Endogenous hypoxia markers: case not proven!

The pivotal role of hypoxia within the pathophysiological framework of solid malignant tumors is now considered to be indisputable. The fact that hypoxia can cause resistance to various cancer therapies and promote malignant progression is reflected in its adverse impact on prognosis which is repeatedly shown for various tumor entities. Knowledge in this area is based on direct assessment of the oxygenation status using O2-sensitive microsensors. However, weaknesses of this standard method are its invasiveness and limitation to accessible tumor entities. Hypoxia-inducible factor (HIF)-1alpha, the master transcriptional regulator of the hypoxic response, as well as certain downstream genes, e.g., glucose transporter (GLUT)-1 and carbonic anhydrase (CA) IX, have been considered to be suitable as surrogate biomarkers of hypoxia due to their tight regulation by O2 levels under certain, well-defined in vitro conditions. The fact that statistical correlations between the expression of these proteins and direct pO2 measurements in the clinic have been sporadically reported seemed to support their role as "endogenous hypoxia markers". Remaining disparities were mainly attributed to the influence of tumor heterogeneity. In a series of studies, we have addressed this question by examining the expression of HIF-1alpha, GLUT-1 and CA IX in tissue micro-areas where direct O2 measurements had previously been carried out, so that the influence of tumor heterogeneity could be reduced to a minimum. Using this methodology, no correlation between the expression of "endogenous hypoxia markers" and direct pO2 measurements could be found. In conclusion, while there may be a stringent association between these markers and the oxygenation status under standardized in vitro conditions, this is not transferable to the clinical assessment of oxygenation status in patients. The term "endogenous hypoxia markers" should therefore be avoided, at least in the clinical oncology setting.