術前3T 磁振造影影像預測T3a癌症：與機器手臂根除性攝護腺切除術病理標本相關性之研究
林冠榮1、黃逸修1,3,4、沈書慧2、黃奕燊1,3,4、林志杰1,3,4、林子平1,3,4、鍾孝仁1,3,4、盧星華1,3,4、黃志賢1,3,4、吳宏豪1,3,4、張延驊1,3,4、林登龍1,3,4
台北榮民總醫院 泌尿部1；台北榮民總醫院 放射線部2：國立陽明大學醫學院 泌尿學科3書田泌尿科學研究中心4
The prediction of T3a disease on pre-operative 3T mp-MRI: a correlation study with the pathology of robotic assisted radical prostatectomy
Kuan-Jung Lin1, Eric YH Huang1,3,4, Shen Shu-Huei2, I-shen Huang1,3,4, Chi-Cheh Lin1,3,4, Tzu-Ping Lin1,3,4, Hsiao-Jen Chung1,3,4, Shing-Hwa Lu1,3,4, William JS Huang1,3,4, Howard HH Wu1,3,4, Yen-Hwa Chang1,3,4, Alex TL Lin1,3,4
Department of Urology1 and Radiology2, Taipei Veterans General Hospital; School of Medicine2 and Shu-Tien Urological Institute3, National Yang-Ming University, Taiwan
Purpose:
Radical prostatectomy is the surgical treatment of choice for localized prostate cancer, which has shown favorable cancer control in patients with organ-confined disease. Patients with locally advanced disease may have worse outcome but still can benefit from the surgery. Whether to evaluate the surgical candidates with 3T multiparametric magnetic resonance imaging (mp-MRI) pre-operatively can differentiate patients with localized disease from locally advanced prostate cancer is an interesting issue. Hence, we retrospectively reviewed the patients who received robotic-assisted radical prostatectomy (RARP) in our hospital to correlate pre-operative 3T mp-MRI with the surgical pathology.
Materials and Methods:
From 2010 to 2016, 230 patients with prostate cancer received pre-operative 3T mp-MRI and robotic assisted radical prostatectomy were identified. The clinical data, PSA, Gleason score, MRI stage, extraprostatic extension (EPE) location on MRI, pathological stage, and location of EPE were retrospectively reviewed. The correlation of EPE on pre-operative mp-MRI and pathology was performed.
Results:
The mean age of the patients was 65.9 ± 6.2 years (range 50-87). Pre-operative PSA was 11.1 ± 9.1 ng/ml (range 0.4-76.6). D’amico risk group was 28.3% in low risk, 40.4% in intermittent risk, and 31.3% in high risk. Pathological stage was 56.1% in T2 stage, 36.5% in T3a stage and 7.4% in T3b stage. EPE was noticed in 33.5% patients on 3T mp-MRI and 36.5% on final pathology. The sensitivity and specificity of predicting EPE on 3T mp-MRI were 60.7%, and 82.2%, respectively. The inter-rater agreement between MRI stage and pathological stage was moderate (weighted Kappa=0.54). The factors influencing the agreement of MRI stage and pathological stage included positive cores of TRUS biopsy, the radiological size of prostate, and the surgical size of prostate. The accuracy of correct prediction of EPE (presence and location) on 3T mp-MRI was 69.6%. The correlation of the location of EPE on MRI and surgical specimen were better at anterior and lateral aspect of prostate and worse at apex, anterior fibromuscular stroma and bladder neck. The factors influencing the accuracy of correct prediction of EPE (presence and location) on 3T-MRI included higher clinical stage, higher D’amico risk group, and smaller radiological size of the prostate. (Table)
Conclusion:
3T mp-MRI had fair sensitivity, specificity and diagnostic accuracy on EPE of prostate cancer. The factors influencing the accuracy of correct prediction of EPE (presence and location) on 3T mp-MRI included higher clinical stage, higher D’amico risk group, and smaller radiological size of the prostate.