ABSTRACT
Background and Objectives
Pneumothorax (PTX) is the most prevalent and serious adverse event associated with endobronchial valve (EBV) treatment in chronic obstructive pulmonary disease (COPD) with emphysema. This study aimed to compare preoperative characteristics and 30‐day outcomes in patients with COPD with and without PTX.
Method
This retrospective nationwide cohort study included patients treated with EBV between 2017 and 2023 at all EBV centres in Denmark. Data were retrieved from medical records.
Results
A total of 228 patients were included, of whom 46 (20.2%) had PTX. The median time to PTX was 7 h (IQR 2–22) and the median drain treatment duration was 13 days (IQR 8.5–22). The risk of PTX was significantly higher when EBV treatment was performed in the upper lobes than in the lower lobes (adjusted RR 6.32, 95% CI 2.56–15.60). High target lobe volume (adjusted OR pr 100 mL increase 1.12, 95% CI 1.02–1.22) and high residual volume (adjusted OR pr 10%pt. increase 1.11, 95% CI 1.01–1.23) were independent risk factors for PTX. More PTX patients had atelectasis after 2 h compared to patients without PTX. The risk of pneumonia (RR 4.5, 95% CI 1.7–11.8), empyema (p = 0.0047) and ICU admission (RR 9.89, 95% CI 1.98–49.4) were significantly higher in the PTX patients than in the patients without PTX.
Conclusions
PTX was more prevalent when EBV treatment was performed in the upper lobes and with high residual volume and target lobe volume, leading to prolonged hospital admission and risk of hospital‐acquired pneumonia.
Keywords: bronchoscopy and interventional techniques, COPD, emphysema, pneumothorax
Short abstract
This is the first national multicentre cohort study investigating the risk factors of pneumothorax after endobronchial valve treatment of COPD. The independent risk factors for pneumothorax were upper lobe treatment, high residual volume and high target lobe volume. This should lead to individualised patient information before the procedure.
See related editorial
1. Introduction
Chronic Obstructive Pulmonary Disease (COPD) is a debilitating disease, which in severe cases causes disabling dyspnoea despite adequate inhalation therapy and rehabilitation. Some of these patients present with emphysema and are suitable for bronchial lung volume reduction (BLVR) with endobronchial valves (EBV). In optimal cases, the treatment causes atelectasis of the target lobe and thereby reduces residual volume (RV) and improves lung function, symptoms and physical ability [1, 2, 3].
However, BLVR also entails a risk of complications. One of the most common and potentially life‐threatening complications is pneumothorax (PTX) with a frequency of approximately 20% [4, 5]. The mechanism of PTX is believed to be due to rapid volume shift with the formation of atelectasis, causing rupture of the diseased lung tissue in the expanding ipsilateral lobe [6]. PTX often requires intrapleural drain treatment, and some patients require temporal removal of one or more valves to improve the healing of the alveolopleural fistula [6]. Patients with PTX can still experience relevant improvement from the treatment [7, 8], but PTX causes prolonged hospitalisation with immobilisation and risk of further complications including hospital‐acquired infections, which might result in further deconditioning. Therefore, PTX is a significant complication of BLVR, and optimisation of the treatment and possible reduction of the frequency of PTX based on risk prediction is important.
The risk factors and consequences of PTX have been investigated in both a nested case–control study [9] and a post hoc analysis of a randomised controlled trial [10]. It was found that low forced expiratory volume in 1 s (FEV1), high residual volume (RV), low attenuation volume on computed tomography (CT), centrilobular emphysema and the presence of pleural adhesions were associated with an increased risk of PTX. To the best of our knowledge, there are no real‐life cohort studies investigating this topic. In this retrospective nationwide multicentre cohort study, the aims were to (1) investigate the risk factors for PTX after treatment of emphysema with BLVR and (2) the association between PTX and the risk of other complications during the first 30 days after the procedure.
2. Methods
2.1. Setting
COPD patients treated with BLVR in Denmark from 2017 to 2024 were included in this study. In Denmark, BLVR is performed at three hospitals (Aarhus University Hospital (AUH), Odense University Hospital (OUH) and Rigshospitalet (RH)). Patients were included at AUH from 2017 to March 2024, RH from 2017 to 2023 and at OUH from 2022 to March 2024. All patients were discussed at a local multidisciplinary team meeting to ensure treatment indication and selection of the target lobe prior to BLVR. All patients were treated with Zephyr EBV (PulmonX Inc., Palo Alto, Calif., USA).
2.2. Participants
Patients ≥ 18 years old treated with BLVR were identified through the ICD10 procedure codes ‘KZLB01 application of valves’ and ‘KUGC19 flexible bronchoscopy with lung volume reduction’.
