Covert airflow obstruction dominates the overt ones in interstitial lung disease: An appraisal

Material & Methods

This study was undertaken at the department of Pleuro-parenchymal Diseases, Institute of Pulmocare and Research, Kolkata, India after clearance from the Institute Ethics Committee between September ‘2021 to October’ 2023. Criteria for defining ILD were based on the concurrence between the independent observations of one experienced pulmonologist and a radiologist diagnosing ILD on HRCT chest. Other investigations (lung function test with assessment of hypoxemia and best possible aetiological evaluation depending on the real-world feasibility) were incorporated in the diagnostic process. All the study participants underwent spirometry (performed observing the ATS/ERS guideline)

Study participants characteristics:

Inclusion/exclusion criteria

Individuals with ILD with age between 30-70 yr (n=235) were consecutively included in this study after obtaining a written informed consent from each. The inability to perform spirometry, the presence of any other concomitant lung disease or complications of the ILD revealed on necessary investigations (radiological, microbiological, cytological or histological) in real-world practice, and the presence of any significant systemic comorbidity that can influence the performance of spirometry according to the investigator were excluded from the study (Figure).

Close

Figure.

Flowchart of participants selection and research methodology.

Export to PPT

Study subject classification

We used the GOLD criteria (FEV1/FVC<0.7) to identify airflow limitation. When applied, it divided the population into two groups, namely (i) ILD ‘unmixed’ and (ii) ILD with OAD or the ‘overlaps’. The two groups were compared statistically using the common spirometric variables such as FEV1, FVC, FEV1/FVC, FEF_25-75 and the surrogate derived from these [(%-predicted value), FEV1 to FEF_25-75 distance (%-predicted values), reversibility of FEV1 and FEF_25-75 to salbutamol inhalation (described in %-predicted values) and FEF_25-75 reversibility (in ml) as shown in Figure. The derived variables considered were FEF_25-75 reversibility and FEV1-FEF_25-75 distance.

Statistical analysis

Parametric un-paired t-test was used for variables like FEV1 and FVC, while Mann-Whitney test was used for the rest of the non-parametric variables. Subsequently, ROC (receiver operating characteristics) curves were drawn with all those selected parameters that showed highly significant difference (P=0.0001) between the two groups (overlap-ILD and unmixed ILD. The identification of the best discriminative parameters was attempted through calculating cut-off values for optimal specificity and sensitivity for all of them. Those chosen cut-offs were applied independently to both the groups to see their performance to unveil airflow limitation in both of these (Figure).

The sample size was checked ‘post- using G power software version 3.1 incorporating the average prevalence of airflow obstruction made with different parameters from available studies. Firstly, apriori analysis was performed to generate the sample size; followed by performing the same using post-hoc analysis to achieve the power as 0.9. This revealed the sample size of 225 to yield the desired power (Supplementary Table I).

Supplementary Table I

Download

Results

Discussion

Thirty out of 235 (12.76%) of the study participants (all had ILD) showed airflow obstruction according to the GOLD criterion⁸. The prospective surrogates of FEV1/FVC<0.7 were chosen using either FEV1 [%-predicted and reversibility to salbutamol inhalation], or FEF_25-75 [%-predicted and salbutamol-reversibility] and the both as FEV1-FEF_25-75 distance. These surrogates have highly significant difference (P≤0.0001) between the groups. ROC curves drawn with the best discriminants showed that FEF_25-75 (%-predicted), had an optimal sensitivity (79.47%) and good specificity (93.33%) to identify the presence of airflow obstruction in ILD with a cut-off value as 41.5 (Figure, Table II). An available information showed FEF_25-75 (<50%-predicted) correlated well with RV/TLC ratio to suggest airflow obstruction in ILD¹¹.

The FEF_25-75, was apparently the best-evolved surrogate of FEV1/FVC<0.7 (the GOLD narrated marker of airflow obstruction), which successfully identified airflow obstruction in about 93.3 per cent (28 of 30) of overlap ILDs and 34 out of 205 participants in unmixed- ILD group independently (Table III). The other two parameters FEV1-FEF_25-75 distance, and FEF_25-75 reversibility could identify an extra 15 and 36 study participants not picked up by the FEF_25-75 cut-off value. Thus, altogether using the derived surrogate criteria, we diagnosed 34, 15, 36 and 7 adding up to 92 (44.87%) participants with airflow limitations in our initial ‘unmixed’ ILD group making the total frequency of airflow obstruction including those diagnosed by the original GOLD criterion (n=30) as 122 out of 235 (51.91%).

The surrogate parameters were mainly chosen from FEF_25-75 that measured the expiratory airflow during the mid-expiratory phase of FVC. Since the FEF_25-75 is thought to mirror the impairment of airflow in small airways¹², all the surrogate parameters were either FEF_25-75 or its derivatives likely to represent the same. The involvement of FEF_25-75 is reportedly reduced in early stages of diseases of small airways¹³ and is the site for origin of primary airflow obstruction in many individuals of OAD (obstructive airway disease)¹⁴. Low FEF_25-75 in individuals with otherwise normal lung function has emerged as a useful predictor for the development of COPD¹⁵. In COPD, decreased FEF_25-75 (%-predicted) is observed frequently¹⁶; reflecting the impact of inflammation of small airways including remodelling as a cardinal feature of cigarette smoking induced COPD¹⁷. The reduction of FEF_25-75 also suggests severe form of asthma or airway disease¹⁸.

