Body mass index, prostate weight and transrectal ultrasound prostate volume accuracy.

PURPOSE

Transrectal ultrasound can be used to calculate prostate volume, which has implications for benign and malignant prostate disease. We hypothesized that obesity may represent a technical challenge when performing transrectal ultrasound that decreases the accuracy of estimating prostate volume.

MATERIALS AND METHODS

We examined the records of men with previously untreated prostate cancer who underwent radical prostatectomy between 1995 and 2006 and who were in the Shared-Equal Access Regional Cancer Hospital database. Transrectal ultrasound volume calculations were correlated with radical prostatectomy specimen weight using the Spearman coefficient. We calculated the percent and absolute error, and evaluated the relationship between them and transrectal ultrasound volume, body mass index, age, prostate specific antigen and race using multivariate linear regression.

RESULTS

A total of 497 patients with preoperative transrectal ultrasound volume, specimen weight and body mass index data were identified in the Shared-Equal Access Regional Cancer Hospital database. Transrectal ultrasound volume correlated modestly with specimen weights (r = 0.692, p <0.001). The median +/- SD absolute error was 9.6 +/- 11.4 gm and the median +/- SD percent error was 22.9% +/- 20.6%. Body mass index was not significantly related to absolute or percent error (p = 0.91 and 0.71, respectively). In addition, patient age, prostate specific antigen and race were not significantly related to absolute or percent error (p >0.05). However, percent error but not absolute error was significantly predicted by transrectal ultrasound volume (p <0.001 and 0.34, respectively). Smaller prostate size was associated with greater percent error, especially when transrectal ultrasound volume was less than 20 cc.

CONCLUSIONS

Transrectal ultrasound volume correlates with specimen weight but it is an imperfect substitute with significant variation in error. The accuracy of transrectal ultrasound depends on measured volume but neither body mass index nor other patient specific variables had a significant impact.