Spirometry plays a crucial role in the diagnosis, severity assessment, and longitudinal follow-up of patients with Chronic Obstructive Pulmonary Disease (COPD) [1]. The two key spirometry parameters are the forced expiratory volume in one second (FEV1) and the Forced vital capacity (FVC), which together reflect the ability to exhale air over time and serve as indicators of airway obstruction [2]. In studies investigating bronchoscopic lung volume reduction therapies, FEV1 is frequently selected as the primary endpoint, or at least as major secondary outcome measure.

Beyond statistical significance, it is crucial to assess whether observed changes in lung function are clinically meaningful. The concept of the Minimal Important Difference (MID) is used for this purpose and represents the smallest difference in an outcome that patients perceive as beneficial [3]. The MID can be applied to classify patients as responders or non-responders. To date, for spirometry only an absolute MID estimate for FEV1 has been established [4]. To our knowledge, no MID estimate exists for FVC. Furthermore, in the field of bronchoscopic interventions, most clinical trials have used a relative MID for FEV1, typically ranging from 10 to 15%, despite this threshold not being supported by the existing literature [5–8]. For example, two ongoing trials use a relative MID for FEV1 of ≥12% as primary outcome (BREATHE-3 study: NCT06891755) or as an important secondary outcome (CONVERT-II study: NCT06035120). A task force suggested a MID for FEV1 of 100–140mL but also concluded that a rigorous determination of the MID for spirometry has not yet been performed and remains warranted [9].

An imprecise MID may lead to overestimation or underestimation of a clinical trial's benefit, particularly when it is used as a primary endpoint. Therefore, the aim of this study was to determine the MID for FEV1 and FVC in patients with severe COPD undergoing bronchoscopic lung volume reduction treatment.

To our knowledge, this is the first study to calculate a minimal important difference (MID) for both FEV1 and FVC, as well as to estimate a relative MID in a population of severe COPD patients. The estimated MID for FEV1 was 139mL or 19% at 6 months and 127mL or 17% at 12 months. For FVC, the MID was 349mL or 16% at 6 months and 379mL or 17% at 12 months. As the MIDs at both timepoints were very similar, we propose using an average MID of 133mL or 18% for FEV1 and 364mL or 16% for FVC.

In COPD research, FEV1 is frequently used as a key primary endpoint, often required by the regulatory agencies such as the US Food and Drug Administration. The MID is valuable not only for evaluating treatment effects but also for determining appropriate sample sizes. The most commonly cited absolute MID for FEV1 in COPD has been 100mL, which is lower than our current calculated estimates [4]. It should be noted, however, that this 100mL threshold was not empirically derived, but rather triangulated from several analyses not primarily designed to determine an MID, making direct comparisons difficult. An ERS/ATS task force has suggested a range of 100–140mL based on professional and patient opinion and our calculated MID indeed falls within this range [9].

A relative MID, expressed as a percentage of baseline values, takes into account differences in disease severity and baseline function, making it potentially more meaningful and easier to compare across populations than an absolute MID and particularly appropriate for patients with low baseline values. To date, a relative MID for FEV1 has not been empirically established, the ERS/ATS task force has suggested a range of 4–12% which is lower than our estimates [9]. However, this was not based on a clinically relevant improvement after a treatment, but rather on differences in dyspnea compared with other patients or on reversibility testing. In the absence of a calculated MID, many studies in bronchoscopic lung volume reduction research, have used higher, self-selected relative MIDs of 10–15% and our results show that the relative MID may be even slightly higher. The choice of MID value can substantially affect the proportion of patients classified as responders. We feel it is therefore important to compare trial outcomes to facilitate consensus and promote the use of a standardized MID across studies.

It is well known that FEV1 does not always correlate strongly with patient-reported outcomes [26]. Nevertheless, in our study, we observed appreciable associations between changes in FEV1 and corresponding changes in 6MWD and SGRQ, consistent with findings from a pooled analysis of 23 clinical trials [27]. Consequently, we also investigated, for the first time, the MID for FVC. To our knowledge, FVC has rarely been used as a primary endpoint. However, FVC has been shown to be a more sensitive variable for assessing reversibility in COPD, to better reflect static hyperinflation, and to correlate more closely with improvements in quality of life compared with FEV1[28–30]. These findings suggest that FVC could serve as a valuable additional endpoint in future studies, including the MID estimates.

