Ann Thorac Surg 2003;76:376-380
© 2003 The Society of Thoracic Surgeons
Original article: general thoracic
Predicted versus observed maximum oxygen consumption early after lung resection
Alessandro Brunelli, MDa*,
Marco Monteverde, MDa,
Alessandro Borri, MDa,
Michele Salati, MDa,
Majed Al Refai, MDa,
Aroldo Fianchini, MDa
a Department of Thoracic Surgery, "Umberto I°" Regional Hospital, Ancona, Italy
Accepted for publication February 14, 2003.
* Address reprint requests to Dr Brunelli, Via S Margherita 23, Ancona 60129, Italy
e-mail: alexit_2000{at}yahoo.com
 |
Abstract
|
|---|
BACKGROUND: The objective of this study was to identify the predictors of underestimation and overestimation of postoperative maximum oxygen consumption (
O2max).
METHODS: A prospective analysis was performed on 229 patients who had 38 pneumonectomies, 171 lobectomies, and 20 segmentectomies. All patients performed a preoperative and postoperative (on average 9.2 days after surgery) maximal stair-climbing test. Predicted postoperative O2max (ppoO2max) was calculated on the basis of the number of functioning segments removed during operation. The patients were divided into three groups: group A (158 cases), patients with a ppoO2max within 1 standard deviation of the observed postoperative O2max; group B (56 cases), patients with a difference between the observed postoperative O2max and ppoO2max greater than 1 standard deviation (underestimation); and group C (15 cases), patients with a difference between ppoO2max and the observed postoperative O2max greater than 1 standard deviation (overestimation). Univariate and multivariate analyses were performed.
RESULTS: The only significant predictor of underestimation was a high percentage of functional parenchyma removed during operation (p < 0.0001). The significant predictors of overestimation were a low percentage of functional parenchyma removed during operation (p = 0.01) and a high preoperative O2max (p = 0.002).
CONCLUSIONS: The prediction of postoperative O2max was not accurate in all patients. Those with a large amount of functional lung tissue removed during operation tended to have a postoperative O2max greater than expected. Conversely, those patients with a small amount of functional lung tissue resected tended to have a postoperative O2max lower than predicted.
|
Introduction
|
|---|
Lung resection reduces the aerobic capacity compared with preoperative values [1]. However, such reduction is variable depending on the extent of resection, and on the respiratory, cardiovascular, neuromuscular, motivational, and hematologic conditions of the patient at the time of the exercise test.
Prediction of the amount of reduction in aerobic capacity early after lung resection may assist in determining whether the patient will tolerate surgery and its ensuing complications [1, 2].
The objectives of this study are to assess the accuracy of predicting postoperative maximum oxygen consumption (O2max) and to identify the characteristics of those patients in whom postoperative O2max is underestimated or overestimated.
|
Patients and methods
|
|---|
Two hundred fifty-two patients performed a preoperative stair-climbing test and subsequently underwent lung resection for non–small-cell lung carcinoma from January 2000 through September 2002, and were prospectively enrolled in the study after giving informed consent.
Twenty-three patients did not perform the postoperative exercise test and were, therefore, excluded from the analysis (7 died, 1 had a stroke, 15 refused to perform the test). The remaining 229 (178 men, 51 women) formed the database of the study.
The same surgical team performed 38 pneumonectomies, 171 lobectomies, and 20 segmentectomies through a muscle-sparing lateral thoracotomy.
