Key points are not available for this paper at this time.
OBJECTIVE: To evaluate the accuracy of endorectal magnetic resonance imaging (MRI) with fast spin-echo sequences in the local staging of clinically localized prostate cancer. PATIENTS AND METHODS: Seventy-one patients with a clinical T1 (18 patients) or T2 tumour (53 patients) underwent endorectal MRI 2-12 weeks before radical prostatectomy. Extraprostatic tumour was diagnosed if MRI showed signs of capsular penetration and/or invasion of the seminal vesicle and/or distal urethra or bladder neck. If the pathological examination showed a single positive margin with no periprostatic tissue, the tumour was classified as indeterminate and not as a T3 tumour. RESULTS: Of the 25 cases of capsular penetration. MRI correctly identified 10 (sensitivity 42%, specificity 100%). Of the 14 cases with seminal vesicle invasion, MRI correctly identified six (sensitivity 43%, specificity 100%), but showed other signs of extraprostatic tumour spread in seven of the eight unidentified cases. Overall, MRI identified 16 of the 30 patients (53%) with occult extraprostatic spread of tumour; there was only one false-positive result. The sensitivity, specificity and accuracy of MRI were 53, 96 and 74%, respectively. CONCLUSION: Endorectal MRI can reduce the rate of preoperative understaging from 42% to 22% and it can be used for a given individual because it can detect extraprostatic invasion with 96% specificity, ensuring that very few, if any, patients will be deprived of curative surgery.
Cornud et al. (1996) studied this question.