Dynamic Contrast-Enhanced MRI In Renal Tumors With Pharmacokinetic Modeling

Part I:Reproducibility of Dynamic Contrast-enhanced MRI in Renal Cell Carcinoma with Pharmacokinetic ParametersPurpose:To investigate the intra-and inter-observer as well as scan-rescan reproducibility of dynamic contrast-enhanced MRI (DCE-MRI) in evaluating the pharmacokinetic parameters of renal cell carcinoma (RCC) under non-continuous scanning mode.Materials and Methods:Our Institutional Review Board approved this prospective study. Written informed consent was obtained from all subjects. Patients underwent paired non-continuous DCE-MRI studies on a 3.0 T MR system. The extended-Tofts model and population-based arterial input function were used to calculate kinetic parameters. The whole-tumor region of interest at the slice with the maximum diameter of the renal cell carcinoma was defined by three observers. Intra-observer and inters-can differences were assessed using paired-t tests, while inter-observer differences one-way analysis of variance (ANOVA). Intra- and inter-observer reproducibility and scan-rescan reproducibility were evaluated using within-subject coefficient of variation (wCoV) and intraclass correlation coefficient (ICC).Results:Twenty-one clear cell RCCs (17 men,4 woman; age 37-69 years, mean age 54.6 years, mean size,5.0±2.2 cm) were recruited from September 2012 to November 2012. There were no significant intra-, inter-observer, or scan-rescan differences in parameters (all P>0.05). All ICCs for intra-and inter-observer agreements were>0.75 (P<0.05), while the scan-rescan agreement was moderate to good; Ve (0.798,95%CI: 0.438-0.937) and Kep (0.906,95% CI:0.299-0.914) had high ICCs (P<0.05). In intra-and inter-observer variability analyses, all parameters except Vp had low wCoV values. Ktrans and Ve had lower intra-observer variation compared to Kep, whereas all three of these parameters had similar inter-observer wCoV values. Regarding scan-rescan variability, Ktrans, Kep and Ve showed similar variation (15.6%,15.4% and 10.1%, respectively). Vp had the largest wCoV in all variability analyses.Conclusions:DCE-MRI demonstrated good reproducibility in the assessment of RCC using Ktrans,Kep and Ve as parameters under non-continuous scanning mode. Vp showed poor reproducibility, and thus may not be suitable for this scanning protocol.Part II:Reproducibility of Dynamic Contrast-enhanced MR Imaging in Renal Cell Carcinoma:histogram analysis of pharmacokinetic parametersPurpose:To investigate the intra- and inter-observer as well as scan-rescan reproducibility of dynamic contrast-enhanced MRI (DCE-MRI) in evaluating the pharmacokinetic parameters of renal cell carcinoma (RCC) using histogram method.Materials and Methods:Our Institutional Review Board approved this prospective study. Written informed consent was obtained from all subjects. Patients underwent paired non-continuous DCE-MRI studies on a 3.0 T MR system. The extended-Tofts model and population-based arterial input function were used to calculate kinetic parameters of renal cell carcinoma. Then histogram metrics of each pharmacokinetic parameter were generated automatically using ImageJ software. Intra-observer and inters-can differences were assessed using paired-^ tests, while inter-observer differences one-way analysis of variance (ANOVA). Intra-and inter-observer reproducibility and scan-rescan reproducibility were evaluated using intraclass correlation coefficients (ICCs).Results:Twenty-one clear cell RCCs (17 men,4 woman; age 37～69 years, mean age 54.6 years, mean size,5.0±2.2 cm) were recruited from September 2012 to November 2012. There were no significant intra-, inter-observer, or scan-rescan differences in parameters (all P>0.05). All ICCs for intra-observer agreements were>0.75 (0.814～0.999; P<0.001), while the inter-observer agreement