2.3. Baseline Characteristics
Information on baseline characteristics like demographics, pulmonary function test, 6 min walk test (6MWT), Medical Research Council dyspnoea scale (MRC), target lobe, number of valves and risk of mortality through BODE‐index [11] was retrieved from medical records. Information on lobe volume and voxel index < −950 HU was retrieved from either the quantitative high‐resolution CT (HRCT) analysis StratX (PulmonX, Drive Redwood City, CA, USA) or a quantitative HRCT analysis based on the HRCT done with the ventilation/perfusion single‐photon emission computed tomography (SPECT) (Hermes Medical Solutions, Stockholm, Sweden). Heterogeneous emphysema was defined as a difference in voxel index < −950 HU ≥ 15% points as described in previous literature [12].
2.4. Fissure Assessment
Fissure completeness assessment differed based on local guidelines; at AUH and OUH, fissure completeness was assessed with quantitative HRCT analysis (StratX; PulmonX, Drive Redwood City, CA, USA), and in case of fissure completeness < 95%, measurements with the Chartis system (Pulmonx Inc., Redwood City, CA, USA) were performed. At RH, all patients were evaluated with both quantitative HRCT analysis (Intellispace Portal CT COPD application, Phillips Healthcare, Andover, USA) and Chartis.
2.5. Pneumothorax Management
Patients were observed for 24–48 h before discharge after BLVR, depending on local practice. A chest x‐ray was performed app. Two hours after the procedure, just before discharge and if clinical deterioration occurred. In case of PTX, the treatment decision, for example, drain treatment and revision bronchoscopy, was made by the treating physician. Time to PTX was defined as the time from the bronchoscope retraction until PTX was diagnosed.
2.6. Outcomes After the Procedure
Hospital length of stay, prevalence of pneumonia, haemoptysis, pleural empyema, intensive care unit (ICU) admission and death were retrieved from the medical records. Days in hospital were defined as the time from procedure until discharge. Pneumonia was defined as a respiratory infection treated with antibiotics, either with or without confirmation by x‐ray or CT scan. Haemoptysis was registered if mentioned in the medical records. Pleural empyema was diagnosed by a CT scan and included the occurrence of purulent fluid in the pleural space during drain treatment. The development of partial or complete atelectasis within the first 2 h as a risk factor for PTX was investigated in a single‐centre post hoc analysis.
2.7. Ethics
Retrospective data extraction from patients included at AUH was approved by the Region of Central Denmark (1‐45‐70‐63‐23). This approval required informed consent from all live patients; deceased patients or patients vacated to another region were excepted. At OUH, all patients were included prospectively after informed consent. At RH, retrospective access to the patient data without informed consent was approved by the Capital Region Denmark (31‐1521‐229).
2.8. Statistical Analysis
Continuous data were evaluated for normal distribution using Q–Q‐plots and histograms and presented as either mean with standard deviation (SD) and 95% confidence interval (CI) or median with interquartile range (IQR). Categorical values are presented as the number of all patients and percentage. A p‐value < 0.05 was considered significant.
To identify risk factors for PTX, univariate logistic regression was used to assess the difference between the patient characteristics in patients with PTX and without PTX (non‐PTX). Forward stepwise selection was used to select significant characteristics for the multivariate logistic regression; the adjusted risk ratios (RR) are presented with 95% CI. The maximum number of variables included in the analysis was chosen to be the number of PTX events divided by 10. Since more than the appropriate number of variables were statistically significant in the univariate analysis, the authors prioritised the variables before the multivariate analysis based on findings from previous literature and clinical relevance.
The frequency of complications and repeated bronchoscopies was compared between the PTX and non‐PTX patients using Student's t‐test for normally distributed continuous variables, and Mann–Whitney in case of non‐normal distribution, Fisher's exact test for binominal parameters in two groups and Chi2 test for binominal parameters with multiple groups.
3. Results
All identified patients were included except for one patient, who refused to provide informed consent. Patient characteristics are presented in Table 1. In total, 228 patients were included in the study (50.9% male, mean age 64 years). A total of 46 patients developed PTX (20.2%, 95% CI 15–26). The median time to PTX was 7 h (IQR 2–22 h) ranging from 10 min to 4 days. PTX was diagnosed after 2 days (50–104 h) in 5/46 (11%) patients and after 3 days (98 and 104 h) in 2/46 (4%) patients. Drain treatment was required in 38/46 (82%) patients, and the median duration of drain treatment was 13 days (IQR 8.5–22 days).
TABLE 1.
