Figure 5

Fibroblast derived insulin-like growth factor-I (IGF-I) signature divides lung cancer patients into two groups with significantly different outcome. (A) Unsupervised hierarchical clustering of fibroblast derived IGF-I signature in Garber lung cancer dataset. The expression values of genes in the fibroblast derived IGF-I signature were extracted from a published expression study by Garber [50]. Genes are presented as rows and the experiments are presented as columns. Although some gene expression data are missing, the fibroblast derived IGF-I signature stratifies lung cancer patients into two groups with high (blue) or low (yellow) expression levels of genes representing the signature. (B) Relationship of expression level of genes building fibroblast derived IGF-I signature with overall survival in Garber data. Kaplan-Meier curves denoting the clinical outcomes of the indicated tumours exhibiting high (blue curve) and low (yellow curve) expression levels of the signature. (C) Unsupervised hierarchical clustering of fibroblast derived IGF-I signature in Bhattacharjee lung cancer dataset. The expression values of genes in the fibroblast derived IGF-I signature were extracted from a published expression study by Bhattacharjee [51]. Genes are presented as rows and the experiments are presented as columns. Fibroblast derived IGF-I signature stratifies adenocarcinoma patients into two groups with high (blue) or low (yellow) expression levels of genes representing the signature. (D) Relationship of expression level of genes building fibroblast derived IGF-I signature with overall survival and disease specific survival in Bhattacharjee dataset. Kaplan-Meier curves illustrating the clinical outcomes of the indicated tumours exhibiting high (blue curve) and low (yellow curve) expression levels of the signature.