Published ahead of print on December 3, 2004, doi:10.1164/rccm.200407-961OC
© 2005 American Thoracic Society doi: 10.1164/rccm.200407-961OC
Effect of Bronchoscopic Lung Volume Reduction on Dynamic Hyperinflation and Exercise in EmphysemaRespiratory Muscle Laboratory, and Departments of Respiratory Medicine, Radiology, and Cardiothoracic Surgery, Royal Brompton Hospital; and Respiratory Muscle Laboratory, Guy's, King's and St Thomas' School of Medicine, King's College Hospital, London, United Kingdom Correspondence and requests for reprints should be addressed to Nicholas Hopkinson, M.A., M.R.C.P., Respiratory Muscle Laboratory, Royal Brompton Hospital, Fulham Road, London SW3 6NP, UK. E-mail: n.hopkinson{at}ic.ac.uk
Endobronchial valve placement improves pulmonary function in some patients with chronic obstructive pulmonary disease, but its effects on exercise physiology have not been investigated. In 19 patients with a mean (SD) FEV1 of 28.4 (11.9)% predicted, studied before and 4 weeks after unilateral valve insertion, functional residual capacity decreased from 7.1 (1.5) to 6.6 (1.7) L (p = 0.03) and diffusing capacity rose from 3.3 (1.1) to 3.7 (1.2) mmol · minute–1 · kPa–1 (p = 0.03). Cycle endurance time at 80% of peak workload increased from 227 (129) to 315 (195) seconds (p = 0.03). This was associated with a reduction in end-expiratory lung volume at peak exercise from 7.6 (1.6) to 7.2 (1.7) L (p = 0.03). Using stepwise logistic regression analysis, a model containing changes in transfer factor and resting inspiratory capacity explained 81% of the variation in change in exercise time (p < 0.0001). The same variables were retained if the five patients with radiologic atelectasis were excluded from analysis. In a subgroup of patients in whom invasive measurements were performed, improvement in exercise capacity was associated with a reduction in lung compliance (r2 = 0.43; p = 0.03) and isotime esophageal pressure–time product (r2 = 0.47; p = 0.03). Endobronchial valve placement can improve lung volumes and gas transfer in patients with chronic obstructive pulmonary disease and prolong exercise time by reducing dynamic hyperinflation.
Key Words: bronchoscopic lung volume reduction • chronic obstructive pulmonary disease • diaphragm • dynamic hyperinflation Patients with advanced chronic obstructive pulmonary disease (COPD) frequently experience exertional breathlessness despite optimal medical therapy. In selected patients lung volume reduction surgery (LVRS) has been shown to improve mortality, exercise capacity, and quality of life (1–3). However, it is associated with significant morbidity and an early mortality rate of about 5% (1, 2). For these reasons and because the procedure poses an unacceptable risk in patients with the most severe disease (1, 4), alternatives have been sought including bronchoscopic lung volume reduction (BLVR). This involves obstructing the airways that supply the most hyperinflated, emphysematous parts of the lung. The rationale for this approach is that endobronchial obstruction should cause these areas to collapse as a result of absorption atelectasis. By reducing lung volumes, symptoms could be improved without recourse to surgery. The technique was first performed with airway blockers (5) and subsequently our group (6) and others (7, 8) have described early experience with the use of endobronchially placed valves. However, our experience suggests that lobar collapse is not necessary for clinically apparent benefit to occur and we reasoned therefore that other physiologic mechanisms must operate. A key element in ventilatory limitation of exercise in COPD is the development of dynamic hyperinflation, in which expiratory flow limitation leads to a progressive increase in end-expiratory lung volume during exercise and consequently restricts the tidal volume that can be achieved (9). Reductions in dynamic hyperinflation have been demonstrated after treatment with bronchodilators (10–12) and after lung volume reduction surgery or bullectomy (13). BLVR could be expected to reduce dynamic hyperinflation either by causing the worst affected areas of lung to collapse or by excluding them from ventilation. In the presence of significant atelectasis BLVR should lead to a better matching of lung and chest wall dimensions, thus increasing available vital capacity as occurs after LVRS (14). By collapsing the most compliant areas of lung this should lead to an increase in lung elastic recoil at any given lung volume, reducing airflow obstruction. In the absence of atelectasis BLVR might still have benefits, first by reducing physiological dead space, which would improve the efficiency of ventilation, and second by reducing the dynamic hyperinflation that occurs at higher levels of ventilation by diverting airflow to less obstructed areas of lung. Therefore the aim of the present study was to investigate the effect of endobronchial valve placement on exercise capacity in patients with emphysema and to relate this to changes in dynamic hyperinflation assessed through changes in end-expiratory lung volumes. Some of the results of these studies have been reported previously in abstract form (15).
