The impact of heterogeneity correction on dosimetric parameters that predict for radiation pneumonitis

Purpose: To determine if heterogeneity correction significantly affects commonly measured dosimetric parameters predicting pulmonary toxicity in patients receiving radiation for lung cancer. Methods and Materials: Sixty-eight patients treated for lung cancer were evaluated. The conformal treatment technique mostly employed anteroposterior/posterior-anterior fields and off-cord obliques. The percent total lung volume receiving 20 Gy or higher (V-20) and mean lung dose (MLD) were correlated with the incidence of radiation pneumonitis. Parameters from both heterogeneity-corrected and heterogeneity-uncorrected plans were used to assess this risk. Results: Univariate analysis revealed a significant correlation between the development of radiation pneumonitis and both V-20 and MLD. A best-fit line to a plot of V-20 from the homogeneous plan against the corresponding V., heterogeneous value produced a slope of 1.00 and zero offset, indicating no difference between the two parameters. For MLD, a similarly significant correlation is seen between the heterogeneous and homogeneous parameters, indicating a 4% difference when correcting for heterogeneity. A significant correlation was also observed between the MLD and V-20 parameters (p < 0.0001). Conclusions: A high degree of correlation exists between heterogeneity-corrected and heterogeneity-uncorrected dosimetric parameters for lung and the risk of developing pneumonitis. Either V20 or MLD predicts the pneumonitis risk with similar effect. (c) 2006 Elsevier Inc.