INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung volume reduction surgery (LVRS) has emerged over the past decade as an adjunct to conventional medical treatment for end-stage emphysema (1). This technique has produced significant physiologic and functional improvement in approximately 75% of recipients (3). Cohort (2, 3, 7) and randomized studies (8) from several centers, including our own, have demonstrated that lung function, exercise capacity (11), and health-related quality of life all improve after LVRS (12). Widespread enthusiasm for using LVRS in the treatment of advanced emphysema combined with considerable controversy about its therapeutic value has led to the organization of two multicenter randomized clinical trials in the United States. These studies are designed to better define selection criteria, duration of response, impact on longevity, and physiologic basis for improvement after surgery (13, 14). It is anticipated that LVRS will play an important role in the therapeutic management of this debilitating disease for years to come (6). There are approximately 1.5 to 2 million patients with symptomatic emphysema in the United States and perhaps 3 to 4 times that number worldwide who could potentially benefit from LVRS. At the present state of the art, however, only approximately 15 to 20% of this population is deemed eligible for operative volume reduction therapy (7).

Emphysema is characterized by destruction of lung tissue, which leads to hyperexpansion and loss of elastic recoil (15). Available medical therapies, such as bronchodilators and anti-inflammatory agents, afford only modest palliation but do not arrest or reverse the progression of the disease. By contrast, resection of damaged lung through volume reduction surgery addresses the issue of hyperexpansion directly, and has proven to be effective in improving respiratory mechanics and reducing symptoms in many patients.

In a study published in 1998, Fessler and Permutt suggested that LVRS improves lung function in emphysema by increasing the percentage and absolute volume of functioning lung remaining within the chest (16). These changes result from removal of overinflated, nonfunctional lung in which the ratio of residual volume (RV) to total volume is greater than in the lung left behind in the chest. This "resizing" of the overexpanded lung in a more normal-sized chest cage results in a net increase in vital capacity (VC).

Stimulated by the findings of Fessler and Permutt, it occurred to us that obliteration of dysfunctional regions of hyperinflated lung may not require a surgical procedure but could be achieved nonoperatively by transbronchoscopic generation of sustained collapse through the following: (1) filling target regions of the lung with oxygen, an absorbable gas that promotes atelectasis; (2) rinsing target airways with biocompatible "antisurfactant" solution to promote destabilization; (3) applying suction to remove residual surfactant and cause rapid collapse; and (4) injecting a biocompatible fibrin-based glue to produce "sealing" and maximize atelectasis. It was anticipated that this approach will, over a period of weeks, collapse and scar the dysfunctional target region and reduce overall lung volume without infection or excess inflammation.

The present communication summarizes the results obtained from the application of this novel technique in an experimental sheep model of emphysema produced by inhalation of papain.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Protocol

Twelve female sheep (weighing 54 ± 12 kg, range 43 to 72 kg) were studied in accordance with a protocol approved by our Institutional Animal Studies Committee and tested to ensure that they were free of zoonoses (Chlamydia psittaci, Mycobacterium pseudotuberculosis, Brucella, and Coxiella burnetti) infections. Detailed baseline lung function measurements were performed in all animals (Norm). After determination of baseline values the animals were exposed to inhaled papain (Sigma Chemical, St. Louis, MO) 7,000 units/wk for four consecutive weeks using a parallel system of dual nebulizers (Pari LC Plus; Pari Respiratory Equipment, Midlothian, VA) together with a nonrebreathing circuit. It has been shown that this method produces physiologic and anatomic abnormalities consistent with emphysema (17). Physiologic measurements were repeated 2 wk after completion of papain inhalation regimen (Emph) and then the sheep were divided into three treatment groups of four animals: surgical volume reduction (SVR), bronchoscopic volume reduction (BVR), and bronchoscopy alone (Sham-BVR).