All patients in the study completed a preoperative (1 day before surgery) and postoperative (mean 9.2 days after surgery) symptom-limited stair-climbing test. The test was performed on room air for all patients. Our hospital has 16 flights of stairs; each flight has 11 steps. Each step is 0.155 m high. The patients were asked to climb, at a pace of their own choice, the maximum number of steps and to stop only in case of exhaustion, limiting dyspnea, leg fatigue, or chest pain. During the exercise the pulse rate and capillary oxygen saturation were monitored by means of a portable pulse oximeter. For each patient, the number of steps climbed and the time taken to complete the test were recorded to calculate the following ergometric variables [3]:
The following spirometric variables were considered for this study: forced expiratory volume in 1 second (FEV1), forced expiratory volume in 1 second to forced vital capacity ratio (FEV1/FVC), carbon monoxide lung diffusion capacity (Dlco), predicted postoperative FEV1 (ppoFEV1) calculated by the formula (preoperative FEV1/number of preoperative functional pulmonary segments times the number of postoperative pulmonary functional segments), predicted postoperative Dlco (ppoDlco) calculated by the formula (preoperative Dlco/number of preoperative functional pulmonary segments times the number of postoperative pulmonary functional segments). All the spirometric variables with the exception of FEV1/FVC were expressed as percentages of predicted values for age, sex, and height. Carbon monoxide lung diffusion capacity was measured by the single-breath method and corrected for the lung volume. Spirometry was performed according to the American Thoracic Society criteria. Computed tomographic scan and bronchoscopy findings were used to estimate the number of functional pulmonary segments. In those patients with a calculated ppoFEV1 less than 50% of predicted and in all pneumonectomy candidates, a quantitative perfusion lung scan was performed [4]. The simple calculation of ppoFEV1 was shown to be as reliable as the lung perfusion scan [4]. The percentage of functional lung parenchyma removed during operation (Func loss%) was calculated on the basis of computed tomographic scan, bronchoscopy, and, when performed, quantitative lung perfusion scan findings [5].
Likewise, predicted postoperative O2max (ppoO2max) was calculated on the basis of the number of functional segments removed during operation (preoperative O2max/number of preoperative functional pulmonary segments times the number of postoperative pulmonary functional segments) [6, 7]. The number of functional segments was estimated by means of computed tomographic scan, bronchoscopy, and, when performed, quantitative lung perfusion scan. Each functional segment was considered equal in mass for this calculation.
The patients were divided into three groups: group A (158 patients), patients with a value of ppoO2max within 1 standard deviation (SD) of the observed postoperative O2max; group B (56 patients), patients with a value of ppoO2max lower than the observed postoperative O2max of 1 SD or more (underestimation); and group C (15 patients), patients with a value of ppoO2max greater than the observed postoperative O2max of 1 SD or more (overestimation).
Univariate comparison was made between the groups by means of the Student’s t test (numerical variables) and the 
2 test (categorical variables). The Bonferroni/Dunn procedure was applied as a correction for multiple comparisons. The following variables were used for comparison: age, oxygen arterial tension, carbon dioxide arterial tension, FEV1, FEV1/FVC, Dlco, ppoFEV1, ppoDlco, Func loss%, total height climbed at preoperative exercise (m), preoperative O2max (in milliliters per minute, per kilogram of body weight), concomitant cardiac disease (previous myocardial infarction or cardiac surgery, history of coronary artery disease, current treatment for cardiac failure, arrhythmia, hypertension), days of postoperative hospital stay, and presence of postoperative cardiopulmonary complications. The significant variables at univariate analysis were then used as independent variables in a logistic regression analysis performed on two distinct populations (group A + group B and group A + group C). The dependent variable for both models was the capacity to predict postoperative O2max within 1 SD of the observed postoperative O2max. To avoid multicollinearity, only one variable in a set of variables with a correlation coefficient more than 0.5 was used in the multivariable analyses. All tests were two-tailed, with a significance level of 0.05. Statistical analysis was performed on the statistical software StatView 5.0 (SAS Inc, Cary, NC).
|
Results
|
|---|
The mean postoperative O2max was 22.5 mL · kg-1 · min-1 with an SD of 4.0 mL · kg-1 · min-1. In 56 patients (24.5%), the postoperative O2max was underestimated by 1 SD or more. In 15 patients (6.5%), the postoperative O2max was overestimated by 1 SD or more. In group A, mean ppoO2max was 21.55 mL · kg-1 · min-1 with an SD of 4.0 mL · kg-1 · min-1. In group B, the mean ppoO2max was 17.86 mL · kg-1 · min-1 with an SD of 3.5 mL · kg-1 · min-1. In group C, the mean ppoO2max was 26.32 mL · kg-1 · min-1 with an SD of 5.9 mL · kg-1 · min-1.
