The application of 3-d treatment planning and dose volume histograms in the treatment of intrahepatic malignancies

Abstract

Radiation therapy has played a minor role in the treatment of intrahepatic malignancies, chiefly because the liver cannot tolerate more than approximately 30 Gy when the entire organ is irradiated with 1.8-2 Gy fractions. Doses in this range would not be expected to control gross disease. We have initiated a clinical protocol which delivers 45-60 Gy to regions of gross tumor by using carefully tailored boost fields. The prescribed dose is based on the hypothesis that the tolerance of the liver to external beam irradiation depends on the fraction of the normal liver treated. If >50% of the normal liver would be treated to encompass the tumor, no boost treatment is given and patients receive 33 Gy to the whole liver, otherwise 30 Gy is given to the whole liver with boost treatment. Boost doses are 15 Gy if 26-50% of the normal liver would be treated and 30 Gy if ~25% of the normal liver can be included. All radiation is delivered as 1.5 Gy fractions twice a day with concurrent fluorodeoxyuridine (FdUrd). To deliver such treatment effectively and safely, we hypothesized that treatment fields needed to be designed using 3-D concepts and accounting for liver motion during respiration. Therefore, we initiated a series of studies as part of our treatment planning process in this protocol which included: 1) assessment of liver motion during treatment; 2) determining if the volume of normal liver were relatively constant regardless of tumor size; 3) the use of 3-D treatment planning to minimize the dose to critical normal structures; and 4) the routine generation of the dose-volume histogram @VI-I) of the normal liver to compare the relative merits of competing plans. The motion of the liver during treatment conditions (quiet respiration) was assessed in two ways. First, the movement of the dome of the diaphragm at the time of simulation was observed by fluoroscopy . Second, films of the liver were taken at the time of simulation after the patient was instructed to hold his breath at maximal quiet inspiration and expiration. studied in this fashion, the average movement of the liver was estimated to be 8 + 3 mm. In the first 10 patients There was good agreement between these methods. This suggests that liver motion can be estimated at the time of simulation and can be included in determining the treatment volume. Patients then had a treatment planning CT scan. The volume of the entire liver (L) and the tumor(s) (T) were demarcated, and the volume of normal liver was calculated as L-T. Twentv three of 38 uatients entered onto this studv as of 3/l/89 have had definable tumors. Although L varied from 1346-4731 cm3 (mean * SD of 2401+ 933 cm3 ) and T varied from’13-2204 cm3 (699 * 714 cm3), L-T (normal liver) showed a much narrower range of 1396-2552 cm3 (1844 + 318 cm3). This size resembles that of notmal liver. These studies support the hypothesis that the normal functioning liver is not replaced by tumor. 3-D treatment planning was then used to determine the optimal field arrangements. Non-axial, non coplanar beam arrangements often prove to be superior to standard axial plans. The use of DVH for kidneys and the normal liver allows a quantitative comparison of competing plans. In some cases, a “standard” coplanar plan treated >50% of the normal liver, while successive refinements reduced the volume of normal liver treated to <25%, permitting a 30 Gy boost to be delivered. The results of the clinical trial, described elsewhere, showed these plans produced tolerable, effective treatment. In summary, we have shown that fully 3-D treatment planning incorporating an assessment of liver motion and DVH analysis can be used on a routine basis to deliver potentially tumoricidal doses of radiation to intrahepatic malignancies.