Patient baseline characteristics of patients with and without pneumothorax.
| All patients | PTX | Non‐PTX | p | |
|---|---|---|---|---|
| N = 228 | N = 46 (20.2%) | N = 182 (79.8%) | ||
| Age (years) | 64.0 (63.0–65.0) | 62.7 (60.1–65.2) | 64.3 (63.3–65.4) | 0.181 |
| Gender (male) | 116 (50.9%) | 24 (52.2%) | 92 (50.5%) | 0.844 |
| BMI (kg/m2) | 23.5 (23.0–24.1) | 23.8 (22.6–25.0) | 23.5 (22.8–24.1) | 0.632 |
| LTOT | 34 (15.2%) | 9 (19.5%) | 25 (13.9%) | 0.252 |
| Smoking history | 0.134 | |||
| Former | 215 (94.3%) | 44 (95.5%) | 176 (96.7%) | |
| Current | 1 (0.4%) | 1 (2.2%) | 0 (0.0%) | |
| Never | 7 (3.1%) | 1 (2.2%) | 6 (3.4%) | |
| Pack years | 38.7 (35.7–41.7) | 33 (27.3–38.7) | 40 (36.8–43.7) | 0.055 |
| Lung function | ||||
| FEV1 (l) | 0.76 (0.73–0.79) | 0.73 (0.67–0.79) | 0.77 (0.73–0.80) | 0.279 |
| FEV1 (%) | 28.1 (27.2–29.1) | 26.5 (24.6–28.4) | 28.5 (27.5–29.6) | 0.086 |
| FVC (l) | 2.2 (2.1–2.3) | 2.2 (2.0–2.5) | 2.2 (2.1–2.3) | 0.805 |
| FVC (%) | 65.7 (63.5–67.9) | 64.8 (59.6–70.0) | 65.9 (63.5–68.3) | 0.678 |
| Ratio (FEV1/FVC) | 35.2 (34.3–36.1) | 35.6 (31.9–36.0) | 33.9 (34.5–36.6) | 0.163 |
| TLC (%) | 135.7 (133.7–137.6) | 138.8 (133.7–143.9) | 134.9 (132.7–137.0) | 0.117 |
| RV (%) | 249.9 (244.1–255.6) | 268.4 (253.3–283.5) | 245.2 (239.2–251.2) | 0.002 |
| DLCO (%) | 34.6 (33.3–35.9) | 32.0 (29.4–34.6) | 35.3 (22.8–36.8) | 0.047 |
| Exercise capacity | ||||
| 6MWTD (m) | 314.7 (301.9–327.6) | 287.7 (258.0–317.5) | 321.5 (307.2–335.7) | 0.041 |
| 6MWT desaturation | 9.5 (8.7–10.2) | 9.4 (7.9–10.9) | 9.5 (8.6–10.4) | 0.882 |
| Dyspnoea and survival | ||||
| MRC | 4.0 (3.9–4.1) | 3.9 (3.7–4.1) | 4.0 (3.9–4.1) | 0.653 |
| BODE index | 6.1 (5.9–6.3) | 6.4 (6.0–6.8) | 6.1 (5.8–6.2) | 0.225 |
| Number of valves | 3.9 (3.7–4.0) | 3.8 (3.5–4.2) | 3.9 (3.7–4.1) | 0.892 |
| Chartis performed | 116 (52.2%) | 20 (44.4%) | 96 (54.2%) | 0.242 |
| CT values | N = 178 | N = 36 | N = 142 | |
| Heterogeneous emphysema | 70 (39.1%) | 15 (41.6%) | 54 (38.7%) | 0.748 |
| Lobe volume (mL) | ||||
| Target lobe(s) | 1805 (1732–1878) | 2020 (1859–2181) | 1751 (1670–1831) | 0.003 |
| Ipsilateral lobe(s) | 1718 (1645–1791) | 1688 (1524–1852) | 1725 (1643–1808) | 0.682 |
| Volume ratio | 1.13 (1.05–1.20) | 1.32 (1.05–1.59) | 1.08 (1.01–1.14) | 0.089 |
| Voxel Index < −950 HU | ||||
| Target lobe(s) | 48.2 (46.3–50.2) | 49.5 (45.8–53.2) | 47.9 (45.6–50.2) | 0.508 |
| Ipsilateral lobe(s) | 35.1 (33.2–37.1) | 36.0 (31.4–40.6) | 34.9 (32.7–37.1) | 0.675 |
| Voxel index ratio | 1.7 (1.4–2.0) | 2.3 (0.8–3.7) | 1.6 (1.4–1.7) | 0.327 |
Note: Data are reported as numbers (n) (%) and mean (95% confidence interval).
