Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan, Republic of China; Department of Radiology, Tri-Service General Hospital, Taipei, Taiwan, Republic of China.
Magn Reson Imaging. 2013 Oct;31(8):1278-84. doi: 10.1016/j.mri.2013.03.018. Epub 2013 May 8.
To investigate parotid perfusion in early-to-intermediate stage after parotid-sparing radiation dose using fat-saturated DCE-MRI, and to verify whether the perfusion alteration was related to radiation dose and the PSV.
Thirty-two parotid glands from 16 consecutive patients with pathologically proven nasopharyngeal carcinoma treated by IMRT were examined. The parotid glands received a radiation dose of 28.9±3.9Gy with a PSV of 43.1%±13.9%. Perfusion parameters were calculated using time-shifted Brix model from fat-saturated DCE-MRI data before (pre-RT) and in early-to-intermediate stage after (post-RT) IMRT. Paired t-test was used to evaluate perfusion changes, while Pearson's correlation test was used to examine perfusion dependency on radiation dose and PSV. For multiple comparisons Bonferroni correction was applied.
Successful fat saturation was achieved in 29 of 32 parotid glands. Compared with pre-RT, the post-RT parotid glands showed significantly higher A, peak enhancement, and wash-in slope, plus a lower Kel, suggesting a mixed effect of increased vascular permeability and acinar loss. Linear regression showed that peak enhancement was positively associated with radiation dose in post-RT parotid glands. Kel and slope were negatively associated with PSV, while time-to-peak was positively associated with PSV significantly.
Our results suggest that time-shifted Brix model is feasible for quantifying parotid perfusion using DCE-MRI. The perfusion alterations in early-to-intermediate stage after IMRT might be related to a mixed effect of increased vascular permeability and acinar loss with dose and PSV dependencies.
利用脂肪饱和 DCE-MRI 研究腮腺保留放疗后早中期腮腺灌注情况，并验证灌注改变是否与剂量和 PSV 相关。
对 16 例经病理证实的鼻咽癌接受调强放疗的连续患者的 32 个腮腺进行了检查。腮腺接受了 28.9±3.9Gy 的放射剂量，PSV 为 43.1%±13.9%。使用时间移位 Brix 模型从脂肪饱和 DCE-MRI 数据中计算灌注参数，这些数据来自调强放疗前（放疗前）和早中期（放疗后）。使用配对 t 检验评估灌注变化，使用 Pearson 相关检验评估灌注对剂量和 PSV 的依赖性。对于多次比较，应用 Bonferroni 校正。
在 32 个腮腺中，成功实现了 29 个腮腺的脂肪饱和。与放疗前相比，放疗后腮腺的 A、峰值增强和洗脱斜率明显升高，Kel 降低，提示血管通透性增加和腺泡丢失的混合效应。线性回归显示，放疗后腮腺的峰值增强与剂量呈正相关。Kel 和斜率与 PSV 呈负相关，而达峰时间与 PSV 呈正相关。
我们的结果表明，时间移位 Brix 模型可用于使用 DCE-MRI 定量腮腺灌注。调强放疗后早中期的灌注改变可能与血管通透性增加和腺泡丢失的混合效应有关，与剂量和 PSV 有关。