Patients with COPD consistent with the GOLD guidelines (16) entered the study if they had significant dyspnea despite optimal medical therapy including pulmonary rehabilitation; presented a heterogeneous pattern of disease with a target area identified by computed tomography (CT) scanning and ventilation perfusion scintigraphy (6); and were either considered too great a risk for LVRS (1, 4), or declined the surgery. The Royal Brompton Hospital (London, UK) Research Ethics Committee approved the study and patients gave their informed consent. Some data from the first eight patients in our series have been published previously (6). Endobronchial occlusion was performed with one-way valves (Emphasys Medical, Redwood City, CA) made of nitinol and silicone (see Figures E1 and E2 in the online supplement), placed to occlude segmental bronchi leading to the most affected area of lung. All procedures were unilateral. Initially, valves were inserted on a single occasion under general anesthesia (6, 17). Subsequently some procedures were performed with sedation only and some of these were staged, with valves being inserted on two separate occasions 1 to 2 weeks apart. Measurements were made in the week preceding and 4 weeks after valve insertion had been completed. A radiologist blinded to clinical outcome assessed CT evidence of atelectasis, defined as changes in the position of interlobar fissures adjacent to the targeted area. Quality of life was assessed on the basis of St George's Respiratory Questionnaire and the Short Form-36.
Pulmonary and Respiratory Muscle Function In all subjects we measured maximal static inspiratory (PImax) and expiratory (PEmax) mouth pressures (18) as well as maximal sniff nasal pressure (Pn,sn) (19). When patients consented to and were able to tolerate the placement of catheter-mounted balloons, esophageal and gastric pressures were determined and transdiaphragmatic pressure was calculated (19). In these patients sniff transdiaphragmatic pressure (Pdi,sn) and the response to bilateral anterolateral magnetic phrenic nerve stimulation (Pdi,tw) were also determined (20).
Exercise Testing In some patients we recorded esophageal and gastric pressures during exercise, calculating esophageal and diaphragmatic pressure–time product (PTP) (21, 22). Further methodologic details are given in the online supplement.
Statistical Analysis
Procedures and Complications Nineteen subjects (16 men) were studied. Baseline characteristics are given in Table 1. Details of the procedure performed in each subject, as well as the presence of radiologic atelectasis and individual changes in cycle endurance time and resting inspiratory capacity, are given in Table E1. All but 1 patient were taking inhaled steroids, 12 were taking long-acting ß2 agonists, 7 were taking oral theophyllines, 3 were taking regular oral prednisolone (less than 10 mg/day), 11 used a nebulizer, and 2 were receiving long-term oxygen therapy. The median (range) number of exacerbations requiring antibiotics in the preceding year was 2 (0–7).
There were no immediate complications related to the procedure itself. Two patients, both of whom had radiologic evidence of volume reduction, developed ipsilateral pneumothoraces: one at 2 days, which required intercostal drainage, and one at 4 weeks, which was small and resolved without intervention. There were no episodes of obstructive pneumonia. In five patients there was a transient worsening of symptoms consistent with an acute exacerbation in the early period after the procedure; these patients were treated with oral antibiotics. One subject developed Clostridium difficile diarrhea, presumably resulting from prophylactic antibiotic treatment, and one patient tripped at home, sustaining a rib fracture. In these patients postprocedure tests were delayed until they had recovered from these events. Radiologic evidence of atelectasis was present in five subjects (26%). In the group as a whole there were significant improvements in airflow obstruction, lung volumes, and transfer factor for carbon monoxide (TLCO) (Table 2).
Exercise Performance In the group overall there was a 39% improvement in mean cycle endurance exercise time from 227 (129) to 315 (195) seconds (p = 0.03), giving a mean  Tlim of +88 (167) seconds (Figure 1). Nine patients (47%) met the 60-second and 30% increase criteria considered to represent a clinically significant benefit and were defined as improvers.
Whole group mean changes in peak and isotime exercise parameters are given in Table E2. At peak exercise, end-expiratory lung volume (EELV) was reduced from 7.60 (1.6) to 7.18 (1.7) L (p = 0.03) and at isotime from 7.47 (1.5) to 6.97 (1.7) L (p = 0.05) with nonsignificant trends toward improvement in other parameters. If patients with atelectasis were excluded, the improvement in Tlim among the remaining 14 patients was no longer significant: preprocedure, 246 (143) seconds; postprocedure, 285 (179) seconds (p = 0.2).
