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

Demystification of artificial intelligence for respiratory clinicians managing patients with obstructive lung diseases

INTRODUCTION

Asthma and chronic obstructive pulmonary disease (COPD) are leading causes of morbidity and mortality worldwide. Despite all available diagnostics and treatments, these conditions pose a significant individual, economic and social burden. Artificial intelligence (AI) promises to support clinical decision-making processes by optimizing diagnosis and treatment strategies of these heterogeneous and complex chronic respiratory diseases. Its capabilities extend to predicting exacerbation risk, disease progression and mortality, providing healthcare professionals with valuable insights for more effective care. Nevertheless, the knowledge gap between respiratory clinicians and data scientists remains a major constraint for wide application of AI and may hinder future progress. This narrative review aims to bridge this gap and encourage AI deployment by explaining its methodology and added value in asthma and COPD diagnosis and treatment.

AREAS COVERED

This review offers an overview of the fundamental concepts of AI and machine learning, outlines the key steps in building a model, provides examples of their applicability in asthma and COPD care, and discusses barriers to their implementation.

EXPERT OPINION

Machine learning can advance our understanding of asthma and COPD, enabling personalized therapy and better outcomes. Further research and validation are needed to ensure the development of clinically meaningful and generalizable models.

Cluster Analysis of Finnish Population-Based Adult-Onset Asthma Patients

BACKGROUND

Phenotypes of adult asthma have been identified in previous studies but rarely in population-based settings.

OBJECTIVE

To identify clusters of adult-onset asthma in a Finnish population-based study on subjects born before 1967.

METHODS

We used population-based data from 1350 asthmatics with adult-onset asthma (Adult Asthma in Finland) from Finnish national registers. Twenty-eight covariates were selected based on literature. The number of covariates was reduced by using factor analysis before cluster analysis.

RESULTS

Five clusters (CLU1-CLU5) were identified, 3 clusters with late-onset adult asthma (onset ≥40 years) and 2 clusters with onset at earlier adulthood (<40 years). Subjects in CLU1 (n = 666) had late-onset asthma and were nonobese, symptomatic, and predominantly female with few respiratory infections during childhood. CLU2 (n = 36) consisted of subjects who had earlier-onset asthma, were predominantly female, obese with allergic asthma, and had recurrent respiratory infections. Subjects in CLU3 (n = 75) were nonobese, older, and predominantly men with late-onset asthma, smoking history, comorbidities, severe asthma, least allergic diseases, low education, many siblings, and childhood in rural areas. CLU4 (n = 218) was a late-onset cluster consisting of obese females with comorbidities, asthma symptoms, and low education level. Subjects in CLU5 (n = 260) had earlier onset asthma, were nonobese, and predominantly allergic females.

CONCLUSIONS

Our population-based adult-onset asthma clusters take into account several critical factors such as obesity and smoking, and identified clusters that partially overlap with clusters identified in clinical settings. Results give us a more profound understanding of adult-onset asthma phenotypes and support personalized management.

Cluster Analyses From the Real-World NOVELTY Study: Six Clusters Across the Asthma-COPD Spectrum

BACKGROUND

Asthma and chronic obstructive pulmonary disease (COPD) are complex diseases, the definitions of which overlap.

OBJECTIVE

To investigate clustering of clinical/physiological features and readily available biomarkers in patients with physician-assigned diagnoses of asthma and/or COPD in the NOVEL observational longiTudinal studY (NOVELTY; NCT02760329).

METHODS

Two approaches were taken to variable selection using baseline data: approach A was data-driven, hypothesis-free and used the Pearson dissimilarity matrix; approach B used an unsupervised Random Forest guided by clinical input. Cluster analyses were conducted across 100 random resamples using partitioning around medoids, followed by consensus clustering.

RESULTS

Approach A included 3796 individuals (mean age, 59.5 years; 54% female); approach B included 2934 patients (mean age, 60.7 years; 53% female). Each identified 6 mathematically stable clusters, which had overlapping characteristics. Overall, 67% to 75% of patients with asthma were in 3 clusters, and approximately 90% of patients with COPD were in 3 clusters. Although traditional features such as allergies and current/ex-smoking (respectively) were higher in these clusters, there were differences between clusters and approaches in features such as sex, ethnicity, breathlessness, frequent productive cough, and blood cell counts. The strongest predictors of the approach A cluster membership were age, weight, childhood onset, prebronchodilator FEV1, duration of dust/fume exposure, and number of daily medications.

CONCLUSIONS

Cluster analyses in patients from NOVELTY with asthma and/or COPD yielded identifiable clusters, with several discriminatory features that differed from conventional diagnostic characteristics. The overlap between clusters suggests that they do not reflect discrete underlying mechanisms and points to the need for identification of molecular endotypes and potential treatment targets across asthma and/or COPD.

