Clinical correlations between late normal tissue endpoints after radiotherapy: implications for predictive assays of radiosensitivity

Abstract

The discovery that certain genetic syndromes are associated with a high cellular radiosensitivity has stirred interest in the concept that radiosensitivity of persons in general should have a genetic component. This has motivated research into assays for prediction of cellular normal tissue radiosensitivity. If such an intrinsic factor were a major factor in the development of late normal tissue injury, this should be detectable as a correlation between the probability of developing injury in different tissues. This hypothesis is tested in a series of 229 patients treated with postmastectomy radiotherapy and evaluated with respect to a number of late endpoints 16 to 71 months after the end of treatment. In each patient, the presence of marked subcutaneous fibrosis and telangiectasia were evaluated in two different treatment areas: in a photon field underneath a 5-mm wax bolus and in an abutted electron field used for treating the chest wall. The use of two different doses per fraction and the fact that a single anterior photon field was used with the dose prescribed at the level of the mid-axilla, led to a substantial variation in total dose and dose per fraction in these patients. A non-tissue-specific patient-to-patient difference in radiosensitivity would cause higher than expected reactions in one treatment area to be correlated with higher than expected reactions in the other area. For each of the two endpoints, telangiectasia or subcutaneous fibrosis, patients experiencing stronger than expected reactions in one treatment area tended to do so in the other area as well. Thus, a strong host factor appears to exist for a specific endpoint. It is an open question whether this is explained by individual variability in intrinsic radiosensitivity, progression rate of injury or other. Contrary to this, no significant correlation was seen when pairing the two late end-points, fibrosis and telangiectasia. Thus, patients showing stronger than expected fibrosis developed on average marked telangiectasia with a probability well predicted from their total dose and dose per fraction. These findings suggest that an assay for clinical expression of late injury would have to be specific for that type of injury.