Improvement in endurance time was associated with improvements in lung function, measures of dynamic hyperinflation and esophageal pressure–time product, as well as a reduction in static compliance (Table 3). Using stepwise regression analysis including all the variables listed in Table 3, only change in resting inspiratory capacity and
The results of a binary analysis comparing "improvers" with "nonimprovers" are given in Table 4. At isotime the improvers had a reduction in dyspnea and respiratory rate as well as highly significant reductions in lung volumes, with a decrease in EELV of 1.01 (0.8) L compared with an increase of 0.14 (0.5) L (p = 0.002) among the nonimprovers.
Changes in exercise capacity were not significantly associated with change in isotime heart rate or oxygen pulse (  O2/heart rate), or with changes in inspiratory muscle strength. Pdi,tw rose by 28.0 (31.6)% in the improvers but fell by 2.3 (23.9)% in nonimprovers, but this difference did not quite reach significance (p = 0.07). Improvement in Tlim was weakly associated with reduction (improvement) in the activity domain of the St. George's Respiratory Questionnaire (r2 = 0.23; p = 0.04). Other parameters were not significantly correlated.
Pressure measurements during exercise were available in 10 patients, that is, 4 improvers (1 with atelectasis) and 6 nonimprovers (1 with atelectasis). In the group as a whole, isotime ventilatory efficiency tended to improve:
Role of Atelectasis The only parameter measured at baseline that predicted the development of radiologic atelectasis was the body mass index (BMI), which was higher among those in whom it occurred: 28.3 (2.9) kg · m–2 versus 21.6 (2.9) kg · m–2 (p = 0.0004).
Static Lung Compliance
Respiratory Muscle Strength
Factors Predicting Improvement in Exercise Capacity A similar analysis was performed to determine which factors at baseline predicted improvement after BLVR (Table 5). Stepwise regression analysis was applied to a model containing BMI, VC % predicted, FEV1, RV/TLC, peak minute ventilation, respiratory rate and tidal volume, Pn,sn, Borg dyspnea and leg fatigue scores, as well as percentage reduction in inspiratory capacity. Only percent predicted vital capacity and respiratory rate were retained: Tlim = 24.5 (RR) – 3.3 (VC % predicted) – 294 (r2 = 0.62; p = 0.0004).
Although the difference was not significant it is interesting to note that the mean change in endurance time in those reporting limitation by leg fatigue alone or a mixture of leg fatigue and breathlessness (n = 6) was –1.7 (70.1) seconds compared with an increase of 129.9 (183.4) seconds in those reporting limitation by breathlessness alone (p = 0.1).
In this study we found that endobronchial valve placement improved mean exercise capacity and reduced dynamic hyperinflation in patients with COPD. The response was heterogeneous, with 47% of patients having a clinically significant improvement in exercise performance. Radiologic atelectasis occurred in only five patients. Improvements in exercise time were independently associated both with improvements in diffusing capacity and with reductions in static lung volumes, both in the presence and absence of atelectasis. Improvement post-BLVR was associated with a lower vital capacity and higher respiratory rate during peak exercise measured at baseline. Before discussing the significance of our findings a number of methodologic issues need to be addressed. The main outcome measure in this study was change in cycle endurance time at 80% of peak workload, which was analyzed as a continuous variable. Cycle endurance time has been shown to be highly reproducible in patients with COPD (11, 23), although there is no widely accepted definition of what constitutes a clinically significant change in endurance time at any given workload. However, on the basis of other studies that have looked at the effect of bronchodilators or LVRS in COPD (11, 12, 24) the scale of improvement observed is unlikely to have occurred as a consequence of natural variation or a placebo effect. The comparison of physiological parameters measured at isotime also makes a placebo effect unlikely. Because the response was heterogeneous, we have also presented the data in a binary fashion comparing changes in those who did or not improve. This definition, although arbitrary, was determined a priori on the basis of changes described in other studies (11, 12, 24) and serves to highlight differences between the two populations. A further limitation of the present study is that it was not possible to obtain a full range of invasive measurements in all subjects. This is to be expected because these were patients with significant disease and some people find passage of esophageal pressure catheters too uncomfortable to tolerate. However, we think that this is unlikely to have rendered these data untypical of the group as a whole because in the subgroup where pressure measurements during exercise were available, the proportion both of improvers (4 of 10) and of patients developing atelectasis (2 of 10) was similar to the group as a whole. The absence of atelectasis could have occurred because there was incomplete blockage of the relevant segment, or because of incomplete valve closure, but this did not appear to be the case when the valves were inspected visually. Because atelectasis did not occur in the majority of patients despite the apparent occlusion of all segments this suggests that there must have been significant interlobar collateral ventilation. A frequent incidence of incomplete fissures in postmortem studies of patients with emphysema has been noted previously (25) and increased collateral ventilation in this condition, which is likely in the presence of relatively lower resistance collateral channels and marked time constant inequalities (26), has been measured by a helium dilution technique (27). Interlobar collaterals have been demonstrated in 12 of 21 explanted lungs from patients with severe emphysema undergoing lung transplantation (28). No clinical factor appeared to predict the presence of these collaterals, which is consistent with our observation that no pre-BLVR factor except for BMI predicted atelectasis. It is possible also that a degree of atelectasis did occur that was too subtle to be apparent on CT. The observation that atelectasis was more likely to occur in patients with a higher BMI is intriguing. It was not retained as an independent correlate of improvement in exercise capacity. It is not obvious by what mechanism this could influence the presence of collateral ventilation. It may be a chance finding, but will be an important factor to consider in future studies.
