnosis of asthma with an assessment of its severity, followed by a review of comorbidities, risk factors, triggers and an assessment of whether treatment is commensurate with severity, whether the prescribed treatments have been adhered to and whether inhaled therapy has been properly administered. Phenotyping of severe asthma has been introduced with the definition of a severe eosinophilic asthma phenotype characterized by recurrent exacerbations despite being on high dose ICS and sometimes oral corticosteroids, with a high blood eosinophil count and a raised level of nitric oxide in exhaled breath. This phenotype has been associated with a Type-2 (T2) inflammatory profile with expression of interleukin (IL)-4, IL-5, and IL-13. Molecular phenotyping has also revealed non-T2 inflammatory phenotypes such as Type-1 or Type-17 driven phenotypes. Antibody treatments targeted at the T2 targets such as anti-IL5, anti-IL5Rα, and anti-IL4Rα antibodies are now available for treating severe eosinophilic asthma, in addition to anti-immunoglobulin E antibody for severe allergic asthma. No targeted treatments are currently available for non-T2 inflammatory phenotypes. Long-term azithromycin and bronchial thermoplasty may be considered. The future lies with molecular phenotyping of the airway inflammatory process to refine asthma endotypes for precision medicine.

Patient characteristics, biomarkers, and exacerbation risk in severe, uncontrolled asthma

BACKGROUND

Greater precision in asthma exacerbation risk prediction may improve outcomes. We sought to identify clinical characteristics and biomarkers associated with elevated exacerbation risk in patients with severe, uncontrolled asthma.

METHODS

Data were pooled from seven similarly designed Phase II and III randomized controlled clinical trials of biologic therapies for the treatment of severe, uncontrolled asthma that enrolled comparable patient populations. Annualized asthma exacerbation rates (AAERs) for patients randomized to placebo were assessed by baseline clinical characteristics and by biomarker concentrations at baseline and over the study duration.

RESULTS

The AAER for the 2016 patients in the combined placebo group was 0.91 (95% CI 0.84‒0.98). Baseline characteristics associated with greater AAER were frequent or severe exacerbations within the prior 12 months, nasal polyposis, maintenance oral corticosteroid use, Asian race, and Asian or Western European region. AAER increased with baseline blood eosinophil counts and fractional exhaled nitric oxide (FeNO) concentration, with the greatest AAER occurring for patients with eosinophils ≥300 cells·μL^-1 and FeNO ≥50 ppb. No relationship was observed between baseline serum immunoglobulin E concentration and AAER. Combining type 2 inflammation criteria for eosinophils and FeNO had greater prognostic value than either biomarker alone. Persistent eosinophil and FeNO elevations throughout the study period were associated with greater AAER.

CONCLUSIONS

Exacerbation history, maintenance corticosteroid use, nasal polyposis, Asian race, geographic region, and elevations in blood eosinophil counts and FeNO concentrations (particularly when combined and/or persistently achieving type 2 inflammation criteria) were associated with increased exacerbation risk in patients with severe, uncontrolled asthma.

Effect of nocturnal Temperature-controlled Laminar Airflow on the reduction of severe exacerbations in patients with severe allergic asthma: a meta-analysis

Background: Allergen avoidance is important in allergic asthma management. Nocturnal treatment with Temperature-controlled Laminar Airflow (TLA) has been shown to provide a significant reduction in the exposure to allergens in the breathing zone, leading to a long-term reduction in airway inflammation and improvement in Quality of life (QoL). Allergic asthma patients symptomatic on Global Initiative for Asthma (GINA) step 4/5 were found to benefit the most as measured by Asthma Quality of Life Questionnaire (AQLQ). However, the effect of TLA on severe asthma exacerbations is uncertain and therefore a meta-analysis was performed.

Methods: Patients with severe allergic asthma (GINA 4/5) were extracted from two 1-year randomised, double-blind, placebo-controlled trials conducted with TLA. A meta-analysis of the effect on severe exacerbations was performed by negative binomial regression in a sequential manner, defined by baseline markers of asthma control (symptoms and QoL scores).

Results: The pooled dataset included 364patients. Patients with more symptoms at baseline (ACT<18 or ACQ7>3; N=179), had a significant mean 41% reduction in severe exacerbations (RR=0.59 (0.38-0.90); p=0.015) in favour of TLA. Higher ACQ7 cut-points of 3.5-4.5 resulted in significant reductions of 48-59%.More uncontrolled patients based on AQLQ total and symptom domains ≤3.0 at baseline also showed a significant reduction in severe exacerbations for TLA vs. placebo ((47% (p=0.037) and 53% (p=0.011), respectively). The meta-analysis also confirmed a significant difference in AQLQ-responders ((Minimal Clinically Important Difference)≥0.5; 74% vs. 43%, p=0.04).

Conclusion: This meta-analysis of individual patient data shows a beneficial effect on severe exacerbations and quality of life for TLA over placebo in more symptomatic patients with severe allergic asthma. These outcomes support the national management recommendations for patients with symptomatic severe allergic asthma. The actual effect of TLA on severe exacerbations should be confirmed in a prospective study with larger numbers of patients.

