Purpose: To develop an ultrasound histogram technique capable of evaluating radiation induced acute toxicity in patients receiving radiotherapy (RT) for early-stage breast cancer.
Methods: Thirty-one patients who underwent conservative breast surgery followed by hypofractionated RT were enrolled in the IRB-approved study. RT involves 15 fractions of whole breast irradiation with simultaneous integrated boost. Each participant received longitudinal breast ultrasound at four timepoints: prior to RT, last day of RT, 3 weeks and 10 weeks post RT. Ultrasound scans were performed using a clinical ultrasound scanner with a 12-MHz linear probe on the irradiated breast and contralateral breast at 5 locations (3, 6, 9, 12 o’clock and tumor bed). The scans on the contralateral breast served as controls. RT-induced breast toxicity was graded in 3 levels (11 patients rated as mild, 10 as moderate, and 10 as severe) based in ultrasound B-mode images. Normalized echo-intensity histograms were computed for the regions of interest including the skin and subcutaneous tissue. The histogram difference (Bhattacharyya distance) between irradiated and contralateral breasts was calculated to capture radiation-induced tissue changes.
Results: Bhattacharyya distance increased from mild to severe toxicity in all locations. Bhattacharyya distance also increased from mild to severe toxicity at various timepoints (Mild: 0.28 ± 0.06 Pre-RT, 0.32 ± 0.10 Last-RT, 0.32 ± 0.09 at 3-week post-RT, 0.31 ± 0.13 at 10-week post RT; Moderate: .39 ± 0.12 Pre-RT, 0.37 ± 0.17 Last-RT, 0.38 ± 0.15 3-week post-RT, 0.4 ± 0.14 10-week post-RT; Severe: 0.42 ± 0.11 Pre-RT, 0.47 ± 0.1 Last-RT, 0.42 ± 0.14 3-week post-RT, 0.50 ± 0.14 10-week post-RT).
Conclusion: Radiation-induced breast toxicity remains a major obstacle to effective RT. This study has shown that ultrasound histogram analysis is feasible and may be developed into a low-cost imaging method for radiation-induced tissue toxicity monitoring and assessment.