Bronchoscopic lung volume reduction
Dirk-Jan Slebos, Karin Klooster and Nick H.T. Ten Hacken
Dept of Pulmonary Diseases, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands.
Correspondence: Dirk-Jan Slebos, Dept of Pulmonary Diseases/Interventional Bronchoscopy AA11, University Medical Center Groningen, PO Box 30001, 9700 RB Groningen, The Netherlands. E-mail: firstname.lastname@example.org
Bronchoscopic lung volume reduction (BLVR) is becoming the last-resort treatment option for patients with severe emphysema where current pharmaceutical treatments are not sufficient and surgical treatments are contraindicated. Over the past decade a number of devices and techniques have been developed to accommodate the very different emphysema phenotypes. The treatment options can be divided into “blocking” and “nonblocking” techniques. Blocking techniques use unidirectional valves to induce the collapse of a single lobe. This treatment is currently the most effective and fully reversible option, although it is only possible in emphysema patients with absence of interlobar collateral ventilation. In patients who do not qualify for a blocking technique, nonblocking techniques using nitinol coils or “sclerosing” techniques, such as vapour ablation or sealants, can be used as an alternative. Due to the complexity of the disease, patient selection, treatment logistics and dealing with complications in these very diseased patients, BLVR should only be performed in centres of excellence where multiple options are available and using a multidisciplinary team approach.
Cite as: Slebos D-J, Klooster K, Ten Hacken NHT. Bronchoscopic lung volume reduction. In: Herth FJF, Shah PL, Gompelmann D, eds. Interventional Pulmonology (ERS Monograph). Sheffield, European Respiratory Society, 2017; pp. 276–293 [doi.org/10.1183/2312508X.10004117].
Bronchoscopic lung volume reduction (BLVR) is “the new kid on the block” in interventional pulmonology, and is becoming the last-resort treatment option for patients with severe emphysema where current pharmaceutical treatments are not sufficient and surgical treatments are contraindicated [1, 2]. During the past decade a number of bronchoscopic interventions have been developed to accommodate the very different emphysema phenotypes that exist . By means of elegant and robust trial designs and repetitive solid outcomes, the use of BLVR as a treatment option for emphysema has also been acknowledged in the 2017 Global Initiative for Chronic Obstructive Lung Disease (GOLD) COPD recommendations (grading level B evidence) [4, 5]. Although lung volume reduction surgery (LVRS) still has a solid scientific and guideline position in the treatment of patients with severe emphysema, the actual use of the surgical approach is rather limited . This is driven by limited surgical expertise worldwide, morbidity, the narrow indication window as defined by the National Emphysema Treatment Trial results  and current lack of confidence by referring physicians. The need for advanced treatments in severe emphysema and the small patient group that might benefit from LVRS drove the development of bronchoscopic alternatives. The current innovative bronchoscopic approaches are much less invasive compared with surgical treatments and are applicable in a greater population of patients with very severe COPD, thereby potentially serving a large need. Currently, BLVR is resulting in large patient referral numbers, thereby potentially also contributing to revisiting and reviving the use of LVRS [2, 8].
Hyperinflation in emphysema
Emphysema as part of the umbrella term “COPD” is characterised by a continuous loss of lung tissue. Chronic inflammation develops due to cigarette smoking or inhalation of toxic agents, leading to destruction of elastin and a weakened lung parenchyma. This loss of lung tissue is visualised in the flow–volume diagram by the well-known sudden drop in the first part of expiration, indicating an increased valve mechanism during forced expiration. In emphysema, reduced elastic recoil decreases the alveolar pressure, whereas loss of radial traction on airways decreases airway patency, both components contributing to a low FEV1. The cardinal sign of emphysema is breathlessness during exertion, a phenomenon poorly associated with low FEV1 and closely associated with hyperinflation. Breathlessness may introduce a vicious circle of avoidance of activities, leading to deconditioning, reduced exercise capacity and increased breathlessness. In particular, the work of Cooper and O’Donnell indicated that hyperinflation plays a key role in the feelings of dyspnoea and reduced exercise capacity in emphysema [9, 10].
