MRI、MRS及^11C-胆碱PET／CT对前列腺癌诊断的对比研究

The comparative study of MRI, MRS and ^11C-choline PET/CT imaging in prostate cancer

目的探讨^11C-胆碱PET／CT对前列腺癌的诊断价值，并与常规MRI及直肠线圈。HMR波谱分析（MRS）的诊断结果对比。方法34例前列腺病变患者行^11C-胆碱PET／CT、常规前列腺MRI及直肠线圈MRS检查。^11C-胆碱PET／CT图像于注射5min后获取。对前列腺良性病变和前列腺癌部位分别计算标准摄取值（SUV）最大值及平均值（SUVmax、SUVmax）。将PET／CT诊断结果与MRI、1HMRS及病理检查结果对比。分析SUV与血清前列腺特异抗原（PSA）的相关性。病理检查结果为在开放式MR光学导引系统下穿刺活组织检查或经直肠指诊穿刺获得。转移灶经局部穿刺活组织检查或影像学随访证实。统计学处理采用SPSS11．0软件。结果（1）原发性前列腺癌SUVmax5．72±2．31，SUVmax4．78±2．20。SUV比较：低分化腺癌〉中分化腺癌〉高分化腺癌。前列腺良性病变的SUVmax2．49±1．43，SUVmax 1．89±1．24。不同病理类型及分化程度的前列腺病变SUVmax、SUVmax之间差异无统计学意义（FSUVmax=2．401，FSUVmax=1．744；P〉0．05）。前列腺癌SUV高于前列腺良性病变，但二者存在交叉重叠现象，如果以SUVmax2．5为分界，则二者之间有25．8％（8／31）交叉。因此，前列腺的形态、前列腺内放射性分布的均匀程度对前列腺癌的诊断价值高于SUV。SUVmax与血清PSA之间呈正相关（等级相关系数rs=0．81986，P〈0．01）。（2）影像诊断与病理检查结果符合的例数^11C-胆碱PET／CT为25例，MRI为21例，MRS为22例。^11C-胆碱PET／CT、MRI及MRS诊断前列腺癌的灵敏度、特异性、阳性预测值、阴性预测值分别为93．75％，66．67％，75．00％，90．91％；87．50％，46．67％，63．64％，77．78％；87．50％，53．33％，66．67％，80．00％。3种方法比较差异无统计学意义（r=1．422，P〉0．05）。（3）^11C-胆碱PET／CT能区别前列腺癌治疗后复查患者复发区与治疗有效区，并可检出远处转移灶。结论^11C-胆碱PET／CT可用于前列腺癌的诊断、分期及检测复发和转移灶；其对前列腺癌的诊断灵敏度、特异性、阳性预测值、阴性预测值均略高于MRS及MRI；SUVmax与血清PSA之间呈正相关。

Objective Many studies had already demonstrated the usefulness of imaging modalities in prostate cancer diagnosis. The aim of the study was to compare the value of ^11C-choline PET/CT imaging with that of MBI and proton magnetic resonance spectroscopy（MBS） in diagnosis of prostate cancer. Methods^11C-choline PET/CT, MRI and MRS were pedormed in a total of 34 consecutive patients with prostate diseases. Pelvic scan started at 5 min after the injection of ^11C-eholine and visual analysis, maximum standardized uptake value （ SUVmax ） and mean value of SUV （ SUVmax ） were used for image evaluation. The resuits of ^11C-eholine PET/CT, MRI and MRS were compared with pathologic findings which obtained from radical prostatectomy or biopsy specimens. Possible correlation between serum prostate specific antigen （PSA） and SUVmax were investigated with linear correlation. Results （ 1 ） For primary prostate cancer, SUVmax and SUVmax in ^11C-eholine PET/CT were 5.72 ± 2.31 and 4.78 ±2.20 respectively. There were no statistically significant differences of SUV among different tumor grade by nonparametrie test （ FSUVmax = 2. 401 ,FSUVmax =1. 744;P 〉 0.05 ） although the SUV in poorly differentiated adenocarcinoma was higher than moderate and well differentiated ones. For benign prostate disease, SUVmax and SUVmax were 2.49 ±1.43 and 1.89 ± 1.24. There was an overlap of SUV between prostate cancer and benign disease, and the overlap was noticed in 25.8% （8/31） ff SUVmax=2.5 was taken the threshold for differentiation. The characteristics of choline＇s distribution in prostate were more valuable than the SUV alone. Spearman rank correlation test showed a linear correlation between SUVmax and serum PSA （rs=0. 819 86, P 〈0.01 ）. （2）In accordance with pathological findings, 25 cases in ^11C-choline PET/CT, 21 cases in MRI and 22 cases in MRS were correctly detected. The sensitivity, specificity, positive predictive value and negative predictive value were 93.75%, 66.67%, 75.00%, 90.91% for^11C-choline, 87.50%, 46.67%, 63.64%, 77.78% for MRI, and 87.50%, 53. 33%, 66.67%, 80.00% for MRS, respectively. No significant differences were found among the three modalities（χ^2 = 1. 422,P 〉0. 05）. （3） ^11C-choline PET/CT could effectively detect the recurrence and distant metastasis of prostate cancer after clinical management. Conelusions ^11C-choline PET/CT played an important role in diagnosis and staging of primary prostate cancer. It could also effectively detect the tumor recurrence. There were no statistically significant differences among n C-choline PET/CT, MRI and MRS diagnostic performances.