Histogram analysis of dynamic contrast enhancement MRI in predicting long-term treatment effect of cervical cancer after chemoradiotherapy
10.3760/cma.j.issn.1005-1201.2019.03.011
- VernacularTitle:MRI动态增强成像参数直方图分析对局部晚期宫颈癌放化疗后的预后评估价值
- Author:
Bo ZHAO
1
;
Kun CAO
;
Xiaoting LI
;
Xiaofan LI
;
Hong ZHENG
;
Haitao ZHU
;
Yingshi SUN
Author Information
1. 北京大学肿瘤医院暨北京市肿瘤防治研究所医学影像科恶性肿瘤发病机制及转化研究教育部重点实验室 100142
- Keywords:
Uterine cervical neoplasms;
Long-time survival;
Magnetic resonance imaging;
Histogram analysis
- From:
Chinese Journal of Radiology
2019;53(3):218-223
- CountryChina
- Language:Chinese
-
Abstract:
Objective To investigate the potential value of pre-treatment dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in predicting long-term survival of cervical cancer treated by concurrent chemoradiotherapy. Methods Uterine cervical cancer patients who received chemo-radiation therapy during 2008 to 2014 in a single institute were retrospectively collected and followed up. Pre-treatment DCE-MRI images were retrieved to generate four hemodynamic related parameters including positive enhancement integral (PEI), maximum slop of increase (MSI), maximum slope of decrease (MSD) and signal enhancement ratio (SER). Region of interests were drawn manually on T2WI along the rim of tumors on each slice, and were then registered to DCE-MRI parametric maps. Histogram analysis software was used to calculate the mean, median, maximum, minimum,10th percentile, 90th percentile, kurtosis and skewness values for each DCE-MRI parameters. Median follow-up time was 54.7 months (range 2.1 to 94.6 months). Uni-and multivariable Cox regression analyses were used to evaluate correlation between the above values and the disease free survival (DFS). Kaplan-Meier curve was used to evaluate survival time. Results Of the 75 patients, 16 of them died from cervical cancer, 3 patients had metastasis, and 1 patient continued to progress. Median, mean, 10th and 90th percentile from MSI, and minimum, kurtosis, skewness were the influencing factors of disease-free survival of cervical cancer with concurrent chemoradiotherapy (P<0.1) Pearson relationship analysis and multivariable Cox regression analysis was performed which indicated PEI kurtosis value of cervical cancer was an independent influencing factor for cervical cancer without disease survival (hazard ratio 1.658, P=0.001). Conclusion PEI kurtosis was an independent factor for DFS of locally advanced cervical cancer treated with concurrent chemoradiotherapy.