Change in Pulmonary Function after BLVR Transfer factor for carbon monoxide increased significantly after valve insertion, and change in this parameter was a powerful independent predictor of improvement in exercise capacity. The mechanism of improvement is unclear. Single-breath assessment of TLCO involves a rapid inspiratory maneuver that may mimic dynamic flow effects during exercise more closely than measurement of static lung volumes. By excluding the worst affected segments of lung from ventilation, endobronchial valve placement would be expected to reduce inhomogeneity of gas mixing, diverting inspired gas to areas that are better perfused. In addition, where atelectasis occurs it may also permit the recruitment of previously "compressed" alveolar units in relatively healthier lung. There may also have been an increase in pulmonary capillary volume due to an improvement in cardiac function (as has been observed after LVRS [29]) either through a reduction in intrathoracic pressure swing or, again, because of a recruitment of additional capillaries in areas of compressed lung that are able to reexpand. Finally, the target area of lung, being the most emphysematous, would be expected to have the slowest time constants and might therefore have been contaminating the "alveolar" part of the sample during expiration preprocedure. This final mechanism, although reflecting emphysematous pathophysiology, might more properly be thought of as a measurement error. LVRS, possibly because the technique inevitably requires the removal of some relatively normal lung tissue, has not been found to improve TLCO in a number of studies (1, 30) although this has not been a universal finding (31, 32). This distinction may represent an important area of difference between the two techniques. Dynamic hyperinflation is central to exercise limitation in COPD (10, 11). In the present study, improvement in exercise capacity was associated with lower lung volumes, less dyspnea, and reduced respiratory pressures at isotime, consistent with previous work examining the effects of bronchodilators (33). Similar improvements in mechanical coupling have been demonstrated at rest after LVRS (34–36) and after exercise training (37). Changes in parameters of dynamic hyperinflation were not retained as independent predictors of improvement in exercise capacity when changes in resting inspiratory capacity and TLCO were included in the model, however. Change in resting IC has previously been shown to be the strongest correlate of improvements in exercise capacity after treatment with bronchodilators (38). The increased IC at rest allows the recruitment of functional airways and allows an increased tidal volume. Dynamic hyperinflation depends on minute ventilation and airflow obstruction. Among improvers, minute ventilation at isotime was unchanged but the pattern of ventilation was altered, with a reduction in respiratory rate and an increase in tidal volume. In the subgroup of patients with invasive measurements we observed an association between reduction in isotime PTPes and improvement in Tlim. PTPdi did not change significantly. This is consistent with the previous observation that in patients with COPD the diaphragm is mechanically disadvantaged, such that PTPdi tends to plateau despite increases in neural drive at higher levels of ventilation during exercise (22, 39). In particular there was a reduction in expiratory pressure generation. Expiratory pressure generation cannot improve ventilation in patients with expiratory flow limitation and the change observed is likely to improve overall efficiency both by reducing the oxygen cost of breathing and possibly by improving cardiac function. On the latter point it should be noted that we did not find any change in either heart rate or oxygen pulse after BLVR. These are, however, both relatively crude markers of cardiac function and it is possible that pulmonary artery measurements would have revealed differences. Interestingly, applying noninvasive ventilatory support only in the inspiratory phase of breathing has previously been shown to reduce expiratory pressure generation during exercise in patients with COPD (22). We did not find that BLVR changed global inspiratory muscle strength. Nor were changes in these parameters correlated with change in exercise time, which suggests that this is not an important mechanism, at least in the short term. In this study an increase in Pdi,tw was correlated with reductions in FRC consistent with the known length–tension relationship of the diaphragm as previously demonstrated by our group and others (40, 41). Shortening muscles moves their force–frequency curve to the right so that low-frequency (i.e., single-stimulus) techniques are most sensitive to the change (42). Volitional maneuvers are known to be less sensitive to changes in lung volume (43), which may explain the apparent discrepancy in the results. We would argue that the increase in Pdi,tw is to an extent an epiphenomenon of resting lung volume change rather than a mechanism of improvement as such. Nevertheless, mechanical