Lung cancer resection: the prediction of postsurgical outcomes should include long-term functional results

To assess (1) the possibility of predicting long-term postoperative lung function, and (2) the usefulness of maximal oxygen consumption (O(2)max) as a criterion for operability and as a predictor of long-term disability.

Prospective study.

Outpatients and inpatients of a university hospital.

Sixty-two consecutive patients (mean +/- SD age, 62 +/- 8 years; 51 male and 11 female patients) were preoperatively evaluated for lung cancer resection (pneumonectomy or bilobectomy [n = 14] and lobectomy [n = 48]).

Clinical examination and recorded respiratory symptoms and spirometry results before surgery and 6 months after surgery. If predicted postoperative FEV(1) (ppoFEV(1)) was < 40%, patients underwent exercise testing; if O(2)max was between 10 mL/kg/min and 20 mL/kg/min, patients underwent a split-function study.

All the patients with ppoFEV(1) > or = 40%-even those patients (26%) with FEV(1) < 80%-underwent thoracotomy without further tests. Seven patients with ppoFEV(1) < 40% underwent exercise testing, and three of them underwent a split-function study. Nine patients (15%; including six patients with COPD and one patient with asthma) had immediate postoperative complications (pneumonia [n = 5] and respiratory failure [n = 4]); seven of these patients had ppoFEV(1) > or = 40%. ppoFEV(1) significantly underestimated the actual postoperative FEV(1) (poFEV(1); p < 0.001) 6 months after pneumonectomy or bilobectomy but was reliable for actual poFEV(1) after lobectomy. Two patients with predicted postoperative O(2)max > 10 mL/kg/min became oxygen dependent and had marked limitation of daily living.

ppoFEV(1) > or = 40% reliably identifies patients not requiring further tests and not at long-term risk of respiratory disability. O(2)max, effective for defining the immediate surgical risk, is not useful in predicting long-term disability.