TUMOR-MARKERS IN THE DIAGNOSIS OF BRONCHIAL-CARCINOMA - NEW OPTIONS USING FUZZY LOGIC-BASED TUMOR-MARKER PROFILES

The diagnosis of lung cancer and early knowledge of its histological t ype are very important; however, this is still a difficult subject for the physician. The aim of this study was to improve the diagnostic ef ficiency of tumour markers in the diagnosis of bronchial carcinoma by mathematical evaluation of a tumour marker profile employing fuzzy log ic modeling. A panel of five tumour markers, including CYFRA 21-1, CEA , NSE, and five additional parameters was determined in 281 patients w ith confirmed primary diagnosis of bronchial carcinoma of different hi stology and stage. A further 131 persons, who had acute and chronic be nign lung diseases, served as a control group. A classificator was dev eloped using a fuzzy-logic rule-based system. The diagnostic value of the combined tumour markers was significantly better than that of the individual markers and of a combination of CYFRA 21-1, CEA, and NSE. T he discrimination of malignant vs benign diseases was realized with a sensitivity of 87.5% and specificity of 85.5%. The rate of correct cla ssification of small-cell vs non-small-cell lung carcinoma was 90.6% a nd 91.1%, respectively; for squamous cell carcinoma vs adenocarcinoma it was 76.8% and 78.8%, respectively. Our detailed analysis has shown that the fuzzy logic system improves diagnostic accuracy up to a rate of 20%, especially in early stages and in patients with all marker lev els in the grey area. Our concept proved to be more powerful than meas urement of single markers or the combination of CEA, CYFRA 21-1, and N SE. Its use may help in distinguishing between malignant and benign di sease and make it possible to define different subgroups of patients e arlier in the course of their disease.