Serial lung function and elastic recoil 2 years after lung volume reduction surgery for emphysema.

STUDY OBJECTIVE
To evaluate serial lung function studies, including elastic recoil, in patients with severe emphysema who undergo lung volume reduction surgery (LVRS). To determine mechanism(s) responsible for changes in airflow limitation.


METHODS
We studied 12 (10 male) patients aged 68+/-9 years (mean+/-SD) 6 to 12 months prior to and at 6-month intervals for 2 years after thoracoscopic bilateral LVRS for emphysema.


RESULTS
At 2 years post-LVRS, relief of dyspnea remained improved in 10 of 12 patients, and partial or full-time oxygen dependency was eliminated in 2 of 7 patients. There was significant reduction in total lung capacity (TLC) compared with pre-LVRS baseline, 7.8+/-0.6 L (mean+/-SEM) (133+/-5% predicted) vs 8.6+/-0.6 L (144+/-5% predicted) (p=0.003); functional residual capacity, 5.6+/-0.5 L (157+/-9% predicted) vs 6.7+/-0.5 L (185+/-10% predicted) (p=0.001); and residual volume, 4.9+/-0.5 L (210+/-16% predicted) vs 6.0+/-0.5 L (260+/-13% predicted) (p=0.000). Increases were noted in FEV1, 0.88+/-0.08 L (37+/-6% predicted) vs 0.72+/-0.05 L (29+/-3% predicted) (p=0.02); diffusing capacity, 8.5+/-1.0 mL/min/mm Hg (43+/-3% predicted) vs 4.2+/-0.7 mL/min/mm Hg (18+/-3% predicted) (p=0.001); static lung elastic recoil pressure at TLC (Pstat), 13.7+/-0.5 cm H2O vs 11.3+/-0.6 cm H2O (p=0.008); and maximum oxygen consumption, 8.7+/-0.8 mL/min/kg vs 6.9+/-1.5 mL/min/kg (p=0.03). Increase in FEV1 correlated with the increase in TLC Pstat/TLC (r=0.75, p=0.03), but not with any baseline parameter.


CONCLUSION
Two years post-LVRS, there is variable clinical and physiologic improvement that does not correlate with any baseline parameter. Increased lung elastic recoil appears to be the primary mechanism for improved airflow limitation.

D espite aggressive medical therapy, including physical rehabilitation, the prognosis and palliative relief of dyspnea in COPD due to emphysema is poor. When the FEV 1 falls below 0. 75 L or 30% *From the Pulmonary Division, Departm ent of M e dicine, Lake- predicted, survival at 3 y ears is only 50 to 60%. 1 • 2 Furthermore, patients admitted to a hospital ICU for exacerbation of COPD have a 1-year mortality rate of 30% and in patients aged >65 years, the 1-year mortality rate doubles. 3 During the past several years, there has been emphasis in thoracic surgical procedures that attempt to provide palliative relief for markedly dyspneic patients with severe, diffuse (nongiant bullous) e mphysema. Unilateral and bilateral video-assisted thoracoscopic 4 • 5 or median stemotomy&-12 incisions are made, and the worst-targeted emphysematous a reas are excised, ie, lung volume reduction surgery (LVRS ). Following bilateral LVRS, results indicate variable improvement in relief from dyspnea, oxygen d ependency, lung function , and exercise tolerance at 6 months,4,5,7-ll,I3.1 4 1 year,6,I5 and ;:::2 years following surgery.6,12 The present study evaluates the clinical and physiologic changes, including lung elasticity, in 12 markedly symptomatic patients with severe emphysema who have been followed up, both preoperatively and every 6 months, for 2 years post-LVRS . It extends our previously published results obtained immediately, 13 6 months, 1 4 and 12 months 15 after bilateral LVRS. The thrust for this study was to analyze the physiologic mechanism(s) responsible for serial changes in lung function 2 years following LVRS.