Abbreviations: 6MWT, 6 min walk test; 6MWTD, 6 min walk test distance; BODE, body mass index/obstruction/dyspnoea/exercise capacity; CI, confidence interval; DLCO, diffusion capacity of the lungs for carbon monoxide; FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; HRCT, high‐resolution computed tomography; HU, Hounsfield Units; LTOT, long term oxygen therapy; MRC, Medical Research Council dyspnoea scale; non‐PTX, patients without pneumothorax; PTX, patients with pneumothorax; RV, residual volume; TLC, total lung capacity.
The most frequent target lobe was the left lower lobe (98/228) followed by the left upper lobe (52/228). The PTX frequency stratified by target lobes can be seen in Figure 1a.
FIGURE 1.
Pneumothorax frequency stratified by target lobe. (a) The number of pneumothorax patients by each target lobe combination compared to the total number of patients, followed by %. (b) The risk of pneumothorax in patients receiving treatment of upper lobes ± middle lobe and lower lobes ± middle lobe (95% confidence interval).
The PTX patients had significantly higher RV (mean 268.4% vs. 245.2%, p = 0.002), lower DLCO (32.0% vs. 35.3%, p = 0.047), lower 6MWD (287.7 m vs. 321.5 m, p = 0.041) and higher target lobe volume (TLV) (2020 mL vs. 1751 mL, p = 0.003) than the non‐PTX patients, respectively. Table 1 shows the difference between patients with PTX and non‐PTX.
The PTX risk was significantly higher after treatment of the upper lobes (33/94, 35.1%) than the lower lobes (13/133, 9.7%), unadjusted RR 3.58 (95% CI 1.99–6.43). Figure 1b displays the frequency of PTX stratified by treatment of upper ± middle lobe and lower lobes ± middle lobe.
The multivariate regression analyses with the four selected variables (upper lobe treatment, RV, TLV and DLCO) showed that treatment of upper lobe ± middle lobe (RR 3.37, p ≤ 0.001), increase in RV (RR per 10%pt. increase 1.065, p = 0.001) and increase in TLV (OR per 100 mL increase 1.075, p ≤ 0.001) were independent risk factors for PTX as displayed in Table 2. The risk of PTX in different intervals of RV and TLV stratified by treatment lobe is shown in Figure 2a,b.
TABLE 2.
Adjusted multivariate analysis.
| Adjusted risk ratio | 95% confidence interval | p | |
|---|---|---|---|
| Treatment of upper lobe ± middle lobe | 3.37 | 1.92–5.94 | ≤ 0.001 |
| Target Lobe volume (100 mL increase) | 1.075 | 1.03–1.12 | ≤ 0.001 |
| RV (10%pt. increase) | 1.065 | 1.03–1.10 | 0.001 |
| DLCO (1%pt. increase) | 0.98 | 0.94–1.01 | 0.282 |
Note: Data are presented as risk ratio with 95% confidence intervals and p‐values, each adjusted for the remaining three variables.
Abbreviations: DLCO, diffusion capacity of the lungs for carbon monoxide; RV, residual volume.
FIGURE 2.
Proportion of pneumothorax stratified by treatment of upper lobes ± middle lobe and lower lobes ± middle lobe in different intervals of (a) residual volume (% of expected) and (b) Target lobe volume (mL).
In three patients, a two‐step procedure was performed. All these patients received treatment of the upper lobes, two in the left upper lobe and one in the right upper lobe. None of these patients had a PTX.
In a post hoc analysis of 116 patients with 26 PTXs (22.4%), partial or total atelectasis was found to develop faster in patients experiencing PTX. A total of 92.3% of the PTX patients experienced partial or total atelectasis 2 h after the procedure compared to 47.2% of the patients without PTX (RR 8.91, 95% CI 2.21, 35.9, p < 0.001).
There was no correlation between time to PTX and target lobe (p = 0.1620), TLV (p = 0.7984) or RV (p = 0.2231).
There was no difference in the PTX risk with or without Chartis (17.2% and 23.5%, respectively, p = 0.2403, risk difference −6.3%, 95% CI −17.0 to 4.2) and no difference in the PTX risk between the three centres (p = 0.382).
No deaths occurred within 30 days or during the hospital stay, whichever was longest.
The median hospital stay was significantly longer for patients with PTX than for non‐PTX patients (13 days (IQR 7–25) vs. 2 days (IQR 1–2), p ≤ 0.001). Use of antibiotics for respiratory infections (RR 2.4, 95% CI 1.6–3.5, p ≤ 0.001), pneumonia confirmed by imaging (RR 4.5, 95% CI 1.7–11.8, p ≤ 0.001), pleural empyema (p = 0.0047) and ICU admission (RR 9.89, 95% CI 1.98–49.4, p = 0.006) were significantly more common in the patients with PTX than the non‐PTX patients. There was no difference in the frequency of haemoptysis (RR 1.2, 95% CI 0.35–4.3). The frequency of complications during the first 30 days after the procedure is shown in Table 3.