SVR animals were anesthetized, placed on a mechanical ventilator, and underwent stapled resection of approximately 20 to 30% of upper and tracheal lobes (SVR) through a ministernotomy. Chest tubes (number 24 French) were placed in both pleural spaces, and connected to Heimlich valves. The chest tubes were removed within 72 h of surgery in all animals. BVR animals were also anesthetized and intubated, and were maintained on mechanical ventilation throughout the procedure. Initially the animals were given 100% oxygen for 15 min. A 7 mm × 1 m fiberoptic bronchoscope, designed to reach and wedge into second- or third-generation bronchi (Instrument Technology, Inc., Westfield, MA), was introduced through the endotracheal tube and advanced into the orifice of each of five target subsegments. Ventral lobe subsegments were selected for bronchoscopic reduction in this study. After positioning of the bronchoscope in the target bronchial branch, a multichannel balloon-tipped occlusion catheter (1.9 mm outer diameter) was introduced through the suction channel of the instrument, and advanced under direct visualization to the desired location. The balloon was inflated and 50 ml of washout solution was injected distally followed by application of continuous suction (120 to 140 mm Hg) for 3 to 5 min to remove as much washout solution as possible, and to promote collapse. The balloon was then deflated, and the catheter gently advanced until fully wedged. The fibrin-based "sealant" (BistechSeal; Bistech Inc., Boston, MA) was injected through the end hole of the catheter. The injection was maintained as the catheter was slowly withdrawn for a distance of 1 to 2 cm, to back-fill and seal each targeted subsegment. The location of the five subsegments targeted for reduction in each animal was recorded. After completion of BVR, the bronchial orifices of all targeted subsegments were inspected visually to ensure adequacy of glue administration and effective sealing. The animals were allowed to recover from anesthesia, observed for 30 to 60 min, and returned to their pens. Sham-BVR animals underwent bronchoscopy without any other procedure during anesthesia.

All treated animals (Sham-BVR, BVR, and SVR) were monitored for hypoxemia, wound infection, fever, clinical signs suggesting the need for analgesia, and failure to eat or drink appropriately. Physiologic measurements were repeated 8 wk after completion of treatment (Post-Rx). The animals were then euthanized and autopsied. The lung territories that had undergone reduction were inspected for gross changes and processed for histologic examination.

Physiologic Measurements

All Norm, Emph, and Post-Rx measurements included assessment of static and dynamic lung physiology. After placement of an intravenous catheter, animals were induced with ketamine (Ketaset; Fort Dodge Pharmaceuticals, Fort Dodge, IA, 70 mg/kg) and diazepam (Valium, 0.25 mg/kg), and intubated under fiberoptic guidance with a 12-mm endotracheal tube. Anesthesia was maintained using continuous intravenous propofol. An esophageal balloon was introduced and the desired position was verified by observing transpulmonary (PL) and transrespiratory system pressures, and cardiac oscillations on PL profiles. Heart rate and arterial oxygen saturation were monitored throughout the study period using an oximeter and tongue probe. FRC was measured by whole body plethysmography during spontaneous breathing. Intermittent airway occlusions were triggered remotely (outside the box) using a pneumatic shutter (Pneumatic controller valve; Hans Rudolf Inc., Kansas City, MO), allowing for measurements of inspiratory efforts against a closed shutter. This facilitated simultaneous measurements of airway pressure and lung volume changes, and determination of intrathoracic gas volume based on Boyle's law (11). Quasi-static pressure-volume relationships (QSPV) during lung deflation from TLC were determined in triplicate. For the purpose of this study, TLC was defined as the lung volume achieved at 30 cm H₂O transpulmonary distending pressure. Dynamic lung function was then assessed during forced oscillatory ventilation using a computer-controlled pneumatic ventilator system applying the optimal ventilator waveform (OVW) technique of Lutchen and coworkers (18). This approach involves the application of a broad-band pseudorandom volume signal [V(t)] which represents a sum of 6 sine waves each with a different frequency. The frequencies and sine wave amplitudes have been selected to minimize harmonic distortion (i.e., signal distortion related to nonlinear effects of the respiratory system) while maintaining flows in the laminar range to allow linear analysis techniques to be applied. Lung impedance, the ratio of pressure (P) to flow ( = dV/dt), is determined by performing Fourier analysis on measured pressure, volume, and flow signals, and fitting the results to the following frequency domain equation of motion (19) using multivariate regression:

(1)

This approach allows for assessment of airway resistance (Raw), tissue resistance (Rti = G/), and tissue dynamic elastance (EL = H¹) over the entire range of breathing frequencies from a single measurement. Changes in the frequency dependence of the dissipative (G) and elastic (H) moduli are sensitive indicators of heterogeneity in lung function. Therefore, this measurement technique provides a method for assessing both the magnitude and spatial distribution of lung dysfunction associated with papain-induced emphysema, as well as any changes in lung physiology that might follow volume reduction therapy.

