Our study established a model combine the in silico biology and empirical methods to identify clinically useful biomarker from information of microarray gene expression profiles in public domain. Through this method, we found trophinin as a novel prognostic marker for early stage lung cancer, which provide insight towards the process of cancer metastasis and the potential target for treatment.

This study was to combine the in silico biology and empirical methods to identify clinically useful biomarker of lung cancer from microarray gene expression profiles in public database.

Using the expression profile of a known metastasis suppressor, CRMP-1, as a template to search for and compare transcriptome expressions in publicly accessible microarray datasets, trophinin (TRO) was identified as the best correlated gene. The expression pattern of trophinin in a lung cancer cell lines of invasion model was investigated using quantitative RT-PCR. The role of trophinin in cancer metastasis was investigated by measuring the change of invasion activity in trophinin gene-transfected cell lines. The relationship of trophinin and the clinicopathological characteristics of lung cancer were further examined by immunohistochemical staining.

Inclusion Criteria:

• patients who underwent surgery for non-small-cell lung cancer at the National Taiwan University Hospital from 1991 to 2000 with available stored tissue blocks were included as potential subjects in this study.

Exclusion Criteria:

-