Cigarette Smoking and Asthma

Globally, around half the adult asthma population are current or former cigarette smokers. Cigarette smoking and asthma interact to induce an "asthma-smoking phenotype(s)," which has important implications for diagnosis, pathogenic mechanisms, and management. The lack of progress in understanding the effects of smoking on adults with asthma is due in part to their exclusion from most investigative studies and large clinical trials. In this review, we summarize the adverse clinical outcomes associated with cigarette smoking in asthma, highlight challenges in diagnosing asthma among cigarette smokers with chronic respiratory symptoms, particularly in older individuals with a long-standing smoking history, and review pathogenic mechanisms involving smoking- and asthma-related airway inflammation, tissue remodeling, corticosteroid insensitivity, and low-grade systemic inflammation. We discuss the key components of management including the importance of smoking cessation strategies, evidence for the effectiveness of the Global Initiative for Asthma recommendations on treatment in cigarette smokers, and the role of treatable traits such as type 2 eosinophilic airway inflammation. Lastly, we provide an algorithm to aid clinicians to manage current and former smokers with asthma. In the future, controlled and pragmatic trials in real-world populations should include cigarette smokers with asthma to provide an evidence base for treatment recommendations.

Asthma Phenotyping in Primary Care: Applying the International Severe Asthma Registry Eosinophil Phenotype Algorithm Across All Asthma Severities

BACKGROUND

We developed an eosinophil phenotype gradient algorithm and applied it to a large severe asthma cohort (International Severe Asthma Registry).

OBJECTIVE

We sought to reapply this algorithm in a UK primary care asthma cohort, quantify the eosinophilic phenotype, and assess the relationship between the likelihood of an eosinophilic phenotype and asthma severity/health care resource use (HCRU).

METHODS

Patients age 13 years and older with active asthma and blood eosinophil count or 1 or greater, who were included from the Optimum Patient Care Research Database and the Clinical Practice Research Datalink, were categorized according to the likelihood of eosinophilic phenotype using the International Severe Asthma Registry gradient eosinophilic algorithm. Patient demographic, clinical and HCRU characteristics were described for each phenotype.

RESULTS

Of 241,006 patients, 50.3%, 22.2%, and 21.9% most likely (grade 3), likely (grade 2), and least likely (grade 1), respectively, had an eosinophilic phenotype, and 5.6% had a noneosinophilic phenotype (grade 0). Compared with patients with noneosinophilic asthma, those most likely to have an eosinophilic phenotype tended to have more comorbidities (percentage with Charlson comorbidity index of ≥2: 28.2% vs 6.9%) and experienced more asthma attacks (percentage with one or more attack: 24.8% vs 15.3%). These patients were also more likely to have asthma that was difficult to treat (31.1% vs 18.3%), to receive more intensive treatment (percentage on Global Initiative for Asthma 2020 step 4 or 5: 44.2% vs 27.5%), and greater HCRU (eg, 10.8 vs 7.9 general practitioner all-cause consultations per year).

CONCLUSIONS

The eosinophilic asthma phenotype predominates in primary care and is associated with greater asthma severity and HCRU. These patients may benefit from earlier and targeted asthma therapy.

Quality of life and asthma control related to hormonal transitions in women's lives

Objectives: The aim was to investigate if menstruation and use of exogenous sex hormones influence self-reported asthma related quality of life (QoL) and asthma control.Methods: The study is based on two asthma cohorts randomly selected in primary and secondary care. A total of 622 female patients 18-65 years were included and classified as premenopausal ≤46 years (n = 338) and peri/postmenopausal 47-65 years (n = 284). Questionnaire data from 2012 and 2014 with demographics, asthma related issues and sex hormone status. Outcome measures were Mini Asthma Quality of Life Questionnaire (Mini-AQLQ) and asthma control including Asthma Control Test (ACT) and exacerbations last six months.Results: Premenopausal women with menstruation related asthma worsening, perimenstrual asthma (PMA) (9%), had a clinically relevant lower Mini-AQLQ mean score 4.9 vs. 5.8 (p < 0.001), lower asthma control with ACT score <20, 72% vs. 28% (p < 0.001) and higher exacerbation frequency 44% vs. 20% (p = 0.004) compared with women without PMA. Women with irregular menstruation had higher exacerbation frequency than women with regular menstruation (p = 0.023). Hormonal contraceptives had no impact on QoL and asthma control. Peri/postmenopausal women with menopausal hormone therapy (MHT) had a clinically relevant lower Mini-AQLQ mean score compared to those without MHT, 4.9 vs 5.4 (p < 0.001), but no differences in asthma control.Conclusion: Women with PMA had lower QoL and more uncontrolled asthma than women without PMA. Peri/postmenopausal women with MHT had lower QoL than women without MHT. Individual clinical management of women with asthma may benefit from information about their sex hormone status.

Risk factors for severe adult-onset asthma: a multi-factor approach

BACKGROUND

The aim was to identify risk factors for severe adult-onset asthma.

