Functional Consequences of Lung Volume Reduction Surgery for COPD

	INTRODUCTION

TOP INTRODUCTION INTERPRETATION OF MEAN DATA INTERPRETATION OF SUMMARY REFERENCES

The paper by Oswald-Mammosser and colleagues (1) examines the effects of lung volume reduction surgery (LVRS) on indices of lung volume/flow rates, hemodynamics, gas exchange and O₂ at rest and during exercise. At first sight, the results appear rather modestdespite gains in FEV₁ and in peak O₂ during exercise that both reached nearly 40%, most other variables failed to significantly improve. Thus, arterial PO₂, PCO₂, AaPO₂ difference, and pulmonary artery pressure were all unchanged after LVRS, both at rest and during exercise. This differs, but only slightly, from other data which have shown small improvements in arterial PO₂ and PCO₂, at least at rest (2, 3).

In the paper under discussion, at constant submaximal power output, neither cardiac output nor minute ventilation were altered post-LVRS; unfortunately, data for these two key variables at peak exercise were not measured. Taken at face value, the paper might lead one to conclude that LVRS improves airflow and exercise capacity, but has no effect on gas exchange or pulmonary hemodynamics. Looking at the mean data, however, appears to obscure some important observations about how LVRS may affect not only lung mechanical properties, but also those pertaining to gas exchange and to blood flow in the lungs.

What catches the eye and leads to this assertion are strong and distinctive correlations between certain pairs of variables, not all of which appear to be addressed in the paper. These deserve some discussion, and may even merit some speculation. The data demonstrate that mean results of a group of patients undergoing a treatment do not tell the whole story examination of individual data can be very revealing. These relationships not only indicate that LVRS is indeed affecting gas exchange and hemodynamics, they suggest possible physiologic mechanisms of change brought about by the surgery. Some day they may even be of value in establishing patient selection criteria for LVRS.

Of the variables presented in the paper, FEV₁ reflecting air flow rates through the bronchial tree, and pulmonary artery pressure (Ppa) symmetrically indicating vascular conductance (at a particular level of blood flow) are of interest. There is a strong relationship between gains in FEV₁ and in O₂ peak after LVRS (r² = 0.77), and between improvements in AaPO₂ and reductions in Ppa after LVRS (r² = 0.89). There is, however, little or no relationship between gains in FEV₁ and reduction in Ppa (r² = 0.05), gains in O₂ peak and fall in Ppa (r² = 0.21), or gains in FEV₁ and fall in AaPO₂ (r² = 0.06). These relationships are all shown in Figure 1.

View larger version (27K): [in this window] [in a new window]   Figure 1.   Subject-by-subject correlations between peak O2, FEV1, pulmonary artery pressure, and alveolar arterial PO2 difference (AaPO2). Note that peak O2 after LVRS is closely related to gains in FEV1 but not to changes in pulmonary artery pressure. In contrast, AaPO2 is unrelated to FEV1 but closely tied to changes in pulmonary artery pressure. Finally, changes in both pulmonary artery pressure and FEV1 are not related. See text for discussion.

There is, therefore, a neat separation between relationships which link O₂ peak to FEV₁, and those that link gas exchange to pulmonary hemodynamics with essentially no apparent interaction. The potential mechanistic implications of these findings is the principal topic of this editorial, but first the overall group data are briefly discussed.

	INTERPRETATION OF MEAN DATA

TOP INTRODUCTION INTERPRETATION OF MEAN DATA INTERPRETATION OF SUMMARY REFERENCES

The finding that FEV₁ was the only reported physiologic variable to increase along with O₂ peak after LVRS suggests that the ability to increase ventilation during exercise was enhanced after surgery. In fact, the percentage gains in the two variables were virtually the same (38% and 36%, respectively). Constancy of arterial PCO₂ with exercise before and after surgery further supports this notion in that along with the higher O₂ peak (and CO₂ peak), ventilation at peak O₂ was increased in rough proportion to metabolic rate. That this appears to have been the major reason for improved peak O₂ is suggested by lack of change in arterial PO₂ (and thus saturation) and the presumed constancy of hemoglobin levels. These in turn imply no increase in arterial O₂ concentration to augment peak O₂. We do not know what happened to peak cardiac output, but an educated guess is that it increased similarly with O₂ peak, based on data that show the cardiac output- O₂ relationship closely follows that of normal subjects in patients with COPD (4). Such an increase in cardiac output would likely be necessary for such a large increase in peak O₂ to occurthe Fick principle underlies this suggestion. But the increase in peak O₂ would seem to hinge primarily on the implied newfound ability to increase ventilation during exercise. This concept is quite compatible with the work of Benditt and colleagues (5) who in fact found that the increase in peak O₂ post-LVRS could be attributed to improvement in ventilation.

	INTERPRETATION OF BETWEEN-SUBJECT RELATIONSHIPS

TOP INTRODUCTION INTERPRETATION OF MEAN DATA INTERPRETATION OF SUMMARY REFERENCES

Not only do the mean group data point to increased ventilation as the primary benefit of LVRS in terms of increasing O₂, the strong correlation between peak O₂ and FEV₁ subject by subject strengthens the argument (see Figure 1, top left panel).

Although gas exchange and hemodynamics both failed to improve as a whole, the data show that gas exchange improvement was closely related to changes in pulmonary artery pressure with an r² of almost 0.9 over a wide range of both gas exchange and pressure (Figure 1, middle right panel). On the other hand, gas exchange was unrelated to changes in FEV₁ (r² = 0.06). This is good evidence that LVRS can affect gas exchange, and that improvement in gas exchange relates not to the ability to move air, but to the ability to move blood through the pulmonary circulation (in a manner that improves ventilation/perfusion matching). Note that LVRS had effects on both FEV₁ and pulmonary artery pressure, but that these effects were completely unrelated (Figure 1, bottom panel). Thus, the gas exchange-pulmonary artery pressure relation is not just a correlation of effects of LVRS on FEV₁. It seems to be an independent result.