Static hyperinflation in emphysema develops over many years (decades), and is due to the continuous loss of lung tissue and destruction of elastin. In contrast, dynamic hyperinflation is an acute process (minutes) due to the higher breathing frequency and reduced time to exhale completely during exercise. During exercise the dynamic hyperinflation on top of static hyperinflation leads to a reduced inspiratory capacity, which induces feelings of dyspnoea. These feelings result from the higher, yet less effective efforts of the inspiratory muscles to reward the increasing ventilatory needs. Inspiratory muscles in a hyperinflated state operate at a shortened resting length in an unfavourable position on the length–tension curve, thus the force generation capacity of the muscles is decreased. Additionally, the load against which the inspiratory muscles must operate is increased due to hyperinflation. In particular, when the inspiratory capacity approaches 0.5 L, this imbalance translates to high-intensity feelings of dyspnoea .
Different interventions are available that may improve hyperinflation and dyspnoea during exercise. Bronchodilators have been shown to reduce hyperinflation during exercise and improve exercise tolerance, yet the effect sizes are small [11, 12]. Their method of action is to decrease expiratory airflow resistance, leading to a better emptying of the lung at shorter expiration times. Supplemental oxygen during exercise improves exercise capacity in another way. Due to a lower ventilator drive the breathing frequency is lowered, also allowing more time to exhale between breaths and thus reducing dynamic hyperinflation. Pulmonary rehabilitation may decrease the ventilatory drive during exercise and slow down the respiration rate. Additionally, emphysema patients may learn to use “pursed lip breathing”, which results in a longer expiration time, as well as improved airway patency [9, 10].
The most direct way to reduce hyperinflation is to deflate the lungs by bronchoscopic or surgical techniques. LVRS improves lung function primarily by better matching the size of the lungs to the thorax . This increases the vital capacity, i.e. the major determinant of FEV1, and consequently improves static and dynamic hyperinflation [13, 14]. BLVR techniques also attempt to improve static and dynamic hyperinflation, yet without the need for invasive surgery. Different bronchoscopic techniques have been developed, e.g. airway bypass stents, valve implants, coil implants, bronchial thermal vapour ablation and biological lung volume reduction, to accommodate different emphysema phenotypes. All these techniques aim to deflate the lung, but the mechanisms of action show important differences. Airway stents bypass the obstructed airways, one-way valves deflate target lobes during expiration (not allowing inspiration), coils lead to torqueing of the bronchi and increase the radial tension of the adjacent airway network, and thermal vapour ablation and biological lung volume reduction induce a local inflammatory and fibrotic reaction. The efficacy and adverse reactions of these techniques also show important differences, and depend on the different mechanisms of actions [2, 3].
Blocking techniques (“valves”)
Endobronchial valve treatment
One-way endobronchial valve treatment is a bronchoscopic procedure designed to collapse an entire lobe to induce lung volume reduction in severe emphysema patients. These valves mimic the effects observed with surgical lung volume reduction by excluding the most diseased lobe of the lung and the result is potentially reversible. The endobronchial valves (Zephyr endobronchial valves; PulmonX, Redwood, CA, USA) are small self-expandable nitinol (shape memory wire) stents with a silicone coating and a unidirectional Heimlich valve incorporated in the body of the device (figure 1). The valves are available in different sizes to accommodate different airway dimensions. Endobronchial valves are delivered using a dedicated delivery catheter and should be placed in all segmental bronchi of the target lobe (figure 2). It is crucial that all valves are perfectly placed in order to allow the lobe to fully empty. Furthermore, the valve treatment will not be effective in situations where interlobar collateral ventilation is present between the target lobe and the adjacent lobe. Interlobar collateral ventilation can be detected by using the Chartis system (PulmonX) (figure 3). Pre-selection of potential treatment candidates can be done using chest CT fissure integrity analysis .
