Ticular cell type. Using laser capture microdissection, Roepman et al. managedTicular cell type. Using laser

Ticular cell type. Using laser capture microdissection, Roepman et al. managed
Ticular cell type. Using laser capture microdissection, Roepman et al. managed to show that the genes expressed in the stroma are highly correlated with metastasis formation [64]. Specifically, they showed that 12 of the genes associated with lymph node metastasis in head-neck squamous cell carcinoma (HNSCC) are predominantly expressed in the stroma, 25 are tumour cell specific and the other genes are equally expressed in the tumour and the stroma. We speculate that the involvement of stroma-derived information might also be of importance in breast and in lung cancer. In our signatures, we found several of the genes that have been identified by Roepman et al. as being predominantly expressed in the stroma (ACTA1, TPM2, CDH2, COL5A1, COL5A2, HNRPL, TCF3). These segregated the patients into two groups with significantly different prognosis. The IGF-I induced GW9662 biological activity signatures in primary breast and in lung fibroblasts are similar to each other (Additional file 10) and to important, previously published signatures (Figure 4). The high reverse correlation of the IGF-I signature and the good-risk 70-genes signature supports the power of the IGF-I derived signature as a negative prognosticator in breast cancer. While the `good-risk 70-genes signature’ [49] was developed to predict freedom from metastasis in a top-down manner and validated in the same dataset of breast cancer patients from the NKI, the IGF-I induced signature is a marker for poor prognosis and is well connected to a defined in vitro biological system. The IGF-I induced signature is also highly correlated to the wound signature [60], another strong prognostic signature in NKI dataset outperforming all known prognostic parameters so far. This is interesting, since a single growth factor, such as IGF-I, is able to induce a gene expression programme similar to the mix of undefined factors inherent in FBS. Using a fully defined stimulus in a concentration within the physiological range provides a simple and well-controlled in vitro model that enables specific experimental interventions to be made. Its effects can then be tested in vivo. Considering the notion by Sotirou [65] that proliferation is a main driver of the strong prognostic signatures such as the good-risk 70-genes signature and the wound signature facilitates speculation that IGF-I is one of the important factors responsible for the induction of proliferation. This does not exclude other, equally or more important,growth factors from inducing proliferation and up-regulation of proliferation associated genes. We observed that both IGF-I signatures derived from lung and breast fibroblasts are exchangeable prognostic factors for the other cancer type, which allowed us to speculate that we could generalize this finding to other types of human solid cancer. The consistent response of fibroblasts (our data and [34,35,55]) to IGF-I might also help to explain the worse outcome of patients with elevated IGF-I levels in different cancer types [4-8], a finding that is not necessarily explained by the cancer cells themselves based on PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/27532042 their IGF-receptor expression status on the cell surface. Specifically, since the correlation of the IGF-IR expression and patient outcome in human breast cancer is conflicting [66], the IGF-I induced gene expression signature showing the functional effects of IGF-I axis stimulation, which is correlated with the patients’ clinical outcome, might be of interest when selecting patients who might benefit best f.