BACKGROUNDIt is still unclear whether pulmonary function tests (PFTs) are sufficient for predicting perioperative risk, and whether all patients or only a subset of them need a cardiopulmonary exercise test (CPET) for further assessment. Thus, this study was designed to evaluate the CPET and compare the results of CPET and conventional PFTs to identify which parameters are more reliable and valuable in predicting perioperative risks for high risk patients with lung cancer.

METHODSFrom January 2005 to August 2008, 297 consecutive lung cancer patients underwent conventional PFTs (spirometry + single-breath carbon monoxide diffusing capacity of the lungs (DLCOsb) for diffusion capacity) and CPET preoperatively. The correlation of postoperative cardiopulmonary complications with the parameters of PFT and CPET was retrospectively analyzed using the chi-square test, independent sample t test and binary Logistic regression analysis.

RESULTSOf the 297 patients, 78 did not receive operation due to advanced disease stage or poor cardiopulmonary function. The remaining 219 underwent different modes of operations. Twenty-one cases were excluded from this study due to exploration alone (15 cases) and operation-related complications (6 cases). Thus, 198 cases were eligible for evaluation. Fifty of the 198 patients (25.2%) had postoperative cardiopulmonary complications. Three patients (1.5%) died of complications within 30 postoperative days. The patients were stratified into groups based on VO(2)max/pred (≥ 70.0%, < 70.0%); VO(2)max×kg(-1)×min(-1) (≥ 20.0 ml, 15.0 - 19.9 ml, < 15.0 ml) and FEV1 (≥ 2.0 L, 1.2 - 1.99 L, < 1.2 L), respectively. The rate of postoperative cardiopulmonary complications was significantly higher in the group with VO(2)max/pred< 70.0% or VO(2)max×kg(-1)×min(-1) < 15.0 ml or FEV1 < 1.2 L than that in the group with VO(2)max/pred ≥ 70.0% or VO(2)max×kg(-1)×min(-1) ≥ 15.0 ml or FEV1 ≥ 1.2 L, respectively. Logistic regression analysis revealed that postoperative cardiopulmonary complications were significantly correlated with age, comorbidities, and poor PFT and CPET results.

CONCLUSIONSFEV1 in spirometry, VO(2)max×kg(-1)×min(-1) and VO(2)max/pred in cardiopulmonary exercise tests can all be used to stratify the patients' cardiopulmonary function status and to predict the risk of postoperative cardiopulmonary complications for the high risk patients with lung cancer. FEV1 and VO(2)max×kg(-1)×min(-1) are better than VO(2)max/pred in predicting perioperative risk. If available, cardiopulmonary exercise testing is strongly suggested for high-risk lung cancer patients in addition to conventional pulmonary function tests, and both should be combined to assess cardiopulmonary function status.