The risk of metastasis, or “tumor spread”, is estimated based on clinical and pathological criteria referring to the primary tumor. However, this risk does not consider the potential expansion of micrometastatic cancer cells, which are too small to be detected. Therefore, current prognosis estimates are likely to be partly incorrect, which poses a challenging issue. The review by Bidard et al. (2013) focuses on this life-threatening hallmark of cancer.
Preclinical research has focused mostly on identifying individual genes and susceptible pathways that affect cancer cell dissemination. While numerous steps in the metastatic cascade have been unravelled in mouse models, most do not recapitulate tumor progression in humans. In particular, studies carried out in immunocompromised or transgenic mouse models do not accurately reflect the evolution of human cancers. Not surprisingly, only a few genes identified in these approaches have been validated in clinical studies. Most problematic is the use of prognostic signatures (expression profiling) to estimate the metastatic risk; the sensitivity and specificity of these signatures are insufficient for routine clinical use.
Future hopes surround the use of accurate signatures to predict risk, involving miRNAs, features of genomic alteration and, ultimately, protein signatures. In addition, two cell types— circulating tumor cells (CTCs) and disseminated tumor cells (DTCs) —should also help to improve prognosis and provide surrogate markers of tumor evolution. Nonetheless, the potential of these two cell types still remains to be determined. Classically, tumors are staged according to the TNM classification i.e. tumor size, node involvement and metastasis and graded to estimate dedifferentiation and proliferation; it is anticipated that more accurate biomarkers (such as with CTCs from venous blood draws) can complement these traditional methods and help to evaluate patient treatment responses in adaptive clinical trials.
Recent studies confirm the presence of CTCs with intermediate or mesenchymal phenotypes (Yu et al. Science 2013). This finding supports our hypothesis that epithelial-to-mesenchymal transition (or EMT) is one mechanism driving carcinoma cell dissemination (Thiery et al. Ciba Foundation Symposium 1988; Thiery Nature Review Cancer 2002; Thiery et al. Cell 2009; Thiery and Lim Cancer Cell 2013). Carcinoma cells that have partially or fully undergone the EMT process are thought to be more refractory to conventional and targeted therapeutics and can escape immune surveillance. The phenotypes of these CTCs during neoadjuvant and adjuvant therapies warrant further assessment.
Metastasis is complex, and despite the identification of hundreds of metastasis regulators, patient survival rates remain poor. A better integration of clinical and biological data could help improve this.