Should direct measurements of tumor oxygenation relate to the radiobiological hypoxic fraction of a tumor?