TABLE 3.
Length of hospital stay and postoperative complications the first 30 days after bronchial lung volume reduction.
| PTX | Non‐PTX | Risk ratio (95% CI) | p | |
|---|---|---|---|---|
| N = 46 | N = 182 | |||
| Days in hospital | 13 (7–25) [1–78] | 2 (1–2) [1–13] | — | ≤ 0.001 |
| Use of antibiotics for respiratory infection | 24 (52) | 40 (22) | 2.4 (1.6–3.5) | ≤ 0.001 |
| Pneumonia | 8 (17.4) | 7 (3.8) | 4.5 (1.7–11.8) | ≤ 0.001 |
| Haemoptysis | 3 (6.8) | 10 (5.5) | 1.2 (0.35–4.3) | 0.7416 |
| Empyema | 2 (4.3) | 0 (0.0) | — | 0.0047 |
| ICU admission | 5 (10.9) | 2 (1.1) | 9.89 (1.98–49.4) | 0.006 |
| Death | 0 (0.0) | 0 (0.0) | — | — |
Note: Data are presented as median (IQR) [range] for continuous variables and number (%) for categorical variables. CI, confidence interval; ICU, intensive care unit; non‐PTX, patients without pneumothorax; PTX, patients with pneumothorax.
Ten of the patients with PTX (22.2%) required revision bronchoscopy during the first 30 days compared to 3 (6.4%) of the non‐PTX patients (RR 3.45, 95% CI 1.57–7.62, p = 0.0015). Six patients with PTX required partial valve removal, which was the most common reason for revision bronchoscopy in the patients with PTX. Five of the six patients experienced reinflation of the lung after the partial valve removal, and valves were reinserted after the 30 days, while the last patient required lung volume reduction surgery as a salvage procedure. After valve reinsertion, four of the five patients had complete atelectasis at the first outpatient visit, while the last patient never achieved atelectasis or clinical improvement after reinsertion and therefore had all the valves removed. The reasons for primary total valve removal in the three PTX patients were severe infection, severe subcutaneous emphysema and severe PTX unresponsive to drain treatment. The causes and frequency of revision bronchoscopy within 30 days can be seen in Figure 3.
FIGURE 3.
Risk of revision bronchoscopy in patients with and without pneumothorax. Numbers in the first boxes are in the number of total and percentage.
4. Discussion
In this real‐life study, we have found an increased risk of PTX after upper lobe treatment, especially with increasing RV and TLV. PTX led to a longer hospital stay, an increased risk of ICU admission, an increased risk of infection, including pneumonia and empyema, and revision bronchoscopy.
Upper lobe treatment appears to be a risk factor for PTX in this study. This was also found in a new cohort study investigating the risk factors for severe pneumothorax [13] and in a post hoc analysis of the STELVIO‐RCT [1, 10]. Here, a non‐significant risk of PTX was found to be higher after treatment of the upper lobes than the lower lobes (25% vs. 18%). Their study included fewer patients than ours, which may explain the non‐significant result. However, this difference was not found in three other studies [7, 8, 14]. Gompelmann et al. (2014) [8] included many patients with incomplete fissures and is therefore not comparable to our study. The next study by Gompelmann et al. (2017) [7] does not mention fissure assessment but included patients from before the importance of fissure integrity was known, as discovered in the VENT trial in 2010 [15]. The study by Kemp et al. [14] only included Chartis‐assessed patients. However, assessment of PTX risk by treatment lobe includes all patients with PTX within 6 months. The median time of PTX onset was 1 day. Still, it is not known how many patients had an onset of PTX of < 1 week, or if this group of patients had another distribution of target lobes, and larger studies with Chartis‐assessed patients are needed.
A theoretical reason for the increased risk of PTX after upper lobe treatment could be that the volume shift must be transferred through the lower lobe, while the volume shift after treatment of the lower lobes can be readily transferred to both the upper lobe and the diaphragm. Our results do not imply that upper lobe treatment should be avoided since many patients only have one good target. However, our results identify patients at an increased risk and allow for closer observation and individualised patient information before the procedure. We did not show any association between the time of PTX onset and the treatment lobe in our study.
