INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It is well known that patients with severe bullous emphysema significantly improve exercise tolerance after excision of the bullae, the mechanism of such improvement being the expansion of underlying compressed lung. In one study (1) even an improvement of pulmonary hemodynamics after the procedure was reported. In contrast, in the case of reduction pneumoplasty for giant bullous emphysema, the basis of lung volume reduction surgery (LVRS) is to withdraw functionless yet nonbullous lung tissue in order to improve the lung's mechanical properties by increasing its elastic recoil (2) and to increase diaphragmatic motion, thus lowering the work of breathing (5). Improvement in quality of life, dyspnea, pulmonary function, and exercise tolerance after LVRS has been reported recently by several investigators (2). Nevertheless, to our knowledge, no study has investigated the outcome of pulmonary hemodynamics at rest and during exercise after LVRS. Indeed, during exercise, pulmonary hypertension (PH) is a common feature in patients with chronic obstructive pulmonary disease (COPD) of the emphysematous type, whereas mean pulmonary artery pressure () often remains normal at rest (15). The increase in exercise capacity after LVRS is obvious in most studies. Nevertheless, the reduction of lung volume may result in an increase of , as is the case in pneumonectomy. This is seen particularly during exercise because of reduction of the vascular bed. The aim of our study was to assess the course of pulmonary hemodynamics at rest and also during exercise after the LVRS procedure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Twelve patients (11 men and one woman 34 to 69 yr of age; median, 57 yr) underwent LVRS. One patient died 3 mo after surgery from myocardial infarction; one was tracheotomized after LVRS and also required cholecystectomy 3 wk after surgery; one (the woman) refused right heart catheterization (RHC) after LVRS, yet her dyspnea and incremental exercise test capacity had greatly improved. The nine patients (34 to 69 yr of age; median, 58 yr) all having RHC before and after LVRS were included in our study. All of them complained of severe dyspnea upon minimal exertion and, less commonly, of cough and sputum. The diagnosis of emphysema was based on physical examination, chest radiographs, high resolution CT scan, lung perfusion scan, and pulmonary function tests. Three patients suffered from ₁-antitrypsin deficiency. Four patients were receiving long-term oxygen therapy (LTO): two of them had only nocturnal LTO, and two had LTO for 16 of 24 h (diurnal oxygenotherapy mainly for exercise purposes). All patients were receiving optimal medical treatment with inhaled steroids and bronchodilators. Lung function was evaluated 1 to 2 wk before surgery and again 3 to 12 mo after LVRS (mean, 4.5 mo after the procedure).

Dyspnea was graded using a five-level scale (16) ranging from 0 (no dyspnea) to 5 (dyspnea at rest). Spirometry was performed to measure VC, FEV₁, and FRC using the helium dilution technique. TLC and residual volume (RV) were then calculated. The reference values were those of the European Respiratory Society (17). In six patients thoracic gas volume (Vtg) could be assessed by means of a total body plethysmograph (Bodyscope; Ganshorn Medizin Electronic GMBH, Münnerstadt/Niderlauer, Germany).

Lung volume diffusion capacity was measured by the single-breath method in all patients and expressed as an absolute value (DL_CO). The reference values were those of our laboratory (18).

In all patients we simultaneously performed gasometric and hemodynamic investigations at rest and during steady-state exercise on an electrically braked bicycle; the load was set at 30 W, or less if dyspnea was major, and the duration of exercise was 5 to 6 min (the load was the same for a given patient before and after LVRS). RHC is performed routinely in our laboratory as a part of the functional investigation for COPD in patients who are candidates for LVRS. Pulmonary variables are usually controlled after 3 to 6 mo, provided that an informed consent has been obtained orally from the patient and was approved by our institutional review board. RHC was performed in the morning, without premedication 2 h after a light breakfast, the patient being in the supine position. We used small-diameter Grandjean catheters (19) (4F; Plastimed, Saint-Leu-La-Forêt, France). The catheter was inserted through a brachial vein, if possible, or a femoral vein, if necessary. We measured and pulmonary artery diastolic pressure (Pd) (Figure 1). Arterial blood was sampled in room air using a Cournand needle inserted into a humeral artery during heart catheterization. Minute ventilation (E), O₂ consumption (O₂) and CO₂ production (CO₂) were measured by means of a closed ventilation system (Medisoft Partn'air 5400; Dinant, Belgium). Alveolar PO₂ was calculated from the equation for alveolar air (20). Cardiac output was measured using the Fick's principle applied to oxygen.