METHODS

We used data from a population-based sample (Adult Asthma in Finland) of 1350 patients with adult-onset asthma (age range 31-93 years) from Finnish national registers. Severe asthma was defined as self-reported severe asthma and asthma symptoms causing much harm and regular impairment and ≥ 1 oral corticosteroid course/year or regular oral corticosteroids or waking up in the night due to asthma symptoms/wheezing ≥ a few times/month. Sixteen covariates covering several domains (personal characteristics, education, lifestyle, early-life factors, asthma characteristics and multiple morbidities) were selected based on the literature and were studied in association with severe asthma using logistic regressions.

RESULTS

The study population included 100 (7.4%) individuals with severe asthma. In a univariate analysis, severe asthma was associated with male sex, age, a low education level, no professional training, ever smoking, ≥ 2 siblings, ≥ 1 chronic comorbidity and non-steroidal anti-inflammatory drug (NSAID)-exacerbated respiratory disease (NERD) (p < 0.05), and trends for association (p < 0.2) were observed for severe childhood infection, the presence of chronic rhinosinusitis with nasal polyps, and being the 1st child. The 10 variables (being a 1st child was removed due to multicollinearity) were thus entered in a multivariate regression model, and severe asthma was significantly associated with male sex (OR [95% CI] = 1.96 [1.16-3.30]), ever smoking (1.98 [1.11-3.52]), chronic comorbidities (2.68 [1.35-5.31]), NERD (3.29 [1.75-6.19]), and ≥ 2 siblings (2.51 [1.17-5.41]). There was a dose-response effect of the total sum of these five factors on severe asthma (OR [95% CI] = 2.30 [1.81-2.93] for each one-unit increase in the score).

CONCLUSIONS

Male sex, smoking, NERD, comorbidities, and ≥ 2 siblings were independent risk factors for self-reported severe asthma. The effects of these factors seem to be cumulative; each additional risk factor gradually increases the risk of severe asthma.

A Systematic Review of Asthma Phenotypes Derived by Data-Driven Methods

Classification of asthma phenotypes has a potentially relevant impact on the clinical management of the disease. Methods for statistical classification without a priori assumptions (data-driven approaches) may contribute to developing a better comprehension of trait heterogeneity in disease phenotyping. This study aimed to summarize and characterize asthma phenotypes derived by data-driven methods. We performed a systematic review using three scientific databases, following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria. We included studies reporting adult asthma phenotypes derived by data-driven methods using easily accessible variables in clinical practice. Two independent reviewers assessed studies. The methodological quality of included primary studies was assessed using the ROBINS-I tool. We retrieved 7446 results and included 68 studies of which 65% (n = 44) used data from specialized centers and 53% (n = 36) evaluated the consistency of phenotypes. The most frequent data-driven method was hierarchical cluster analysis (n = 19). Three major asthma-related domains of easily measurable clinical variables used for phenotyping were identified: personal (n = 49), functional (n = 48) and clinical (n = 47). The identified asthma phenotypes varied according to the sample’s characteristics, variables included in the model, and data availability. Overall, the most frequent phenotypes were related to atopy, gender, and severe disease. This review shows a large variability of asthma phenotypes derived from data-driven methods. Further research should include more population-based samples and assess longitudinal consistency of data-driven phenotypes.

Characterization of Asthma Trajectories from Infancy to Young Adulthood

BACKGROUND

Development of asthma is complicated by the multidimensional nature of the disease.

OBJECTIVE

To identify and characterize trajectories of asthma from infancy to young adulthood, and their associations with lung function and inflammatory and respiratory markers in adolescence and young adulthood.

METHODS

A latent class analysis was performed in a population-based cohort (N = 4089). Parental and self-reported symptoms of asthma were used to investigate asthma development. We characterized background factors, allergic comorbidity, and IgE sensitization and investigated associations with asthma markers.

RESULTS

A 4-class solution of asthma trajectories was identified: never/infrequent (n = 3291 [80.4%]), early-onset transient (n = 307 [7.5%]), adolescent-onset (n = 261 [6.4%]), and persistent asthma (n = 230 [5.6%]). Uncontrolled asthma was equally prevalent in the adolescent-onset and persistent asthma trajectory groups, at both age 16 (41.7% vs 42.4%; P = .90) and 24 years (53.7% vs 52.4%; P = .81). The persistent asthma trajectory group had a higher proportion of eosinophil counts greater than or equal to 0.3 (10⁹ cells/L) at age 24 years compared with the adolescent-onset trajectory group (31.0% vs 18.5%; P < .01).

CONCLUSIONS

The adolescent-onset and persistent asthma trajectory groups had equal burdens of asthma control in adolescence and young adulthood. However, the persistent asthma trajectory group showed more signs of type 2 inflammation than the adolescent-onset trajectory group. This unbiased approach highlights the need of identifying patients with adolescent asthma to optimize care, because they suffer the same lack of asthma control as those with persistent asthma.