clinical investigations Preoperative Severity of Emphysema Predictive of Improvement After Lung Volume Reduction Surgery* Use of CT Morphometry

Transcription

1 clinical investigations Preoperative Severity of Emphysema Predictive of Improvement After Lung Volume Reduction Surgery* Use of CT Morphometry Robert M. Rogers, MD, FCCP; Harvey O. Coxson, PhD; Frank C. Sciurba, MD, FCCP; Robert J. Keenan, MD; Kenneth P. Whittall, PhD; and James C. Hogg, MD, PhD Study objective: To determine how the volume and severity of emphysema measured by CT morphometry (CTM) before and after lung volume reduction surgery (LVRS) relates to the functional status of patients after LVRS. Design: A histologically validated CT algorithm was used to quantify the volume and severity of emphysema in 35 patients before and after LVRS: total lung volume (TLV), normal lung volume (< 6.0 ml gas per gram of tissue), volume of mild/moderate emphysema (ME; 6.0 to 10.2 ml gas per gram of tissue), volume of severe emphysema (> 10.2 ml gas per gram of tissue), surface area/volume (SA/V; meters squared per milliliter), and surface area (SA; meters squared). Outcome parameters included maximal cardiopulmonary exercise (CPX) performance in 21 patients and routine pulmonary function in all patients. We hypothesized that baseline CTM parameters predict response to LVRS and that the change in these parameters may offer insight into mechanisms of improvement. Patients and intervention: Thirty-five patients with severe emphysema who had successful LVRS. Results: The significant decrease in TLV following LVRS was entirely accounted for by a decrease in severe emphysema. The SA/V and the SA both increased significantly following LVRS. The change in maximal CPX in watts following surgery correlated significantly with baseline values of severe emphysema (r 0.60), which was collinear with TLV, and SA/V. The change in diffusing capacity of the lung for carbon monoxide revealed a significant positive linear relationship with preoperative severe emphysema (r 0.37) and a negative relationship with ME (r 0.37). Change in watts revealed a strong relationship with changes in severe emphysema (r 0.75) and weaker but significant relationships with change in TLV, ME, SA/V, and SA. Other measures of pulmonary function revealed significant albeit less dominant relationships with baseline CTM and change in these indexes. Conclusion: Using CTM, we have identified a close relationship between baseline severe emphysema, or change in severe emphysema, and the improvement in CPX after LVRS. These observations support a potential role of CTM in future clinical trials for predicting responders to LVRS and identifying mechanisms of improvement. (CHEST 2000; 118: ) Key words: CT; emphysema; exercise; lung volume reduction surgery Abbreviations: CPX cardiopulmonary exercise; CTM CT morphometry; Dlco diffusing capacity of the lung for carbon monoxide; FRC functional residual capacity; HU Hounsfield units; LVRS lung volume reduction surgery; ME mild/moderate emphysema; NL normal lung; RV residual volume; SA surface area; SA/V surface area of lung per total lung volume; TLC total lung capacity; TLV total lung volume Lung volume reduction surgery (LVRS), first described by Brantigan and Mueller 1 in 1957 and Brantigan et al 2 in 1961, has had resurgence. Despite *From the Division of Pulmonary, Allergy and Critical Care Medicine, Department of Medicine (Drs. Sciurba and Rogers), and the Division of Thoracic Surgery, Department of Surgery (Dr. Keenan) at the University of Pittsburgh Medical Center and School of Medicine, Pittsburgh, PA; and the University of British Columbia Pulmonary Research Laboratory (Drs. Coxson, Whittall, and Hogg), St. Paul s Hospital, Vancouver, Canada. documented improvements in mean physiologic values following the procedure, 3,4 it remains unclear why some patients have a significant improvement in Supported by a grant from the George H. Love research fund. Manuscript received October 25, 1999; revision accepted May 5, Correspondence to: Robert M. Rogers, MD, FCCP, Division of Pulmonary, Allergy and Critical Care Medicine, 440 Scaife Hall, 3550 Terrace St, Pittsburgh, PA; rogersrm@msx.upmc.edu 1240 Clinical Investigations