Table 1
shows the results of the comparison among the three groups. In comparison to the patients in group A, the patients in group B (underestimation) had a lower ppoFEV1 (p = 0.0007), a lower ppoDlco (p = 0.009), and a greater Func loss% (p < 0.0001), and reached a lower height climbed at preoperative exercise test (p = 0.03). Compared with the patients in group A, the patients in group C (overestimation) had a higher preoperative O2max (p = 0.002), a lower Func loss% (p = 0.003), and a higher ppoDlco (p = 0.02).
As shown in Table 2
, 25% of the patients who underwent segmentectomy had their postoperative O2max overestimated (compared with 4.7% in lobectomy and 5.2% in pneumonectomy). Among the pneumonectomy patients, the rate of underestimation was 55.3% (compared with 5% in segmentectomy and 19.9% in lobectomy). Table 3
shows that the risk of underestimation is greater in lower lobectomy with respect to upper lobectomy (p = 0.03), whereas the risk of overestimation is greater in upper lobectomy (p = 0.08).
The first model of logistic regression analysis included as independent variables ppoFEV1, ppoDlco, Func loss%, and total height climbed at preoperative exercise test, and showed that the only significant predictor of underestimation was a high Func loss% (regression coefficient = 0.145; p < 0.0001). The second model included as independent variables preoperative O2max, Func loss%, and ppoDlco, and showed that the significant predictors of overestimation were a low Func loss% (regression coefficient = -0.1; p = 0.01), and a high preoperative O2max (regression coefficient = 0.17; p = 0.002).
|
Comment
|
|---|
Major lung resection reduces the aerobic capacity from 14% to 21% in the early postoperative period [1]. The reduced tissue oxygenation in the face of increased oxygen demand, which is normally observed in the uncomplicated and even more in the complicated postoperative period, may be critical in those patients with a marginal aerobic reserve. An oxygen debt may arise, which, if it remains critically high for a specified period of time, may lead to multiorgan failure [2, 8]. Therefore, the accuracy of the prediction of postoperative maximum aerobic capacity is of utmost importance for risk stratification before surgery.
A good correlation between predicted and observed postoperative O2max was reported in small series in which the postoperative tests were performed at least 3 months after operation [6, 9]. The objective of the present study was to assess the accuracy of the prediction of postoperative O2max in the early period after lung resection, when critical pathophysiologic derangements and adaptations occur that increase the risk of major cardiopulmonary morbidity.
Accurate prediction of the reduction of aerobic reserve in the immediate postoperative period may assist the surgeon in selection and counseling of patients before surgery, and even in planning the extent of the resection itself.
Inasmuch as O2max results from a complex of factors, which enter at various levels in the oxygen transport system (respiratory, cardiovascular, and neuromuscular systems, hemoglobin level, gas exchange, motivation), likewise its postoperative reduction may be influenced by a number of variables. This variability may be the basis of underestimation and overestimation of postoperative O2max in some patients.
We arbitrarily chose the SD to assess the discrepancy between observed O2max and ppoO2max, inasmuch as it is the most commonly used measure of variability of a set of numbers in the same units of measurement as the original observations. One SD was selected as a cutoff value to allow a sufficient number of observations (31%) to fall above or below such value.
We used the stair-climbing test as a form of symptom-limited exercise. This test is routinely applied in our institution for risk stratification before operation, and we regard it as extremely appealing for its simplicity and brevity. It requires minimum equipment and personnel, and it is therefore economical and widely applicable. Climbing stairs is a form of exercise that is familiar to patients and has been shown to be a greater stress than cycling or treadmill walking, yielding higher values of O2max [10, 11]. In this study O2max was calculated rather than measured, to adopt a technically simple, economical, and widely reproducible noninvasive method of exercise. Although the accuracy of the calculation of O2max is not universally accepted, the same equation was used for both the preoperative and postoperative tests. Should a discrepancy exist between measured and calculated O2max, this would be similar for both tests. Because ppoO2max is derived from the preoperative O2max value and then compared with the observed postoperative O2max (both of which are calculated by using the same equation), a possible error of calculation would be compensated.
