Glioma: Application of Histogram Analysis of Pharmacokinetic Parameters from T1-Weighted Dynamic Contrast-Enhanced MR Imaging to Tumor Grading

Article Figures & Data

Download figure
Open in new tab
Download powerpoint
Fig 1.
ROIs manually defined on axial T2WI coregistered with parametric maps in a 43-year-old woman with a grade II glioma (A) and the histogram and cumulative histogram from the entire-tumor volume (B–D) are presented.
Download figure
Open in new tab
Download powerpoint
Fig 2.
ROIs manually defined on axial T2WI coregistered with parametric maps in a 61-year-old man with a grade III glioma (B) and the histogram and cumulative histogram from the entire-tumor volume (B–D) are presented.
Download figure
Open in new tab
Download powerpoint
Fig 3.
ROIs manually defined on axial T2WI coregistered with parametric maps in a 37-year-old woman with a grade IV glioma (C) and the histogram and cumulative histogram from the entire-tumor volume (B–D) are presented.
Download figure
Open in new tab
Download powerpoint
Fig 4.
A histogram and cumulative histogram were derived from the total parametric values of all masses (n = 28) according to each grade. A–F, The histogram and cumulative histogram show that lower grade gliomas present a higher relative frequency at low parametric values and a lower relative frequency at high parametric values compared with higher grade gliomas for all of the parameters. C and F, However, in the Vp histogram, the relative frequency of grade IV gliomas is higher than that of grade III gliomas at a number of low values.

Table 1:

The pharmacokinetic parameters in each grade of glioma^{^a}

	Grade II^{^b}/Low-Grade^{^c}	Grade III^{^b}	Grade IV^{^b}	High-Grade^{^c}
K^trans ^{^d}	0.027 ± 0.041	0.057 ± 0.091	0.158 ± 0.226	0.122 ± 0.195
Ve^{^e}	5.809 ± 10.381	15.469 ± 26.165	45.103 ± 63.888	35.059 ± 55.920
Vp^{^e}	3.858 ± 3.958	5.735 ± 5.772	10.624 ± 12.396	8.826 ± 10.723

↵a All data are means.
↵b There were statistically significant differences among grades (P < .01).
↵c There were statistically significant differences between low- and high-grade (P < .01).
↵d Data for K^trans are in units of minute⁻¹.
↵e Data for Ve and Vp are percentages.

Table 2:

ROC results of the percentiles with the best AUC for glioma grading

Parameter	AUC^{^a}	Sensitivity (%)^{^b}	Specificity (%)^{^b}	Cutoff Value^{^c}	P Value
K^trans
C98	0.912 (0.742–0.985)	76.19 (16/21)	100 (7/7)	0.277	<.01
Ve
C90	0.939 (0.779–0.994)	85.71 (18/21)	100 (7/7)	19.70	<.01
Vp
C84	0.769 (0.572–0.906)	61.9 (13/21)	100 (7/7)	11.71	<.01

↵a Data in parentheses are 95% confidence intervals.
↵b Sensitivity and specificity for identifying high-grade tumors. Data in parentheses are the numbers used to calculate the percentages.
↵c Data for K^trans are in units of minute⁻¹, and data for Ve and Vp are percentages.