2 their major clinical symptoms as a result of LVRS, and others do not. We suggest that some of the variability may reflect our inability to quantify the severity, extent, and location of emphysema prior to surgery. For editorial comment see page 1231 One of the more important tools in assessing the extent of emphysema in LVRS candidates is CT. The CT images are conventionally interpreted using a subjective visual analysis of the image. This visual approach results in inconsistent interpretation, with interobserver agreement of 0.6 and a tendency to overestimate the amount of emphysema when compared to pathologic measurements. 5 Others have shown that by using lung attenuation values as a quantitative evaluation, they had moderately strong correlation with preoperative physiology and outcome. 6 The purpose of this study is to extend these observations by utilizing a recently described quantitative method of analysis that we will refer to as CT morphometry (CTM). 7 CTM is a quantitative analytic technique that measures lung volume, volume of emphysema, surface area (SA; meters squared), and SA per total lung volume (TLV) (SA/V; meters squared per milliliter) from CT density values. This technique was validated in resected lungs of patients with varying degrees of emphysema using standard histologic analyses. In this study, we used CTM to analyze the CT scans from 35 consecutive patients who underwent LVRS. Physiologic measures before and 3 months after LVRS were correlated with the volume and severity of preoperative LVRS emphysema, as identified by CTM, to predict the physiologic and clinical response of this group of patients after LVRS. Our results strongly suggest that CTM analysis yields valuable information about lung structure in terms of lung volume, volume of emphysema, and SA, and these values may be useful in preoperative evaluation and postoperative follow-up of patients undergoing LVRS. Patient Selection Materials and Methods The procedures used in this study were approved by the Institutional Review Board of the University of Pittsburgh Medical Center. All of the patients signed informed consent forms that allowed the use of physiologic data, CT scans, and the surgically resected tissue. Patient selection criteria have been published elsewhere. 3 CT was obtained on all patients and was used to identify surgically accessible emphysematous lesions and to exclude patients with diffuse disease using standard visual assessment. The CTM analysis was performed retrospectively. The patients underwent bilateral LVRS by video-assisted approach (n 29) or median sternotomy (n 6) using previously reported techniques. 4,8 10 This report reflects an analysis of data collected from 35 consecutive patients who successfully completed the baseline and 3 months of follow-up radiologic and physiologic testing necessary for our analysis (group T). Twentyone of these patients had formal cardiopulmonary exercise (CPX) testing (group E). Fourteen patients were excluded from CPX testing because they could not maintain arterial saturations, measured by pulse oximetry of 84% for 3 min of unloaded bicycle pedaling. The patients included in this study were among the 82 patients who underwent bilateral LVRS between October 21, 1994, and February 1, Of the 82 patients, 7 died and the others had incompleted data sets at 3 months. The 35 patients in this analysis were not preselected on the basis of radiologic anatomy or functional status but on availability of data for analysis. Pulmonary Function Studies Spirometry and lung volumes were measured preoperatively and postoperatively using previously described standard techniques and normal reference values. 3 CPX Testing Group E patients performed incremental, symptom-limited CPX testing on an electronically braked cycle ergometer (model KEM III; Mijnhardt BV; Bunnik, Holland). This ergometer produces a resistance equivalent to 14 W during unloaded pedaling. The protocol utilized a ramp increment of 4 W or 8 W following 3 min of unloaded pedaling. Maximal watts were recorded when pedaling frequency fell and remained 40 revolutions/min. Quantitative CT Analysis CTM The subjects in the study received a conventional, noncontrast CT scan (10-mm thick axial slices) on a GE 9800 Highlight Advantage CT scanner (General Electric Medical Systems; Milwaukee, WI) approximately 1 week prior to surgery. All