After the first efforts using one-way endobronchial valves  and the first feasibility trials showing promise [17, 18], the international multicentre (2:1) randomised controlled VENT trial was performed, investigating the safety and efficacy of endobronchial valve treatment in patients with heterogeneous emphysema. This study was undertaken in the USA (321 patients in 31 sites) as well as in Europe (171 patients in 23 sites), with the results being published separately for the two continents [19, 20]. The main inclusion criteria were COPD with a FEV1 15–45% predicted, residual volume (RV) >150% predicted and heterogeneous emphysema on a chest CT scan. In the US trial, FEV1 improved by +6.8%, 6-min walk distance (6MWD) by +19.1 m and the St George’s Respiratory Questionnaire (SGRQ) score by −3.4 points (all p<0.05) .
In the European part of the trial, mean FEV1 improved by +6.5% and 6MWD by +5 m (both nonsignificant), with a change of −4.7 points in the SGRQ score (p<0.05) . Although the overall trial results might have just reached statistical significance, the clinical responder rate was rather disappointing. However, a small group of patients did benefit very significantly from this treatment and post hoc analyses showed that a complete fissure (being a surrogate for absence of interlobar collateral ventilation) and a correct valve placement (meaning a full lobar occlusion with valves) resulted in a clinically significant improvement in outcome [19–21].
These post hoc insights of the VENT trial contributed to the understanding of the importance of having no interlobar collateral flow between the treated lobe and adjacent nontreated lobe. A (near-)complete interlobar fissure on chest CT scan might indicate such a situation, but it is still a surrogate for the real-life situation. Therefore, the development of the Chartis system to actually measure functional collateral ventilation created a breakthrough in patient selection for endobronchial valve treatment . With the introduction of Chartis, responder rates in target lung volume reduction improved to 75% and set the standard for future trials [23, 24].
The first trials (both single-centre trials) prospectively using best-responder criteria for endobronchial valve treatment were the BeLieVeR-HIFi study (using CT fissure analysis to select patients)  and STELVIO trial (using Chartis to select patients) , and were the benchmark for subsequent larger randomised controlled studies. The BeLieVeR-HIFi study was a (1:1) randomised, full sham bronchoscopy controlled study in patients (n=50) with heterogeneous emphysema and a fissure completeness of >90% on CT . Chartis was performed in all patients, but only used for post hoc analysis. At 3-month follow-up, FEV1 improved by mean +24.8% in the endobronchial valve group versus +3.9% in the sham control group. Median results showed an improvement in FEV1 of 8.8% in the treatment group versus 2.9% in the controls (p=0.0326), +25 m in 6MWD versus 3 m for the controls (p=0.0119) and −4.4 points in the SGRQ score versus −3.6 points for the controls (nonsignificant). However, when using the Chartis data, four patients in the treatment group had presence of collateral ventilation and thus no clinical benefit. Excluding these patients with collateral ventilation improved all clinical outcomes of this study .
In the STELVIO trial , 68 patients were (1:1) randomised to treatment with endobronchial valves or usual care, with a 6-month follow-up period. A re-bronchoscopy was allowed in this trial to adjust the initial valve placement in case of lack of target lobar volume reduction. Endobronchial valve treatment demonstrated a +20.9% improvement in FEV1 for the treatment group versus +3.1% for the controls, an improvement in 6MWD of +60 m for the treatment group versus −14 m for the controls and a SGRQ score difference of −14.7 points in favour of the treated versus the control patients (all p<0.001) . This trial furthermore demonstrated the advantage of endobronchial valve treatment, as the valves can be removed or replaced to optimise outcomes and manage pneumothoraces, which was necessary in a third of all patients. The impact of endobronchial valve treatment in well-selected patients was furthermore shown by a significant and clinically very relevant improvement in physical activity as measured in steps per day, which showed a difference of 1340 steps per day between treatment and controls . More recently, the 1-year follow-up results of STELVIO have been reported, including the data of the crossover to treatment results of the initial control patients . The data show persistence of benefit of the treated patients, with significant improvements at 1 year in FEV1 (+17%), RV (−687 mL), 6MWD (+61 m) and SGRQ score (−11 points) (all p<0.001). All these improvements in a disease as severe as end-stage emphysema can contribute to a potential survival benefit. Although never scientifically proven, post hoc data and open-label single-centre experiences show strong signals that support this statement [29, 30].
