骨显像半定量法对骨肉瘤侵袭范围的评估及其与MRI的对照研究

A correlative study of semi-quantitative bone scanning and MRI on osteosarcoma

目的:旨在找到一种评估骨肉瘤侵袭范围的骨显像半定量方法 ,并与MRI进行比较,探讨如何更准确测量肢体骨肉瘤骨侵袭范围。方法:回顾性分析病理证实为骨肉瘤的23例患者全身骨扫描及局部MRI影像。设计骨显像放射性计数变化率阈值法,放射性计数变化率R=(T-NT)/NT×100%,公式转化为T=R×0.01×NT+NT,假设多个R值,获得多个肿瘤边界T值即可获得肿瘤范围测量值,同时采用骨显像目测法及MRI T1WI测量肿瘤范围,以病理范围为标准进行对比研究。采用配对t检验,组内相关系数。结果:骨显像目测法与病理范围有显著性差异(t=-3.041,P=0.006<0.01)。骨显像阈值法R取80%(t=-1.519,P=0.143>0.05)、100%(t=-0.642,P=0.527>0.05)、120%(t=0.192,P=0.850>0.05)、140%(t=1.178,P=0.252>0.05)时与病理范围无显著性差异,R取100%及120%时平均差值最小。MRI T1WI肿瘤范围与病理范围无显著性差异(t=-1.121,P=0.112>0.01)。当MRI T1WI骨髓出现多节段信号变化,选取骨显像半定量法R=100%及R=120%时测量值最接近的MRI信号变化平面为测量平面,所得MRI测量值与病理测量值相符。骨扫描半定量分析法,R取100%及120%,两名测量者及同一测量者不同时间两次测量值之间高度一致(ICC>0.900)。结论:MRI在评估骨肉瘤骨侵袭范围方面占主导地位;骨显像放射性计数变化率阈值法提供了较目视法客观准确的测量值,建议R阈值取100%或120%;MRI骨髓出现多节段信号变化,骨显像放射性计数变化率阈值法获得的测量结果有助于在MRI T1WI图像准确选择测量平面。

Objective： To find a semi-quantitative bone imaging method by comparison with MRI in order to measure bone invasion of limb osteosarcoma accurately. Methods： 23 patients diagnosed with osteosarcoma by histology were included.Preoperative whole body bone scan and partial magnetic resonance imagings（MRI） were analyzed retrospectively. Threshold method of radioactive count changing-rate on bone scan was designed. The radioactive count changing-rate R =（T-NT）/NT ×100% was converted into T=R×0.01×NT＋NT. By assuming multiple R values, different T values of tumor boundary were obtained. Visual measurements on bone scintigraphy and MRI T1 WI were compared with pathological boundaries as the gold standard. Paired t-test was used to analyze the accuracy. The intraclass correlation coefficient was used to evaluate the repeatability in two observers and two times of measurements. Results： There were significant differences between bone scanning visual method and pathologic measurements（t=-3.041, P=0.0060.01）. There was no significant difference between bone scan threshold method and pathologic range, with radioactive count changing-rate R =80%（t =-1.519, P =0.143 〉0.05）, R =100%（t=-0.642, P=0.527〉0.05）, R=120%（t=0.192, P=0.850〉0.05）, R=140%（t=1.178, P=0.252〉0.05）. When R=100% and 120%, average difference was the minimum. The differences between MRI T1 WI measurements and pathologic measurements showed no difference（t=-3.041, P=0.006〈0.01）. In cases with varied signal intensities on MRI, the measurements of MRI and pathology were matched in the plane which was the closest to bone scan semi-quantitative method（R =100% and R =120%）. In bone scanning visual method, there was significant difference of osteosarcoma extents between two different measurers（P〈0.05）, and there was no significant difference between two times of the same observer（P〉0.05）. In bone scan threshold method, measurements of two observers and two different times of the same observer showed high consistency（ICC〉0.900）. Conclusion： MRI is dominant in the evaluation of intramedullary invasion in osteosarcoma. In assessing range of far-articular bone invasion,threshold method of radioactive count changing-rate on bone scan is more objective and accurate than visual method. Bone scintigraphy helps to determine the measurement plane on MRI T1 WI in cases with varied signal intensities.