Patient Selection
From February through June 1995, we evaluated 28 patients aged 67±8 years (mean± SD) who unde1went LVRS. The intent was to obtain preoperative and postoperative lung function studies, including measurements of lung elasticity at 6-month intervals. Following LVRS, five patients died (at 1, 16, 17, 20, and 24 months ) from respiratory failure, one patient was unavailable for follow-up, and three patients refused to be retested. Incomplete data were obtained in 7 of the remaining 19 patients since they refi.1sed repeated measurements of lung volumes, diffusing capacity, and elastic recoil post-LVRS, but did agree to spirometry. However, 12 (10 male ) patients aged 68±9 years (mean±SD) satisfied our criteria and are described in detail. In the seven other patients, as well as th e five patients who died post-LVRS and the one patient unavailable for follow-up and three patients who refused any additional testing, their preoperative clinical status and all lung function studies, including lung elastic recoil, were similar to the results obtained in the 12 presently reported patients. Furthermore, these seven patients demonstrated no significant increase in spirometry at 12 and 24 months post-LVRS compared with their baseline. Data from 8 of the 12 patients in this report were previously published at 6 14 and 12 15 months post-LVRS. The current data obtained at 24 months post-LVRS have not been previo usly reported. The cigarettesmoking history of the patients studied was 50± 15 pack-years (mean±SD).
All the patients who had LVRS were markedly symptomatic \vith grade 2:3 dyspnea, 16 tolerance limited to walking < 100 yd, with severe, fixed airflow limitation that had not improved despite antibiotics, oxygen, corticosteroids, aerosol, and oral bronchodilators. Thin-section (2-mm) high-resolution CT of the lungs 17 · 18 demonstrated visual emphysema scores ranging from 60 to 80 in the upper third lung fields and scores ranging from 40 to 70 in the lower third lung fi elds. Heterogeneity of visually scored emphysema distribution behveen upper and lower lung fields was present in every patient. We obtained standard nuclear medicine-perfusion lung scans (six view) in all patients, and in five patients, 99 mTc macroaggregated albumin single-photon emission CT scans. Results demonstrated vascular distribution abnormalities corresponding to lung CT scans. There was relatively well-preserved perfusion in the lower third of lungs and none or markedly decreased perfusion in the upper third of lungs.

Operative Technique
As previously described, 4 · l3 -l5 after obtaining informed consent and approval of the Institutional Human Investigation Committee at Chapman Medical Center, patients underwent sequential 1498 bilateral upper lung fields video-assisted thoracoscopic stapled lung resectional surge1y at the same operative sitting.

Lung Function Studies
As previously described, 14 · 15 dyspnea evaluation, 16 arterial blood gases at rest, and lung function, including elastic recoil studies, were measured after obtaining informed consent in a pressure-compensated flow plethysmograph (model 6200 Autobox; SensorMedics; Yorba Linda, Calif) and compared with predicted values. Maximum expiratory flow and ai1way conductance (Caw) were plotted against static lung elastic recoil pressure curves, as previously described. 1 4.1 5 Studies in patients were obtained within 6 to 12 months and 2 weeks prior to and repeated every 6 months for 2 years after LVRS. The studies at 6 to 12 months prior to LVRS were originally obtained as yearly follow-up studies in patients with severe airflow limitation.

Exercise Studies
Progressive exercise testing to symptom-limited maximum was obtained using cycle ergom etry (Tunturi; Turku, Finland) with 2-min increases of 10 to 20 W at pedaling cycle of 40 to 50 1pm. Subjects breathed room air through a mouthpiece with nose clips using a low-resistance hvo-way nonrebreathing valve. Expired gases were collected and analyzed (Vmax 29; SensorMedics Inc). A subset of only 7 of the 12 patients agreed to and were evaluated preoperatively and every 12 months post-LVRS. Lung function studies in the 7 patients were not significantly different from the 12 patients, either pre-LVRS or post-LVRS.

Statistical Methods
Comparison of the difference between patients before and after surgery was determined using hvo-tailed paired or twosample unpaired t test with p values <0.05 being significant.
Because of the small sample size, the degree of linear association behveen hvo continuous valiables was assessed using the nonparametric Spearman correlation coefficients based on ranks. Each patient served as his or her own control subject for comparison with end points post-LVRS.

RESULTS
Results of serial complete lung function and resting arterial blood gas studies in 12 patients appear in Table l. Spirometry, lung volumes, and diffusion studies were available in patients 6 to 12 months prior to surgery, and results (data not shown) were similar when compared with 2-week preoperative baseline values, despite aggressive therapeutic intervention, including physical rehabilitation. The average hospital stay was 10.7±1.0 days (mean±SD). Dyspnea 16 was improved in every patient by ::::::1 grade at 12 months post-LVRS and ::::::1 grade in 10 patients 24 months post-LVRS. Oxygen dependence, full or part time, because of resting or postexercise Pa0 2 <59 mm Hg was eliminated in two of seven patients up to 24 months post-LVRS. Up to 2 years preoperatively, two patients each required four hospitalizations for exacerbation of their COPD. Within 2 years post-LVRS, these same patients required one and three hospitalizations.