All trials thus far have primarily focused on treating patients with a heterogeneous emphysema distribution. However, post hoc analysis using quantitative CT (QCT) analysis also showed solid significant results for pulmonary function, exercise and quality of life in patients with a homogeneous emphysema distribution [20, 26]. These findings were prospectively tested in the multicentre (1:1) randomised controlled IMPACT trial, where endobronchial valve treatment was evaluated in patients (n=93) with homogeneous emphysema in the absence of collateral ventilation (using Chartis). At 3 months after treatment, FEV1 improved by +13.7% from baseline in the valve treatment group versus −3.2% in the controls, 6MWD to 22.6 m versus −17 m and quality of life measured by the SGRQ score to −8.6 versus 1.0 points (all p<0.001) . This trial clearly demonstrated the validity of endobronchial valve treatment in patients with a predominant homogeneous emphysema phenotype without collateral ventilation. This is an important finding, especially since the majority of these patients do not qualify for LVRS.
Recently, the data from the TRANSFORM trial, the first multicentre (2:1) randomised controlled trial in patients (n=97) with heterogeneous emphysema and absence of collateral ventilation (using Chartis), were presented . At 6-month follow-up after endobronchial valve therapy, FEV1 improved +20.7% in the treatment group versus −8.6% for controls, with a between-group difference in RV of 700 mL, 6MWD of 78.7 m and SGRQ score of −6.5 points (all p<0.001). These findings confirm the earlier single-centre trial data experience from STELVIO  using best-responder criteria for endobronchial valve treatment .
Intrabronchial valve treatment
The Spiration intrabronchial valve (Spiration/Olympus, Redmond, WA, USA) is also a unidirectional valve mechanism with the same proposed mechanism of action as the endobronchial valves. The intrabronchial valve, however, has an umbrella-shaped design, which is compressed against the airway, thus acting as a valve mechanism. The intrabronchial valve is bronchoscopically placed by a special delivery catheter and comes in different sizes. Very precise airway sizing (using a special balloon sizing catheter kit) is crucial to obtain full lobar occlusion.
After an early pilot study with bilateral total occlusion of both upper lobes, showing a serious safety issue , the use of intrabronchial valves was promoted to be used bilaterally, but nonfully occluding by leaving one segment of a lobe untreated, and thus avoiding pneumothoraces. In a (1:1) randomised sham controlled multicentre study, patients (n=73) with upper lobe predominant heterogeneous emphysema were evaluated using this approach . As anticipated, no pneumothoraces occurred. However, there were also no significant improvements in mean SGRQ, FEV1, RV and 6MWD. A second randomised controlled trial using the same treatment approach also failed to demonstrate significant clinical outcomes . The bilateral nonoccluding approach was abandoned after these trials. Driven by the results of the endobronchial valve trials and clinical expertise, intrabronchial valve treatment was thereafter only used in the same way as for endobronchial valves, i.e. complete lobar occlusion. In order to prove this concept, a very elegant trial was performed by EBERHARDT et al. , who in a single-centre (1:1) randomised controlled trial compared unilateral complete lobar occlusion with intrabronchial valves with the bilateral nonoccluding approach in upper lobe heterogeneous patients (n=22). As expected, the unilateral treatment group showed significant improvements in all clinical outcomes, whereas the bilateral nonoccluding approach resulted in no benefit .