View larger version (42K): [in this window] [in a new window]   Figure 1.   Respiratory swings of the diastolic pressure (Pd) obtained by measuring the difference at rest and during exercise between the highest and the lowest value of the diastolic pressure, for values recorded through three or four respiratory cycles.

Maximal incremental exercise testing (using a 1-min incremental cycle exercise protocol) was performed on a bicycle ergometer (Type 1000 S; Medifit, Utrecht, The Netherlands and Medisoft Partn'air 5400) another day on seven patients. It could not be performed on one patient because of the severity of dyspnea. A second patient did not tolerate the mouthpiece or the facial mask for measuring the ventilatory variables during maximal exercise test. The reference values were those of Hansen and colleagues (21).

Statistics

Comparison of the variables before and after LVRS was made using Wilcoxon's t test for small and paired series. The correlation between two parameters was assessed by calculating Pearson's correlation coefficient. Significance was set at the 5% level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The surgical procedure was bilateral in all patients and consisted in resecting 20 to 30% of each lung.

The results of the dyspnea score, lung function data, exercise testing, blood gas analysis, and pulmonary hemodynamic data before and after LVRS are summarized in Tables 1234567. Dyspnea improved significantly in all but one patient, in whom it remained stable (Patient 6). There was an overall increase in FEV₁ from 705 to 1,005 ml (p < 0.05). In only one patient (Patient 6) was there a decrease (200 ml). RV decreased from a median of 3,840 to 3,140 ml. Median TLC decreased from 7,060 ml before surgery to 6,450 ml after surgery. Vtg was measured in six patients before and after LVRS and decreased in all of them (median value, from 6,560 to 4,940 ml). DL_CO did not change after LVRS. Overall maximal oxygen uptake at peak exercise increased significantly (p < 0.05) except in Patient 6. 

Pa_O2 and Pa_CO2 did not change after LVRS either at rest or during exercise. Neither did change at rest or during exercise. One patient could not perform exercise testing during RHC before LVRS because of his severe dyspnea. There was a significant overall decrease of Pd at rest (p < 0.01) and during exercise (p < 0.05). We also observed a significant correlation between the change in resting ( before minus after LVRS) and the change in resting Pd (r = 0.73, p < 0.03) (Figure 2), and also between the change in exercising and the change in resting Pd (r = 0.80, p < 0.02) (Figure 3). A significant correlation was found between the change in exercising and the change in exercising Pa_O2 (r = 0.70, p < 0.05) (Figure 4), as well as between the change in exercising and exercising Pa_CO2 (r = 0.76, p < 0.03). Furthermore, there were good correlations between the change in FEV₁ and the change in Pa_O2 at rest and during exercise (r = 0.90, p < 0.01, and r = 0.75, p < 0.05, respectively), and also between the change in VC and in Pa_O2 at rest (r = 0.81, p < 0.01).

View larger version (7K): [in this window] [in a new window]   Figure 2.   Correlation between the change (before minus after LVRS) in the respiratory swings of the pulmonary diastolic pressure at rest ( Pdr) and the change in mean pulmonary artery pressure at rest (PAPr).

View larger version (8K): [in this window] [in a new window]   Figure 3.   Correlation between the change (before minus after LVRS) in the respiratory swings of the pulmonary diastolic pressure at rest ( Pdr) and the change in mean pulmonary artery pressure during exercise (PAPe).

View larger version (9K): [in this window] [in a new window]   Figure 4.   Correlation between the change (before minus after LVRS) in the exercising PaO2 (PaO2e) and the change in mean pulmonary artery pressure during exercise (PAPe).