Lung Function Studies
In the 12 patients described in detail at 24 months post-LVRS, there was still significant improvement in most physiologic studies except resting arterial blood gases when compared with preoperative values. Compared with baseline, the FEV 1 , specific airway conductance (SGaw), diffusing capacity, static lung elastic recoil pressure at both functional residual capacity (FRC ) and total lung capacity (TLC), and coefficient of retraction (static lung elastic recoil pressure at TLC!fLC) remained significantly improved 24 months post-LVRS, despite reduction in all static lung volumes. Spirometric and lung volumes were most improved at 6 months post-LVRS .

Exercise Studies
Results of exercise studies appear in Table 2, and pre-LVRS , all patients had severe exercise intolerance. The increase in oxygen consumption, minute ventilation, tidal volume, and resting oxygen saturation peaked at 1 year post-LVRS . However, even at 2 years post-LVRS, exercise performance and resting oxygen saturation remain above pre-LVRS baseline values.

Maximum Flow Volume Loops
Analysis of the mean maximum expiratory and inspiratory flow volume l oops in 12 patients demonstrates severe airflow limitation and hyperinflation at baseline (Fig 1). Compared with preoperative LVRS baseline, there was a continued downward shift on the volume axis toward lower lung volumes even at CHEST / 11 3/6/JUNE, 1998

Static Lung Elastic Recoil Pressure Curves
Preoperatively, there was marked loss of lung elastic recoil (Fig 2). The peak mean increase in static lung elastic recoil occurred 6 to 12 months

Maximum Expiratory Flow-Static Lung Elastic Recoil Pressure Curves
At baseline, the critical transmural pressure in small airway collapsible segment (Ptm 1 ) was shifted toward higher pressures than age-matched normal subjects, and conductance of small airway S segment (Gs ) was markedly reduced (Fig 3). There was a significant increase in small airway Gs, and decrease in Ptm 1 only up to 12 months post-L VRS when compared with baseline. After 1 year post-LVRS, values for Gs and Ptm 1 were similar to baseline preoperative values. However, the increased driving pressure (elastic recoil) increased maximum expiratory flow at isovolume points.

Caw-Lung Elastic Recoil Pressure Curve
Initially, all patients had reduced airway conductance that could not be accounted for solely by loss of lung elastic recoil (Table 1 and Fig 4). Up to 12 months following LVRS , despite the reduction in FRC, total Caw measured at FRC increased significantly due to the significant increase in lung elastic recoil. At 24 months post-LYRS, the increase in FEY 1 correlated best with the increase in coefficient of retraction (r=0.75; p=0.03) and increase in Caw (r=0.89; p=0.001) . This emphasizes post-LYRS the importance of increased lung elastic recoil despite reduction in lung volume to increase maximum expiratory flow and airway caliber. However, relief from dyspnea, oxygen independence, and improved exercise tolerance did not correlate with increased FEY 1 .

DISCUSSION
Results in the present study reveal that at 24 months after targeted bilateral stapled L YRS for severe, nonbullous generalized emphysema, 12 selected patients maintained significant improvements in lung function, with variable relief from dyspnea, improved oxygen independence, and increased exercise tolerance when compared with baseline. This is primarily due to increased lung elastic recoil despite the reduction in lung volume. However, preoperative clinical, physiologic, and CT lung studies could not identify those individual patients who had optimal clinical improvement and increases in FEY 1 post-LYRS.

Lung Elastic Recoil
We have previously reported 19 · 20 that expiratory airflow limitation in clinically unsuspected and early physiologic (normal or near-normal FEY 1 ), but moderately advanced morphologic emphysema (mean visually scored anatomic grade 50) and bullous lung disease 21 without concomitant emphysema could be accounted for by loss of lung elastic recoil. This results in decreased driving pressure and loss of alveolar support to tether the airways during forced exhalation. We 21 and others 22 -25 have also noted the increase in expiratory airflow and Caw following bullectomy in isolated bullous lung disease, and bullous emphysema could be attributed to the increase in lung elastic recoil. The increase in lung elastic recoil described by Sciurba et aJ2 6  following unilateral L VRS for generalized emphysema, and our results immediately, 6 months, and 12 monthsl3-15 following bilateral LVRS, is probably the mechanism for improvement in expiratory airflovv. The present report extends these conclusions 24 months post-LVRS.