QSPV relationships were fit to the exponential equation of Salazaar and Knowles (20), in which lung volume is expressed as a function of pressure according to the equation: V(P) = Vmax  Ae^kP, where V is volume, P pressure, Vmax the volume attained at "infinite" distending pressure, A = (Vmax  Vmin), Vmin is the volume at zero distending pressure (equivalent to RV), and k is a parameter describing the shape of the exponential fit between pressure and volume (the shape factor). In the analysis that follows, RV/TLC ratios were determined as Vmin values divided by the lung volume measured when PL = 30 cm H₂O.

Necropsy and Histopathology Data

After completion of Post-Rx physiologic measurements, the animals were euthanized with intravenous pentobarbital (100 mg/kg) and autopsied. The lungs were removed, examined, and photographed. They were cut into 2- to 2.5-cm sagittal slices and inspected for scar formation, infection, and pleural adhesions. The locations of changes were correlated with previously recorded BVR injection sites. Tissues from Sham-BVR lungs, target BVR injection sites, and from LVR surgical specimens were harvested, fixed in 15% buffered formalin, and stained with hematoxylin and eosin for comparison by light microscopy.

Outcome Variables and Statistical Analysis

Treatment responses were summarized using three groups of outcome variables: procedural complications, static and dynamic lung physiology, and lung anatomy.

Complications were defined as (1) hypoxemia (determined by tongue oximetry measurements) requiring oxygen supplementation (Sa_O2 < 85% for longer than 5 min); (2) wound infection assessed clinically by the veterinary staff; (3) pneumonia assessed by clinical symptoms and chest radiography; (4) fever > 1.5° C above baseline; (5) postprocedural weight loss > 10% of body weight. Treatment complication rates were quantified by multiplying the number of days each animal suffered from complications by the number of animals with complications in each treatment group. Rate of complications among the different treatment groups were compared using Kruskal-Wallis analysis for multiple samples.

Static lung physiology was expressed in terms of k, Vmax, and Vmin values as determined from matching QSPV data to the Salazaar-Knowles relationship as described previously. Dynamic lung function was expressed in terms of Raw, G (the tissue dissipation modulus), and H (the tissue elastic modulus). Comparisons between physiologic parameters at different time points among different treatment groups were performed by analysis of variance (ANOVA). Differences between specific groups were identified using Duncan's post hoc analysis test. Gross anatomic and histologic changes were recorded and photographed. Statistical significance was defined as a p value of less than 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physiologic and Histologic Responses to Papain Exposure

Papain exposure produced changes in static lung physiology and function consistent with emphysema in all 12 study animals. Two weeks after completion of the 4-wk course of papain exposure, lung and trapped gas volumes increased at all transpulmonary pressures. QSPV relationships were shifted upward and to the left, and RV/TLC ratios increased. These changes are consistent with the physiologic patterns reported in human emphysema (Figure 1) (15).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Effects of papain inhalation on the static pressure-volume relationship of the sheep lung. The following data summarize curve fits obtained from linear regression analysis using the log transform of the exponential Salazaar-Knowles equation: TLC: 2.52 ± 0.37 L (preinhalation) versus 3.00 ± 0.29 L (postinhalation); RV: 0.69 ± 0.22 L (pre) versus 1.29 ± 0.42 L (post); RV/TLC: 0.27 (pre) versus 0.43 (post). Fits were excellent for all cases (r > 0.93; for n = 12 Norm data, r = 0.95 ± 0.04; for n = 12 Emphy r = 0.96 ± 0.03). At baseline (filled triangles) TLC and RV are as shown, with RV/TLC ratio equal to 0.27. After papain exposure (filled squares) the static P-V relationship shifts upward and to the left, displaying a pattern consistent with that described in patients with smoking-related emphysema. Both TLC (p = 0.006) and RV (p = 0.009) increased significantly, and the pattern of change was such that RV/TLC ratio also increased 59% to 0.43. The shape factor (k), obtained from curve fits to the Salazaar-Knowles equation, was not significantly increased, although there was a trend toward increasing values (0.11 ± 0.04 to 0.15 ± 0.06, n = 12, p = 0.09) subsequent to papain therapy.