The reason for choosing the FEV1-FEF_25-75 distance as a surrogate has theoretical grounds. Studies have shown that FEF_25-75 (% predicted) values <65 per cent may have clinical relevance especially when FEV1 values are normal¹⁹. A cross-sectional study involving 234 individuals with respiratory symptoms found that while FEV1 and FEV1/FVC did not predict airway hyper-reactivity, FEF_25-75 did²⁰. Reduced FEF_25-75 might precede impairment of FEV1, so indicating early asthma and poor prognosis in early asthma^21,22. Studies in allergic rhinitis have shown that reduced FEF_25-75 could be linked to bronchial hyper-reactivity²³ and positive response to bronchodilator testing²⁴. People regard COPD as a disease of small airways¹¹. Thus, FEF_25-75 can represent primarily the early change in small airways that precedes and exceeds the change in FEV1 and the parameter is thought to reflect a differentially predominant small airway obstruction in a case of OAD¹². Therefore, the FEV1-FEF_25-75 distance is likely to increase in early airway diseases. Theoretically, for any OAD, differential regional involvement in either small (distal) and relatively large (proximal) part of the intrathoracic airways can be FEF_25-75 and FEV1 respectively. In this study, the FEV1-FEF_25-75 distance was studied to understand the early airflow obstruction and a predominantly small airway changes inthe study participants.

We presume that FEF_25-75 represents airflow mostly at bronchi proximal to the respiratory units to manifest reversibility as observed in our practice since for the lack of smooth muscles in the walls; the bronchioles distally should lose the potential for dynamic behaviour of bronchodilator responsiveness. The FEF_25-75 actually represents airflow in a region of smaller airways and not some of the exact divisions of airway ramifications. Given this, change in FEF_25-75 should be contemplated when there is a concomitant change in FEV1 and/ or FVC. The change in FEF_25-75 (%-predicted) correlated well to the bronchodilator responsiveness in asthmatic children²⁴. It helps to appreciate the relevant reversible airflow obstruction²⁵. The reversibility potential of the FEF_25-75 was apparent in a study when the significant bronchodilator reversibility was described by PEF, FEF_25-75, and per cent of sGAW²⁶.

However, literature is lacking on combined asthma and ILD. Both ILD and COPD are diseases of the adulthood and late adulthood; hence, COPD is more likely the OAD in ILD. In this study, the FEF_25-75 related parameters unearthed a far higher number of airflow obstruction than those by the FEV1/FVC criterion. Particularly, participants with, reference standard of airflow obstruction as FEV1/FVC <0.7, had the statistically lesser FEF_25-75 from those without airflow limitation (Table I). Findings of this study suggest that individuals diagnosed based on FEF_25-75 are likely to have a predominant small airway disease to start with as is apparent from both the initial (OAD+ILD) overlap and the unmixed-ILD group having lower FEF_25-75 in common. This observation also supports the concept that that FEF_25-75 (%-predicted) can be an earlier marker for COPD than FEV1¹⁵. The impact of COPD or early COPD in individuals with ILD is hence a research issue worth exploring. We feel that the reduction of FEF_25-75 (%-predicted) and other surrogate markers of FEV1/FVC ratio (<0.7) were determinants of COPD in the study participants.

The airflow changes in ILD are mostly in the peripheral bronchial axis that includes the small airways and we have reported earlier that FEF_25-75 may even be increased compared to FVC in ILD²⁷. In face of the interstitial fibrosis imparting traction over to the cartilage-free small airways, the FEF_25-75 change may be minimal or even reversed. Thus, a reduction in FEF_25-75 is possibly a more plausible marker of airflow obstruction in ILD. Since the chronic hypersensitivity pneumonitis (cHP) remains the frequent-most aetiology of ILD in India²⁸, it is possible that small airway involvement in ILD could have caused the airflow obstruction reflected by change in FEF_25-75 itself or the parameters derived of it.

Bronchodilator responsiveness (in terms of FEV1 reversibility) signifies a dynamic constriction of airway lumen. In our cohort, it is significantly lower in the overlap group compared to the ‘unmixed’ patients of ILD with the former showing more severe airflow obstruction [FEV1/FVC ratio of 41.15±211.9 vs. 57.70±80.77; P=0.04] possibly from depleted reversibility potential because of remodelling effect of OAD. The better reversibility /responsiveness of FEV1 or FEF_25-75 in the ‘unmixed’ ILD patients (with preserved FEV1/FVC ratio) suggests the limitation of the FEV1/FVC as the sole criteria to understand airflow obstruction. The story signifies that the level of affection (peripheral and central), airway remodelling, and the type of affection (predominantly fibrotic or inflammatory/constrictive) concomitantly influence the status of FEF_25-75 in ILD. The issue demands intricate research.

The presence of coexisting OAD appears therapeutically important in ILD. A FEV1-reversibility of 100-ml or the minimum clinically important difference (MCID) in COPD²⁹ may signify a positive and perceptible treatment-effect on wellbeing and the health related quality of life in COPD³⁰. Such a change in FEV1 can improve FVC that acts as prognostic marker and an end-point in therapeutic trials in ILD³¹. Incidentally, the change in FVC was higher in our original ‘overlap’ group (Table I).

This study was not without limitations. This was a single-centre based observation. A concomitant analysis of DLCO and lung volumes in this relatively small number of participants could possibly yield much important information. Similarly, concomitant use of impulse oscillometry could have been a good and insightful adjunct. It would have been worthwhile to look for the a etiological association in the exercise since the aetiology of ILD varies from place to place, the relative frequency of airflow obstruction in ILD can vary in different geographical areas. However, the concept will be applicable everywhere.

Overall, this study opens a new domain for consideration in ILD. The implication of the observation demands further research to guide the decision of treatment of OAD in ILD.