Mechanical Changes
A physiologic consequence of emphysema is loss of lung elastic recoil, causing hyperinflation. This, together with dynamic airway collapse and intrinsic positive end-expiratory pressure, causes a shift in breathing to higher lung volumes. There is shortened diaphragm muscle and reduced surface area 27 with significant functional impairment in muscle strength with hyperinflation. 28 Furthermore, dyspnea may be better correlated with abnormal respiratory muscle dysfunction, breathing patterns, and hyperinflation than expiratory airflow limitation. 29-3l Following LVRS, there is marked reduction in all static lung volumes, eg, residual, FRC, and TLC. After 6 months post-LVRS, there is a progressive increase in these volumes, although they remain 1502 significantly below baseline values even at 24 months. We believe this increasing hyperinflation corresponds to the subsequent loss of the lung elastic recoil that also occurs after 6 months following LVRS.
Despite the overall reduction in all static lung volumes, there are increases in FEVI> FVC, Caw, and maximum expiratory flow compared with baseline at isovolume points that peak at 6 months but persist 24 months post-LVRS. This also reflects the initial increase and subsequent loss of lung elastic recoil that is observed following LVRS. Moreover, even at 24 months post-LVRS, lung elastic recoil pressure at TLC and FRC remain significantly increased when compared with baseline values.

Diffusing Capacity
The increased diffusing capacity 6 to 12 months following LVRS 14 · 15 was maintained after 2 years and probably reflects a greater alveolar-capillary surface area due to less lung tissue compression and increased transpulmonary pressure. may be more even distribution of ventilation/perfusion ratios to the areas with better potential for diffusion.

Mechanism of Airflow Limitation
Analysis of the maximum expiratory airflow-static lung elastic recoil pressure curve and the Caw-static lung e l astic recoil curves (Figs 3 and 4) indicates markedly abnormal airflow and Gaw, both before and after LVRS , that cannot be explained completely by the loss of lung elastic recoil and/or airway collapse. We attribute this to marked intrinsic small airway structural abnormalities that are associated with long-term cigarette smoking in patients with severe COPD and emphysema. 18 • 32 This physiologic profile was also seen in half the patients with a 1antitrypsin deficiency, reported by Black et al, 33 as well as those patients with moderately severe airflow limitation, reported by Duffell et al. 34 Leaver et aP 5 noted in all 16 COPD patients with relatively severe airflow limitation (FEV 1 , 1.02 ± 0.4 L [mean±SD]) that maximum expiratory airflow was disproportionately reduced co mpared with the l oss of lung elastic recoil. They attributed this to a combination of intrinsic airways diseas e and enhanced collapsibility of flow-limiting airways (decreased Gs with increased Ptm' ). Hogg e t aP 6 de monstrated a predominant periphe ral intrinsic small airways site to explain the e l evated airway resistance in patients who died of emphysema. They described mucus plugging, narrowing, fibrosis, distortion, and obliteration of small airways. 36 Moreover, increasing the distending pressure (elastic recoil) failed to decrease the airway resistance , and the re was no difference between inspiratory and expiratory airway resistance. They concluded that despite destruction of alveolar support for the airways and decreased lung elastic recoil in severe e mphysema, airflow limitation is primarily due to intrinsic small airways abnormalities. 36

Dyspnea
The relationship between post-LVRS improvement in FEV 1 and relief of dyspnea is poorly understood. Using multivariate analysis, we have previously noted that the increase in FEV 1 following unilateral stapled LVRS correlated statistically (p<0.05 ) only with smoking history and younger age, but not with preoperative thoracic gas volume, spirometry, or diffusing capacity. 37 Furthermore, analysis of 154 patients undergoing bilateral stapled LVRS noted that only th e presence of a bilateral upper lobe heterogeneous pattern on lung CT and perfusion scan correlated with improvement in revealed that FEY 1 increased a mean of 50% (from 0.64:±::0.27 L [mean±SD] to 1.04:±::0.4 L), but was not correlated with improvement in dyspnea score (r=0.3; p=0.3). 39 The postoperative LYRS improvement in dyspnea score correlated weakly with preoperative plethysmograph calculated TLC (r=0.3; p=0.2), trapped gas volume (plethysmograph TLC-gas dilution TLC) (r=0.2; p=0.7), residual volume (RY)/TLC (r=0.4; p=0.05), and RY (r=0.4; p=0.03).39

