| Purpose: To compare the difference of evaluating the angiogenesis and pathological grade of HCC between dual-input and single-input two-compartment pharmacokinetic model of DCE-MRI.Materials and Methods: Written informed consent was obtained from all subjects.Patients underwent non-continuous DCE-MRI studies on a 3.0-T system. DCE-MRI pharmacokinetic parameters including volume transfer constant (Ktrans), extracelluar extravascular volume fraction (ve), exfflux rate constant (kep), blood plasma volume fraction (vp) were calculated using dual-input and single-input two-compartmental model respectively, and Hepatic perfusion index (HPI) was calculated using a dual-input two-compartmental model. The value of parameters was measured at the slice with the maximum diameter of HCC by using the whole-tumor region of interest (ROI).Pathological results included pathologic diagnosis of HCC classified as grade Ⅰ～Ⅳaccording to the major Edmondson-Steiner’s classification, subgroups of low-grade HCC(≤ grade Ⅱ) and high-grade HCC (> grade Ⅱ),and CD34 as a marker of HCC micro vessel density (MVD) calculation. The correlation between all parameters and MVD of HCC, as well as pathological grade of HCC were assessed by using Spearman or Pearson correlation test. The differences of all parameters between low-grade and high-grade HCC were compared by using independent-samples t test or Mann-Whitney U test. MVD were divided into two groups according to mean value. The performance evaluation of MVD and pathological grade were compared among the different parameters by using receiver operating characteristic (ROC) analysis.Results: Forty-nine HCC patients (40 men, 9 women, age 35～78 years, median age 62 years, mean size of tumors, 5.7±2.9cm) were recruited from June 2016 to January 2017.the pathological grade in 49 HCC patients were divided to grade Ⅱ 20 cases, grade Ⅲ 21 cases, grade Ⅳ 8 cases, and no grade Ⅰ. MVD value in 29 HCC was range 8-94, mean value 45.2±25.5. The parameters of dual-input model (Ktrans,kep and vp) were significantly positive correlated with MVD of HCC (r＝0.838, 0.544, 0.716,P＝0.000, 0.002, 0.000,respectively), and parameters of single-input model (Ktrans, kep ) were significantly positive correlated with MVD of HCC (r = 0.718, 0.3999, P = 0.000, 0.032, respectively), vpof single-input model, ve and HPI were no correlated with MVD of HCC. Among all parameters, only ve of dual-input model was significantly negative correlated with pathological grade (r = -0.489, P = 0.000). In the evaluation of MVD of HCC, the evaluation of performance in descending order was: Ktran of dual-input model >vP of dual-input model > Ktrans of single-input model, kep of dual-input model > kep of single-input model.Conclusion: DCE-MRI pharmacokinetic parameters of dual-input two-compartment model demonstrated advantages of evaluating the angiogenesis and pathological grade of HCC over those of single-input two-compartment model. Dual-input two-compartment model may more suitable for pharmacokinetic analysis of DCE-MRI in HCC.Purpose: To compare the difference of evaluating the angiogenesis and pathological grade of HCC between IVIM-DWI and pharmacokinetic anlysis of DCE-MRI.Materials and Methods: Written informed consent was obtained from all subjects.Patients underwent respiratory triggered DWI with 8 b-values (0～800 s/mmm ) and non-continuous DCE-MRI studies on a 3.0-T system. Perfusion fraction (f),pseudodiffusion coefficient (D*) and pure diffusion coefficient (D) were calculated using segmented biexponential analysis. ADCtot was calculated with monoexponential fitting of all b-values DWI data. DCE-MRI pharmacokinetic parameters including volume transfer constant (Ktrans), extracelluar extravascular volume fraction (ve), exfflux rate constant (kep),blood plasma volume fraction (vp) and Hepatic perfusion index (HPI) were calculated using a dual-input two-compartmental model. The value of parameters was measured at the slice with the maximum diameter of HCC by using the whole-tumor region of interest(ROI). Pathological results included pathologic diagnosis of HCC classified as grade I-IV according to the major Edmondson-Steiner’s classification, subgroups of low-grade HCC(< grade II) and high-grade HCC (>grade II), and CD34 as a marker of HCC MVD calculation. The correlation between IVIM-DWI parameters and DCE-MRI pharmacokinetic parameters, the correlation between all parameters and MVD of HCC and the correlation between all parameters and pathological grade were assessed by using Spearman or Pearson correlation test. The differences of all parameters between low-grade and high-grade HCC were compared by using independent-samples t test. MVD were divided into two groups according to mean value. The performance evaluation of MVD and pathological grade were compared among the different parameters by using receiver operating characteristic (ROC) analysis.Results: Forty-four HCC patients (36 men, 8 women, age 37～74 years, median age 61 years, mean size of tumors, 5.9±2.1cm) were recruited from June 2016 to January 2017.the pathological grade in 44 HCC patients were divided to grade Ⅱ 18 cases, grade Ⅲ 19 cases, grade Ⅳ 7 cases, and no grade Ⅰ. MVD of 29 cases, range 8-94, mean value 45.2±25.5. The parameter f was significantly positive correlated with parameter Ktrans and vp (r = 0.336, 0.564, P = 0.026, 0.000), The parameter D* was significantly positive correlated with parameter Ktrans and vp (r = 0.504, 0.385, P = 0.000, 0.010, respectively),and The parameter D was significantly positive correlated with parameter ve (r = 0.442, P=0.003). The parameters (f, D*, Ktrans,kep,and vp) were significantly positive correlated with MVD of HCC (r = 0.580, 0.509, 0.838, 0.544, 0.716, P = 0.001, 0.005, 0.000, 0.002,0.000, respectively). The parameters (D, ADCtot and ve) were significantly negative correlated with pathological grade (r = -0.546, -0.411, -0.456, P = 0.000, 0.006, 0.001,respectively). In the evaluation of MVD of HCC, the evaluation of performance in descending order was: Ktrans>vp>f>kep>D*. In the evaluation of pathological grade of HCC, the evaluation of performance in descending order was: D>ADCtot>v.Conclusion: IVIM-DWI parameters were correlated with DCE-MRI pharmacokinetic parameters. Pharmacokinetic model of DCE-MRI was better than IVIM-DWI for evaluating angiogenesis of HCC, while IVIM-DWI was better than Pharmacokinetic model of DCE-MRI for evaluating pathological grade of HCC. Combining application of two functional imaging techniques might provide more comprehensive evaluation of angiogenesis and pathological grade of HCC. For the patients who can not receive intravenous gadolinium-based contrast media, IVIM-DWI will be an alternative technique. |
PDF Full Text Request