Measurements of dynamic lung function after papain treatment demonstrated that Raw increased by 162% (0.46 ± 0.19 Norm to 1.21 ± 0.56 cm H₂O/L/s Emph, p = 0.0011 by paired t test). However, Rti; (G) and EL (H) did not increase significantly. These changes in Raw, and lack of frequency dependence in G and H, suggest that papain exposure caused a mild to moderate homogenous form of emphysema characterized by hyperinflation, and airflow obstruction. Microscopic examination of samples from sheep lung with papain-induced emphysema demonstrated significant destructive changes at the level of the respiratory bronchioles, alveolar ducts, and alveoli (Figure 2).

View larger version (74K): [in this window] [in a new window]   Figure 2.   Histomicrographs comparing normal with papain-exposed sheep lung tissue. Papain inhalation caused marked histologic changes in tissue architecture with parenchymal destruction at the level of the alveolar ducts and alveoli.

Comparison of the Physiologic Effects of BVR and SVR

QSPV data from Norm, Emph, and Post-Rx (Sham-BVR, SVR, and BVR) animals are summarized in Figure 3. Baseline static lung function was similar in all groups; Vmax, Vmin, and k values were not significantly different by post hoc analysis. After papain exposure, all animals showed similar physiologic changes consistent with emphysema. Vmax and Vmin increased and there was a leftward overall shift in the QSPV relationships of all groups. Responses to treatment were significantly different among Sham-BVR, SVR, and BVR animals, however. Sham treatment produced little change in QSPV profiles (Figure 3). By contrast, animals in both BVR and SVR groups demonstrated significant reductions in RV and TLC, with larger percentage reductions in RV than in TLC.

View larger version (25K): [in this window] [in a new window]   Figure 3.   Baseline, postpapain, and posttreatment static pressure-volume relationships are summarized for Sham-BVR (left panel ), SVR (center panel ), and BVR (right panel ) animals. Mean values for data fitted to the Salazaar-Knowles equation are shown. Sham treatment had little effect on lung volumes; there was a slight tendency for P-V profiles to return toward baseline, as reflected in TLC (V at 30 cm H2O PL) and RV (Vmin), but changes were not significant. By contrast, SVR and BVR produced large, significant (by post hoc analysis) reductions in TLC and RV.

Changes in dynamic lung function in response to treatment are summarized in Table 1. At baseline Raw, G, and H were similar in Sham-BVR, BVR, and SVR animals. After papain exposure, Raw increased in each group similarly, and changes compared with baseline were significant by post hoc analysis. Values of G and H did not change significantly, however. The initial increases in Raw observed after papain exposure tended to persist in Sham-BVR, as well as in BVR- and SVR-treated animals. G and H, which were unaffected by papain exposure, were similarly unaffected by Sham-BVR, BVR, or SVR.

Comparison of Complication Rates among the Different Treatment Groups

Although the magnitude of reduction in lung volumes in BVR and SVR groups was similar, the minimal invasiveness of BVR relative to SVR was understandably associated with fewer postprocedural complications, as summarized in Table 2. Results are expressed in terms of the total number of days associated with each complication for the individual treatment groups. All SVR animals exhibited fever beyond the first postoperative day, required analgesics for as long as 4 d, oxygen therapy after extubation for several hours, and antibiotics for as long as 7 d as determined by the veterinary staff. Two animals experienced clinically significant wound infections requiring medical intervention. By contrast, in the BVR animals only 2 of 4 developed fever requiring antibiotics for 3 d. Although all four BVR animals needed oxygen therapy in the immediate postprocedure period, Sa_O2 values improved rapidly, and treatment was required for less than 1 h in each instance. Nonparametric statistical analysis (Kruskal-Wallis testing) indicated that SVR was associated with significantly greater morbidity than either BVR or sham therapy (p = 0.05). None of the animals in the BVR or SVR groups experienced respiratory failure requiring reinstitution of mechanical ventilator support, significant weight loss, or death.

