BACKGROUND: There are important treatment and prognostic implications in distinguishing between organ-confined prostate cancer that has spread locally outside the capsule and that which has spread into the seminal vesicles. This study is the first Australian study to report local accuracy for the locoregional staging of prostate cancer with endorectal magnetic resonance imaging (MRI). METHODS: From July 2002 to December 2005, 129 patients were referred for endorectal MRI for all indications. Inclusion criteria were biopsy-proven prostate cancer, minimum 4 weeks from previous biopsy and radical retropubic prostatectomy within 12 months of MRI. This yielded 47 patients. Those with prior hormonal and neoadjuvant radiotherapy or significant postbiopsy haemorrhage were excluded. In addition, those patients examined with our alternate-contrast-enhanced protocol were also excluded. A total of 38 patients met all inclusion criteria. A General Electric 1.5-T whole-body MR imaging unit with an endorectal coil was used with interpretation by two genito-urinary MR radiologists. Final histopathological report was used for correlation. RESULTS: Median age was 60 years with a range 44-72 years. Median prostate-specific antigen was 6.3 with a range of 2-82, and median Gleason score was 6 with a range of 5-8. Sensitivity, specificity and accuracy for extracapsular extension and seminal vesicle invasion were 69, 82 and 76% and 60, 100 and 95%, respectively. For extraprostatic extension, 71, 86 and 79%, respectively. CONCLUSIONS: Staging accuracy is similar to internationally published standards. Improvements in hardware and software and increased reader experience will add value to the local Australian prostate imaging programme.
BACKGROUND: There are important treatment and prognostic implications in distinguishing between organ-confined prostate cancer that has spread locally outside the capsule and that which has spread into the seminal vesicles. This study is the first Australian study to report local accuracy for the locoregional staging of prostate cancer with endorectal magnetic resonance imaging (MRI). METHODS: From July 2002 to December 2005, 129 patients were referred for endorectal MRI for all indications. Inclusion criteria were biopsy-proven prostate cancer, minimum 4 weeks from previous biopsy and radical retropubic prostatectomy within 12 months of MRI. This yielded 47 patients. Those with prior hormonal and neoadjuvant radiotherapy or significant postbiopsy haemorrhage were excluded. In addition, those patients examined with our alternate-contrast-enhanced protocol were also excluded. A total of 38 patients met all inclusion criteria. A General Electric 1.5-T whole-body MR imaging unit with an endorectal coil was used with interpretation by two genito-urinary MR radiologists. Final histopathological report was used for correlation. RESULTS: Median age was 60 years with a range 44-72 years. Median prostate-specific antigen was 6.3 with a range of 2-82, and median Gleason score was 6 with a range of 5-8. Sensitivity, specificity and accuracy for extracapsular extension and seminal vesicle invasion were 69, 82 and 76% and 60, 100 and 95%, respectively. For extraprostatic extension, 71, 86 and 79%, respectively. CONCLUSIONS: Staging accuracy is similar to internationally published standards. Improvements in hardware and software and increased reader experience will add value to the local Australian prostate imaging programme.