We found high RV and high TLV to be independent risk factors for PTX, with a linear increase of PTX risk with an increase in RV and TLV. This could be due to a high RV and TLV representing a higher volume shift with atelectasis formation [6]. A high RV was also found to be a risk factor for PTX in a nested case–control study [9] (2016). However, this association was not found in another study investigating a relatively small number of patients, comparing 14 PTX patients with 50 non‐PTX patients [10]. None of these two studies found TLV to be a risk factor for PTX. In contrast to the study by Gompelmann et al. [9], we did not find a high voxel index < −950 HU to be a risk factor for PTX. This may be due to differences in study design or statistical choices. TLV has been found to increase with increasing severity of pneumothorax in a new large cohort study [13].
We did not include 6MWD in the multivariate regression analysis since the number of events did not allow for the inclusion of more than 4 variables. This choice was made before approaching the multivariate analysis. We expect that the low 6MWD would only indirectly be related to the risk of PTX through high RV. Furthermore, other studies find conflicting results, with Gompelmann et al. [9] finding a higher 6MWD in the PTX patients than the non‐PTX patients. Thus, it is still unknown if 6MWD is an independent risk factor for PTX.
In three cases of treatment of upper lobes, a 2‐step procedure was performed. None of these patients experienced PTX after either the first or the second procedure, but the number of patients was too low for statistical analysis.
Atelectasis was found in significantly more of the PTX patients 2 h after the procedure than in the non‐PTX patients. This was suggested as a risk factor of pneumothorax in the 2020 updated expert statement by Van Dijk et al. [6]. Some of the difference between the groups could be explained by a lower risk of PTX in patients without treatment response, but since the difference between groups is large, a link between the rapid development of atelectasis and PTX is likely. At least three nights of hospitalisation after BLVR has been recommended by an expert panel [6]. In our study, 2/46 (4%) PTXs occurred later than 3 days, while 3 (7%) PTXs occurred between days 2 and 3, suggesting that shorter hospital stays in patients without PTX risk factors could be implemented. We did not look at differences in severity or treatment indication of PTX at different time points after the procedure.
No previous studies have investigated the risk of other complications in patients with PTX compared to non‐PTX patients. We found a higher risk of pneumonia and pleural empyema in patients with PTX, which is reasonable to expect given the prolonged hospital stay, immobility, thoracic pain and drain treatment. We also found an increased use of antibiotic therapy for respiratory infections and imaging‐confirmed pneumonia. Confirmation of a pneumonia diagnosis by imaging resulted in an increased risk ratio between the PTX and non‐PTX patients compared to the risk ratio of antibiotic treatment. Since the non‐PTX patients are discharged earlier than the PTX patients, they mostly receive treatment for pneumonia symptoms through their general practitioner, where x‐ray is rarely used, making the pneumonia diagnosis less confident. Preventive antibiotics are routinely used with BLVR in all three centres, but it is not known whether PTX patients might benefit from further preventive antibiotics. Nevertheless, the focus should be on ordinary preventive measures like mobilisation, respiratory physiotherapy and pain management.
Revision bronchoscopies during the first 30 days after valve placement were very common in the PTX‐group (22.2%), which is consistent with the findings in other studies [1, 14, 16]. However, the prognosis was good after partial valve removal. Five of the six PTX patients were managed with partial valve removal and experienced PTX resolution and later successful valve replacement. Four of these patients furthermore obtained atelectasis at the first outpatient visit. This supports the recommendation from the 2020 Expert statement on the management of PTX [6], that PTX can be successfully managed with temporary partial valve removal.
The strengths of our study are that it is a nationwide cohort study, including all BLVR centres in Denmark, all the patients had a complete follow‐up, and all consecutive patients were included, except for one patient, who did not wish to consent. There are some limitations to our study. The study is retrospective and based on medical records. Accordingly, the quality of the available information depends on the accuracy and completeness of reporting by the treating physicians. The number of events did not allow for the inclusion of all the potential risk factors for PTX in the multivariate model. The presence of ipsilateral pleural adhesions, which is a known risk factor of pneumothorax from a previous study, was not investigated in this study. All patients were assessed with an x‐ray after the procedure, but some small asymptomatic PTX ex vacuo could have been missed. The protocol for assessment of fissure completeness differed between centres and might have introduced information bias due to misclassification for the Chartis procedure that potentially could have made a difference in the prevalence of PTX and reduced the generalisability of the results. However, despite this, no significant differences in PTX risk between centres were observed. In approximately half of the patients, Chartis was only used if fissure completeness was < 95%, which involves a risk of overlooking collateral ventilation [17]. The risk of other complications, like pneumonia, haemoptysis and so on, depends on the accuracy and completeness of reporting by the treating physicians, and since the PTX patients had a longer hospital stay, there is a risk of underreporting in non‐PTX patients.
