The Study Of Diffusion-Weighted Imaging In Sprague-Dawley Rat Osteosarcoma Model

Objective To establish SD rat osteosarcoma model andevaluate the role of DWI in the identification of the necrosis andthe histologic structure of the tumor, and the relationshipbetween the microcirculation and ADC values of the tumor. Methods The tumor cells(UMR-106) in logarithmgrowth stage were transplanted in the back of 16 male SD ratsby subcutaneous injection of a suspension. The planted tumorswere scanned with a clinical 1.5 T ultraconduct MR system.Routine sequences, contrast enhanced T1-SE were obtained.DWI was performed on the platform of the single shot echoplanar imaging(EPI). Two b-values(0 and 1000s/mm2) imageswere acquired with diffusion gradients applied in all 3orthogonal directions. The rats were sacrificed immediatelyafter imaging and the tumors were cut into sections which werestained with hematoxylin and eosin(H&E) and analyzed usingimage analyse system. Alive and necrotic areas of the tumorswere identified on these sections. The signal intensity on T1WI,T2WI,postcontrast T1WI and increasing rate of signal intensitypre-and postcontrast T1WI were measured in the alive andnecrotic areas of the tumors. ADCs,ADCr and ADCp wereobtained in 3 orthogonal directions by calculation, and ADCtr 4英 文 摘 要was calculated in trace images. ADCm was obtained by directmeasurement. These ADC values metioned above werecompared each other. ADCm of the alive and necrotic areas ofthe tumors were compared also. The percentage of the totalblood vessel areas in the axial sections of the tumors wascompared with its corresponding ADCm. P values less than 0.05were considered a statistically significant difference. Results Sixteen rats all grew transplanted tumor in 3 to 6days after transplanting. The SNR of the routine MR imageswas high. Reference to the routine MR images the lesion can belocated in the DWI although its SNR was lower. The SNR of theADCm was somewhat high that the lesion and the normalstructure could be identified combined with the T2WI. There is no significant difference between the signalintensity in the alive and necrotic areas of the tumors on T1WIand T2WI(P＞0.05). There was significant difference betweenthe signal intensity in the alive and necrotic areas of the tumorson postcontrast T1WI(P＞0.05). The signal intensity in aliveareas of the tumors on postcontrast T1WI was higher. There issignificant difference between the increasing rate of signalintensity in alive and necrotic areas of the tumors pre-andpostcontrast T1WI(P ＞0.05). The increasing rate of signalintensity in alive areas of tumors pre-and postcontrast T1WI washigher. It was feasible to identify the necrotic regions by signalintensity on postcontrast T1WI or the increasing rate of signalintensity pre-and postcontrast T1WI, and its sensitivity was 5英 文 摘 要86.7%(13/15), specificity was 38.5%(15/39) and veracity was51.9%(28/54). However, there were overlapping between thetwo parameters of the alive and necrotic areas of the tumors. There was no significant difference among the ADCs,ADCr,ADCp,ADCtr and ADCm of the alive areas of thetumors by analysis of variance(F＝1.98, P＞0.05). There wasno significant difference among the ADCs,ADCr,ADCp,ADCtr and ADCm of the necrotic areas of the tumors byanalysis of variance (F＝2.05, P＞0.05). There was significant difference between the ADCm in thealive and necrotic areas of the tumors(P＞0.05). The ADC inalive tumor areas were lower. There were a little overlappingbetween the ADC of the alive and necrotic areas of the tumors.It was feasible to identify the necrotic regions by ADC values,and its sensitivity was 93.3%(14/15), specificity was94.9%(37/39) and veracity was 94.4%(51/54). The correlation coefficient of the percentage of the totalblood vessel areas in the axial sections of the tumors and itscorresponding ADC was -0.136, P＜0.0001. So A...