Impact of margin on tumour and normal tissue dosimetry in prostate cancer patients treated with IMRT using an endorectal balloon for prostate immobilization.

In IMRT treatment, margin for planning target volume is determined by organ motion and set-up error. The margin width that achieves the desired dose escalation, while minimizing normal tissue exposure is dependent upon patient immobilization and/or organ localization techniques. In this study, we compare the impact of margin width on the dosimetry of tumour and normal tissues using an endorectal balloon filled with 100 cc of air. Plans were generated for ten patients using margin widths of 0, 3, 5, 8 and 10 mm. The prescription dose to prostate and seminal vesicles was 70 Gy in 35 fractions with 15% of bladder allowed to receive above 65 Gy, 15% of rectum above 68 Gy and 10% of femurs above 45 Gy. Margins above 5 mm produced significantly lower mean doses for both prostate and seminal vesicles without affecting TCP. For normal tissues, mean doses, percent volumes above prescription constraints and NTCP increased as a function of margin width, especially when this was 5 mm or above. We conclude that planning with tighter margins of < or =5 mm improves IMRT dosimetry for prostate and normal tissues and is only possible when target localization and/or immobilization devices are routinely used.