5. Conclusion
In conclusion, high residual volume, target lobe volume and upper lobe treatment were independent risk factors of PTX after treatment of emphysema with endobronchial valves. PTX patients experienced faster development of atelectasis, a higher risk of post‐procedural pneumonia, antibiotic use, ICU stay, empyema and longer hospital stay than patients without PTX. Among these patients, focus should be on ordinary preventive measures like mobilisation, respiratory physiotherapy and pain management together with individualised patient information before the procedure.
Author Contributions
Kirstine Hermann Jørgensen: conceptualization (equal), formal analysis (equal), funding acquisition (equal), investigation (equal), validation (equal), writing – original draft (equal), writing – review and editing (equal). Thomas Decker Christensen: conceptualization (lead), funding acquisition (lead), resources (equal), writing – original draft (equal), writing – review and editing (equal). Ingrid Louise Titlestad: investigation (equal), resources (equal), writing – review and editing (equal). Jesper Rømhild Davidsen: investigation (equal), resources (equal), writing – review and editing (equal). Kathrine Bock: resources (equal), writing – review and editing (equal). Kristine Jensen: investigation (equal), resources (equal), writing – review and editing (equal). Morten Bendixen: investigation (equal), resources (equal), writing – review and editing (equal). Ole Dan Jørgensen: conceptualization (equal), writing – original draft (equal), writing – review and editing (equal). Michael Perch: conceptualization (equal), investigation (equal), resources (equal), writing – original draft (equal), writing – review and editing (equal). Elisabeth Bendstrup: conceptualization (lead), funding acquisition (lead), resources (equal), supervision (equal), writing – original draft (equal), writing – review and editing (equal).
Ethics Statement
Retrospective data extraction from patients included at AUH was approved by the Region of Central Denmark (1‐45‐70‐63‐23). This approval required informed consent from all live patients; deceased patients or patients vacated to another region were except. At OUH, all patients were included prospectively after informed consent. At RH, retrospective access to the patient data without informed consent was approved by the Capital Region Denmark (31‐1521‐229).
Conflicts of Interest
K.H.J. reports a grant from PulmonX outside the submitted work. T.D.C. has been on the speaker bureaus for AstraZeneca, Chiesi and GlaxoSmithKline and has been on an Advisory Board for Sanofi and AstraZeneca. J.R.D. reports fees for advisory board participation and lectures from Boehringer Ingelheim and AstraZeneca outside the submitted work. K.B. has nothing to report. K.J. has nothing to report. M.B. received a fee from PulmonX for a focused interview after testing new products outside the submitted work. I.L.T. has received a fee for advisory board participation outside the submitted work. O.D.J. has nothing to report. M.P. has received unrestricted research grants from PulmonX and Therakos and reports advisory board activity for Ryme Medical, Takeda, TFF and Zambon, all unrelated to this manuscript. E.B. has received lecture fees from Boehringer Ingelheim, AstraZeneca and Daiichi‐Sankyo, travel support from Boehringer Ingelheim and a fee for advisory board participation from Simbec‐Orion and Boehringer Ingelheim outside the submitted work.
Jørgensen K. H., Christensen T. D., Titlestad I. L., et al., “Risk Factors for Pneumothorax After Treatment With Endobronchial Valves: A Cohort Study,” Respirology 30, no. 10 (2025): 979–986, 10.1111/resp.70069.
See related editorial
Associate Editor: Ioannis Kalomenidis; Senior Editor: Fanny Wai San Ko
Funding: This study has been supported by the Central Denmark Region (Viborg, Denmark) and the independent fund Karen Elise Jensens Fond (Copenhagen, Denmark).
