Background: There is evidence in the literature that the incidence of pulmonary complications and mortality is fair enough in patients with lower pulmonary function than conventionally accepted. In this article, we validate in patients with low baseline lung function (ie, FEV(1) or diffusing capacity of the lung for carbon monoxide [DLCO] < 80%) an algorithm to evaluate anatomic lung surgery in patients with low predicted postoperative lung function (ie, either FEV(1)-postoperative estimated [ppo] or DLCO-ppo < 40% or both between 30% and 40% predicted) if peak oxygen uptake (VO(2)peak)-ppo > 10 mL/kg/min.
Methods: We prospectively studied 126 consecutive patients evaluated for anatomic resection of lung tumors by thoracotomy.
Results: Ninety-two patients were operated on: age 67 (8 SD) years; FEV(1) 63 (14)% pp; DLCO 71 (19)% pp; VO(2)peak 71 (19)% predicted; and 2-year Kaplan-Meier conditional probability of survival (CPS) 0.62 (0.06). Thirty-day perioperative mortality was 6.4%. Thirty-four patients were not functionally fit, or rejected the procedure: age 69 (8) years; FEV(1) 58 (16)% predicted; DLCO 67 (26)% predicted; VO(2)peak 66 (16)% predicted. In this group, 2-year CPS was 0.18 (0.08), P < .01. Subgroups A (FEV(1)-ppo and DLCO-ppo > 40% predicted) and B (either FEV(1)-ppo or DLCO-ppo < 40% predicted or both between 30% and 40% predicted) were comparable in terms of perioperative morbidity; however, they were different in terms of 30-day mortality (A, 1/53 [1.9%]; B, 5/37 [13.5%]; P = .047; relative risk, 7.2; 95% CI 1.1-27.7). The survival functions of both subgroups were significantly different (P < .01) from nonsurgical subjects.
Conclusions: Adherence to the proposed algorithm results in a reasonable preoperative mortality in patients with low preoperative lung function. Although perioperative mortality is significantly higher when predicted postoperative lung function is low, 2-year survival of patients is better than if such patients had not undergone surgery.