The correlation between observed O2max and ppoO2max was 0.56, a value lower than those reported by Bolliger and associates [7] at 3 months after surgery (0.71) and Puente-Maestu and coworkers [9] at 4 months after surgery (0.68). The different types of exercise used (stair-climbing vs cycle ergometer), the different sample sizes (250 patients in our study versus 25 patients in the study of Bolliger and colleagues [7] and 26 patients in that of Puente-Maestu and colleagues [9]), and the earlier time of postoperative observation in our study compared with the others may explain these differences.
In 24.5% of our patients postoperative O2max was underestimated by 1 SD or more. The only significant predictor of this underestimation was a high Func loss% (mean, 28.1%). In another 6.5% of patients the observed postoperative O2max was overestimated by 1 SD or more. A low Func loss% (mean, 11.3%) was a significant predictor of this overestimation. These findings are consistent with the fact that 55.3% of our pneumonectomy patients had their postoperative O2max underestimated, and 25% of our segmentectomy patients had their postoperative O2max overestimated. Moreover, in lower lobectomy, in which usually a greater amount of functional lung tissue is resected compared with upper lobectomy, postoperative O2max tends to be better than predicted. This may be explained by the fact that O2max is dependent on many other physiologic factors unrelated to pulmonary function, which may render less precise the prediction of postoperative O2max at very low and very high Func loss%. It can be speculated that, after the resection of a small percentage of functional parenchyma, the reduction in postoperative performance is greater than expected owing to other factors such as thoracotomy chest pain, postoperative deconditioning, anemia, and the surgical stress itself. On the other hand, when a large extent of parenchyma is resected, the patients may perform better than expected because cardiovascular and pulmonary mechanisms may partially compensate for the reduction in the area of gas exchange. These observations may become critical when lesser resections (wedge or segmentectomy) are performed for a compromised cardiopulmonary status or other severe comorbidities. In these patients the overestimation of their postoperative maximum aerobic capacity may lead to devastating consequences. The underestimation of postoperative O2max in candidates for resection of a large amount of functional lung parenchyma, such as in pneumonectomy candidates, is no less dangerous. In fact, it can lead to the improper exclusion of the patients from surgery or to a lesser resection, which may compromise radicality.
We also found that those patients with a high preoperative O2max tended to lose more than expected in their postoperative aerobic capacity, irrespective of the extent of the resection. In group C, the patients had a higher preoperative O2max with respect to group A owing to a higher speed (13.2 ± 4.7 m/min versus 11.0 ± 2.3 m/min, respectively; p = 0.002), whereas the altitude climbed at preoperative test was similar between the two groups. However, in the postoperative exercise, these patients reduced their speed by 43.9% (from 13.2 m/min to 7.4 m/min) versus a reduction of only 22.7% (from 11.0 m/min to 8.5 m/min) in the patients in group A. Yet, the height climbed at postoperative exercise test remained similar between the two groups (13.75 m in group A versus 14.75 m in group C; p = 0.5). In the group of patients in which the postoperative O2max was overestimated, this marked reduction in the speed of the exercise determined, in turn, a disproportionate reduction in postoperative O2max (-33.6% with respect to preoperative test). This unexpectedly high reduction in postoperative speed and O2max in a group of patients in which only two pneumonectomies and a high proportion of segmentectomies (33.3%) were performed may have been caused by subjective factors such as poor effort, a modified individual symptom tolerance, or a reduced threshold of discomfort and fatigue perception. Motivation in a symptom-limited exercise test may be a critical factor [12]. In fact, in comparison with group A, patients in group C did not have an increased frequency of cardiopulmonary comorbidities, a longer postoperative hospital stay (which may have caused greater deconditioning), or a lower hemoglobin level before the postoperative test, factors that could have determined a poorer performance. Moreover, the postoperative morbidity rate in group C was more than twice that in group A (33.3% versus 15.2%, respectively; p = 0.06). The occurrence of postoperative complications may have impacted on the postoperative performance of these patients, who showed a lower than expected O2max.
