Abstract
Lung cancer is still a primary cause of cancer death in the western world. Thus, a comprehensive primary diagnostic program which comprises the measurement of biomarkers released by lung cancers is of great importance in the clinical management of lung cancer patients. This review present data for the early lung cancer detection comparing computed tomography (CT), sputum cytology and a tumor marker panel of CYFRA 21-1, CEA, NSE, ProGRP, and SCC which was included into fuzzy classification.Sensitivities in the CT studies were mainly reported with 100% and in sputum analyses with a maximum of 87.5%. The corresponding specificity in sputum cytology analysis was 92.7%. By use of CT, all peripheral squamous or adenocarcinomas of the lung were detected, but there was a lack in detection of SCLC or central localised tumors. In addition, CT detected many non-calcified nodules, of which only a fraction is actual lung cancer leading to a considerable amount of false positive results. At 95{\%}-specificity the sensitivity of a tumor marker panel included into fuzzy modelling for the detection of NSCLC was 50% at stage I. Also SCLC patients could be detected in more than 70% in limited disease stage. Detection rates of advanced tumor stages were even higher. The positive predictive value of biological markers was 80% while the negative predictive value reached around 90%. Modern biological marker approaches in combination with CT and sputum cytology analyses may improve the differential diagnosis especially for the early detection of lung cancer in high-risk populations.