Chronic Obstructive Pulmonary Disease

Risk factors of acute exacerbation and disease progression in young patients with COPD

Abstract

Objective We aimed to elucidate the clinical factors associated with acute exacerbation and disease progression in young patients with chronic obstructive pulmonary disease (COPD).

Methods This retrospective longitudinal observational study included patients with COPD aged between 20 and 50 years with post-bronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC)<0.7. Eligible patients were followed up with ≥2 spirometry examinations at 1 year interval after COPD diagnosis. The primary outcome was moderate-to-severe acute exacerbation in young patients with COPD. Secondary outcomes were early initiation of regular inhalation therapy and accelerated annual post-bronchodilator FEV1 decline.

Results A total of 342 patients were followed up during a median of 64 months. In multivariable analyses, risk factors for moderate-to-severe exacerbation were history of asthma (adjusted HR (aHR)=2.999, 95% CI=[2.074–4.335]), emphysema (aHR=1.951, 95% CI=[1.331–2.960]), blood eosinophil count >300/µL (aHR=1.469, 95% CI=[1.038–2.081]) and low FEV1 (%) (aHR=0.979, 95% CI=[0.970–0.987]). A history of asthma, sputum, blood eosinophil count >300/µL, low FEV1 (%) and low diffusing capacity of the lung for carbon monoxide (DLCO) (%) were identified as clinical factors associated with the early initiation of regular inhalation therapy. The risk factors associated with worsened FEV1 decline were increasing age, female sex, history of pulmonary tuberculosis, sputum, low FEV1 (%) and low DLCO (%).

Conclusions In young COPD patients, specific high-risk features of acute exacerbation and disease progression need to be identified, including a history of previous respiratory diseases, current respiratory symptoms, blood eosinophil counts, and structural or functional pulmonary impairment.

What is already known on this topic

  • Although early detection of clinical deterioration of chronic obstructive pulmonary disease (COPD) is rising in importance, risk factors associated with acute exacerbation and disease progression in young patients with COPD were not clarified.

What this study adds

  • Risk factors for moderate-to-severe exacerbation were history of asthma, emphysema, high blood eosinophil count and low forced expiratory volume in one second (FEV1); similarly, risk factors for early initiation of regular inhalation therapy were a history of asthma, sputum, high blood eosinophil count, low FEV1 (%) and low diffusing capacity of the lung for carbon monoxide (DLCO) (%), while risk factors associated with accelerated FEV1 decline were increasing age, female sex, history of pulmonary tuberculosis, sputum, low FEV1 (%) and low DLCO (%).

How this study might affect research, practice, or policy

  • Young COPD patients with a history of previous respiratory diseases, current respiratory symptoms, high blood eosinophil counts, or structural or functional pulmonary changes need to be distinguished as a high-risk group for worsening the clinical course of COPD.

Introduction

Chronic obstructive pulmonary disease (COPD) is a heterogenous disorder characterised by irreversible and usually progressive airflow limitation due to genetic susceptibility, early-life insults, infection and exposure to noxious particles or gases. Many studies on COPD have been conducted in middle-aged or elderly patients, but increasing evidence suggests that COPD develops from early life and progresses over the years or decades.1 Recently, the concept of ‘COPD in young people’ has been focused on as a distinctive phenotype of COPD, but little is known about its natural course.2 Approximately one-fourth of young people with impaired lung function showed a rapid lung function decline.3 Nearly half of young individuals with COPD often remain undiagnosed, despite already having moderate-severity airflow limitation and ongoing parenchymal destruction, such as emphysema.4 COPD in young people is known to be associated with an over fourfold higher risk of exacerbation and more than a threefold higher risk of mortality compared with healthy controls.5 6 Therefore, it is imperative to thoroughly evaluate the risk factors for exacerbation and the progression of COPD in young individuals.