scans were performed with the subject supine during end-inspiratory breath holding. The image data were transferred to a Silicon Graphics Indy Workstation (Silicon Graphics; Mountainview, CA). CTM was performed using a recently described technique, 7 whereby the TLV was calculated by summing the lung voxels on the CT scan, and multiplying by the voxel dimension. A voxel (volume element) is defined as the individual picture element (pixel) of a CT image, which has dimensions in the x and y plane, and includes the CT slice thickness as the third dimension (z axis). The lung volume was also partitioned into three compartments occupied by severe and mild/moderate emphysematous lesions, as well as normal lung (NL). Briefly, this was performed by converting the radiograph attenuation values, measured in Hounsfield units (HU), into a measurement of the volume of gas per gram of tissue (milliliters per gram) and applying two cutoff values. 7 The first cutoff is at 10.2 ml/g, which corresponds to 910 HU. 11 It has been shown that lung inflated 10.2 ml/g ( 910 HU) represents emphysematous lesions 5 mm in diameter 7,12 and is referred to as severe emphysema. The second cutoff is 6.0 ml/g, which corresponds to 856 HU. Lung inflated between this cutoff and the severe emphysema cutoff (ie, 6.0 to 10.2 ml/g) has been shown to represent emphysematous lesions 5mm in diameter, referred to as mild/moderate emphysema (ME). The remainder of the lung volume inflated 6.0 ml/g was considered to be representative of NL parenchyma. 7 The volume of these three categories (severe emphysema, ME, and NL) was calculated by summing the number of CT voxels identified by each cutoff and CHEST / 118 / 5/ NOVEMBER,

3 multiplying by the voxel dimension. It should be noted that, while CT images obtained using narrow beam collimation (1 mm) and high-resolution reconstruction algorithms produce images that allow the visualization of emphysema, thicker sections and lower-resolution algorithms allow a more reliable discrimination of different densities within the lung. 13 Secondly, the SA of the lung parenchyma and SA/V were estimated using a prediction equation derived by correlating measurements of radiograph attenuation values with quantitative histology. 7 Statistical Analysis Descriptive data were analyzed using a paired two-tailed t test. The relationship between the changes in physiologic and functional indexes as a function of preoperative CTM variables following LVRS was analyzed first using descriptive simple regression analysis, followed by a stepwise forward multiple regression analysis. The primary dependent variable chosen was the change in maximal watts (group E) because it is an objective functional parameter that integrates the myriad of potential physiologic changes following LVRS. 14 Secondary dependent variables used in the analysis were change in FEV 1, FVC, residual volume (RV), total lung capacity (TLC), and diffusing capacity of the lung for carbon monoxide (Dlco); these were analyzed in group T. Independent CTM parameters were preoperative TLV, volume of severe emphysema, volume of ME, volume of NL, lung SA, and SA/V. A similar approach was used to compare the relationship between the changes in physiologic and functional indexes as a function of changes in CT variables following LVRS. Because we considered that the change in Dlco, in particular, might be related more to the preoperative proportion of the TLV occupied by our morphologic categories (ie, severe emphysema/tlv, ME/TLV, NL/TLV) rather than the absolute volume, we performed linear regression analysis to compare these volumeadjusted CT parameters to the change in Dlco. A p value of 0.05 was considered significant. Results Patient Characteristics and Physiologic Response to LVRS Patient characteristics and physiologic indexes before and after LVRS for the total group of 35 patients (group T) are shown in Table 1. At 3 months after Figure 1. Categorization of emphysema before and after LVRS. These histograms represent the preoperative (Pre) and postoperative (Post) volumes of the TLV, severe emphysema (SE), ME, and NL measured from CT scans. Note that most of the reduction in TLV is due to decrease in severe emphysema. surgery, group T demonstrated statistically significant improvement in FVC, FEV 1, RV, TLC, and