The use of molecular imaging to evaluate radiation fields in the adjuvant setting of breast cancer: a feasibility study.

PURPOSE: To evaluate standard radiation fields in locally advanced breast cancer using the information of a preoperative FDG-PET showing lymph node involvement. PATIENTS AND METHODS: In 15 consecutive patients, referred for postoperative radiotherapy, a standard simulation was performed. Planning CT was fused semiautomatically with preoperative FDG-PET, and thoracic wall/breast (clinical target volume [CTV]), lungs, and location(s) of axillary nodal involvement on FDG-PET (PET(ax)) were delineated. A dose computation was performed using the standard field simulation (plan-). If plan-resulted in inadequate dose delivery to PET(ax), a plan adaptation was performed to improve this deficit (plan+). Mean dose-volume histograms (DVHs) were made of the relevant structures for both plan- and plan+. Student's paired t-test was performed on all mean reference values.
RESULTS: In 13 patients an image fusion could be performed. Dose computation showed underdosage to the PET(ax) in eleven out of 13 patients. After plan adaptation, the dose delivered to PET(ax) could be increased, with a statistically significant difference (p < 0.01) in favor of plan+ for all reference values on the DVHs. This was achieved without changes in dose delivery to CTV or lungs.
CONCLUSION: Standard radiation portals in postoperative radiation therapy in breast cancer with lymph node involvement do not automatically result in an adequate dose delivery to the region of highest biological activity. With these preliminary results in a small series it was found feasible to correct this without compromising the dose to the CTV or lungs for patients in whom a preoperative FDG-PET shows nodal involvement.