Rodo-Pin A; Pulmonology Department, Hospital del Mar- Barcelona,Spain.
Martin-Ontiyuelo C; Rodriguez-Fuster A;et al

Respiration. 105(8):864-872, 2026.

INTRODUCTION: To determine whether peak workload (Wpeak), expressed as a
percentage of predicted value during a cardiopulmonary exercise test
(CPET), can predict postoperative cardiopulmonary complications in
high-risk patients undergoing lung cancer resection.

METHODS: A retrospective cohort study was conducted on 226 consecutive
patients who underwent anatomical lung resection (lobectomy or
segmentectomy) for non-small cell lung cancer between 2013 and 2017. All
patients were considered high risk, with preoperative FEV1 and DLco <80%
predicted and underwent full CPET within 30 days prior to surgery. The
main endpoint was the occurrence of cardiopulmonary complications within
30 days postoperatively. The predictive ability of Wpeak (% predicted) for
complications was evaluated using ROC analysis and optimal cutoff was
identified. Outcomes were compared above and below this threshold.

RESULTS: A Wpeak threshold of 52% predicted was identified as the optimal
cutoff, with an area under the ROC curve of 0.59 (95% CI: 0.51-0.67; p =
0.015). Patients with Wpeak <=52% had a significantly higher rate of
postoperative complications (58% vs. 33%, p = 0.001), particularly
respiratory complications (47% vs. 28%, p = 0.008), and longer hospital
stays (mean 24 vs. 15 days, p = 0.002). They also had lower FEV1, DLco,
VO2peak, and heart rate responses compared to patients with Wpeak >52%.

CONCLUSIONS: Wpeak <=52% of predicted was associated with increased risk
of postoperative complications and longer hospitalization. Although the
discriminatory ability was modest (AUC 0.59), Wpeak <=52% predicted was
associated with higher postoperative complications and may serve as a
complementary tool for risk stratification, especially in resource-limited
settings. Further prospective validation is warranted.