目的·评估¹⁸F-MD-PSMA PET/CT在中高风险前列腺癌（prostate cancer，PCa）患者初始分期中的应用价值。方法·对2017年9月至2022年6月在上海交通大学医学院附属新华医院就诊的67例中高危PCa患者采用¹⁸F-MD-PSMA PET/CT进行初步分期；患者在¹⁸F-MD-PSMA PET/CT检查前2周内接受常规成像（conventional imaging，CI），包括多参数磁共振成像（multi-parameter magnetic resonance imaging，mp-MRI）和全身骨显像（bone scintigraphy，BS），其中25例患者同期进行了¹⁸F-FDG PET/CT检查。评估¹⁸F-MD-PSMA PET/CT对初始分期的灵敏度（sensitivity，SEN）、特异度（specificity，SPEC）、阳性预测值（positive predictive value，PPV）、阴性预测值（negative predictive value，NPV）及准确率（accuracy，ACU），并将结果与¹⁸F-FDG PET/CT、mp-MRI和BS的结果进行对比。以术后病理的T、N分期结果及临床随访的骨转移结果为参考标准进行Kappa一致性检验，分析¹⁸F-MD-PSMA PET/CT及CI在诊断原发灶累及范围、区域淋巴结转移、骨转移方面与参考标准的一致性，计算Kappa系数，并进行比较。结果·在67例PCa患者中，38例接受了根治性前列腺切除术并且有完整的病理学诊断资料，其中分别有27例和1例患者同时接受了区域淋巴结清扫术和扩大盆腔淋巴结清扫术，以病理结果作为诊断金标准。mp-MRI和¹⁸F-MD-PSMA PET/CT诊断包膜内病灶的检出率均为100%，诊断双侧腺叶内病灶的SEN分别为26.3%和63.2%，SPEC均为75.0%。与病理结果进行Kappa一致性检验，结果显示¹⁸F-MD-PSMA PET/CT诊断包膜外侵犯（extraprostatic extension，EPE）、精囊腺侵犯（seminal vesicle invasion，SVI）、膀胱颈侵犯（bladder neck invasion，BNI）的一致性均高于mp-MRI。Fisher确切概率法显示，2种检查方法诊断EPE、SVI的SEN（P=0.226，P=0.491）和SPEC（P=1.000，P=0.342），以及诊断BNI的SEN（均P=1.000）比较，差异均无统计学意义。在诊断淋巴结转移方面，基于淋巴结数量分析，¹⁸F-MD-PSMA PET/CT与病理结果的一致性高于mp-MRI（Kappa系数分别为0.555和0.137）；Fisher确切概率法提示2种检查方法的SEN和SPEC差异均无统计学意义（P=0.562，P=0.829）。基于患者分析，¹⁸F-MD-PSMA PET/CT与病理结果一致性高于mp-MRI（Kappa系数分别为0.850和0.313）；两者SEN比较，差异无统计学意义（P=1.000）。在诊断骨转移方面，基于骨病灶数量分析，¹⁸F-MD-PSMA PET/CT与随访结果的一致性高于BS（Kappa系数分别为0.500和0.299）；Fisher确切概率法提示2种检查方法的SEN比较，差异无统计学意义（P=0.219）。基于患者分析，¹⁸F-MD-PSMA PET/CT与随访结果的一致性高于BS（Kappa系数分别为0.953和0.766）；两者的SEN比较，差异无统计学意义（P=1.000）。¹⁸F-MD-PSMA PET/CT检测后，21例（31.3%）患者的风险分层上升，1例（1.5%）风险分层下降；32例（47.8%）患者的初始分期改变，其中27例（40.3%）上调，5例（7.5%）下调。结论·¹⁸F-MD-PSMA PET/CT在诊断中高风险PCa患者的双侧腺叶内病灶、EPE、SVI、区域淋巴结转移及骨转移方面，相比CI具有一定的优势，并在此基础上改变了部分患者临床分期及转移状态的诊断。

Objective ·To evaluate the role of ¹⁸F-MD-PSMA PET/CT in the initial stage of patients with moderate and high risk prostate cancer (PCa). Methods ·A total of 67 patients with moderate and high risk PCa who were treated in Xinhua Hospital, Shanghai Jiao Tong University School of Medicine from September 2017 to June 2022 were initially staged by ¹⁸F-MD-PSMA PET/CT. Conventional imaging (CI), including multi-parameter magnetic resonance imaging (mp-MRI) and bone scintigraphy (BS), were performed within two weeks before ¹⁸F-MD-PSMA PET/CT. Twenty-five patients underwent ¹⁸F-FDG PET/CT at the same time. The sensitivity (SEN), specificity (SPEC), positive predictive value (PPV), negative predictive value (NPV) and accuracy (ACU) of ¹⁸F-MD-PSMA PET/CT in the initial stage were evaluated, and the results were compared with those of ¹⁸F-FDG PET/CT, mp-MRI and BS. The consistency of ¹⁸F-MD-PSMA PET/CT and CI in terms of primary lesion, regional lymph node metastasis and bone metastasis was evaluated by Kappa consistency test refering to the postoperative pathological T and N staging results and bone metastasis results of clinical follow-up. Kappa coefficient