TOPLINE:

For lung resection candidates, adding cardiopulmonary exercise testing parameters to basic clinical data and pulmonary function tests did not yield a significant improvement in machine learning models’ ability to predict postoperative pulmonary complications (PPCs) or postoperative cardiovascular complications (PCCs).

METHODOLOGY:

• Researchers conducted a secondary analysis of two prospective multicenter studies to assess whether cardiopulmonary exercise testing parameters provided additional value to machine learning models for predicting PPCs or PCCs.
• They included 497 adult patients (median age, 66 years; 44.3% women) who underwent lung resection surgery for a proven or highly suspected lung tumor and were also eligible for cardiopulmonary exercise testing.
• Researchers used five types of machine learning models to assess the predictive performance of basic clinical data combined with pulmonary function tests with and without adding cardiopulmonary exercise testing parameters.
• They collected data on PPCs including pneumonia and respiratory failure as well as on PCCs such as arrhythmias, myocardial infarction, and stroke from both unselected and selected patients (ie, who met the American College of Chest Physicians or European Respiratory Society/European Society of Thoracic Surgery criteria).

TAKEAWAY:

• Adding cardiopulmonary exercise testing parameters to standard clinical data and pulmonary function tests did not significantly improve the model-based prediction of PPCs.
• Likewise, when the above-mentioned criteria were considered, adding cardiopulmonary exercise testing parameters did not significantly improve the model-based prediction of PPCs or PCCs.

IN PRACTICE:

“Updated guidelines should consider re-evaluating CPET [cardiopulmonary exercise testing] indication criteria for preoperative assessment of lung resection candidates,” the authors of the study wrote.

SOURCE:

The study was led by Ákos Filakovszky, Johannes Kepler University and Kepler University Hospital, Linz, Austria. It was published online on October 02, 2025, in Thorax.

LIMITATIONS:

The study was limited by missing data on arterial blood gases at rest and peak exercise prior to cardiopulmonary exercise testing. Diffusing lung capacity for carbon monoxide was not measured in 5% of patients, and heart rate data were missing in 2% of patients. Low exercise capacity may have reduced the prognostic value of peak oxygen consumption, and potential chronotropic incompetence in 23% of patients may have affected the prognostic value of the chronotropic response.

DISCLOSURES:

The study was supported by a research grant from the Ministry of Health of the Czech Republic. The authors reported having no competing interests.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.