On the impact of tissue inhomogeneity corrections in clinical thoracic radiation therapy.

The magnitude and impact of the corrections required for calculated absorbed doses due to the presence of real anatomic inhomogeneities were studied in 100 "standard" clinical thoracic cases. Correction factor distributions were generated with the equivalent tissue-air ratio method. The correction factors at the point of target absorbed dose, the average within the high dose area, and the average and maximum within the irradiated lung tissue were determined. Results indicate that among patients undergoing similar treatment techniques and radiation energy, correction factors can vary greatly. The diversity in anatomy, geometry, and density emphasizes the need for patient-specific dose calculations if the appropriate accuracy is to be achieved. Even for simple parallel-opposed fields encompassing the mediastinal and bilateral lung regions, the correction at the point of target absorbed dose was found to be as much as 5 to 16% for photon energies between 60Co and 25 MV. The average correction factor within the high dose region was generally comparable in value. For beams that traverse the lungs obliquely, the correction factor at the point of target absorbed dose was as large as 1.21 with 25 MV photons. These values correlated linearly with the lung pathlength preceding the point. Within lung tissues, the corrections were often more dramatic. Using the dose response curve for normal lung tolerance, the clinical consequence of ignoring the increased dose to lung tissue can be predicted. For 100 patients, the use of uncorrected dose would result in an underestimation of the risk for radiological radiation pneumonitis by 7% on average and by as much as 19% maximally. These data suggest that whenever lung tissue and/or tumor response levels are critical, as is often the case for radical therapy, multi-center trials, and dose response determination, inhomogeneity corrections should be implemented.