was calculated and compared. Results ·Of the 67 patients with PCa, 38 patients underwent radical prostatectomy and had completed pathological data, with 27 patients undergoing regional lymphadenectomy and 1 patient undergoing expanded pelvic lymphadenectomy at the same time. The pathological results were obtained as gold standard. The detection rates of mp-MRI and ¹⁸F-MD-PSMA PET/CT in diagnosing intrathecal lesions were both 100%. The SENs in diagnosing bilateral intralobular lesions were 26.3% and 63.2%, respctively; the SPECs were both 75.0%.The Kappa consistency test showed that the consistency of ¹⁸F-MD-PSMA PET/CT in diagnosis of extracapsular extension (EPE), seminal vesicle invasion (SVI), and bladder neck invasion (BNI) was higher than that of mp-MRI. Fisher′s exact test showed that there were no statistically significant differences in SEN (P=0.226, P=0.491) and SPEC (P=1.000, P=0.342) between the two methods for diagnosing EPE and SVI, as well as SEN (P=1.000) for diagnosing BNI. In terms of diagnosis of lymph node metastasis, based on the analysis of lymph node numbers, the consistency between ¹⁸F-MD-PSMA PET/CT and pathological results was higher than that of mp-MRI (Kappa coefficients of 0.555 and 0.137, respectively). Fisher′s exact test showed that there were no statistically significant differences in SEN and SPEC between the two examination methods (P=0.562, P=0.829). Based on the patients, the consistency between ¹⁸F-MD-PSMA PET/CT and pathological results was higher than that of mp-MRI (Kappa coefficients of 0.850 and 0.313, respectively). There was no statistically significant difference in SEN between the two methods (P=1.000). In terms of diagnosis of bone metastasis, based on the analysis of bone lesion numbers, the consistency between ¹⁸F-MD-PSMA PET/CT and clinical follow-up results was higher than that of BS (Kappa coefficients of 0.500 and 0.299, respectively). Fisher′s exact test showed that there was no statistically significant difference in SEN between the two methods (P=0.219). Based on the patients, the consistency between ¹⁸F-MD-PSMA PET/CT and clinical follow-up results was higher than that of BS (Kappa coefficients of 0.953 and 0.766, respectively). There was no statistically significant difference in SEN between the two methods (P=1.000). The risks of 21 patients (31.3%) were increased after ¹⁸F-MD-PSMA PET/CT detection, with 1 patient (1.5%) decreasing. The initial stage of 32 cases (47.8%) were changed after ¹⁸F-MD-PSMA PET/CT detection, with 27 cases (40.3%) upstaged and 5 cases (7.5%) downstaged. Conclusion ·¹⁸F-MD-PSMA PET/CT is superior to CI in the diagnosis of bilateral intralobular lesions, EPE, SVI, regional lymph node metastasis and bone metastasis in intermediate and high risk PCa, and on this basis, the diagnosis of clinical stage and metastatic status of some patients has been changed.