Recently, the results of the very first (2:1) randomised controlled trial using intrabronchial valves as a unilateral full occluding treatment were reported in abstract form . In this Chinese study (REACH study), heterogeneous emphysema patients (n=101) selected on the basis of having a complete fissure on CT as a surrogate for having no collateral ventilation (no measurement of collateral ventilation was performed) were evaluated. At 6 months after treatment, 67% of the patients showed a significant target lobar volume reduction on CT, with a mean reduction of 779 mL. Also at 6 months, mean FEV1 improved to +12.9% for the treatment group versus −1.7% for controls (p<0.001) and the SGRQ score to −9.1 versus +3.5 points (p=0.0023). The pneumothorax rate was ∼8%. This trial clearly shows that the intrabronchial valve design also works in this patient group. Comparable to the BeLieVeR-HIFi results where endobronchial valves were used, also using CT fissure analysis only as the determinator to select patients for valve treatment, the outcome of valve treatment seems even more pronounced when adding Chartis as a physiological measure of collateral flow to the treatment algorithm. The two approaches have not been evaluated head to head, but comparing the results of the trials using Chartis versus the REACH study clearly shows the additional value of selecting-out patients with presence of collateral flow despite having a complete fissure on CT. To illustrate this, in the STELVIO trial  the target lobar volume reduction on CT was mean 1.3 L with a 88% response rate and in the TRANSFORM trial  the target lobar volume reduction was 1.1 L with a response rate of 89%. Retrospective analysis shows a quite similar predictive value of both dedicated CT fissure analysis and Chartis [38, 39], with an even higher predictive value when combining both techniques .
Choosing between the two commercially available valve designs will depend on local practice and expertise, ease of use, and product availability and marketing strategies. In the more experienced centres they are sometimes even used together in one patient, because each of them has its unique features. Whereas the endobronchial valve will more easily accommodate the different airway dimensions with less sizing issues and due to the design being compliant to the changes in airway diameter in COPD airways, the intrabronchial valve is more suitable for very difficult airway anatomical situations (especially the B6 segments).
Pneumothorax after valve therapy
A post-procedural pneumothorax is the most common complication after valve treatment, but should actually be regarded as being part of the procedure. This is because of the likelihood that it will happen, the experience needed to deal with the different types of pneumothoraces that arise and the fact that patients who developed a pneumothorax have a similar outcome compared with those who do not [15, 41, 42]. The sudden change in lung volume due to the valve-induced atelectasis causes a further decrease of the negative intrapleural pressure, and with subpleural bullae and adhesions being present can cause a pneumothorax. These pneumothoraces can be asymptomatic in the case of “ex vacuo” situations when the treated lobe is affected or highly symptomatic, requiring chest tube insertion when arising from the nontreated lobe. In the early endobronchial valve studies the pneumothorax ratio was ∼4%, although in the most recent studies this increased to 20–25% due to optimal patient selection and more experienced treating physicians. This higher pneumothorax rate correlates perfectly with a much better outcome compared with previous low-rate pneumothorax studies .
A few attempts have been made to identify patients at risk for developing a pneumothorax, with high baseline RV, target lobar volume and presence of significant pleural adhesions associated with a higher pneumothorax rate [43, 44].
This knowledge can be used to discuss the risk of pneumothorax with the individual patient.
The treatment of a valve-induced pneumothorax is described in detail in an expert statement on this topic . As the majority of pneumothoraces occur within the first days, it is recommended to keep these patients for observation in the hospital for 3–5 days with access to emergency chest tube placement after treatment. Dealing with these pneumothoraces is one of the reasons to organise high-volume centres of excellence, in order to be able to make this special pneumothorax management routine practice .
Lung volume reduction using endobronchial coils is another technique developed to treat patients bronchoscopically with advanced emphysema. The RePneu lung volume reduction coil (PneumRx Inc., Santa Clara, CA, USA) is a nitinol device (figure 4) that is delivered bronchoscopically using a special delivery system into subsegmental airways. About 10–12 coils, available in three sizes (100, 125 and 150 mm, to accommodate different airway lengths), are placed in the desired lobes under fluoroscopy to visualise positioning (figure 5). The procedure is preferably performed under general anaesthesia and patients generally stay in hospital for 1 night of observation after the procedure. The lung volume reduction coil procedure is a sequential treatment, with one lobe per procedure treated; the contralateral lobe is treated 4–8 weeks later. Bilateral treatment is needed to obtain optimal results. The coils have to be regarded as an implant and treatment is permanent. However, they are adjustable periprocedurally; long after the procedure, at most one or two coils can be removed when causing local problems, but only in experienced hands [45, 46].