Several risk factors for the development of COPD in young people have been identified across different stages of life.7 The genetic contribution to COPD development relies on susceptible genes related to impaired lung growth.8 The pathogenesis of COPD may begin much earlier, even before birth, as passive fetal smoke exposure in utero is associated with an increased risk of COPD.9 Active and passive exposure to cigarette smoking in childhood or adolescence is a main risk factor for COPD development in young people.10 Additionally, childhood ailments such as asthma and respiratory infections can affect adult lung function and development of COPD.10 Maternal, childhood and adult exposure to air pollution are considered early determinants of the development of COPD.11 Compared with controls, young patients with COPD are characterised by older age, higher smoking exposure, lower forced expiratory volume in one second (FEV1) or forced vital capacity (FVC), and more emphysema.4 However, it is not well known which risk factors are involved in the further progression of COPD in young people. In clinical trials, starting regular inhaled therapy earlier in less advanced stages of COPD leads to a greater decrease in mortality.12 Interventions in the early stages of COPD can modify the natural course of the disease in patients at a high risk of exacerbation of progression. Therefore, it is necessary to identify risk factors for exacerbation or disease progression in young populations with COPD. In an unadjusted analysis, the BODE (Body Mass Index, airflow obstruction, dyspnoea and exercise capacity) Index was reportedly associated with a rapid FEV1 decline in younger patients with COPD.13

This study aimed to identify the clinical factors associated with acute exacerbation and disease progression including the early initiation of regular inhalation therapy and accelerated lung function decline in young patients with COPD.

Methods

Our study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.14 Our study did not involve patients or the public.

Study design and eligibility criteria

This retrospective study reviewed the electronic medical records of young adults with COPD at two teaching hospitals (Boramae Medical Center and Seoul National University Hospital) from January 2005 to December 2020. The inclusion criteria were (1) age between 20 and 50 years, (2) diagnosis of COPD at baseline spirometric evaluation and (3) follow-up with at least two or more spirometric examinations at 1 year interval after COPD diagnosis. COPD diagnosis relies on chronic respiratory symptoms such as dyspnoea, cough, sputum production or exacerbations, along with risk factors, and spirometric criteria of post-bronchodilator FEV1/FVC <0.7. The definition of young patients with COPD was based on a previous reference.15 Exclusion criteria were (1) comorbid chronic lung disease, including concurrent asthma, interstitial lung disease and bronchiolitis obliterans; (2) diagnosis of lung cancer during follow-up; (3) surgical lung resection during follow-up; and (4) end-of-life care setting with life expectancy at <1 year.

Variables

Baseline demographic data were collected, including age, sex, Body Mass Index (BMI), cigarette smoking history, cigarette pack-years, underlying comorbidities such as history of asthma, pulmonary tuberculosis, non-tuberculous mycobacterial pulmonary disease, diabetes mellitus, myocardial infarction, heart failure, chronic liver disease, chronic kidney disease, malignancy, connective tissue disease and peptic ulcer, and symptoms such as cough, sputum and dyspnoea. Smoking history, medical history and symptoms were obtained from electronic medical records, while all other information was derived from clinical assessments. Laboratory findings, radiological findings on chest CT and results of spirometric examination were reviewed from electronic medical records. Laboratory findings include white blood cell count, neutrophil percentage, lymphocyte percentage, neutrophil/lymphocyte ratio, eosinophil percentage, eosinophil count, protein, albumin, blood urea nitrogen, creatinine and total bilirubin. From chest CT, structural abnormalities such as emphysema, bronchiectasis and bronchial wall thickening were checked with reference to radiology reports. Information on post-bronchodilator FEV1, FVC, FEV1/FVC, forced expiratory flow at 25%–75% of forced vital capacity (FEF25-75) and diffusing capacity of the lung for carbon monoxide (DLCO) were obtained from pulmonary function tests. The reference values for the spirometric measurements were calculated using Morris reference values.16

Outcomes

The primary outcome was moderate-to-severe acute exacerbation in young patients with COPD. Acute exacerbation was defined as moderate if the patient was treated with antibiotics and/or oral corticosteroids, and as severe if the patient visited the emergency room or required hospitalisation because of an exacerbation.2 The secondary outcomes were early initiation of regular inhalation therapy and accelerated annual post-bronchodilator FEV1 decline (mL/year). Regular inhalation therapy included long-acting beta-agonist, long-acting muscarinic antagonist and inhaled corticosteroid. As there has been no consensus on the definition of early initiation of regular inhalation therapy for COPD, we defined early initiation as starting treatment within the median time to initiation of regular inhalation therapy during follow-up.

Spirometric data

The spirometry longitudinal data analysis (SPIROLA) software (Centres for Disease Control and Prevention, MorganTown, WV, USA) was used to determine the distribution of annual FEV1 decline rate (mL/year) in young patients with COPD.

Statistical analysis

Categorical variables were analysed using the chi-square test or Fisher’s exact test, and continuous variables were analysed using Student’s t-test or the Wilcoxon rank-sum test. Risk factors associated with moderate-to-severe exacerbations were evaluated using the Cox proportional hazards model. We used the Schoenfeld test to verify the proportionality of hazards assumption. To evaluate the risk factors for treatment initiation within 3 years after COPD diagnosis, we performed multivariable logistic regression analysis adjusted for covariates. We estimated Kaplan-Meier curve and conducted the log-rank test to compare the time to moderate-to-severe exacerbation or regular inhalation therapy initiation according to statistically significant variables. Factors related to annual FEV1 changes (mL/year) were analysed using a multivariable linear mixed effect model. We performed imputation for missing values using the k-Nearest Neighbour imputation method with ‘missForest’ package. The stepwise variable selection method was applied for the multivariable Cox proportional-hazards model or logistic regression model using the R package ‘My.stepwise’. The stepwise variable selection method initially selected variables with p values <0.1 in univariable analyses and excluded variables with p values ≥0.05 by performing multivariable analysis whenever variables were added one by one. In situations where collinearity might occur between variables, we selected the model with the lowest Akaike’s information criterion. A variance inflation factor of >4.0 was considered as significant multicollinearity. Statistical significance was set at p<0.05. All statistical analyses were performed using R statistical software (V.4.1.2; R Core Team, [2021], Vienna, Austria).

Results

Among the 495 young patients with post-bronchodilator FEV1/FVC <0.7, 100 were not eligible because follow-up spirometric examination was not performed. We excluded 53 patients because 35 had comorbid chronic lung diseases, 9 underwent surgical lung resection, 8 were diagnosed and treated for lung cancer during follow-up, and 1 received end-of-life care. A total of 342 patients were thus included in the present study (figure 1). The median follow-up period was 64 (IQR=34–98) months.

Figure 1
Figure 1

Flow diagram of inclusion.

Patient characteristics

The baseline characteristics of the 342 patients with COPD are summarised in table 1. The mean age of the patients was 43.7 years, and male patients were 70.8%. Nearly half of the patients were ever-smokers (54.1%), with a median of 20 pack-years. A history of asthma and pulmonary tuberculosis was found in 37.7% and 39.8% of patients, respectively. Dyspnoea was present in 73.7% of the patients, and one-third of them had severe dyspnoea on the modified medical research council (mMRC) ≥2. The mean percentage and count of blood eosinophils were 4.0% and 283 cells/µL, respectively, and 34.9% of patients had a blood eosinophil count >300 cells/µL (table 2). Chest CT was performed in 287 patients, among whom emphysema and bronchiectasis were detected in 25.7% and 26.0% of patients, respectively. At the baseline spirometric examination, the predictive values of FEV1 (%) and FVC (%) were 65.6% and 74.8% and FEV1/FVC ratio was 59.7%. Impaired FVC (<80%) and DLCO (<80%) were found in 52.3% and 41.4% of patients, respectively.

Table 1
|
Baseline characteristics of the included patients
Table 2
|
Clinical features of the included patients

Risk factors of moderate-to-severe exacerbation

We found that 41.5% of patients experienced moderate-to-severe exacerbation, 39.2% experienced moderate exacerbation and 16.4% experienced severe exacerbation during follow-up. The median time to the first acute exacerbation was 27 (IQR=1–62) months. The moderate-to-severe exacerbation rate was 0.37 /year, moderate exacerbation rate was 0.24 /year and severe exacerbation rate was 0.13/year (online supplemental e-table 1). In univariate Cox regression analysis, history of asthma, history of pulmonary tuberculosis, sputum, dyspnoea with mMRC ≥2, bronchial wall thickening, baseline blood eosinophil count >300/µL, baseline and follow-up blood eosinophil count >300/µL, FEV1 (%), FVC (%), FEV1/FVC (%) and FEF25-75 (%) were risk factors for moderate-to-severe exacerbation. In the multivariable Cox regression analysis, independent risk factors for moderate-to-severe exacerbation were history of asthma (adjusted HR (aHR)=2.999, 95% CI=[2.074–4.335]), emphysema (aHR=1.951, 95% CI=[1.331–2.960]), blood eosinophil count >300/µL (aHR=1.469, 95% CI=[1.038–2.081]) and low FEV1 (%) (aHR=0.979, 95% CI=[0.970–0.987]) (table 3). In the Kaplan-Meier curves, according to significant clinical variables, the risk of moderate-to-severe exacerbation was higher in the subgroups with history of asthma, history of tuberculosis, sputum, mMRC ≥2, FEV1/FVC<60% and FEV1 <50% (online supplemental e-figure 1).

Table 3
|
Cox regression analysis for moderate-to-severe exacerbation

Factors associated with early treatment

We found that 279 (81.6%) patients required regular inhalation therapy; however, only 142 (41.8%) patients started with the therapy at the point of diagnosis. The characteristics of patients who started regular inhalation therapy at diagnosis, those who started during follow-up and those who never initiated regular inhalation therapy were described in online supplemental e-table 2. History of asthma, sputum, a higher blood eosinophil count and lower FEV1 were associated with requirement of earlier treatment. Notwithstanding more frequent exacerbation events, FEV1 decline rate was lower in patients that started regular inhalation therapy immediately at diagnosis than in patients that delayed regular inhalation therapy. The median time to initiation of regular inhalation therapy was approximately 39 [IQR=19–71] months (approximately 3 years) in patients who did not start regular inhalation therapy at the point of diagnosis but eventually required regular inhalation therapy during follow-up. Therefore, we searched for risk factors for regular inhalation therapy initiation within 3 years of the diagnosis of COPD. In the univariable logistic regression analysis, age, ever-smoker, history of asthma, history of pulmonary tuberculosis, sputum, blood eosinophil count >300/µL, blood urea nitrogen and FEV1/FVC were related to regular inhalation therapy initiation within 3 years. In multivariate logistic regression analysis, independent clinical factors related to regular inhalation therapy initiation within 3 years were history of asthma (adjusted OR (aOR)=3.443, 95% CI [1.674 to 7.078]), sputum (aOR=3.253, 95% CI [1.381 to 7.660]), blood eosinophil count >300/µL (aOR=2.058, 95% CI [1.016 to 4.171]), FEV1 (%) (aOR=0.940, 95% CI [0.919 to 0.961]) and DLCO (aOR=0.980, 95% CI [0.962 to 0.999]) (table 4). In the Kaplan-Meier curves according to significant clinical variables, regular inhalation therapy was initiated earlier in subgroups with sputum, history of asthma, blood eosinophil count >300/µL, FEV1/FVC <60% and DLCO <60% (online supplemental e-figure 2).

Table 4
|
Clinical factors related with initiation of regular inhalation therapy within 3 years of COPD diagnosis

Clinical variables affecting FEV1 decline

Spirometric examination was performed a median of 4 (IQR=4–6) times for each patient. Moreover, the median interval between spirometry measurements was 15 (IQR=12–22) months. The distribution of annual FEV1 change (mL/year) is described in online supplemental e-figure 3. The annual FEV1 decline rate in young patients with COPD was −18 (IQR=−83–39.5) mL/year. A rapid FEV1 decline (annual decline of 60 mL/year) was observed in 114 (33.3%) patients. After adjusting for clinical variables related to annual FEV1 change, the annual FEV1 decline rate was accelerated in patients with increasing age, female sex, history of pulmonary tuberculosis, sputum, low FEV1 (%) and low DLCO (%) (table 5). Meanwhile, the annual FEV1 change was less likely to decline in patients with a history of asthma, non-tuberculous mycobacterial pulmonary disease or cough.

Table 5
|
Clinical variables affecting annual FEV1 change in young patients with COPD

Discussion

Given the growing importance of early detection of clinical deterioration, our study identified the risk factors for acute exacerbation and disease progression in young patients with COPD. Compared with other cohorts for young COPD patients, our patients were predominantly male, had a lower BMI, fewer current smokers, less emphysema and a higher blood eosinophil count.4 5 During the 5 years of observation, 41.5% of the patients experienced moderate-to-severe exacerbation, and 80.7% initiated regular inhalation therapy. The annual FEV1 decline rate was a median of −18 mL/year, and rapid FEV1 decline (<−60 mL/year) was found in 33% of young patients with COPD, which was numerically higher than those (−4 mL/year of annual FEV1 decline rate and 25% of rapid FEV1 decline) of the COPD patients with age ≥40 years.17 Young COPD patients with a history of asthma or pulmonary tuberculosis, current respiratory symptoms such as sputum, high blood eosinophil counts, or structural or functional pulmonary impairment need to be distinguished as a high-risk group for worsening the clinical course of COPD. Our findings emphasise the importance of recognising and addressing high-risk features in young COPD patients, especially a syndemic occurrence of respiratory comorbidities, as it can be beneficial in preventing acute exacerbation and disease progression.18

The severity and activity of the disease in young patients with COPD are similar to those in older patients with COPD.13 19 Although there are no clinical trials specifically targeting young patients with COPD, a recent study was conducted on pre-COPD patients—symptomatic smokers without spirometric airflow limitation.20 However, the study could not find the benefit of indacaterol/glycopyrronium in improving symptoms .Therefore, it is still necessary to identify whether young patients with COPD or pre-COPD patients can potentially benefit from early intervention. Further studies including young patients will be needed to investigate whether early clinical management in specific subgroup populations can prevent disease progression of COPD.

The FEV1 decline rate has been considered a conventional indicator of disease progression.21 However, as FEV1 decline is usually not invariably progressive, multidimensional approaches are needed.22 Exacerbation can be regarded as an important marker of clinically important deterioration in patients with COPD. Acute exacerbations become more frequent with progression of COPD.23 Conversely, frequent exacerbations affect COPD severity.24

Initiation of COPD treatment can be an indirect indicator of disease progression. Early initiation of regular inhaled therapy may have benefits in slowing disease progression, although this association is often influenced by confounding by indication.25 26 Among smokers with normal lung function, 12.6% were prescribed regular inhaled treatment, and factors associated with this empiric treatment were respiratory symptoms, presence of emphysema on imaging studies, history of asthma and prior exacerbations.26 COPD patients with mild symptoms and no history of acute exacerbation had an option of using short-acting bronchodilators instead of long-acting bronchodilators.2 In patients who only use short-acting bronchodilators, maintain mild symptoms and do not experience exacerbations, the initiation of regular inhalation treatment might be delayed. In the clinical context, a shortened duration from diagnosis to initiation of treatment for COPD suggests a more rapid progression to clinical COPD.5

Considering our patients’ high blood eosinophil count and history of asthma, misdiagnosing actual asthma as COPD is a concern. In a meta-analysis, a high baseline blood eosinophil count and a history of asthma were more found in young patients with COPD than in general COPD patients.27 In the GOLD 2023 report, a history of childhood asthma was proposed as an important etiotype of COPD and classified as COPD-A, suggesting distinct subtypes of COPD.2 Because 33 (9.6%) patients showed bronchodilator positivity or between-visit variability in FEV1, it is plausible that some of our patients had asthmatic features. However, 333 (97.4%) maintained fully reversible airflow limitation (post-BDR FEV1/FVC <0.7) in follow-up spirometric examination without symptom variability and lung function variability. Therefore, it is more likely that some COPD patients had asthmatic features rather than a majority of the included patients were pure asthmatics.

Current evidence suggests that Th2 or eosinophilic inflammation is related to the severity or activity of COPD. Blood eosinophil count, a Th2 inflammatory biomarker in COPD, was higher in COPD patients with an advanced disease stage or at an increased risk of exacerbation.28 In our study, a high blood eosinophil count was an independent risk factor for moderate-to-severe exacerbation or earlier initiation of regular inhaled therapy in young patients with COPD. This finding suggests that Th2 inflammation may play a pathological role in the progression of COPD in young patients. In contrast, cigarette smoking, which usually exacerbates neutrophilic airway inflammation, was not a significant risk factor for acute exacerbation or disease progression in our study. This finding is concurrent with that of a recent study that reported no association between cigarette smoking and acute exacerbation in young patients with COPD.29

The FVC and DLCO can be adjunctive biomarkers for COPD progression in young patients. FVC decline is associated with a higher risk of acute exacerbation in patients with COPD.30 In particular, FVC was more declined in early stage of COPD.31 In a prospective cohort study, patients with preserved ratio impaired spirometry usually showed a lower FVC and higher mortality than those with COPD GOLD 1.32 Our study showed that FVC was impaired in more than half of the young patients with COPD. Moreover, young patients with COPD with a lower FVC (%) had more moderate-to-severe exacerbations or accelerated FEV1 decline. In addition, it is well known that impaired DLCO is related to the advanced stages of COPD.33 Impaired DLCO has been reported as an independent risk factor for future exacerbation.34 In our study, DLCO had been impaired in 41.4% of young patients with COPD. We found that DLCO was related to earlier initiation of COPD treatment or an accelerated lung function decline rate. DLCO seems an important physiological biomarker of parenchymal, alveolar, and capillary damage in patients with COPD.

A history of tuberculosis is associated with the development of COPD.35 In the GOLD 2023 report, tuberculosis-associated COPD was proposed as an important etiotype of COPD, which is classified as COPD-I, suggesting distinct subtypes of COPD.2 Post-tuberculosis lung damage is related to accelerated lung function decline and the development of airflow limitation.36 37 In our study, a history of tuberculosis was related to an accelerated FEV1 decline rate. Recent studies have reported the benefits of bronchodilators in patients with bronchiectasis and airflow limitation and have proposed a role for inhaled corticosteroids based on blood eosinophil counts.38 39 Further studies are needed to determine the optimal treatment for patients with tuberculosis-associated COPD.

Our study has several limitations. First, this study was performed retrospectively with a small number of young patients with COPD in two teaching hospitals. Because of the nature of study design and data constraints, analysing important indicators such as the BODE Index was not possible. In our medical centres, 6 min walk test was not routinely performed for COPD patients. The information regarding symptoms was obtained based on records from outpatient clinics, which makes the specific details limited, potentially leading to an underestimation of prevalence. Further prospective studies can provide better evidence on the natural course of COPD in young patients. Second, as the data were obtained from teaching hospitals, there is a risk of selection bias in that symptomatic patients were more likely to be included. Therefore, the findings of our study might not be representative to all young COPD population. Third, the majority of identified risk factors related to COPD progression were not treatable traits. Based on previous studies, sputum production, impaired lung function and high blood eosinophil count may be potential treatable traits favouring inhaled corticosteroid use in young patients with COPD.40–42 As for patients at high risk, particularly those with non-treatable factors, they necessitate intensive surveillance and close follow-up for early identification of exacerbations and disease progression. Future clinical trials are imperative to determine the efficacy of therapeutic interventions in preventing exacerbations and disease progression in young COPD patients. Fourth, it may be arbitrary that the early initiation of regular inhalation therapy was within 3 years of COPD diagnosis. The definition of early regular inhalation therapy needs to be optimised through long-term observation of the natural course in asymptomatic COPD patients.

In conclusion, young COPD patients with a history of asthma or pulmonary tuberculosis, current respiratory symptoms, high blood eosinophil count, or structural or functional pulmonary impairment are indicative of potential high-risk factors for the clinical deterioration of COPD such as acute exacerbation, requirement of early regular inhalation treatment and accelerated lung function decline.