For this study, we used the anchor-based method to calculate the minimal important difference (MID). When applying the anchor-based method, it is essential to ensure that the selected anchor is appropriate for the analysis. A suitable anchor should ideally be derived from high-quality studies, originate from a comparable COPD population, and, to some extent, reflect improvements in lung function. For this analysis, we selected quality of life measured by the SGRQ and exercise capacity measured by 6MWD as anchors. An advantage of both anchors is that their respective MIDs were established in similar patient populations. The MID for the SGRQ was determined in patients with severe emphysema undergoing bronchoscopic lung volume reduction, while the MID for the six-minute walk distance (6MWD) was established in patients with severe emphysema undergoing lung volume reduction surgery [21,22]. Moreover, this MID was comparable to the value calculated in patients with COPD undergoing pulmonary rehabilitation [31]. We believe that both quality of life and exercise capacity reflect improvements in lung function. The associations between changes in the anchors and changes in lung function parameters exceeded the predefined threshold of 0.35, in accordance with the literature [25]. However, most correlations were only moderate, and it may therefore be questioned whether the anchors fully capture changes in lung function. A strength of our analysis is that both anchors, although measuring different domains, yielded similar MID estimates. Furthermore, as mentioned previously, it is well established that FEV1 does not always correlate strongly with other clinical outcomes; therefore, a perfect association probably cannot be expected. Unfortunately, only these two anchors were available for our calculations. The inclusion of additional anchors, such as cycle ergometry, dyspnea scores, patient global impression of change or the COPD Assessment Test Questionnaire (CAT) could have further strengthened our analysis.

Strengths of our study include the relatively large sample size and the high level of standardization. All measurements were performed in a single specialized research hospital in the Netherlands, following an identical measurement protocol and sequence, and using the same equipment throughout the study. A general limitation of the MID is that it should not be interpreted as a definite cut-off for determining individual treatment response or clinical usefulness. Nevertheless, we believe it is important to reach a consensus on MID values to enable meaningful comparisons of trial outcomes and to inform appropriate sample sizes.

A limitation of our MID calculation is that our cohort consisted exclusively of patients with very severe COPD undergoing bronchoscopic lung volume reduction using three different treatment modalities: valves, coils, and airway bypass stents. It has been previously demonstrated that different interventions across diverse COPD patient populations may yield different MID estimates [32]. This may limit the generalizability of our findings. To enable the application of our MID estimates to a more heterogeneous COPD population, further research is needed to investigate whether other treatments—such as bronchodilators or pulmonary rehabilitation—or different types of bronchoscopic interventions, as well as other COPD patient populations, yield similar validation of our MID estimates.

In conclusion, this study is the first to estimate a minimal important difference (MID) for both FEV1 and FVC, as well as a relative MID in patients with severe emphysema undergoing lung volume reduction. The estimated MID for FEV1 was 133mL (18%) and for FVC 364mL (16%). While the MID should not be used as a strict cut-off, achieving consensus on these values is essential to facilitate comparability across studies and to guide appropriate sample size estimation. Validation of these estimates in other COPD populations would be valuable.

JEH: Conceptualization; Investigation; Methodology; Validation; Visualization; Writing-original draft; Formal analysis; Project administration; Data curation. SCA: Investigation; Writing-review & editing; Methodology; Formal analysis. KK: Investigation; Writing-review & editing; Data curation. DJS: Conceptualization; Investigation; Writing-review & editing; Methodology; Supervision; Data curation.

During the preparation of this work the authors used ChatGPT in order to improve the readability and language of the manuscript. After using this tool/service, the authors reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

JEH and SCA have nothing to disclose. DJS reports to have received grants from PulmonX Corp, Pulmair, Apreo, FreeFlowMedical, Morair and Nuvaira (all USA) for clinical trial expenses outside this study, and received consulting fees from Nuvaira, MoreAir, Apreo, PulmonX (all USA) that was paid to institution outside this study, and received payment for lectures from PulmonX, Nuvaira (both USA), that was paid to institution, outside this study. KK reports speaker fees from PulmonX outside the submitted work.

The data that support the findings of this study are available from the corresponding author upon reasonable request.