Diffusion lung capacity for carbon monoxide (DLCO) is an independent prognostic factor for long-term survival after...

We examined the early and late prognostic significance of DLCO and forced expiratory volume in 1 sec (FEV1) in patients who underwent surgical resection of lung cancer.

From 1997 to 2004, 462 patients underwent successful complete resection of their lung cancer and had full pulmonary function testing including DLCO performed. Mean follow-up was over 5 years (64.8 months--range: 0-158 months).

Postoperative 90-day mortality was 2.6% (12/462). At last follow-up, of the remaining 450 patients, 182 patients were alive, 130 had died of cancer, and 138 have died of other causes and did not have recurrent cancer. Mean DLCO values were 69.4%, 66.8%, and 53.9%, respectively. Mean FEV1 values were 81.3%, 78.1%, and 71.5%, respectively. Mean DLCOs and FEV1s between patients who died of cancer versus other causes were significantly different (P < 0.0001 and P = 0.0157). When cause-specific survival was analyzed for both DLCO and FEV1 simultaneously, DLCO had a very significant effect on survival from other causes (HR 0.966, P < 0.0001) when adjusted for FEV1. However, when adjusted by DLCO, FEV1 had no significant effect. A DLCO <40% best predicted decreased survival from causes other than cancer within stage I lung cancers (stage IA HR 0.953, P < 0.0001; stage IB HR 0.968, P < 0.0001).

DLCO was found to be a significant prognostic factor for long-term survival after lung cancer surgery. This may serve as a surrogate for competing morbidities with declining values predicting a higher risk of late non-cancer-related death.