Cardiopulmonary exercise testing for preoperative risk assessment before hepatic resection

Abstract

Background: Contemporary liver surgery practice must accurately assess operative risk in increasingly elderly populations with greater co-morbidity. This study evaluated preoperative cardiopulmonary exercise testing (CPET) in high-risk patients undergoing hepatic resection. Methods: In a prospective cohort referred for liver resection, patients aged over 65 years (or younger with co-morbidity) were evaluated by preoperative CPET. Data were collected prospectively on functional status, postoperative complications and survival. Results: Two hundred and four patients were assessed for hepatic resection, of whom 108 had preoperative CPET. An anaerobic threshold (AT) of 9·9 ml O2 per kg per min predicted in-hospital death and subsequent survival. Below this value, AT was 100 per cent sensitive and 76 per cent specific for in-hospital mortality, with a positive predictive value (PPV) of 19 per cent and a negative predictive value (NPV) of 100 per cent: no deaths occurred above the threshold. Age and respiratory efficiency in the elimination of carbon dioxide (V̇E/V̇CO2) at AT were statistically significant predictors of postoperative complications. Receiver operating characteristic (ROC) curve analysis showed that a threshold of 34·5 for V̇E/V̇CO2 at AT provided a specificity of 84 per cent and a sensitivity of 47 per cent, with a PPV of 76 (95 per cent confidence interval (c.i.) 58 to 88) per cent and a NPV of 60 (48 to 72) per cent for postoperative complications. Long-term survival of those with an AT of less than 9·9 ml O2 per kg per min was significantly worse than that of patients with a higher AT (hazard ratio for mortality 1·81, 95 per cent c.i. 1·04 to 3·17; P = 0·036).