RESUMO
BACKGROUND: Non-small cell lung cancer elevates serum carcinoembryonic antigen (CEA). CEA determinations are not recommended currently. This study aims to identify the correlation between reducing serum CEA levels with progression-free survival (PFS) and overall survival. METHODS: This study assessed at baseline and in every scheduled visit serum CEA levels throughout first-line therapy. A sensitivity and specificity analysis identified the best cut-off point and correlated it with progression-free survival and overall survival. Multivariate Cox proportional hazard models were conducted. RESULTS: We assessed 748 patients with elevated serum CEA levels at diagnosis. A ≥20% decrease from baseline was associated with a 2-fold median survival compared with patients with lower decreases (20.5 months vs 9.1 months; hazard ratio, 0.53; 95% confidence interval, 0.44 to -0.64; P < .001). CEA sensitivity and specificity to predict survival was 79.8% and 59.8%, respectively. A ≥10% decrease in CEA concentrations was associated with longer progression-free survival (7.7 months vs 5.9 months; hazard ratio, 0.71; 95% confidence interval, 0.57 to -0.88; P = .001) in those treated with chemotherapy, and in patients under tyrosine kinase inhibitors (11.9 months vs 7.3 months; hazard ratio, 0.63; 95% confidence interval, 0.47 to -0.83; P = .0001) and a ≥20% decrease. CONCLUSION: In patients with metastatic non-small cell lung cancer with an elevated baseline CEA level, the percentage decrease of CEA concentrations above the threshold during the first-line therapy was associated with more prolonged survival and progression-free intervals. Serum CEA determinations are a feasible, noninvasive option for monitoring and prognosis.