End-tidal carbon dioxide in the early phase of cardiopulmonary exercise testing prior to major colorectal cancer surgery associates with postoperative outcome.

Abstract

Cardiopulmonary exercise testing (CPET) objectively informs preoperative risk stratification prior to major surgery. CPET facilities are resource intensive and therefore more cost-effective triage methods are desirable for scalability. We tested two dynamic CPET parameters (end-tidal CO<inf>2</inf> (P<inf>et</inf>CO<inf>2</inf>) and heart rate (HR)) to early phase exercise and resting diffusion capacity (DLCO) as potential point of care assessments that could be used outside of formal CPET testing facilities. We conducted a retrospective cohort study including 84 patients who underwent CPET prior to elective major abdominal cancer surgery. Data were analyzed for P<inf>et</inf>CO<inf>2</inf> and HR in response to early phase (2 minutes) exercise as well as resting DLCO against postoperative complications and two-year survival. Complications were classified according to Clavien-Dindo classification into less severe and severe (CD>IIIb) grades. Optimal cut points for predicting outcomes were determined using the Youden Index of receiver operating characteristic (ROC) curves. Multivariate regression modeling including both logistic and Cox proportional-hazards model adjusted to age and comorbidity burden was used to analyse the association between the selected parameters and postoperative outcomes. P<inf>et</inf>CO<inf>2</inf>, in response to two minutes of loaded exercise was higher in survivors than in non-survivors (median (IQR) 40.0 (4.2) mmHg vs. 34.5 (5.2) mmHg, P<0.001). There was no association between chronotropic response and postoperative outcome. The optimal cut point for predicting postoperative complications and survival was 38 mmHg and 37.1 mmHg for P<inf>et</inf>CO<inf>2</inf>, respectively. Low P<inf>et</inf>CO<inf>2</inf> was associated with considerably lower odds of survival (OR 0.12; 95% CI 0.03, 0.47; P=0.003) and additionally increased odds of severe postoperative complications (OR 6.77; 95% CI 1.45, 38.4; P=0.019). Reduced age-predicted DLCO% <80% was associated with increased mortality (HR 5.27; 95% CI 1.09, 25.5; P=0.039). Assessment of DLCO at rest and dynamic assessment of P<inf>et</inf>CO<inf>2</inf> during the early phase of exercise may potentially be developed as inexpensive point-of-care triage tools to scale objective preoperative risk assessment.