Purpose: To prospectively measure the total positioning error present in lateral pelvic fields of patients undergoing prostatic irradiation, and to evaluate the effect of a rigid table insert and soft immobilization on the magnitude of the measured error.
Materials and methods: Sixty-one consecutive patients receiving radical prostatic irradiation with a four field technique underwent a total of 234 lateral portal films during the first, third, fifth and seventh week of treatment. The position of the isocentre was compared to the isocentre on the corresponding simulator films and the magnitude and direction of deviations recorded. The patients were divided in to three cohorts of 15 patients, 15 patients and 31 patients. The first cohort was treated on a standard treatment couch, the second cohort treated with the table top stiffened using a 1 cm polycarbonate insert, and the third cohort treated with a soft immobilization device supporting the lower legs, and the polycarbonate insert.
Results: There was no difference in the mean deviation of the vector of the isocentre displacement in the y and z directions identified at any of the four times when measurements were taken during therapy between the cohorts treated with or without the polycarbonate insert, but without immobilization. The overall mean deviation for these first two cohorts of patients was 3.9 mm. The positioning of patients treated with immobilization was compared to those treated without, and the immobilized patients had a significantly improved overall mean deviation of 2.6 mm (P = 0.002). This was a result of improvement in both the random and systematic components of the total error. In addition, the proportion of errors greater than 5 mm was reduced from 17% of set-ups to 8% of set-ups. The time during the course of treatment when the measurement was taken had no effect on positioning error for any of the treatment groups.
Conclusion: Stiffening the treatment couch with a 1 cm thick polycarbonate insert had no effect on reducing total positioning error, but immobilization with an inexpensive and non-customized foam rubber leg support reduced total positioning error in a statistically significant way.