Quantitative clinical radiobiology of early and late lung reactions

To quantify the response of human lung to a course of fractionated radiotherapy based on a literature review of published clinical data.

Quantitative clinical radiobiology is concerned with the estimation of parameters that describe the clinical outcome of radiotherapy as a function of patient and treatment characteristics. Here, parameters describing the steepness of the dose-response curve, the response to a change in dose per fraction and to a change in overall treatment time for early and late lung injury are compiled based on published clinical studies.

Two phases of lung injury are seen, radiation pneumonitis and lung fibrosis. The first signs of early lung changes are seen almost immediately after irradiation. This reaction peaks after 5 to 6 months, and settles partially before 9-10 months. Around that time, the late changes become manifest and these are stable in most cases. There is an important distinction between lung injury and radiotherapy-related morbidity, as even severe changes in a small volume may not give rise to any clinical symptoms. Many assays have been developed for lung damage, and these highlight various clinical and biological aspects of lung damage. Here, the literature on steepness of dose-response curves and fractionation sensitivity is reviewed and quantified by the alpha/beta ratio of the linear-quadratic model for both radiation pneumonitis and lung fibrosis. For the early phase a significant time factor exists. Current best estimates for these radiobiological parameters are derived. Other external factors affecting these estimates are briefly discussed.

Quantitative estimates of radiobiological characteristics of human lung are available for the pneumonitis phase where the fractionation sensitivity is in the same range as for most late-responding normal tissues. Short intensive schedules may also bear an added risk for pneumonitis as the dose recovered per day is around 0.5 Gy. For the later phase of lung fibrosis, the estimates are fewer and generally less precise. It is clear though, that the alpha/beta ratio is low, possibly 2-3 Gy. No time factor has been demonstrated for the late reaction. Due to the considerable physiological reserve capacity in the normal human lung, the relationship between damage and morbidity depends strongly on the lung volume affected. It therefore seems likely that for small volumes irradiated to high doses, the dose-limiting complications may not be due to restriction of lung function, but rather to haemorrhage and formation of fistulae.