Necropsy and Histopathology Results

Necropsy performed 8 to 12 wk after treatment demonstrated no gross abnormalities in the lungs of Sham-BVR control animals other than overall lung enlargement. By contrast, SVR was consistently associated with pleural adhesions, and parenchymal scarring at sites of surgical stapling. In the BVR group, two of the four animals developed pleural adhesions with subpleural scarring, and showed sterile parenchymal abscesses (three abscesses in two animals). In a single animal, BVR was specifically limited to the right lung only. As shown in Figure 4, the treated lung is grossly smaller than that on the untreated contralateral side.

View larger version (182K): [in this window] [in a new window]   Figure 4.   Gross pathology from an animal that had undergone five bronchoscopically directed volume reduction procedures all within the right lung. Four of the five were successful in that associated scarring was present. The figure shows the lung to be grossly smaller, demonstrating the ability of BVR to reduce volumes on a fairly large scale.

Of the 20 BVR target areas in four animals (5 per animal), only 11 sites exhibited sustained collapse (2.75 ± 0.5 subsegments/animal). An example of the gross changes of atelectasis and scarring, and histologic changes of collagen deposition, which follow BVR is shown in Figures 5A and 5B. Three of the 11 target sites exhibited sterile abscesses (Figure 5C). They were encapsulated by scar tissue without evidence of infection either by histology or Gram stain.

View larger version (104K): [in this window] [in a new window]   View larger version (127K): [in this window] [in a new window]   View larger version (197K): [in this window] [in a new window]   Figure 5.   (A) Gross pathology of a target region from an animal after BVR (left panel ) compared with that of a section of lung taken from a control animal that had undergone sham bronchoscopy in the location shown. BVR produced a well-developed peripheral scar within the distal portion of the target zone. (B) Histomicrographs from this region, in which extensive collagen deposition is demonstrated, suggestive of a permanent scar. (C ) An area in a different target region where a peripheral sterile abscess developed as a result of glue injection. There was no evidence of bacterial infection on Gram stain, or evidence of pneumonitis in surrounding areas.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates that in an animal model of mild to moderate emphysema, BVR was as effective as SVR in causing sustained reductions in TLC and RV at 2 to 3 mo postprocedure. Because BVR is not associated with surgical trauma, and requires only a simple bronchoscopy that can be performed in the outpatient setting, the procedure should cause significantly less morbidity, and be less costly than SVR in clinical practice. Although the potential application of BVR remains to be determined, this approach holds the promise of providing an effective nonsurgical alternative for reducing lung volume, and holds far-reaching clinical implications for patients with end-stage emphysema. The ability to reduce lung volume by a nonoperative method with less mortality, morbidity, and cost may broaden existing indications for volume reduction therapy and allow performance of staged, or even repeated procedures. BVR offers the promise of helping many patients currently not considered appropriate candidates for SVR.

The emphysema animal model used in the present study is an adaptation of the papain inhalation dog model originally proposed by Barnas and coworkers (21), and generated physiologic changes consistent with mild to moderate homogenous emphysema. Raw increased by 163%, RV by 87%, and TLC by 20%. Animals in the Sham-BVR group showed no significant change in Raw, TLC, or RV during the 8-wk follow-up observation period, suggesting that papain exposure produces permanent remodeling of airways and parenchyma, rather than mere transient inflammation.

