Radiation pneumonitis after hypofractionated radiotherapy: evaluation of the LQ(L) model and different dose parameters.

PURPOSE To evaluate the linear quadratic (LQ) model for hypofractionated radiotherapy within the context of predicting radiation pneumonitis (RP) and to investigate the effect if a linear (L) model in the high region (LQL model) is used. METHODS AND MATERIALS The radiation doses used for 128 patients treated with hypofractionated radiotherapy were converted to the equivalent doses given in fractions of 2 Gy for a range of alpha/beta ratios (1 Gy to infinity) according to the LQ(L) model. For the LQL model, different cut-off values between the LQ model and the linear component were used. The Lyman model parameters were fitted to the events of RP grade 2 or higher to derive the normal tissue complication probability (NTCP). The lung dose was calculated as the mean lung dose and the percentage of lung volume (V) receiving doses higher than a threshold dose of xGy (V(x)). RESULTS The best NTCP fit was found if the mean lung dose, or V(x), was calculated with an alpha/beta ratio of 3 Gy. The NTCP fit of other alpha/beta ratios and the LQL model were worse but within the 95% confidence interval of the NTCP fit of the LQ model with an alpha/beta ratio of 3 Gy. The V(50) NTCP fit was better than the NTCP fit of lower threshold doses. CONCLUSIONS For high fraction doses, the LQ model with an alpha/beta ratio of 3 Gy was the best method for converting the physical lung dose to predict RP.