Exercise and Dyspnea
At 1 year post-LYRS, the significant increase in maximum oxygen consumption and work performance we noted was achieved with increased minute ventilation and tidal volume with decreased respiratory frequency. At 2 years post-LYRS, the increase in maximum oxygen consumption may be related, in part, to improved cardiac output due to less mechanical constraints. However, the observed less-thanrobust improvement in gas exchange and exercise tolerance at 2 years following LYRS emphasizes that there may be disproportionate improvement(s) in lung mechanics, exercising ability, perception of dyspnea, and gas exchange. Similar observations have been noted postlung transplantation 40 and LYRS. 41 A recent study by O'Donnell and colleagues 42 investigated the potential mechanisms for short-term (3 months) relief of dyspnea in eight patients following unilateral LYRS. They attributed it to a combination of reduced thoracic hyperinflation, decreased breathing frequency, reduced mechanical constraints on tidal volume, and increased FYC. Keller et al 43 noted in 25 patients 4 months after LYRS that increased maximal oxygen consumption was accomplished by increased inspiratory and expiratory flows with larger minute ventilation and tidal volume, but no change in respiratory frequency and no correlation with clinical relief from dyspnea. Benditt et al 44 noted that following LYRS, improved exercise performance was associated with increased maximal ventilation.
Benditt et al 45 evaluated eight patients 3 months after bilateral LYRS and noted improvement in ventilatory muscle recruitment. There was a reduction in both end-expiratory resting and exercise esophageal and gastric pressures. Results were consistent with less recruitment of the abdominal and accessory muscles and a relatively greater contribution of the diaphragm in inspiratory muscle generation. Bloch et al 46 made similar observations studying patients before and after LYRS using respiratory inductive plethysmography. Martinez et al 47 reached similar conclusions after evaluating 17 patients at 1504 least 3 months after bilateral LYRS. In addition to increased lung recoil at TLC, they noted variable increases in maximal inspiratory mouth and transdiaphragmatic pressure, FEY t> work performance, and less dynamic hyperinflation and breathlessness. Similar to Keller et al, 43 they noted little change in the ratio between inspiratory time and total respiratorycycle duration. Teschler et al 48 also reported increases in transdiaphragmatic pressures at 3 months after unilateral stapled LYRS. However, the increase in transdiaphragmatic pressures reported42.45,47,48 post-LYRS were not measured at lung isovolumes pre-LYRS.

Improvement in FEV 1
Roue et al 12 noted clinical and functional improvement in 4 of 11 patients at 2 years post-LYRS, in three patients at 3 years post-LYRS, and in none at 4 years post-LYRS. Cooper et al 6 followed up their initial20 patients (mean age, 56 years) for a mean of 30 months post-LYRS (range, 25 to 39 months) and noted persistent clinical and physiologic improvement. This is in contrast to our experience in the present study. Furthermore, in 90 patients who had bilateral stapled LYRS, we noted a mean increase in FEY 1 of 0.39:±::0.03 L (mean±SEM) at 3 to 6 months postsurge1y with subsequent decline in FEY 1 per year of0.255±0.057 L (mean±SEM) over 420:±:: 15 days (mean± SEM) follow-up time. 49 A weak correlation was noted between 3-and 6-month post-LYRS incremental gain in FEY 1 and subsequent decline in FEY 1 (r=0.292; p=0.003), and individual response could not be predicted. 49

Lung CT and Perfusion Lung Scans
All of the patients in the present study had a heterogeneous distribution of emphysema on lung CT with upper-third predominance and matching perfusion scan abnormalities. Wang et al 50 and Weder et al5 1 have reported modest lung CTSI and scintigraphic correlation of FEY 1 improvement with upper-lobe predominance (r=0.38; p=.001) 50 and heterogeneity (r=0.31; p=0.002) .so CONCLUSIONS The results in the present study extend our earlier experience 13 -15 and document the variable clinical and physiologic improvement in lung elastic recoil and expiratory airflow limitation observed 2 years after bilateral LYRS in 12 selected symptomatic patients with severe, generalized emphysema who had exhausted medical therapy. The increase in lung elastic recoil peaked at 6 months post-LYRS. We Clinical Investigations urge caution in the interpretation and extension of the data because of lack of a control group and the small number of patients studied.
ACKNOWLEDGMENT: Robeti Hyatt, MD, provided critical review of this manuscript and Andy Newsom, CPFT, RCP provided technical assistance.