functional residual capacity (FRC), but no significant changes were found in the Dlco. In group E, physiologic variables responded similarly to group T following surgery. The characteristics and maximal cycle ergometry watts before and after LVRS of the subgroup of 21 patients capable of performing CPX testing (group E) are shown in Table 1. Maximal watts significantly increased following surgery. Changes in CTM Indexes Following LVRS TLV, measured by CTM, decreased significantly following LVRS (Fig 1 and Table 2) and was similar to the decrease in TLC determined using plethys- Table 1 Descriptive Physiology of Study Groups Before and After LVRS* Group T Group E Variables Baseline 3-mo Postoperative Baseline 3-mo Postoperative FVC, L FEV 1,L RV, L TLC, L FRC, L Dlco, ml/min/mm Hg Watts *Group T (n 35; age, years; 22 men and 13 women); group E (n 21; age, years; 16 men and 5 women). p p p Clinical Investigations

4 Table 2 CT Parameters Before and After LVRS (n 35)* Parameters Baseline 3 mo Postoperative p Values TLV, ml 6, , , , Vol NL, ml/g 1, , Vol ME, ml/g 1, , Vol SE, ml/g 3, , , SA/V, m 2 /ml SA, m *Data are presented as mean SD. mography (888 ml vs 1,004 ml, respectively). The change in TLV was entirely due to a decrease in the size of the compartment representing severe emphysema; the changes in NL and ME were not statistically significant, although the mean ME value tended to decrease and the NL tended to increase (Fig 1). Predicting Change in Functional Status and Physiology as a Function of Baseline CTM Indexes The change in our primary outcome parameter (ie, watts) was predicted by baseline values of TLV, severe emphysema, and SA/V (Table 3). Multiple regression analysis determined significant collinearity among TLV, SA/V, and severe emphysema, such that inclusion of severe emphysema in the model completely accounted for the effect of the other two variables. The relationship between severe emphysema and the improvement in CPX is shown graphically in Figure 2. As can be seen in Figure 2, the higher the volume of severe emphysema preoperatively, the higher the change in watts postoperatively. The baseline CT parameters correlated with the change in several secondary parameters (ie, FVC, FEV 1, RV, TLC, Dlco) analyzed in group T. Many showed a significant statistical relationship, but all R values were 0.46 (Table 4). The relationship between baseline severe emphysema and the change Table 3 Change in Watts as a Function of Baseline Parameters* Baseline Variables Partial Correlation p Value TLV, L NL ME SE SA/V, m 2 /ml SA, m *n 21; this relationship is expressed graphically in Figure 2. See Figure 1 legend for abbreviation. Figure 2. Change in watts vs preoperative volume of severe emphysema. The individual data points and the linear regression line are shown (r 0.595). The greater the amount of severe emphysema preoperatively, the greater the improvement in exercise performance as a result of LVRS. See Figure 1 legend for abbreviation. in FEV 1 is significantly weaker (r 0.366) than that observed with change in watts shown in Figure 2. Of interest, the change in Dlco revealed a significant positive linear relationship with preoperative severe emphysema (r 0.37) and a negative relationship with ME (r 0.37). Logistic multiple regression analysis determined that both parameters independently contribute to the model. The linear relationship between change in Dlco with preoperative severe emphysema adjusted for TLV (SE/TLV; p 0.002) was greater than with the absolute severe emphysema (Fig 3). Similarly, simple regression analysis determined a strong negative relationship between change in Dlco and preoperative ME/TLV (r 0.58, p ; Fig 4). The graph of the change in Dlco vs preoperative ME/TLV (Fig 4) suggests that if this ratio is 0.3, the Dlco does not change or increases, while 0.3 there is a tendency for the Dlco to decrease, suggesting that capillaries are being removed. Change in Functional Status and Physiology as a Function of Changes in CT Anatomy Analysis of the primary dependent variable (change in watts) in group E revealed a strong relationship with change in severe emphysema, and weaker but significant relationships with change in TLV, ME, SA/V, and SA (Table 5). Multiple regression analysis determined collinearity between all significant variables, such that inclusion of change in severe emphysema in the model completely accounted for the effect of the other variables. Figure CHEST / 118 / 5/ NOVEMBER,