Physiologic responses to BVR and SVR were similar and significantly different from those observed among Sham-BVR control animals. Collapse and scarring of 2.75 subsegments per animal by BVR was, on average, equivalent to surgical resection of 20 to 30% of lung tissue. The quasi-static pressure- volume relationships displayed significant decreases in TLC and RV in both BVR and SVR treatment groups (Figure 3). The physiologic model developed by Fessler and Permutt, in which changes in FEV₁ relate directly to changes in FVC, suggests that BVR and SVR are equally effective as volume reduction therapies. Neither BVR nor SVR produced significant changes in airflow resistance, indicating that in this animal model, the isovolume increases in recoil and tethering had little effect on airway caliber. However, these changes should improve expiratory flows by increasing elastic recoil driving pressures. Our observations, and these physiologic projections, are consistent with findings of Hoppin (22), Fessler and Permutt (16), and Gelb and coworkers (2) regarding how LVRS works to improve lung function in human subjects.

The functional measurements and anatomic findings reported in this sheep model indicate that the washout and glue systems (BistechSeal) used to achieve BVR produced localized collapse without causing lasting inflammation, or atelectasis in adjacent regions that had not been targeted for therapy. Physiologic information regarding response patterns is obtained from the frequency dependence of pulmonary resistance (RL) and EL. A response characterized by spatial heterogeneity due to partial collapse or inflammation in regions adjacent to primary target areas would have caused a rightward shift in EL versus frequency relative to SVR responses, and an increase in the magnitude of the dissipation modulus, G. No such changes were observed.

Gross inspection and histologic examination suggest that the favorable physiologic responses observed after BVR are a result of replacement of air-containing pulmonary parenchyma by atelectasis and scar formation, in most cases, with minimal inflammation (Figure 5). However, only 11 of 20 regions targeted for BVR demonstrated sustained collapse and the remaining nine subsegments were not altered by the procedure, giving a 55% success rate. Furthermore, three of the 11 "successful" BVR sites developed sterile abscesses that could have resulted either from liquefaction necrosis induced by fibrin degradation products of the sealant matrix (23), or by ischemic necrosis as a consequence of mucosal distension and generation of local pressures at the site of gel polymerization that exceed capillary perfusion pressures. None of the three animals with sterile abscesses displayed persistent fevers, anorexia, weight loss, or other systemic signs suggestive of active infection. In addition, none developed empyema. Nevertheless, it is not known whether these sterile abscesses would have evolved into volume-reducing scars or into infected lung abscesses. Clearly these complications are not acceptable, and prevent BVR from being employed clinically as it has been described here. However, new washout solution and sealant preparations that address these important limitations are currently being developed and tested.

BVR, as described in this study, has several additional important limitations that must be addressed before this approach can be employed safely in humans. Regions targeted for BVR in this pilot study were all dependent. This was done specifically to reduce the tendency of fibrin gels to migrate within the airways as a consequence of gravitational forces. Although the gels tend to "anchor" securely to tissues within target regions of lung by polymerizing in a distal-to-proximal fashion and forming a castlike structure within the small airways and alveoli, we have not yet demonstrated the safety and effectiveness of this approach in apical lung zones. Furthermore, although the changes in static and dynamic lung function observed in these papain-treated sheep are consistent with emphysema, the physiology of this model differs significantly from that observed in the majority of patients who are presently considered for LVRS. Patients generally suffer from more severe lung dysfunction (> 125% increases in TLC, > 175% increases in RV, and 300 to 500% increases in RL) than was present in animals in this study. Finally, most patients selected for LVRS harbor heterogenous rather than the homogeneous disease produced by papain. The responses to BVR observed in this sheep model with mild to moderate homogeneous disease may not replicate results obtained in patients with more severe, heterogeneous disease.

Despite important limitations, the initial results presented in this report are encouraging and suggest that BVR holds the promise of an attractive alternative to SVR for treatment of end-stage emphysema. We anticipate that ongoing development and modification of BVR will soon make this technique suitable for clinical use. By virtue of its less invasive nature, BVR has the potential for substantially reducing the morbidity, mortality, and cost associated with volume reduction therapy. A more highly refined approach to BVR will allow utilization by many patients, who currently, are not considered for the procedure because of prohibitive operative risk, homogenous disease, or a prior history of thoracic or volume reduction surgery.