Surgical clips for observation of respiratory motion of the breast Via 2D fluoroscopy and 3D CT scans

Abstract

Many breathing techniques have been applied for an accurate treatment of the breast. A normal breathing technique that could provide an equally accurate treatment with less intervention and shorter treatment time is worth investigating by thoroughly observing the respiratory motion of the breast via 2D fluoroscopy and 3D CT scans with surgical clips for position verification of the tumor bed. A group of 22 patients with surgical clips implanted in the lumpectomy cavity during breast-conserving surgery as effective internal markers of the length, width, and depth of the tumor bed were involved. The group was divided into two subgroups according to data acquisition methods; namely, 2D fluoroscopy and 3D CT scan. Seven out of twenty two patients were involved in the 2D fluoroscopic study, and the rest were involved in the 3D CT study. No breathing control device was used in this study. To compensate for the sample size and to get a more favorable result for direct clinical application, we introduced the standard error of the mean (SEM), including 95% confidence interval based on t-distribution. The low-level in-house software for identifying the individual clip locations and setup adjustment was used to improve accuracy of the breast irradiation treatment. For the 2D fluoroscopic study, the average movement and standard deviation(SD) of the clip were 1.30 ± 0.65 (median: 1.23, 95% CI: 1.01 - 1.59), 0.77 ± 0.48 (median: 0.55, 95% CI: 0.56 - 0.98), and 1.20 ± 0.46 (median: 1.06, 95% CI: 1.03 - 1.37) for the anterior-posterior (AP), lateral, and tangential directions, respectively. Our results were compatible with those of Kinoshita et al.'s 3D study. In our 3D CT study, excursion lengths during a cycle of breathing were measured. The average excursion length and SD of the center-of-mass (COM) of the clips were 5.1 ± 3.5 (median: 4.7, 95% CI: 4.34 - 5.86), 0.8 ± 0.8 (median: 0.5, 95% CI: 0.62 - 0.98), and 4.9 ± 4.3 (median: 3.1, 95% CI, 3.97 - 5.83) for the AP, lateral, and superior-inferior (SI) directions, respectively. Both the isocenter and the diaphragm movements were measured in the 3D CT study. The difference between the COM and the isocenter is considered to be not statistically significant (p > 0.7) for the three orthogonal directions. However, the p value for the difference between the clip and the lung is less than 0.001. We saw no strong correlation between the clip and the diaphragm movements. No immobilization devices or techniques were needed to control the movement of the breast tumor bed because our results suggest that the tumor bed is not significantly moved by breathing motion compared to the diaphragm during normal breathing. The motions of the breast in the AP and the SI directions were significantly greater than that in the lateral direction, which suggests that anisotropic and direction-dependent clinical margin may be more favorable in normal breathing clinical situations, although the reduction of internal margins should be performed with great care due to the individual tumor motion.