5 Table 4 Partial Coefficients of Change in Physiologic Parameters as a Function of Baseline CT Parameters* Variables TLV NL ME SE SA/V SA FVC, L p 0.04 p 0.02 p 0.05 FEV 1,L p 0.03 p 0.01 RV, L p 0.03 p 0.02 p 0.03 TLC, L Dlco, ml/min/mm Hg p 0.03 FRC, L p 0.02 p 0.02 *See Figure 1 legend for abbreviation. Dominant independent variable. Significant but codependent. 5 shows the relationship between change in watts as a function of change in severe emphysema. Results of secondary outcome variables (change in FVC, FEV 1, RV, TLC, Dlco) were derived from the total group T. The change in FVC reveals no significant relationship with the change in any CT parameters. Changes in FEV 1, RV, and TLC revealed significant linear relationships only with change in severe emphysema (R for FEV 1, 0.54; R for RV, 0.46; and R for TLC, 0.44). The change in Dlco reveals no significant relationship with change in any CT parameter in absolute volumes or when adjusted for TLV. Discussion LVRS for end-stage emphysema was introduced 40 years ago. 1 The physiologic basis for the surgery was sound, 15 and while it was shown to improve airway conductance postoperatively, 16 this change was transient 17 and the mortality was high. 2 Several articles have rekindled interest in the procedure and report improvement in physiologic parameters such as FEV 1, FVC, FRC, RV, TLC, 4,10,18 and lung elastic recoil, 3,19,20 as well as lower surgical operative mortality. Other fundamental physiologic changes identified after LVRS, beyond the improvements in lung elastic recoil, include improvement in diaphragmatic strength, decrease in central respiratory drive, and changes in pulmonary vascular resistance and gas exchange. 3,26 The rationale for the use of LVRS is based on two major concepts: (1) very severe, hyperinflated emphysematous lung is essentially nonfunctional and can be removed with impunity, and (2) the remaining lung subsequently expands in the vacated thoracic Figure 3. Change in Dlco vs volume of severe emphysema/ TLV preoperatively (SE/TLV Pre). The individual data points and the linear regression line are shown (r 0.34). The higher the volume of severe emphysema/tlv preoperatively, the more the Dlco improves; the lower the value for severe emphysema/ TLV, the more likely the Dlco will decrease. Figure 4. Change in Dlco vs volume of ME/TLV preoperatively (ME/TLV Pre). The individual data points and the linear regression line are shown (r 0.58). The graph shows that the higher the value of ME/TLV preoperatively, the more likely it was for the Dlco to decrease Clinical Investigations

6 Table 5 Change in Watts as a Function of Change in CT Parameters (n 21)* Variables Partial Correlation p Value in TLV, L in NL in ME in SE in SA/V, m 2 /ml in SA, m *See Figure 1 legend for abbreviation. cavity, resulting in increased elastic recoil and the consequent improvement in lung mechanics. As noted above, 23 there is recent evidence that LVRS improves the function of the respiratory muscles and decreases respiratory central drive. A major problem in assessing candidates for LVRS is the characterization of their emphysema, which is a pathologic diagnosis that can only be suspected clinically by using physiologic surrogates such as the Dlco 27 and pulmonary mechanical indicators of hyperinflation. The chest roentgenogram and CT of the chest, however, give us visual evidence of the pathologic findings of emphysema. All these studies can assess the presence or absence of emphysema, but are inadequate to precisely localize or quantify the emphysematous lesions in the lung. Early studies on LVRS combined the chest roentgenogram and 131 I macroaggregated lung scan to locate the very poorly perfused areas of the lung 16,28 ; however, this failed to give a three-dimensional estimate of the Figure 5. Change in watts vs change in severe emphysema. The individual data points and the linear regression line are shown (r 0.749). The graph shows that with a decrease in severe emphysema, there was a greater tendency for an increase in watts. lung. A major advance in emphysema research was the advent of quantitative CT scans using either a density mask 11 or a density histogram analysis. 