Comparison of FDG-PET/CT and CT for delineation of lumpectomy cavity for partial breast irradiation

Purpose: The success of partial breast irradiation critically depends on proper target localization. We examined the use of fluorodeoxyglucose-positron emission tomography (FDG-PET)/computed tomography (CT) for improved lumpectomy cavity (LC) delineation and treatment planning. Methods and Materials: Twelve breast cancer patients underwent FDG-PET/CT on a GE Discovery scanner with a median time-from surgery to PET/CT of 49 days. The LC was contoured on the CT scan by a radiation oncologist and, together with a nuclear medicine physician, on the PET/CT scan. The volumes were calculated and compared in each patient. Treatment planning target volumes (PTVs) were calculated by expanding the margin 2 cm beyond the LC, maintaining a 5-mm margin from the skin and chest wall, and the treatment plans were evaluated. In addition, a study with a patient-like phantom was conducted to evaluate the effect that the window/level settings might have on contouring. Results: The margin of the LC was well visualized on all FDG-PET images. The phantom results indicated that the difference between the known volume and the FDG-PET-delineated volume was <10%, regardless of the window/level settings. The PET/CT volumes were larger than the CT volumes in all cases (median volume ratio, 1.68; range, 1.24-2.45; p = 0.004). The PET/CT-based PTvs were also larger than the CT-based PTV (median volume ratio, 1.16; range, 1.08-1.64; p = 0.006). In 9 of 12 patients, a CT-based treatment plan did not provide adequate coverage of the PET/CT-based PTV (99% of the PTV received <95% of the prescribed dose), resulting in substantial cold spots in some plans. In these cases, treatment plans were generated which were specifically designed to cover the larger PET/CT-based PTV. Although these plans showed an increased dose to the normal tissues, the increases were modest: the non-target breast volume receiving >= 50 Gy, lung volume receiving >= 30 Gy, and heart volume receiving >= 5 Gy increased by 5.7%, 0.8%, and 0.2%, respectively. The normal tissue dose-volume objectives were still met with these plans. Conclusion: The results of our study have shown that FDG-PET/CT can be used to define the LC volume. The increased FDG uptake was likely a result of postoperative inflammation in the LC. The targets defined using PET/CT were significantly larger than those defined with CT alone. Our results have shown that treatment plans can be generated to cover these larger PET/CT target volumes with only a modest increase in irradiated tissue volume compared with CT-determined PTVs. (C) 2008 Elsevier Inc.