Application of Data Mining Technology on Analyzing a Series of Biomarkers for Diagnosis of Bone Metastasis from Different Cancers

Abstract

The aim of this study was to evaluate the value index of tumor biomarkers for diagnosis and prediction of bone metastasis by combining pyridinoline cross-linked carboxy-terminal telopeptide of type I collagen (ICTP) in serum with other biochemical markers in patients with several cancers. All of 1253 patients and biomarkers data with cancers and the related benign diseases and 312 healthy controls were included in the study. Bioinformatics, data mining technology and statistics were carried out to calculate sum of direct positive index (SDPI) and sum of indirect positive index (SIPI). The average ICTP was higher in patients with bone metastasis from liver, gastric, lung, colorectal and pancreatic cancer, than in cancer patients without bone metastasis. For those cancers patients with bone metastasis, which could not be differentiated by single ICTP, combining with SDPI and SIPI would distinguish types of carcinama in situ from similar symptoms. The measurement of multiple tumor markers was very useful in complementary diagnosis and prediction of different carcinoma with bone metastasis bottom.