Mei-Lin Gong1,2, Lian Li3, Kang Li2, Shi-Jian Li4, Ming Wen1. 1. Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China. 2. Department of Radiology, Chongqing General Hospital / Chongqing Hospital of the University of Chinese Academy of Sciences, Chongqing 400013, China. 3. Department of Radiology, The First Affiliated Hospital of Army Medical University, Chongqing 400038, China. 4. Department of Urology, Chongqing General Hospital / Chongqing Hospital of the University of Chinese Academy of Sciences, Chongqing 400013, China.
Abstract
OBJECTIVE: To investigate the relationship between the apparent diffusion coefficient (ADC) histogram parameters based on the whole tumor and the pathological grade and lymph node metastasis (LNM) of PCa. METHODS: This retrospective study included 82 cases of PCa confirmed pathologically and subjected to MRI preoperatively. We obtained a series of ADC histogram parameters, such as ADCmean, ADCmedian, ADC25%, ADC75%, entropy, and histogram width, by processing the ADC images via the Firevoxel Post-Processing and the SPSS24 software. We compared the parameters between the high-risk and low- or moderate-risk groups as well as between the LNM-positive and LNM-negative groups of the patients, and analyzed the diagnostic performance of the parameters with statistically significant differences. RESULTS: The high-risk group, compared with the low- or moderate-risk one, showed a significantly lower ADCmean (［590 ± 120］ vs ［837 ± 142］ ×10－6 mm2/s, P < 0.01), ADCmedian (［560 ± 117］ vs ［804 ± 139］ ×10－6 mm2/s, P < 0.01), ADC25% (［446.5 ± 98］ vs ［717 ± 118］ ×10－6 mm2/, P < 0.01) and ADC75% (［667 ± 132］ vs ［931 ± 167］ ×10－6 mm2/s, P < 0.01). The ADCmean manifested the highest diagnostic performance, with an AUC of 0.907, a sensitivity of 0.933 and a specificity of 0.796. No statistically significant difference was found between the high-risk and the low- or moderate-risk one in entropy (3.58 ± 0.39 vs 3.63 ± 0.42, P = 0.238) or the histogram width (［540 ± 73］ vs ［520 ± 65］ ×10－6 mm2/s, P = 0.086). Both entropy and the histogram width were remarkably higher in the LNM-positive than in the LNM-negative group (3.95 ± 0.41 vs 3.12 ± 0.45, P < 0.01; ［578 ± 59］ vs ［455 ± 68］ ×10－6 mm2/s, P < 0.01), and the former had an even higher diagnostic performance, with an AUC of 0.836, a sensitivity of 0.887 and a specificity of 0.781. There were no statistically significant differences between the LNM-positive and LNM-negative groups in the ADCmean (［768 ± 135］ vs ［790±128］ ×10－6 mm2/s, P = 0.402), ADCmedian (［759 ± 110］ vs ［775 ± 121］ ×10－6 mm2/s, P = 0.225), ADC25% (［643 ± 91］ vs ［657 ± 89］ ×10－6 mm2/s, P = 0.654) or ADC75% (［895 ± 127］ vs ［872 ± 129］ ×10－6 mm2/s, P = 0.926). CONCLUSIONS: ADC histogram parameters are related to pathological grade and LNM of PCa, and the analysis of the ADC histogram based on the whole tumor has an important value for preoperative evaluation and prognostic estimation of the malignancy.
OBJECTIVE: To investigate the relationship between the apparent diffusion coefficient (ADC) histogram parameters based on the whole tumor and the pathological grade and lymph node metastasis (LNM) of PCa. METHODS: This retrospective study included 82 cases of PCa confirmed pathologically and subjected to MRI preoperatively. We obtained a series of ADC histogram parameters, such as ADCmean, ADCmedian, ADC25%, ADC75%, entropy, and histogram width, by processing the ADC images via the Firevoxel Post-Processing and the SPSS24 software. We compared the parameters between the high-risk and low- or moderate-risk groups as well as between the LNM-positive and LNM-negative groups of the patients, and analyzed the diagnostic performance of the parameters with statistically significant differences. RESULTS: The high-risk group, compared with the low- or moderate-risk one, showed a significantly lower ADCmean (［590 ± 120］ vs ［837 ± 142］ ×10－6 mm2/s, P < 0.01), ADCmedian (［560 ± 117］ vs ［804 ± 139］ ×10－6 mm2/s, P < 0.01), ADC25% (［446.5 ± 98］ vs ［717 ± 118］ ×10－6 mm2/, P < 0.01) and ADC75% (［667 ± 132］ vs ［931 ± 167］ ×10－6 mm2/s, P < 0.01). The ADCmean manifested the highest diagnostic performance, with an AUC of 0.907, a sensitivity of 0.933 and a specificity of 0.796. No statistically significant difference was found between the high-risk and the low- or moderate-risk one in entropy (3.58 ± 0.39 vs 3.63 ± 0.42, P = 0.238) or the histogram width (［540 ± 73］ vs ［520 ± 65］ ×10－6 mm2/s, P = 0.086). Both entropy and the histogram width were remarkably higher in the LNM-positive than in the LNM-negative group (3.95 ± 0.41 vs 3.12 ± 0.45, P < 0.01; ［578 ± 59］ vs ［455 ± 68］ ×10－6 mm2/s, P < 0.01), and the former had an even higher diagnostic performance, with an AUC of 0.836, a sensitivity of 0.887 and a specificity of 0.781. There were no statistically significant differences between the LNM-positive and LNM-negative groups in the ADCmean (［768 ± 135］ vs ［790±128］ ×10－6 mm2/s, P = 0.402), ADCmedian (［759 ± 110］ vs ［775 ± 121］ ×10－6 mm2/s, P = 0.225), ADC25% (［643 ± 91］ vs ［657 ± 89］ ×10－6 mm2/s, P = 0.654) or ADC75% (［895 ± 127］ vs ［872 ± 129］ ×10－6 mm2/s, P = 0.926). CONCLUSIONS: ADC histogram parameters are related to pathological grade and LNM of PCa, and the analysis of the ADC histogram based on the whole tumor has an important value for preoperative evaluation and prognostic estimation of the malignancy.