Preoperative evaluation of the ‘hostile pelvis’ in radical prostatectomy with computed tomographic pelvimetry

Authors:
Publisher website
Google Scholar
Abstract
OBJECTIVE: To determine whether preoperative pelvimetry based on computed tomography (CT) can be used to predict technical difficulties during open radical prostatectomy (RP). PATIENTS AND METHODS: An open RP database accrued prospectively between January 1997 and June 2005 was used to identify 450 patients with preoperative pelvic imaging. Of these, 165 had adequate imaging of the pelvis with CT to allow pelvimetry using software provided with the medical imaging records. Several pelvic measurements were recorded in conjunction with body mass index and transrectal ultrasonographic estimates of prostatic volume. Outcome measures used to reflect technical surgical difficulties included operative duration, blood transfusion requirements within 30 days of RP, the pathological positive surgical margin and prostatic capsular breech rate. Logistic and linear regression analyses were used to determine the relationship between variables before and after RP RESULTS: The selected pelvimetric measurements failed to predict either operative duration or the peri‐operative blood transfusion requirement. Prostatic volume was predictive of operative duration; for every increase of 20 mL in prostate volume the duration increased by 8.4 min. Although pelvimetric measures failed to predict positive surgical margins at pathology, the transverse diameter predicted the likelihood of a positive margin due to capsular breech. With every 8.6 mm (1 sd ) decrease in transverse diameter, the odds of a capsular breech resulting in positive surgical margins increased 5.3 times (95% confidence interval 2.1–20.0, P = 0.002). CONCLUSIONS: Although the ‘hostile pelvis’ influences the likelihood of prostatic capsular breech resulting in positive surgical margins, CT pelvimetric screening of patients before RP is unlikely to be cost‐effective. Routine pelvic CT in the evaluation of patients before RP is not supported.

This publication has 15 references indexed in Scilit: