Forced expiratory volumes in 3 s is a sensitive clinical measure for assessment of bronchodilator reversibility in elderly Chinese with severe lung function impairment

Introduction

Therapeutic effectiveness of inhaled drugs in patients with chronic airway inflammatory diseases such as COPD and asthma is usually assessed by the changes in pulmonary function. FEV₁ is the most commonly used indicator of pulmonary function. Clinical trials have revealed that changes in FEV₁ before and after treatment are not sensitive enough to reflect the effect of bronchodilators in patients with serious airflow limitation, especially the elderly.¹ FVC was reported to be more sensitive than FEV₁ in detecting bronchial reversibility in COPD patients.² Patients who did not show bronchial responsiveness in FEV₁ might show changes in lung volume measurements called “volume response”.³

According to the American Thoracic Society/European Respiratory Society (ATS/ERS) criteria, FVC testing requires a forced expiratory time of 6 s or a plateau in the volume–time curve.⁴ This is often a relatively long time to exhale, especially for patients with severe airway limitation, aged people and patients with diseases such as severe cough and heart failure.^5,6 In a study on elderly population, 25% of the subjects could complete FEV₃ but not FVC.⁷ Another study demonstrated that among 2,928 lung function tests only 47% could exhale for <4 s.⁶

The FEV₃ is the rapidly exhaled volume during the first 3 s of a forced expiratory maneuver, which starts at the level of total lung capacity. It is reproducible, needs shorter expiratory effort and provides an exact outcome.⁸ Studies have reported that FEV₃ could be considered as a possible surrogate for FVC⁹ as well as an alternative to FEV₁.¹⁰ However, whether FEV₃ can detect bronchodilator responsiveness as sensitively as FVC in the elderly population remains controversial.

In the present study, we studied a large unselected Chinese elderly population presenting to our hospital defined by lung function characteristics in relation to bronchodilator response. We attempted to evaluate the changes in FEV₁, FVC and FEV₃ in response to a bronchodilator and reveal if changes in FEV₃ post-bronchodilator test could be an acceptable and sensitive indicator to evaluate bronchodilator response in the elderly patients with severe airflow limitation. In addition, we attempted to find some sensitive indicators that would reflect the improvement in lung function accurately in patients who in spite of their best ability could not breathe continuously for 6 s – a problem that needs clinical resolution.

Materials and methods

Elderly subjects who had undergone pulmonary function examinations in Beijing hospital from July 2003 to April 2015 were analyzed with the following characteristics: age 60 years or older, completion of post-bronchodilator spirometry that included FEV₃ in lung function test.

Measurement of pulmonary function: All subjects underwent pulmonary function tests in Beijing Hospital on a Vmax 622 pulmonary function instrument (Sensormedics, USA) and plethysmograph (Sensormedics, USA) according to pulmonary function test guidelines.⁵ Each lung function test was repeated 3 times and the best value was considered for analysis. All subjects underwent a bronchodilator test which required the administration of 400 μg of salbutamol via metered dose inhaler. Spirometry was repeated 15 mins after short-acting β₂-agonist administration. Data from patients were collected anonymously. The enrollees were divided into 6 groups according to ERS criteria for categorizing the severity of lung function impairment based on FEV₁% predicted value:⁴ Group A, FEV₁≥80% predicted value; Group B, 70% predicted value ≤ FEV₁<80% predicted value; Group C, 60% predicted value ≤ FEV₁<70% predicted value; Group D, 50% predicted value ≤FEV₁<60% predicted value; Group E, 35% predicted value ≤FEV₁<50% predicted value and Group F, FEV₁<35% predicted value.⁴ Changes (Δ) in FEV₁, FEV₃ and FVC after the bronchodilator test were expressed in milliliters, percentage of predicted value or percentage of baseline. To minimize any bias due to age, height and weight, ΔFEV₁, ΔFEV₃ and ΔFVC were also z score normalized. Correlation between lung function impairment and changes in FEV₁, FEV₃ and FVC after bronchodilator responsiveness test was estimated. The study protocol was approved by the Beijing Hospital ethics committee.

Data was analyzed using the SPSS 17.0 software. Normally distributed data was expressed as , and non-normally distributed data was expressed as M (Q1, Q3). Chi-square test was used to compare the relative frequencies of patients among the groups. Mann–Whitney U test was used for comparing data that was not normally distributed. P<0.05 was regarded as statistically significant.

Ethics approval and informed consent

The study was conducted in accordance to the Declaration of Helsinki and approved by the ethics committee of Beijing hospital. Informed consents were waived due to the retrospective nature of the study, and all patient data was anonymized.

Results

In the present study, a total of 7745 tests in subjects aged 60 years or older with a median age of 71 years were reviewed. Out of the total, 61.7% were males (4781) and 38.3% were females (2964). Out of the 7745 patients, 2985 (38.5%) were in Group A, 1129 (14.6%) were in Group B, 1066 (13.8%) were in Group C, 837 (10.8%) were in Group D, 1183 (15.3%) were in Group E and 545 were in Group F (7%). There was no significant difference in age and gender distribution between the groups (P>0.05) (Table 1). In order to facilitate statistical analysis, Groups E and F were classified as severely lung function–impaired groups and Groups A–D were classified as not severely lung function–impaired groups.

Table 1 Baseline characteristics of the study population in subgroups stratified by severity of lung function impairment

According to FVC criteria in evaluating positive bronchodilator response (a change post-bronchodilator test in FVC ≥12% of baseline and ≥200 mL), the cutoff value of FEV₃ in evaluating positive bronchodilator test has been counted. A ROC curve was used to determine the best corresponding cutoff for FEV₃ (Figure 1). The area under the ROC curve was 0.969 (95% CI: 0.965–0.973, P<0.001). Post-bronchodilator test a change in FEV₃≥10% of baseline and ≥165 mL had the best sensitivity (92.5%) and specificity (90.4%), and there was an excellent agreement between the two diagnostic cutoffs (κ=0.737; P<0.001).

Figure 1 Receiver operating characteristic (ROC) curve showing the performance of FEV3 in evaluating bronchodilator test based on FVC evaluation criteria. Abbreviation: AUC, area under the curve.

Post-bronchodilator FEV₁ increased by ≥12% of baseline in 21% (1627) of subjects, but the percentage was reduced to 14.6% (1131) according to the criteria that both FEV₁ changed ≥12% of baseline and ≥200 mL. Similar results were observed with FEV₃ and FVC (Table 2). Out of the 496 subjects with post-bronchodilator increase in FEV₁ by ≥12% of baseline but <200 mL, 389 were in the severely lung function–impaired group (213 cases, 18% in Group E; 176 cases, 32.3% in Group F). There was a significant difference in the frequency of patients between the severely lung function–impaired group and the not severely lung function–impaired group (χ2=962.77, P<0.001) (Table 3, Figure 2).

Table 2 Characteristics of bronchodilator response in the study population

Table 3 Characteristics of post-bronchodilator response in subgroups stratified by level of lung function impairment

Figure 2 Frequency distribution of subjects in each group showing post-bronchodilator change in FEV1 of more than 12% of baseline but ≤200 mL in various subgroups stratified by level of lung function impairment. Note: ∆= change.

Significant bronchodilation is regarded as a change in post-bronchodilator FEV₁ or FVC ≥12% of baseline and ≥200 mL according to the ATS/ERS guidelines.⁴ There were 314 subjects – mainly in the severely lung function–impaired groups (129 cases, 10.9% in Group E; 122 cases, 22.4% in Group F) – whose bronchodilator test was negative when evaluated with FEV_1, but became positive with FVC. There was a significant difference in the frequency of patients between the severely lung function–impaired group and not-severely lung function–impaired group (χ2=626.97, P<0.001) (Table 3, Figure 3). There were 593 subjects, mainly in the severely lung function–impaired groups, whose bronchodilator test was negative when evaluated with FEV₁ but was positive when evaluated with FEV₃ (196 cases, 16.6% in Group E; 136 cases, 25% in Group F). There was a significant difference in the frequency of patients between severely lung function–impaired group and not severely lung function impaired (χ2=420.14, P<0.001). (Table 3, Figure 4).

Figure 3 Frequency distribution of subjects in each group whose post-bronchodilator responses become positive with ∆FVC response in various subgroups stratified by level of lung function impairment. Note: ∆= change.

Figure 4 Frequency distribution of subjects in each group whose post-bronchodilator responses become positive with ∆FEV3 response in various subgroups stratified by level of lung function impairment. Note: ∆= change.

The post-bronchodilator changes in the z score of FEV₁, FVC and FEV₃ were related to the severity of lung function impairment (Table 4, Figure 5). As the severity increased, ∆zFEV₁ after the bronchodilator test increased gradually and peaked in Groups D and E, but decreased in Group F. In contrast, ∆zFVC continued to increase with the increase in the severity of ventilatory impairment, which peaked in Group F. Post-bronchodilator changes in zFVC were less than zFEV₁ in Groups A, B and C (Group A z= –3.16, Group B z= –2.86, Group C z= –2.86, all P<0.05) and similar to ΔzFEV₁ in Group D (z= –0.27, P>0.05). But in Groups E and F, ΔzFVC significantly exceeded ΔzFEV₁ (Group E z= – 4.15, Group F z= –6.96, all P<0.001). Changes in zFEV₃ exhibited similar trends. Post-bronchodilator ∆zFEV₃ was less than or equal to the change in zFEV₁ for subjects in Groups A to D (Group A z= –2.03, P<0.05, Group B z= –1.51, Group C z= –0.98, Group D z= –0.68, P>0.05), but in Groups E and F, ΔzFEV₃ significantly exceeded ΔzFEV1 (Group E z= –3.16, P<0.05, Group F z= –4.41, P<0.001). Similar results were obtained after subgroup analysis of male and female subjects (Tables 4–6).

Table 4 Spirometric Indexes expressed as change post-bronchodilator and z scores in subgroups stratified by level of lung function impairment

Table 5 Spirometric Indexes in male subjects expressed as change in post-bronchodilator and z scores in subgroups stratified by level of lung function impairment

Table 6 Spirometric Indexes in female subjects expressed as change in post-bronchodilator and z-scores in subgroups stratified by level of lung function impairment

Figure 5 Post-bronchodilator change in z score of FEV1 (left) is larger in 70% pred≤FEV1<80% pred group, 60% pred≤FEV1<70% pred group and 50% pred≤FEV1<60% pred group compared with FEV1≥80% pred group and declines as obstruction becomes most severe. Conversely, the post-bronchodilator effect on FVC (middle) increased with the severity of lung function impairment. The patterns for FEV3 and FVC are similar. Thin lines, 5th and 95th percentiles; central thick line, median.Note: ∆= change; % pred, percentage of the predicted value.

Discussion

Elderly subjects – who are also the main population of COPD – are more prone to severe and fixed airways resistance. Increased airway resistance increases intrathoracic gas volume and results in diminished lung elastic recoil pressure.¹¹ After treatment with a bronchodilator, elderly patients – especially those with serious airflow limitation – report improvement in symptoms and exercise tolerance with little change in FEV₁.¹² Although FEV₁ is a widely used efficacy end point, it cannot sensitively reflect the real change in lung function due to the improvement in airway obstruction.¹³