Data-Influence Analytics in Predictive Models Applied to Asthma Disease

Asthma is one of the most common chronic diseases around the world and represents a serious problem in human health. Predictive models have become important in medical sciences because they provide valuable information for data-driven decision-making. In this work, a methodology of data-influence analytics based on mixed-effects logistic regression models is proposed for detecting potentially influential observations which can affect the quality of these models. Global and local influence diagnostic techniques are used simultaneously in this detection, which are often used separately. In addition, predictive performance measures are considered for this analytics. A study with children and adolescent asthma real data, collected from a public hospital of São Paulo, Brazil, is conducted to illustrate the proposed methodology. The results show that the influence diagnostic methodology is helpful for obtaining an accurate predictive model that provides scientific evidence when data-driven medical decision-making.

Data-driven questionnaire-based cluster analysis of asthma in Swedish adults

The aim of this study was to identify asthma phenotypes through cluster analysis. Cluster analysis was performed using self-reported characteristics from a cohort of 1291 Swedish asthma patients. Disease burden was measured using the Asthma Control Test (ACT), the mini Asthma Quality of Life Questionnaire (mini-AQLQ), exacerbation frequency and asthma severity. Validation was performed in 748 individuals from the same geographical region. Three clusters; early onset predominantly female, adult onset predominantly female and adult onset predominantly male, were identified. Early onset predominantly female asthma had a higher burden of disease, the highest exacerbation frequency and use of inhaled corticosteroids. Adult onset predominantly male asthma had the highest mean score of ACT and mini-AQLQ, the lowest exacerbation frequency and higher proportion of subjects with mild asthma. These clusters, based on information from clinical questionnaire data, might be useful in primary care settings where the access to spirometry and biomarkers is limited.

The Journal of Allergy and Clinical Immunology: In Practice 2019 Highlights

This article provides highlights of the clinically impactful original studies and reviews published in The Journal of Allergy and Clinical Immunology: In Practice in 2019 on the subjects of anaphylaxis, asthma, dermatitis, drug allergy, food allergy, immunodeficiency, immunotherapy, rhinitis/sinusitis, and urticaria/angioedema/mast cell disorders. Within each topic, practical aspects of diagnosis and management are emphasized. Treatments discussed include lifestyle modifications, allergen avoidance therapy, positive and negative effects of pharmacologic therapy, and various forms of immunologic and desensitization management. We designed this review to help readers consolidate and use this extensive and practical knowledge for the benefit of their patients.

Development and validation of an asthma exacerbation prediction model using electronic health record (EHR) data

Objective: Asthma exacerbations are associated with significant morbidity, mortality, and cost. Accurately identifying asthma patients at risk for exacerbation is essential. We sought to develop a risk prediction tool based on routinely collected data from electronic health records (EHRs).Methods: From a repository of EHRs data, we extracted structured data for gender, race, ethnicity, smoking status, use of asthma medications, environmental allergy testing BMI status, and Asthma Control Test scores (ACT). A subgroup of this population of patients with asthma that had available prescription fill data was identified, which formed the primary population for analysis. Asthma exacerbation was defined as asthma-related hospitalization, urgent/emergent visit or oral steroid use over a 12-month period. Univariable and multivariable statistical analysis was completed to identify factors associated with exacerbation. We developed and tested a risk prediction model based on the multivariable analysis.Results: We identified 37,675 patients with asthma. Of those, 1,787 patients with asthma and fill data were identified, and 979 (54.8%) of them experienced an exacerbation. In the multivariable analysis, smoking (OR = 1.69, CI: 1.08-2.64), allergy testing (OR = 2.40, CI: 1.54-3.73), obesity (OR = 1.66, CI: 1.29-2.12), and ACT score reflecting uncontrolled asthma (OR = 1.66, CI: 1.10-2.29) were associated with increased risk of exacerbation. The area-under-the-curve (AUC) of our model in a combined derivation and validation cohort was 0.67.Conclusion: Despite use of rigorous methodology, we were unable to produce a predictive model with an acceptable degree of accuracy and AUC to be clinically useful.

Future Risks in Patients With Severe Asthma

A major burden of severe asthma is the future risk of adverse health outcomes. Patients with severe asthma are prone to serious exacerbation and deterioration of lung function and may experience side effects of medications such as oral corticosteroids (OCSs). However, such future risk is not easily measurable in daily clinical practice. In particular, currently available tools to measure asthma control and asthma-related quality of life incompletely predict the future risk of medication-related morbidity. This is a significant issue in asthma management. This review summarizes the current evidence of future risk in patients with severe asthma. As future risk is poorly perceived by controlled asthmatics, our review focuses on the risk in patients with ‘controlled’ severe asthma. Of note, it is likely that long-term OCS therapy may not prevent future asthma progression, including lung function decline. In addition, the risk of drug side effects increases even during low-dose OCS therapy. Thus, novel treatments are highly desirable for reducing future risks without any loss of asthma control.