With regard to LTO, two patients did not require further treatment after LVRS (one with nocturnal therapy and one who had oxygen for 16 of 24 h); one patient continued on nocturnal oxygen therapy; another one required only nocturnal therapy instead of 16 h/d.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

LVRS has become an extended surgery for emphysema for improving the dyspnea of severely affected patients (2). Most of the mechanical abnormalities observed in emphysema are due to a loss of the lung elastic recoil that results in decreased expiratory airflow (because of a decreased driving pressure) and collapse of the peripheral airways. When LVRS is performed, the resection of the most distended areas of the lung reduces hyperinflation, thus enabling better functioning of the diaphragm and other respiratory muscles (5, 6, 8). As a consequence, the work of breathing is decreased (7). The decrease in hyperinflation is associated with an increase of the elastic recoil of the lung with enhancement of expiratory airflow. Furthermore, during exercise, there is less dynamic hyperinflation (2, 8). From a clinical point of view, LVRS improves exercise tolerance with less dyspnea, hence increasing the quality of life scores (9). However, lung resection, even with regard to distended tissue with poor vascularization, may further reduce the vascular bed, and therefore favor PH in emphysematous patients, especially during exercise. Indeed, emphysematous patients have no or only mild PH at rest (15, 22- 24), whereas moderate to severe PH is observed during exercise in most cases (15).

In our study, median did not change at rest after LVRS. We observed a decrease of of  5 mm Hg in two patients only, and the change was less than 5 mm Hg in the others. Opposing results have been observed by Weg and colleagues (25) at rest with the development of PH in five of six patients after LVRS (mean increase of 7 mm Hg). Contrary to Weg and colleagues (25), Sciurba and colleagues (2) have reported an improvement in right ventricular systolic function at rest, as estimated by two-dimensional echocardiography after LVRS. These investigators explained their results by a reduction in pulmonary vascular resistance by capillary recruitment, which would be secondary to less compression of the vessels by hyperinflated lung areas or by the tethering of extraalveolar vessels because of the improvement of elastic recoil. In our opinion, LVRS may have two consequences on pulmonary hemodynamics in emphysematous patients. On the one hand, the resection of lung tissue could reduce the vascular bed and increase pulmonary vascular resistance, but on the other hand, better mechanical properties of the respiratory system with less hyperinflation and improvement of elastic recoil may counterbalance the effect of a reduced vascular bed as suggested by Sciurba and colleagues (2). We also suggest that smaller intrathoracic pressure variations could play a role in inducing less functional compression of the pulmonary vascular bed (15, 26), especially during exercise. Large intrathoracic pressure swings are generated by emphysematous patients to overcome the effects of the loss of lung elastic recoil. These intrathoracic pressure variations, caused by an increased airway resistance, may partly explain PH in patients with COPD (27) with an increase in the pulmonary vascular resistance during expiration but no corresponding decrease in the pulmonary vascular resistance during inspiration (the pulmonary resistance vessels are maximally distended in the recumbent position in emphysematous patients and fail to distend with an increasing transmural pressure during inspiration). In the present study we observed a significant decrease of the respiratory swings of the diastolic pressure at rest and during exercise. Moreover, we observed a good correlation between the change in after LVRS at rest, as well as during exercise, and the change in the resting Pd (r = 0.73, p < 0.03 and r = 0.80, p < 0.02, respectively). The decrease of the respiratory swings of the pulmonary diastolic pressure, which reflects the intrathoracic pressure, can be explained by better mechanical properties of the respiratory system after LVRS with an increased elastic recoil and less hyperinflation. Dynamic hyperinflation may also explain PH in emphysematous patients as suggested by Butler and colleagues (28) who emphasized the importance of the rise of wedge pressure (Pw), and hence that of during exercise because of the rise in pressure in the cardiac fossa associated with lower lobe gas trapping in patients with COPD. Montes de Oca and colleagues (29) stressed the hemodynamic consequences (with an increased left ventricle afterload, which could lead to increased Pw) of deranged ventilatory mechanics, which led also to a reduced oxygen pulse in their patients with severe COPD during exercise. Dynamic hyperinflation has been shown to improve after LVRS by Martinez and colleagues (8), and this improvement may also counterbalance the effect of a resection of the vascular bed in emphysematous patients. An important and significant reduction of the vascular bed is unlikely after LVRS since in our study, as in the study by Sciurba and colleagues (2), the diffusion capacity at rest after LVRS was unchanged. Diffusion capacity was even found to be largely increased in the study of Gelb and colleagues (4). These results suggest that the alveolo-capillary gas exchange surface is not impaired. This can be explained by either a well-targeted resection, causing only minimal damage to the vascular bed, or possibly by a recruitment of capillaries after LVRS, to counterbalance anatomic vascular resection.