29,30 These techniques, combined with the spatial resolution of the CT, have provided the investigator with truly quantitative three-dimensional data on the lung structure. Further, CTM allows the quantification of lung parenchymal features, such as SA, which were previously only available after removal of the organ from the body. Furthermore, our CTM analysis extends previous quantitative CT work, by translating the more obscure density measurements (in our experience, a term not easily understood by nonradiologists) into clinical and pathophysiologically meaningful indexes. The purpose of this investigation is to utilize a newly described CTM technique 7 to retrospectively analyze the volume and severity of emphysematous lesions in the preoperative CT and the subsequent changes following a clinically successful LVRS. We compared these CTM measurements to the changes in pulmonary function in 35 patients, and formal CPX testing in 21 patients. Because of the complexity of the pulmonary physiologic response, we chose to use CPX as an objective measure of improvement because it is an integration of the entire cardiopulmonary system. 14 These patients were not selected on the basis of clinical improvement, but merely on the basis of having completed the above studies. Our patients responses to LVRS are similar to those of other published reports with statistically significant improvement of FVC, FEV 1, RV, FRC, TLC, and variable response of the Dlco. 9,10,18,20 We confirm the results of others 31,32 who have shown that there is a reduction of the TLV after LVRS, both by CT and plethysmography measurements. 32 We have further shown that this decrease is predominantly due to reduction in the lung volume that is considered to have the worst emphysematous lesions, ie, inflated 10.2 ml/g, 7 which we have called severe emphysema. Because the surgical technique removes not only severely diseased lung, but also mild/moderate disease and normal lung, the fact that there is no change in the latter two categories suggests that there is some reexpansion of the remaining lung. This is also demonstrated by the increase in both SA and SA/V of the lung parenchyma. We speculate that, following LVRS, the remaining lung expands and may change categories from either the mild/moderate to the severe range, or from the normal to the mild/moderate range. The increase in SA and SA/V is probably due to this reexpansion of underinflated alveoli, particularly in the NL compartment, thereby increasing the SA of the lung. The clinical importance of the CTM technique is CHEST / 118 / 5/ NOVEMBER,

7 reflected in the close relationship between baseline emphysema volume, or change in emphysema volume, and the improvement in functional performance demonstrated with change in CPX performance. These observations support a potential role of CTM in predicting responders to surgery and in identifying mechanisms of improvement. When one examines the subgroup of patients on whom we had formal CPX testing (Table 3), it is clear that the baseline values for TLV, SA/V, and severe emphysema in the preoperative CT scan predict improvement in CPX response, although the former two parameters are codependent on the severe emphysema value. This suggests that the preoperative volume of severe emphysema has predictive value with respect to improvement in CPX response following LVRS. A low Dlco has been used as the standard physiologic marker for emphysema when associated with reduced expiratory flow and an increase in RV. 27 This relationship has been viewed as a reflection of capillary destruction by the emphysematous process. However, a low Dlco could also be a result of compression of pulmonary vessels or cardiopulmonary interaction. When we examined change in Dlco, it improved in 15 patients, decreased in 9 patients, and did not change in 11 patients (Fig 4), although the preoperative LVRS and postoperative LVRS mean values were not significantly different. The magnitude of change of Dlco showed a strong positive relationship with the preoperative volume of severe emphysema (Fig 3) and a strong, independent, negative relationship with the preoperative volume of ME (Fig 5). We propose that these findings may reflect a greater removal of