SHOULD DIRECT MEASUREMENTS OF TUMOR OXYGENATION RELATE TO THE RADIOBIOLOGICAL HYPOXIC FRACTION OF A TUMOR

Purpose: Numerous previous studies have attempted to relate the radiob iological hypoxic fraction (HF) to direct measures of tumor oxygenatio n such as HbO(2) saturations, tumor pO(2) levels, or hypoxic cell labe ling, Although correlations have been found within tumor lines, no ove rall relationships were seen across tumor lines. The current objective was to examine the effect on HF of changes in the fractions of the ox ygenated and anoxic tumor cells that remain clonogenic. Methods and Ma terials: A mathematical model was developed that relates the HF to dir ect measures of tumor oxygenation, The primary assumptions were that: (a) the tumor is divided into distinct compartments of either fully ox ygenated or fully anoxic cells, and (b) the survival of the oxygenated cells is negligible compared to that of the anoxic cells, Based on th ese assumptions, the HF is plotted as a function of the fractions of c lonogenic or nonclonogenic, and oxygenated or anoxic cells. Results: I f all cells are clonogenic, then the HF equals the fraction of anoxic cells, If a higher fraction of anoxic than oxygenated cells are nonclo nogenic, then the HF will be overestimated by the fraction of the tumo r measured to be anoxic using direct measuring techniques, If a higher fraction of the oxygenated than anoxic cells are nonclonogenic, the H F will be underestimated by the fraction of anoxic cells. Conclusion: Correlations between the HF and direct measures of tumor oxygenation h ave been described within tumor lines evaluated under different physio logical condition, However, such relationships can be totally unpredic table between different tumors if the fraction of the anoxic cells tha t is clonogenic varies substantially, Clearly, if tumor anoxia cannot be detected using direct measures, this is an accurate indication that the tumor is well oxygenated, When tumor anoxia is present, however, the conclusions are ambiguous, Even when a small fraction of the tumor is measured as anoxic, direct measures of tumor oxygenation may not b e representative of the HF if a substantial proportion of nonclonogeni c cells is present.