Potential advantages of modulated conformal therapy in treatment of prostate cancer

Purpose/Objective: The recent observation that adjuvant para-aortic lymph node irradiation was associated with a statistically significant improvement in the survival of patients with high-risk cancer of the cervix (JAMA 274:387-93) has revived interest in such approaches in other pelvic neoplasms, such as prostatic cancer. This study was undertaken to analyze the potential advantages and limitations of dose distributions calculated by simulated annealing inverse planning for primary prostate target volumes and secondary lymphatic target volumes. Material and Methods: A prototype NOMOS inverse planning system was used to contour primary prostate target volumes and secondary lymphatic volumes using CT scans of patients whose plans were being developed for conventional treatment. Simulated annealing was used to compute beam modulation patterns targeted for the treatment volumes and was constrained by acceptable doses for the bladder, the rectum, the bowel, and the spinal cord. The lymphatic target volume followed the iliac vessels along the pelvic side walls and united at the sacral promontory to follow the para-aortic and paracaval lymph nodes anterior to the vertebral bodies. The dose distributions were analyzed using dose-volume histograms (DVHs) and by comparison of the isodose curves. The modulated beam conformal plans were compared to the conventional plans. Plans computed for modulated fan-beam arcs were compared with plans computed for nine fixed-gantry modulated cone-beams. The prostate Planning Target Volume (PTV) dose aim was 7000 cGy and the lymphatic PTV dose was 5500 cGy. The normal tissue structures were limited to various levels between 4000 cGy and 6000 cGy. The modulated cone-beam plans were delivered to a phantom using a dynamic multi leaf collimator (D-MLC) on a 23 MV linear accelerator (Varian) to test the feasibility of the delivery process. Results: Conformal target doses were shown to be achievable for the prostate, seminal vesicles, and the iliac and para-aortic lymph nodes. The conformal plans were characterized by DVHs that demonstrated significant improvement over the conventional treatment plans. There was no significant difference between the fixed-gantry technique and the fan-beam arc technique. Doses above 2800cGy were confined to less than 50% of the bladder and rectal walls. All plans were deliverable using the D-MLC in less than 15 minutes. Conclusion: Modulated cone-beam conformal dose distributions can be computed using simulated annealing for clinically relevant prostate target volumes. The computations for the technique indicate significant reduction of dose to critical organs when compared with conventional approaches. The technique is technically feasible. This investigation was supported in part by PHS grant number CA4380 awarded by the National Cancer Institute