functional capillaries in the latter group and the removal of fewer capillaries in regions of poorly perfused severe emphysema. Pulmonary capillary volume can theoretically increase or decrease following LVRS. What determines an increase or decrease in capillary volume is the balance between the changes in regional lung mechanics and physical removal of capillaries. Our data support the concept that capillary volume can be increased in patients who have greater proportions of severe emphysema lung surgically removed. These data are consistent with recent report of weight gain after LVRS that was statistically significantly related to improvement in Dlco and not measurement of mechanics at 12 months. 33 The physiologic data on patients undergoing LVRS presented in this article are very similar to those reported by many other large centers, ie, statistically significant improvement in the mean values for standard clinical physiologic parameters except for Dlco. Hence, while these improved mean values are statistically significant, it should be noted that there is significant variation from individual to individual in the magnitude of the change in each of these parameters. Thus far, no physiologic parameter has been found that predicts successful clinical outcome of LVRS, suggesting to us that the variability in response may rest in the amount and distribution of severe emphysema. A recent study 20 of patients who had LVRS who were followed up to 4 years confirmed the early reports 16,17 that the clinical and physiologic improvement is not sustained in the majority of patients (less then one third of their patients). These authors showed that the patients who had a sustained improvement had a statistically higher FVC and vital capacity at baseline. We would offer as an additional hypothesis that it may be related to the volume and distribution of severe emphysema. In summary, CTM allows an objective assessment of lung structure and shows that the preoperative volume of emphysema predicts which patients will have the greatest improvement in cardiopulmonary function following LVRS. Further, it has the added advantage that well-trained technicians can perform this analysis across multiple centers, thereby increasing the sample size and minimizing any institutional bias. It is the authors hypothesis that CTM will give us better insight into the preoperative and postoperative abnormalities of the lungs of patients who have LVRS, and will allow us to predict, with quantitative methods, which patients will benefit from LVRS. Because the number of patients in this study is small and we do not have a control group, we view this study as hypothesis generating, and suggest that this analysis could be applied to a larger group of subjects, including patients who were rejected for surgery and those who do not survive the procedure. In addition, the appeal of CTM includes the following: (1) that it can be utilized on a standard personal computer, (2) that it can be analyzed across centers without the need for subjective visual analysis, (3) that it easily lends itself to blinding the person conducting the analysis, and (4) that future development could potentially develop a pathologic map to guide the surgery. ACKNOWLEDGMENT: The authors wish to express our sincere thanks to Laurie Silfies, who coordinated the data analysis, Claude Lavallée who prepared the manuscript, William Slivka, who conducted the physiologic testing, William Bradford Rogers, who helped in many important ways to get this study completed, and a special thanks to Dr. Carl Fuhrman, for his thoughtful suggestions. References 1 Brantigan O, Mueller E. Surgical treatment of pulmonary emphysema. Am Surg 1957; 23: Brantigan OC, Kress M, Mueller E. The surgical approach to 1246 Clinical Investigations