Another relationship of interest was that between reductions (after LVRS) in arterial PCO₂ and pulmonary artery pressure during exercise (r² = 0.58). The relationship between changes in arterial PCO₂ and FEV₁ was nonexistent (r² = 0.003). This further supports the notion that reduction in pulmonary arterial pressure after LVRS improves ventilation/perfusion relationships, although this conclusion would be stronger if maximal exercise ventilation values were, like FEV₁, unrelated to arterial PCO₂. Unfortunately, these data were not reported.

Both of the above gas exchange-pulmonary artery pressure relationships in turn suggest that despite airway obstruction being the dominant physiological defect in COPD, perfusion and its distribution are key to gas exchange inefficiency, and to whether LVRS affects gas exchange. This can be nicely explained if LVRS preferentially removes poorly perfused areas which have a high ventilation/perfusion (A/) ratio. It is well known that even with adequate total ventilation, CO ₂ retention can occur when high A/ areas exist ("physiological dead space" or "wasted ventilation"). If these areas are removed, arterial P CO₂ and AaPO₂ should fall without change in total ventilation. Pulmonary vascular pressures can then fall as well if these removed areas permit remaining, better perfused, but less expanded lung regions to regain volume, and/or if the surgery has preferentially removed high vascular resistance pathways.

	SUMMARY

TOP INTRODUCTION INTERPRETATION OF MEAN DATA INTERPRETATION OF SUMMARY REFERENCES

What have the data in this paper told us? First, a word of caution may be warranted. The relationships depicted in Figure 1 are seductively strong, but come from a small group of patients. Would the correlations be as good if the study were repeated? Hopefully yes, but possibly no. But if one allows the above mix of data analysis and speculation, the following conclusions may be offered:

1. LVRS can improve peak exercise substantially (by almost 40% here), likely by allowing an increase in maximal exercise ventilation. This probably reflects less dynamic compression, thus increasing expiratory flow rates (6, 7). Whether any factors downstream of the lungs (muscle blood flow, muscle O₂ transport, or muscle metabolic capacity) have improved and contributed to improved peak O₂ cannot be deduced from the present paper.
2. Increases in FEV₁ do not, however, explain the improvement in pulmonary gas exchange efficiency that some patients enjoy. That such improvement is closely related to improvements in pulmonary artery pressure suggests that distribution of blood flow in the lung is a key to the effects of LVRS on gas exchange. It is compatible with the suggestion that LVRS improves gas exchange best when areas of high ventilation/perfusion ratio are preferentially removed.
3. Finally, the success of LVRS itself as a procedure should not be gauged only by group mean responses. The paper of Oswald-Mammosser and colleagues (1) shows that examination of individual patient data appears valuable in assessing the technique and its physiologic outcomes.

PETER D. WAGNER

Department of Medicine

University of California/San Diego

La Jolla, California

	References

TOP INTRODUCTION INTERPRETATION OF MEAN DATA INTERPRETATION OF SUMMARY REFERENCES

1. Oswald-Mammosser, M., R. Kessler, G. Massard, J.-M. Wihlm, E. Weitzenblum, and J. Lonsdorffer. 1998. Effect of lung volume reduction surgery on gas exchange and pulmonary hemodynamics at rest and during exercise. Am. J. Respir. Crit. Care Med. 158: 1020-1025 [Abstract/Free Full Text].

2. Sciurba, F. C., R. M. Rogers, R. J. Keenan, W. A. Slivka III, J. Gorscan, P. F. Ferson, J. M. Holbert, M. L. Brown, and U. J. Landreneau. 1996. Improvement in pulmonary function and elastic recoil after lung reduction surgery for diffuse emphysema. N. Engl. J. Med. 334: 1095-1099 [Abstract/Free Full Text].

3. Cooper, J. D., G. A. Patterson, R. S. Sundaresan, E. P. Trulock, R. D. Yusen, M. S. Pohle, and S. S. Lefrak. 1996. Results of 150 consecutive bilateral lung volume reduction procedures in patients with severe emphysema. J. Thorac. Cardiovasc. Surg. 112: 1319-1330 [Abstract/Free Full Text].

4. Agusti, A. G. N., J. Roca, and P. D. Wagner. 1997. Responses to exercise in lung diseases. In J. Roca and B. Whipp, editors. European Respiratory Monography on Clinical Exercise Testing. Eur. Respir. J. 2(Monograph 6):32-50.

5. Benditt, J. O., S. Lewis, D. E. Wood, L. Klima, and R. K. Albert. 1997. Lung volume reduction surgery improves maximal O₂ consumption, maximal minute ventilation, O₂ pulse, and dead space-to-tidal volume ratio during leg cycle ergometry. Am. J. Respir. Crit. Care Med. 156: 561-566 [Abstract/Free Full Text].

6. Gelb, A. F., R. J. McKenna, M. Brenner, R. Fischel, A. Baydur, and N. Zamel. 1996. Contribution of lung and chest wall mechanics following emphysema resection. Chest 110: 11-17 [Abstract/Free Full Text].

7. Gelb, A. F., N. Zamel, R. J. McKenna, and M. Brenner. 1996. Mechanism of short-term improvement in lung function after emphysema resection. Am. J. Respir. Crit. Care Med. 154: 945-951 [Abstract].