Cardiopulmonary exercise testing before lung resection surgery: still indicated? Evaluating predictive utility using...

Area of Science:

• Pulmonary Medicine
• Thoracic Surgery
• Cardiology

Background:

• Cardiopulmonary exercise testing (CPET) criteria for lung resection risk stratification are over a decade old.
• Advances in patient care necessitate updated risk assessment tools.
• Hypothesis: CPET offers no additional predictive value for postoperative complications.

Purpose of the Study:

• To evaluate the added predictive value of CPET in risk stratification for lung resection.
• To assess if CPET improves machine learning model performance in predicting postoperative complications.

Main Methods:

• Secondary analysis of two prospective, multicentre studies.
• Included lung resection candidates assessed with CPET, pulmonary function tests (PFTs), and arterial blood gas analysis.
• Machine learning models trained to predict pulmonary and cardiovascular complications using nested cross-validation.

Main Results:

• 497 patients included; 14% experienced pulmonary complications (PPCs) and 18% cardiovascular complications (PCCs).
• CPET parameters did not enhance machine learning model prediction accuracy for PPCs or PCCs across unselected or guideline-selected subgroups.
• Area under the receiver operating characteristic curve (AUC-ROC) showed no significant improvement with CPET inclusion.

Conclusions:

• CPET did not improve the predictive performance of machine learning models for PPCs or PCCs in contemporary surgical practice.
• The utility of CPET in preoperative risk stratification for lung resection warrants re-evaluation.
• Current guidelines for CPET use in lung resection may require updating.