8 pulmonary emphysema. Dis Chest 1961; 39: Sciurba FC, Rogers RM, Keenan RJ, et al. Improvement in pulmonary function and elastic recoil after lung-reduction surgery for diffuse emphysema. N Engl J Med 1996; 334: Cooper JD, Trulock EP, Triantafillou AN, et al. Bilateral pneumectomy (volume reduction) for chronic obstructive pulmonary disease. J Thorac Cardiovasc Surg 1995; 109: Bankier A, De Maetelaer V, Keyzer C, et al. Pulmonary emphysema: subjective visual grading versus objective quantification with macroscopic morphometry and thin section CT densitometry. Radiology 1999; 211: Gierada DS, Slone RM, Bae KT, et al. Pulmonary emphysema: comparison of preoperative quantitative CT and physiologic index values with clinical outcome after lung-volume reduction surgery. Radiology 1997; 205: Coxson HO, Rogers RM, Whittall KP, et al. A quantification of the lung surface area in emphysema using computed tomography. Am J Respir Crit Care Med 1999; 159: Cooper JD, Patterson GA. Lung-volume reduction surgery for severe emphysema. Chest Surg Clin North Am 1995; 5: Cooper JD, Patterson GA, Sundaresan RS, et al. Results of 150 consecutive bilateral lung volume reduction procedures in patients with severe emphysema. J Thorac Cardiovasc Surg 1996; 112: Keenan RJ, Landreneau RJ, Sciurba FC, et al. Unilateral thoracoscopic surgical approach for diffuse emphysema. J Thorac Cardiovasc Surg 1996; 111: Muller NL, Staples CA, Miller RR, et al. Density mask. An objective method to quantitate emphysema using computed tomography. Chest 1988; 94: Miller RR, Muller NL, Vedal S, et al. Limitations of computed tomography in the assessment of emphysema. Am Rev Respir Dis 1989; 139: Kemerink GJ, Kruize HH, Lamers RJ, et al. Density resolution in quantitative computed tomography of foam and lung. Med Phys 1996; 23: Sciurba FC. Early and long-term functional outcomes following lung volume reduction surgery. Clin Chest Med 1997; 18: Rogers RM, Sciurba FC, Keenan RJ. Lung reduction surgery in chronic obstructive lung disease. Med Clin North Am 1996; 80: Rogers RM, DuBois AB, Blakemore WS. Effect of removal of bullae on airway conductance and conductance volume ratios. J Clin Invest 1968; 47: Rogers RM. Stress-relaxation in pulmonary emphysema and its relation to airway conductance. Am Rev Respir Dis 1970; 101: McKenna RJ Jr, Brenner M, Gelb AF, et al. A randomized, prospective trial of stapled lung reduction versus laser bullectomy for diffuse emphysema [see comments]. J Thorac Cardiovasc Surg 1996; 111: ; discussion Gelb AF, McKenna RJ Jr, Brenner M, et al. Contribution of lung and chest wall mechanics following emphysema resection. Chest 1996; 110: Gelb A, McKenna RJ Jr, Brenner M, et al. Lung function 4 years after lung volume reduction surgery for emphysema. Chest 1999; 116: Criner G, Cordova FC, Leyenson V, et al. Effect of lung volume reduction surgery on diaphragm strength. Am J Respir Crit Care Med 1998; 157: Laghi F, Jubran A, Topeli A, et al. Effect of lung volume reduction surgery on neuromechanical coupling of the diaphragm. Am J Respir Crit Care Med 1998; 157: Lahrmann H, Wild M, Wanke T, et al. Neural drive to the diaphragm after lung volume reduction surgery. Chest 1999; 116: Celli BR. Pulmonary rehabilitation for patients with advanced lung disease. Clin Chest Med 1997; 18: Celli BR, Montes de Oca M, Mendez R, et al. Lung reduction surgery in severe COPD decreases central drive and ventilatory response to CO 2. Chest 1997; 112: Weg IL, Rossoff L, McKeon K, et al. Development of pulmonary hypertension after lung volume reduction surgery. Am J Respir Crit Care Med 1999; 159: Gelb AF, Gold WM, Nadel JA. Mechanisms limiting airflow in bullous lung disease. Am Rev Respir Dis 1973; 107: Rogers RM, Kuhl DE, Hyde RW, et al. Measurement of the vital capacity and perfusion of each lung by fluoroscopy and macroaggregated albumin lung scanning: an alternative to bronchospirometry for evaluating individual lung function. Ann Intern Med 1967; 67: Hayhurst MD, MacNee W, Flenley DC, et al. Diagnosis of pulmonary emphysema by computerised tomography. Lancet 1984; 2: Gould GA, MacNee W, McLean A, et al. CT measurements of lung density in life can quantitate distal airspace enlargement: an essential defining feature of human emphysema. Am Rev Respir Dis 1988; 137: Bae KT, Slone RM, Gierada DS, et al. Patients with emphysema: quantitative CT analysis before and after lung volume reduction surgery; work in progress. Radiology 1997; 203: Becker MD, Berkmen YM, Austin JH, et al. Lung volumes before and after lung volume reduction surgery: quantitative CT analysis. Am J Respir Crit Care Med 1998; 157: Christensen P, Paine R, Curtis J, et al. Weight gain after lung volume reduction surgery is not correlated with improvement in pulmonary mechanics. Chest 1999; 116: CHEST / 118 / 5/ NOVEMBER,