disadvantage and shortening of the inspiratory muscles will reduce their ability to overcome intrinsic positive end-expiratory pressure and the greater elastic load at higher lung volumes as the pressure–volume curve of the lung flattens. In the presence of dynamic hyperinflation, the function of the respiratory muscle pump is limited by mechanical constraints on ventilation rather than by an inability of the respiratory muscles to sustain contraction as evidenced by the fact that diaphragm fatigue does not occur after exhaustive exercise in patients with COPD (44). It is not clear by what mechanism PEmax improved because there was a reduction in TLC. The absence of a change in cough gastric pressure suggests that abdominal muscle strength was not significantly improved. Improvements in maximal expiratory mouth pressure might have arisen because of an improvement in the alignment of thoracic expiratory muscles, or because of better transmission of pressure. The association between changes in FRC and Pdi,tw was similar in magnitude to those observed in other studies examining acute change either in a laboratory setting (40) or after surgical intervention (31, 32, 45, 46). In particular, the fact that a decrease in FRC was associated with an increase in Pdi,tw suggests that the optimal length for the diaphragm in patients with COPD lies below the FRC. We chose to target the most affected lobe of lung in patients with heterogeneous emphysema because this approach appears to be most successful in LVRS (2). Selecting the most appropriate patients for BLVR remains problematic. It is clear that patients with the mildest disease are unlikely to benefit from the intervention whereas in patients with the most severe disease it may either be ineffective or hazardous, as it could further reduce the lung available for gas transfer and expose patients to the risk of pneumothorax (5, 7, 8). We found that improvement was greatest in those with the highest respiratory rates during baseline exercise and would therefore support the use of cardiopulmonary exercise testing as part of the selection criteria as well as pulmonary function. From a clinical perspective it is worth noting that patients whose exercise was wholly or partly limited by leg discomfort were less likely to benefit. This is consistent with the concept that interventions to improve exercise capacity by improving ventilation may not be effective if exercise is limited by leg fatigue (47, 48).
The targeting strategy may also be important. Snell and coworkers studied patients with upper lobe–predominant emphysema and aimed to produce bilateral upper lobe obstruction. This produced improvements in gas transfer but no evidence of atelectasis or significant improvements in lung volumes (7). By contrast, Yim and coworkers, using a predominantly unilateral approach guided by Finally, this study adds to the body of literature demonstrating that BLVR is a relatively safe technique, which produces significant benefit in some patients, particularly those with the most restricted ventilation. A further question remains as to the time course over which improvements may occur as well as their duration. Clarification of its role in the management of COPD will depend on the results of larger randomized controlled trials, which are now in progress.
Supported by the Wellcome Trust (N.S.H.) and the European Union (QLK6-CT-2002-02285; M.I.P.). Emphasys Medical provided valves free of charge and an unrestricted educational grant to the hospital that has partially funded Dr. Toma's salary. The company had no part in the analysis of data or in the preparation of this manuscript. At the presubmission stage, Emphasys Medical reviewed the manuscript for factual errors and to ensure that there were no patentable disclosures that they had not already had an opportunity to cover or that were considered proprietary by the company. This review led to no changes in the manuscript. This manuscript has an online supplement, which is accessible from this issue's table of contents at www.thoracic.org Conflict of Interest Statement: N.S.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; T.P.T. received support from Emphasys Medical, which provided an unconditional educational grant to the Royal Brompton Hospital that partially funded Dr. Toma's salary and also supported his attendance at international conferences (ERS and ATS); D.M.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; P.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; D.M.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; M.I.P. has received fees for lecturing from AstraZeneca, GlaxoSmithKline, Nutricia, and Cephalon for a total of $3,700 between 2000 and 2003 and also attended professional meetings (ERS, ATS, and APSS) as a guest of various companies including GlaxoWellcome, AstraZeneca, Cephalon, 3M, and Ciba. Received in original form July 29, 2004; accepted in final form November 30, 2004
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
scan as well as CT produced greater than 75% collapse in 4 of 20 lobes targeted (
