Correlation of new two-dimensional geometrical parameters to lung and heart dose-volume parameters in breast cancer radiation therapy.

Abstract

Objective: To develop new two-dimensional geometric parameters for pulmonary and cardiac dose estimation in left-sided breast cancer radiation therapy without dose-volume histogram (DVH).

Methods: On the CT image of 90 patients with left breast cancer, treatment planning was performed using two opposed tangent fields with/without supraclavicular. The field-in-field technique and 6MV photons were used. From DVH dosimetric parameters of mean dose, Vx (x (Gy) =5, 10, 15, 20, 30, 40, 50) were calculated, and from heart and lung outlines on the beam's eye view, new geometric parameters of percent of lung area in tangent and supraclavicular fields (%area of the lung in the tangent (ALT), %ALS) and percent of heart in tangent field (%area of the heart in the tangent (AHT)) were measured. Correlation, regression, and diagnostic performance by receiver operating characteristic curve (ROC) were investigated for statistical analysis.

Results: The Pearson coefficient between %ALT and Vx (x = 10, 15, 20, 30, 40) show strong correlation in patient treatment with only opposed tangents (>0.85) and weaker in treatment by opposed tangents with supraclavicular (0.56-0.88), the %ALS indicate weak correlation (<0.5) and %AHT show strong correlation (0.93-0.98). The regression analysis shows a positive relation between %ALT and mean dose (R2 = 0.8), V20Gy (R2 = 0.9) in the lung (tangent treatment), and between %AHT and mean dose (R2 = 0.9), V20Gy (R2 = 1.0) in the heart. The ROC analysis shows by %ALT <20.3 for treatment by just opposed fields, %ALT <22.1% for treatment tangents with supra, and %AHT <11.6%, practical lung and heart dose constraints are addressed.

Conclusion: The proposed geometric parameters could replace previous one-dimensional maximum and central distances for predicting doses to lung and heart.

Advances in knowledge: This study presents simple geometric parameters that could estimate pulmonary and cardiac dose in left breast cancer treatment from a 2D radiograph.