Analysis of late effects data using dose-response models: application to human skin telangiectasia data

Area of Science:

• Radiation Oncology
• Radiobiology
• Medical Physics

Background:

• Skin telangiectasia is a critical late effect in radiotherapy.
• Understanding dose-response relationships is crucial for optimizing radiation therapy.
• Previous studies provide clinical data for reanalysis.

Purpose of the Study:

• To reanalyze clinical data on skin telangiectasia using linear-quadratic (LQ) and NSD isoeffect models.
• To evaluate the sensitivity of vessel endothelium to radiation dose per fraction.
• To compare the predictive accuracy of LQ and NSD models across different fractionation schemes.

Main Methods:

• Reanalysis of prospective clinical data on skin telangiectasia.
• Application of linear-quadratic (LQ) and National Council on Radiation Protection and Measurements (NCRP) dose-response models.
• Statistical analysis to determine model parameters and significance.

Main Results:

• An alpha/beta value of 2.75 Gy was obtained for the LQ model, indicating high sensitivity of the vessel endothelium.
• The NSD model showed the time factor was not significant for treatment times up to 68 days.
• Discrepancies between LQ and NSD models were observed for very small and large numbers of fractions, with NSD predicting higher isoeffective doses in certain scenarios.

Conclusions:

• The vessel endothelium exhibits high sensitivity to radiation dose per fraction.
• The LQ and NSD models provide comparable predictions for 10-35 fractions but diverge significantly for fewer than 5 fractions.
• Accurate dose delivery uncertainty is estimated between 3-4.5% for telangiectasia as a decisive late effect.