Search

The radiobiological data obtained from a multicentre clinical trial of the British Institute of Radiology, which compared the treatment of carcinoma of the laryngo-pharynx by 3 fractions per week (3F/wk) with 5 fractions per week (5F/wk) radiotherapy, have been studied. The trial involved an intake of 734 patients between 1966 and 1975. The number of fractions, overall treatment time and total doses used by different treatment centres ranges from 9 to 40 fractions, 18 to 70 days and 3880 to 7800 cGy, respectively. An 11-13% reduction in the total radiation dose was applied for treatments with 3F/wk as compared with 5F/wk in centres treating over 6 weeks and 3 weeks, respectively. All patients were followed for 10 years from the start of treatment. Different types of early and late normal-tissue reactions were investigated, ranging from a low percentage incidence of perichondritis to 95% for slight early reactions. Greater than 80% of the late normal-tissue reactions seen were observed within the first year after the start of treatment, and 96% were observed within the first 5 years. There was no statistically significant difference in the normal-tissue event-free rates between the 3F/wk and 5F/wk treatment groups. This finding did not differ when different major treatment centres were studied separately. For a number of end-points, alpha/beta ratios and N- and T-exponents of a modified nominal standard dose (NSD) formula have been calculated.

A programmed graphics terminal has been connected on-line to a large time-shared computer for calculating dose distributions in radiotherapy treatment planning and provides a viable alternative to dedicated systems and batch working. The terminal equipment is based on a mini-computer and includes a function-key devise for outline input, a large-screen refresh oscilloscope for viewing results and an X-Y plotter for hard-copy. Radiotherapy dose computation programs in standard Fortran are stored and run on a large remote computer with graphical interaction at the terminal. External beam programs can calculate dose distributions for most commonly used treatment situations and can compute in off-axis planes. Data input is fully interactive and easy to understnad. Dose distributions are displayed as isodose contours. Advantages of the system include accuracy, speed, ease of use and maintenance, and transferability of the programs between different host computers.

A method of computing megavoltage dose distributions using an empirical mathematical model is described. The dose at a point in a medium is calculated as the product of a central axis percentage depth dose, a crossplot factor and a wedge factor. For a given set of conditions (i.e. type of machine, source-skin distance and wedge) 16 parameters are required to calculate the values of the three factors and hence the dose at any point. These parameters can be obtained from a relatively small number of experimental measurements. Dose distributions from a variety of different treatment machines have been computed by the method, and some comparisons of the calculated and measured doses have been made using "Goodness of Fit" score values. These show that the use of the method as a basis for computer treatment planning should be clinically acceptable.