Distinction between postoperative recurrent glioma and delayed radiation injury using MR perfusion weighted imaging.

INTRODUCTION: Distinction between postoperative recurrent glioma and radiation injury remains a tough diagnostic problem for routine imaging methods. The purpose of this study is to evaluate the differentiated effectiveness of perfusion weighted imaging (PWI) for the two entities. METHODS: PWI was performed using Siemens 3.0-T MR system for 35 patients with new contrast-enhancing lesions at the site of treated glioma. Regions of interest (ROIs) were manually drawn at the contrast-enhancing lesion and peri-lesion edema areas. For calculation of standardised relative cerebral blood volume (rCBV) ratios, the same size ROIs were drawn at the area of contralateral hemisphere normal white matter on rCBV maps. At least five ROIs were selected at each lesion. The rCBV values were measured and the rCBV ratios were calculated. The maximum rCBV (rCBV(max)) ratio at each region was chosen for analysis. The patients were divided into two groups: tumour recurrence and radiation injury. The mean rCBV(max) ratios were compared between the two groups. RESULTS: The mean rCBV(max) ratio in the contrast-enhancing lesion was significantly higher in the tumour recurrence (4.36 ± 1.98) compared with that (1.28 ± 0.64) in the radiation injury (P < 0.01). The mean rCBV(max) ratio in the peri-lesion edema was also significantly higher in the tumour recurrence (1.79 ± 0.51) compared with that (0.85 ± 0.28) in the radiation injury (P < 0.05). A recurrent tumour was suggested when the rCBV(max) ratio >2.15 based on the receiver operating characteristic curve. Four patients with recurrent tumour and three with radiation injury were misclassified. CONCLUSION: PWI is a useful method to distinguish tumour recurrence and radiation injury.

Comparison of volumetric methods for tumor measurements on two and three dimensional MRI in adult glioblastoma.

INTRODUCTION: Change in tumor size is a frequent endpoint in cancer clinical trials, but whether change in size should be measured using volume on two-dimensional (2D) or three-dimensional (3D) images is not certain. We compared volumetric measurements on post-contrast 2D and high-resolution 3D T1-weighted MR images (T1WI) in evaluating tumor response in glioblastoma multiforme (GBM). METHODS: Tumor volume measurements were performed on 86 MRI studies from 37 adult patients with GBM on post-contrast 5 mm 2D T1WI and isotropic high-resolution T1WI. The means of the two volumes were compared and their association was analyzed. RESULTS: There is no significant difference between volumes measured on 2D and 3D in 86 scans (Z = 0.63, p = 0.53), and a high correlation was revealed between them (r = 0.95, 95% CI: 0.93-0.97, p < 0.001). When the percentage changes were categorized into traditional tumor response criteria (complete response/partial response/stable disease/progressive disease), the kappa coefficient between the volume on 2D and volume on 3D was 0.80 (95% CI: 0.57-1.03, p < 0.05) with an overall agreement of 84%. CONCLUSIONS: Volume on post-contrast 2D T1WI appears comparable to volume on 3D T1WI and should be a practical alternative to volume on 3D in evaluating tumor response.

Distinction between postoperative recurrent glioma and radiation injury using MR diffusion tensor imaging.

INTRODUCTION: This study aims to evaluate the differentiated effectiveness of MR diffusion tensor imaging (DTI) to postoperative recurrent glioma and radiation injury. METHODS: Conventional MRI and DTI examination were performed using Siemens 3.0 T MR System for patients with new contrast-enhancing lesions at the site of treated tumor with postoperative radiotherapy. The region of interest was manually drawn on ADC and FA maps at contrast-enhancing lesion area, peri-lesion edema, and the contra-lateral normal white matter. Then ADC and FA values were measured and, the ADC ratio and FA ratio were calculated. Twenty patients with recurrent tumor and 15 with radiation injury were confirmed by histopathologic examination (23 patients) and clinical imaging follow-up (12 patients), respectively. The mean ADC ratio and FA ratio were compared between the two lesion types. RESULTS: The mean ADC ratio at contrast-enhancing lesion area was significantly lower in patients with recurrent tumor (1.34 ± 0.15) compared to that with radiation injury (1.62 ± 0.17; P < 0.01). The mean FA ratio at contrast-enhancing lesion area was significantly higher in patients with recurrent tumor (0.45 ± 0.03) compared to that with radiation injury (0.32 ± 0.03; P < 0.01). Neither mean ADC ratio nor FA ratio in edema areas had statistical difference between the two groups. A recurrent tumor was suggested when either ADC ratio <1.65 or/and FA ratio >0.36 at contrast-enhancing lesion area according to the receiver operating characteristics curve analysis. Three patients with recurrent tumor and two with radiation injury were misclassified. CONCLUSION: DTI is a valuable method to distinguish postoperative recurrent glioma and radiation injury.