was good except for kurtosis of Ktrans (ICC,0.728); as for scan-rescan analysis, all histogram metrics of Ve showed good agreement (ICC,0.758～0.798) and skewness and kurtosis of Ktrans showed poor agreement (ICC,0.352,0.308, respectively).Conclusions:DCE-MRI in renal cell carcinoma demonstrated good intra-and inter-observer reproducibility using histogram metrics of pharmacokinetic parameters, while in scan-rescan evaluation, only all histogram metrics of Ve showed good reproducibility, and skewness and kurtosis of Ktrans showed poor reproducibility.Part III:Subtype Differentiation of Renal Tumors Using Dynamic Contrast-enhanced MRI Pharmacokinetic ParametersPurpose:To investigate the value of dynamic contrast-enhanced MRI (DCE-MRI) pharmacokinetic parameters (Ktrans & Ve) on the differential diagnosis of renal tumors.Materials and Methods:Our Institutional Review Board approved this prespective study. Written informed consent was obtained from all subjects.110 non-continuous DCE-MRI studies of renal tumors including clear cell renal cell carcinoma (ccRCC) (n=65), papillary renal cell carcinoma (pRCC) (n=12), chromophobic enal cell carcinoma (cRCC) (n=9), uroepithelial carcinoma (UEC) (n=14), fat poor angiomyolipoma (fpAML) (n=10) were evaluated retrospectively on a 3.0 T MR system from September 2012 to December 2013. The extended-Tofts model and population-based arterial input function were used to calculate kinetic parameters. The whole-tumor region of interest at the slice with the maximum diameter of the renal tumors was defined by one radiologist. DCE-MRI parameters differences among five types of renal tumors were analyzed using one-way analysis of variance (ANOVA). The difference between renal benign and malignant tumors was handled using independent t test. So were difference between ccRCC and non-clear cell RCC, between fpAML and non-clear cell RCC, and between UEC and RCCs. If yes, receiver operation characteristics (ROC) curve was used to evaluate the efficacy of DCE-MRI parameters and sensitivity, specificity and Youden’s Index were calculated.Results:There were significant differences in Ktrans and Ve among five renal tumors (P <0.001 and P=0.044 respectively). fpAML demonstrated highest Ktrans (0.511±0.159 min-1) followed by ccRCC (0.459±0.190 min-1), for Ve, ccRCC showed highest value (0.526±0.173), then fpAML (0.459±0.190 min-1); pRCC showed lowest Ktrans (0.206 ±0.127 min-1) and Ve (0.385±0.157). The difference of Ktrans and Ve between renal benign and malignant rumors was not significant (P=0.064 and P=0.721, respectively), while that between ccRCC and non-clear cell RCC was significant (P <0.001 and P=0.002 respectively) and the area under ROC curve (AUC) when distinguishing ccRCC and non-clear cell RCC was 0.819 and 0.716 respectively, and a threshold Ktrans value of 0.33 min-1 permitted distinction with 76.9% sensitivity,71.4% specificity and Youden’s Index 0.483, and when Ve was greater than 0.317,86.2% sensitivity,52.4% specificity and Youden’s Index 0.386 can be obtained. The difference of Ktrans between fpAML and non-clear cell RCC was significant (P<0.001), so was between RCC and UEC (P=0.015); and the AUC was 0.924 and 0.766 respectively; Ktrans greater than 0.365 min-1 can distinguish fpAML and non-clear cell RCC with 100% sensitivity,71.4% specificity and Youden’s Index 0.762 and Ktrans greater than 0.228 min-1 can distinguish RCC and UEC with 84.9% sensitivity and 71.4% specificity and Youden’s Index 0.563. The difference of Ve between fpAML and non-clear cell RCC was not significant (P=0.069), so was between RCC and UEC (P=0.396).Conclusions:DCE-MRI kinetic parameters demonstrated promising value in the differential diagnosis of renal tumors, especially for RCC subtypes characterization, distinction between fpAML and non-clear cell RCC and between UEC and RCC.