Comparison Of 3d Ultrasound, Ct, And Mri For Target Definition In The Post Lumpectomy Breast

Definition of the lumpectomy cavity is a critical step in planning breast boost and partial breast irradiation. Several studies have shown limitations of CT-based planning. CT/MRI fusion can improve visualization, but is expensive and requires a separate imaging session. Fused CT/3D ultrasound (3DUS) is a lower cost alternative, with both images performed in one session. A recent study has demonstrated that fused CT/3DUS improves interobserver variability compared to CT alone. We compared cavity visualization of fused CT, 3DUS and MRI and investigated whether 3DUS assists in the visualization of borders not seen on CT, using MRI as a baseline. Twenty women, post lumpectomy, were enrolled in the IRB approved study, 18 of which had evaluable images. CT scans were obtained with patients supine, arms extended behind the head. 3DUS was performed during the CT session and co-registered with the CT. A 3T MRI without contrast was performed immediately afterward in the same position using a flexible six-element body RF matrix coil. Three contourers assessed image clarity from 0-5 using the BCCA Seroma Clarity Scale. Image fusion tools were used to compare the modalities for utility in treatment planning, and the frequency with which 3DUS contributed to contour definition of the CT image. The mean single-modality clarity scores were 3.76 for 3DUS, 3.59 for CT, and 4.44 for MRI. The CT showed a homogenous area, with minimal distinction between fluid/soft tissue, between post-operative stranding and normal breast, and indistinct borders between the seroma and chest wall. The 3DUS showed the dark anechoic fluid filled excision cavity plus the cavity wall, as well as clear demarcation between seroma/ soft tissue and seroma/chest wall, with no stranding. The 3DUS contributed to identification of CT image borders in 17/18 patients. T2 MRI had the highest clarity, demonstrating a concentric dark area surrounding a bright well-defined seroma cavity. The anechoic area of the 3DUS corresponded with the bright area on T2 MRI in size and configuration in 17/18 patients. This correlation was excellent in 12/18 patients, good in the remaining patients. CT did not isolate this bright T2 region from the surrounding postoperative complex. The use of CT as a single modality for identification of the post lumpectomy cavity is limited. The 3DUS has a clarity score higher than CT but not as high as MRI. There is a clear correlation between the seroma seen as an anechoic area on the 3DUS and as a bright area on T2 MRI, both of which provide information not seen CT. Fused CT/3DUS images provide complementary information to the radiation oncologist for defining the lumpectomy cavity in the majority of patients, and is a low-cost alternative to CT/MRI fusion without the need for a separate imaging session.