Correlation of whole-volume apparent diffusion coefficient histogram parameters of rectal adenocarcinoma with clinico-pathological characteristics
10.11855/j.issn.0577-7402.2019.09.09
- Author:
Jia-Rong CHENG
1
Author Information
1. Imaging and Nuclear Medicine, Gansu University of Traditional Chinese Medicine
- Publication Type:Journal Article
- Keywords:
Apparent diffusion coefficient;
Histogram analysis;
Magnetic resonance imaging;
Prognosis;
Rectalcancer
- From:
Medical Journal of Chinese People's Liberation Army
2019;44(9):773-778
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To investigate the correlation between the histogram parameters of apparent diffusion coefficient (ADC) with clinico-pathological characteristics of rectal adenocarcinoma. Methods: The preoperative magnetic resonance imaging data of 64 patients diagnosed by pathology as rectal adenocarcinoma were collected. OmniKinetics software was employed to draw the outline of the ADC image on each layer of the tumor. The whole-volume ADC histogram parameters were automatically calculated with the post-processing software, including the minimum (ADCmin), maximum (ADCmax), median (ADCmedian), mean (ADCmean), 10th percentile (ADC10), 25th percentile (ADC25), 75th percentile (ADC75), 90th percentile (ADC90), skewness, and kurtosis. The differences in ADC parameters of tumor markers, immunohistochemical index, tumor size and lymph node metastasis were compared, and multiple linear regression was performed to analyze the correlation between the ADC histogram parameter and the clinico-pathological characteristics of rectal adenocarcinoma. Results: The ADC25 values were significantly lower in the groups of lymph node metastasis, Ki-67 ≥50%, CA199 ≥37 U/ml and CA72-4 ≥8.2 U/ml than those in the groups of lymph node non-metastasis, Ki-67 <50%, CA199 <37 U/ml and CA72-4 <8.2 U/ml (P<0.05 or P<0.01). The ADC75 values were lower in CD31 positive group and CD56 positive group than those in their negative groups (P<0.05). There was no correlation between skewness and clinical pathological features (P>0.05). Multiple linear regression analysis reveled that positive CD31 was significantly correlated with ADCmean, ADC25 and ADC75 (P<0.05), and the ADC25 showed the greatest effect on CD31 positive (standard regression coefficient was 0.210). Conclusion: ADCmax, ADCmean, ADCmedian, ADC10, ADC25 and ADC75 possess high reliability in diagnosis of rectal adenocarcinoma. Among them, ADCmean, ADC25 and ADC75 have higher correlation with tumour markers and immunohistochemical indicators, may be used as important imaging biomarkers for evaluating the prognosis of rectal adenocarcinoma.