After LVRS we observed no significant change of Pa_O2 or Pa_CO2 in our patients at rest or during exercise. Our results at rest agree with those of Miller and colleagues (10) who reported no change of these variables after LVRS for the 40 patients included in their study. Cooper and colleagues (9) observed no significant change for Pa_CO2 at rest, but a significant improvement for Pa_O2 in 20 patients after LVRS. Sciurba and colleagues (2) found a significant improvement for Pa_CO2 in the 20 patients of their study and suggested that this is a result of improved alveolar ventilation. In our study we observed good correlations between the change in FEV₁ and Pa_O2 at rest as well as during exercise, and also between the change in VC and Pa_O2 at rest. After LVRS, there is an increased elastic recoil, which explains at least partly the increase in VC and in FEV₁ and also improvement in A/ (ventilation to perfusion ratio) abnormalities, which are reflected by a paralleled improvement in Pa_O2. We observed no correlation between the change of Pa_O2 and the change of at rest. This is not surprising since, in contrast with patients with COPD of the "bronchitic type" (22), alveolar hypoxia (hypoxemia being related to the degree of alveolar ventilation in such patients) seems less important in emphysematous patients in whom PH is mostly due to the loss of vascular bed. This idea is supported indirectly by Biernacki and colleagues (30) who have shown that the correlation between the morphologic grade of emphysema (assessed by a quantitative CT scan) and the degree of hypoxemia, hypercapnia and PH is rather poor.

We observed no significant change in blood gases either at rest or during exercise. Despite the small number of patients in our series, our results are in agreement with those of Benditt and colleagues (31) who in 21 patients did not observe any improvement in Pa_O2 or in Pa_CO2 at peak exercise after LVRS. Moreover, in their study, alveolar-arterial oxygen pressure difference (AaPO₂) increased from rest to exercise before and also, without significant change, after LVRS. Our results for AaPO₂ agree with theirs, but it must be emphasized that we could measure AaPO₂ at rest and during exercise in only six patients. Nevertheless, it cannot be ruled out that less hypoxemia during exercise may contribute to lessen the in the present study. Indeed, we observed a significant correlation between the change in after LVRS and the change in Pa_O2 (r = 0.70, p < 0.05) during exercise, although exercising and Pa_O2 were not changed significantly by LVRS. However, this correlation might be explained by the fact that exercising Pa_O2 partially reflects the degree of amputation of the pulmonary capillary bed, which leads to a shortened transit time of the red cells in the pulmonary capillaries, leading to hypoxemia during exercise (15).

A decreased cardiac output or change in ventilation could also influence the level of but we observed no change of these variables after LVRS at rest or during steady-state exercise. Moreover, we observed no correlation between the change in ventilation and the change in at rest or during exercise in our series. Our results on ventilation are in agreement with those recently reported by Bloch and colleagues (32). At rest, as compared with preoperative values, these investigators reported no difference for tidal volumes or for minute ventilation after LVRS. Benditt and colleagues (31) reported no difference in the level of ventilation after LVRS at isowatt exercise (at the maximal level of work attained during an incremental exercise test before LVRS).

In our series, one patient (Patient 6) did not improve his dypnea or his lung function after LVRS, although we observed a reduction in respiratory swings at rest and during exercise. It may be that resection was not well targeted in this case. This patient also had an ₁-antitrypsin deficiency and could therefore be less improved, but the two other patients (Patients 2 and 4) with marked ₁-antitrypsin deficiency did well. This case underlines the need for better patient selection. Indeed, the selection criteria are evolving according to growing experience and to the analysis of the long-term outcome of the patients as emphasized by Russi and colleagues (33).

In conclusion, our study has shown that in most cases LVRS has no adverse effect on pulmonary hemodynamics at rest or during exercise. The possible effect of an anatomically reduced vascular bed after LVRS may be counterbalanced by a decrease of pulmonary vascular resistance. The latter may result from a lower hyperinflation, increased elastic recoil, and hence from capillary recruitment and from better mechanical properties of the lung with less functional compression of the pulmonary vessels.