The bronchodilator test is a better guide to disease progression and can identify patients who might benefit from treatment with a bronchodilator.¹⁴  In this study, we evaluated changes in lung function parameters before and after bronchodilator treatment to find appropriate indicators for evaluating lung funtion improvement in the elderly. The findings showed that there were notably more subjects in the severely lung function–impaired group exhibiting significant improvement in FEV₃ or FVC with less change in FEV₁ in response to the bronchodilator test (FEV₁ less than 50% predicted value). To minimize any bias due to age, height and weight, we also expressed values or changes as z scores. The z scores indicate the standard deviation of a measurement that differed from, the average predicted value, which was free of any bias. Hence, this result showed that an increase in airflow obstruction increased ΔFEV₁ followed by a decline, while ΔFVC and ΔFEV₃ increased progressively. In subjects with mild respiratory defect, ΔFEV₃ or ΔFVC was less than or equal to ΔFEV₁, but ΔFEV₃ or ΔFVC exceeded the relative change in FEV₁ in severely lung function–impaired group. Results also indicated that FEV₃ and FVC are sensitive indicators of bronchodilation in severe airway obstruction, whereas FEV₁ is more sensitive in assessing bronchodilation in mild ventilatory dysfunction.

It has been reported that acute FVC response to bronchodilators was significantly more correlated with health-related quality of life than FEV₁ response.¹⁵ Using the criterion of “clinically significant” to describe an increase in FVC, FEV₁ (≥12% of pre-bronchodilator and ≥0.2 L ) or FEV₃ (≥10% of pre-bronchodilator and ≥0.165 L ), we observed that FVC and FEV₃ were better measurements for assessing reversibility than FEV₁, especially in patients with severely impaired lung function. Also, there was a strong correlation between FEV₃ and FVC in assessing the responsiveness to bronchodilator. Repeated FVC can be stressful in aged patients with severe obstruction.⁵ Thus, it seems that when FVC is not reliable, FEV₃ might have special clinical application value in detecting bronchial reversibility.

Our study presents new data on FEV₃ implying its clinical applicability for better interpretation of reversibility monitoring, particularly in severely impaired patients who cannot blow for ≥6 s even after their best attempts. Larger FEV₃ and FVC responses can provide useful information regarding reduction of hyperinflation with beneficial effects on dyspnea, exercise tolerance and function of small airways,¹⁶ which has important clinical application value.

FEV₆ has been proposed as an alternative to FVC. Our previous research¹⁷ found that FEV₆ strongly correlated with FVC which did not vary with the forced expiratory time and it could be used as a valid alternative to FVC in diagnosing airflow obstruction in elderly males. FEV₆ is closer to FVC in terms of expiratory time, whereas FEV₃ is between FEV₁ and FEV₆ in expiratory time. FEV₃ also makes spirometry easier, faster and safer than FVC measurement and does not correlate with forced expiratory time. In addition, FEV₃ has unique application advantages in patients who cannot exhale for 6 s. In such subjects, addition of FEV₃ in the daily practice of pulmonary medicine will help physicians to find clinically important relief for hyperinflation.

One of the strengths of our study is that we discussed unselected data on bronchodilator response across various lung function impairment in an aging population. Also, the subjects involved in this study were not limited to any disease or a diagnosis of airflow obstruction. The result reflected the characteristics of common clinical practices in the real world, thus highlighting their clinical usefulness. However, the present study also has some limitations. Being a retrospective study, it failed to provide information about clinical symptoms and disease severity of the subjects. It also lacked information about the use of medication before spirometry. The lung function tests were part of the routine clinical maneuvers, and some subjects might not have withheld medications before testing which may have led to an underestimation of bronchodilator responsiveness. Another significant caveat is the lack of specific diagnosis-related subgroup analyses, which requires further study in the future.

Currently, FEV₃ can be recorded with many spirometers. Thus, according to the result of our study, we encourage to report and analyze FEV₃ for better assessment of spirometry. Compared with the score of the clinical symptom and life quality, parameters of spirometry such as FEV₃ are more objective. Sensitive assessment of improvement in pulmonary function is helpful for disease evaluation and treatment options. However, correlations between the characteristic changes in FEV₃ and subsequent clinical manifestations remains to be further studied.

In summary, this population-based retrospective study showed that in elderly subjects, especially those with severe lung function impairment, FEV₃ combined with FVC can be a clinically effective and sensitive outcome measure to detect bronchial reversibility. In those elderly subjects who cannot complete ≥6 s of forced expiration and whose FVC is not reliable, FEV₃ combined with FEV₁ can be clinically valuable in detecting bronchial reversibility.