In conclusion, we showed that the prediction of O2max early after surgery, calculated on the basis of the number of functional segments resected, was not accurate in all patients. In those in whom a large amount of functional lung tissue was resected, postoperative O2max tended to be underestimated. Conversely, in those patients in whom a small amount of functional lung was resected, postoperative O2max tended to be lower than predicted. Because both situations have important clinical implications in the selection of patients and in planning the extent of surgery, we recommend interpreting ppoO2max cautiously when used for risk stratification before surgery.
|
Acknowledgments
|
|---|
We thank Arrey-Tabot Orok Henshaw, MS, for editing the language of this manuscript.
|
References
|
|---|
- Brunelli A., Monteverde M., Salati M., Borri A., Al Refai M., Fianchini A. Stair climbing test to evaluate maximum aerobic capacity early after lung resection. Ann Thorac Surg 2001;72:1705-1710.[Abstract/Free Full Text]
- Gilbreth E.M., Weisman I.M. Role of exercise testing in preoperative evaluation of patients for lung resection. Clin Chest Med 1994;15:389-403.[Medline]
- Olsen G.N., Bolton J.W.R., Weiman D.S., Hornung C.A. Stair climbing as an exercise test to predict the postoperative complications of lung resection. Two years’ experience. Chest 1991;99:587-590.[Abstract/Free Full Text]
- Markos J., Mullan B.P., Hillman D.R., et al. Preoperative assessment as a predictor of morbidity, and mortality after lung resection. Am Rev Respir Dis 1989;139:902-910.[Medline]
- Brunelli A., Al Refai M., Monteverde M., Sabbatini A., Xiumé F., Fianchini A. Predictors of early morbidity after major lung resection in patients with and without airflow limitation. Ann Thorac Surg 2002;74:999-1003.[Abstract/Free Full Text]
- Corris P.A., Ellis D.A., Hawkins T., Gibson G.J. Use of radionuclide scanning in the preoperative estimation of pulmonary function after pneumonectomy. Thorax 1987;42:285-291.[Abstract]
- Bolliger C.T., Wyser C., Roser H., Soler M., Perruchoud A.P. Lung scanning and exercise testing for the prediction of postoperative performance in lung resection candidates at increased risk for complications. Chest 1995;108:341-348.[Abstract/Free Full Text]
- Shoemaker W.C., Appel P.L., Kram H.B. Tissue oxygen debt as a determinant of lethal and non-lethal postoperative organ failure. Crit Care Med 1988;16:1117-1120.[Medline]
- Puente-Maestu L., De Lucas P., Arnedillo A., Calle M., Tatai E., Monturiol M. Prediction of maximal oxygen uptake after thoracic surgery by radionuclide perfusion scanning. Am Rev Respir Dis 1994;149:A785.
- Holden D.A., Rice T.W., Stelmach K., et al. Exercise testing, 6-min walk, and stair climb in the evaluation of patients at high risk for pulmonary resection. Chest 1992;102:1774-1779.[Abstract/Free Full Text]
- Pollock M., Cousar J., Martinez F., et al. Comparison of stir climbing to cycle ergometry in patients with chronic airflow obstruction (CAO). Chest 1989;96(Suppl):231.
- Killian K.J., LeBlanc P., Martin D.H., et al. Exercise capacity and ventilatory, circulatory, and symptom limitation in patients with chronic airflow limitation. Am Rev Respir Dis 1992;146:935-940.[Medline]
This article has been cited by other articles:
|
|
|
|
 
A. Brunelli, F. Xiume, M. Refai, M. Salati, R. Marasco, V. Sciarra, and A. Sabbatini
Evaluation of Expiratory Volume, Diffusion Capacity, and Exercise Tolerance Following Major Lung Resection: A Prospective Follow-up Analysis
Chest,
January 1, 2007;
131(1):
141 - 147.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