This study was previously presented at the European Respiratory Society, Vienna, 2024.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Klooster K., ten Hacken N. H. T., Hartman J. E., et al., “Endobronchial Valves for Emphysema Without Interlobar Collateral Ventilation,” New England Journal of Medicine 373, no. 24 (2015): 2325–2335, 10.1056/NEJMoa1507807. [DOI] [PubMed] [Google Scholar]
- 2. Davey C., Zoumot Z., Jordan S., et al., “Bronchoscopic Lung Volume Reduction With Endobronchial Valves for Patients With Heterogeneous Emphysema and Intact Interlobar Fissures (The BeLieVeR‐HIFi Study): A Randomised Controlled Trial,” Lancet 386, no. 9998 (2015): 1066–1073, 10.1016/S0140-6736(15)60001-0. [DOI] [PubMed] [Google Scholar]
- 3. Criner G. J., Sue R., Wright S., et al., “A Multicenter Randomized Controlled Trial of Zephyr Endobronchial Valve Treatment in Heterogeneous Emphysema (LIBERATE),” American Journal of Respiratory and Critical Care Medicine 198, no. 9 (2018): 1151–1164, 10.1164/rccm.201803-0590OC. [DOI] [PubMed] [Google Scholar]
- 4. Fiorelli A., D'Andrilli A., Bezzi M., et al., “Complications Related to Endoscopic Lung Volume Reduction for Emphysema With Endobronchial Valves: Results of a Multicenter Study,” Journal of Thoracic Disease 10 (2018): 3315–3325, 10.21037/jtd.2018.06.69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Koster T. D., Klooster K., Ten Hacken N. H. T., van Dijk M., and Slebos D. J., “Endobronchial Valve Therapy for Severe Emphysema: An Overview of Valve‐Related Complications and Its Management,” Expert Review of Respiratory Medicine 14 (2020): 1235–1247. [DOI] [PubMed] [Google Scholar]
- 6. van Dijk M., Sue R., Criner G. J., et al., “Expert Statement: Pneumothorax Associated With One‐Way Valve Therapy for Emphysema: 2020 Update,” Respiration 100 (2021): 969–978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Daniela Gompelmann F. J. F. H., Slebos D. J., Valipour A., Ernst A., Criner G. J., and Eberhardt R., “Pneumothorax Following Endobronchial Valve Therapy and Its Impact on Clinical Outcomes in Severe Emphysema,” Respiration 87 (2014): 485–491. [DOI] [PubMed] [Google Scholar]
- 8. Gompelmann D., Benjamin N., Kontogianni K., et al., “Clinical and Radiological Outcome Following Pneumothorax After Endoscopic Lung Volume Reduction With Valves,” International Journal of Chronic Obstructive Pulmonary Disease 11 (2016): 3093–3099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gompelmann D., Lim H. J., Eberhardt R., et al., “Predictors of Pneumothorax Following Endoscopic Valve Therapy in Patients With Severe Emphysema,” International Journal of Chronic Obstructive Pulmonary Disease 11 (2016): 1767–1773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. van Geffen W. H., Klooster K., Hartman J. E., et al., “Pleural Adhesion Assessment as a Predictor for Pneumothorax After Endobronchial Valve Treatment,” Respiration 94 (2017): 224–231. [DOI] [PubMed] [Google Scholar]
- 11. Bartolome R., Celli C. G. C., Marin J. M., et al., “The Body‐Mass Index, Airflow Obstruction, Dyspnea, and Exercise Capacity Index in Chronic Obstructive Pulmonary Disease,” New England Journal of Medicine 350 (2004): 1005–1012. [DOI] [PubMed] [Google Scholar]
- 12. Roodenburg S., Klooster K., Slebos D., and Hartman J., “The Impact of Emphysema Heterogeneity on Treatment Response After Endobronchial Valve Treatment,” ERJ Open Research 9, no. 4 (2023): 00279‐2023, 10.1183/23120541.00279-2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Brock J. M., Dittrich S. A., Eichhorn F., Schlamp K., Kontogianni K., and Herth F. J., “Lessons Learned: Risk Factors and Clinical Impact of Severe Pneumothorax After Endoscopic Lung Volume Reduction With Endobronchial Valves,” Chest 167, no. 4 (2024): 1012–1023, 10.1016/j.chest.2024.10.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kemp S. V., Slebos D.‐J., Kirk A., et al., “A Multicenter Randomized Controlled Trial of Zephyr Endobronchial Valve Treatment in Heterogeneous Emphysema (TRANSFORM),” American Journal of Respiratory and Critical Care Medicine 196, no. 12 (2017): 1535–1543, 10.1164/rccm.201707-1327OC. [DOI] [PubMed] [Google Scholar]
- 15. Sciurba F. C., Ernst A., Herth F. J. F., et al., “A Randomized Study of Endobronchial Valves for Advanced Emphysema,” New England Journal of Medicine 363, no. 13 (2010): 1233–1244, 10.1056/NEJMoa0900928. [DOI] [PubMed] [Google Scholar]
- 16. Criner G. J., Delage A., Voelker K., et al., “Improving Lung Function in Severe Heterogenous Emphysema With the Spiration Valve System (EMPROVE). A Multicenter, Open‐Label Randomized Controlled Clinical Trial,” American Journal of Respiratory and Critical Care Medicine 200, no. 11 (2019): 1354–1362, 10.1164/rccm.201902-0383OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Klooster K., Koster T. D., Ruwwe‐Glösenkamp C., et al., “An Integrative Approach of the Fissure Completeness Score and Chartis Assessment in Endobronchial Valve Treatment for Emphysema,” International Journal of Chronic Obstructive Pulmonary Disease 15 (2020): 1325–1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.