ENDORECTAL MRI OF LIMITED VALUE FOR PREDICTING T3 PROSTATE CANCER

NEW YORK (Reuters Health) Sep 28 - Endorectal MRI has little clinical value for predicting pT3 (locally invasive) prostate cancer preoperatively, researchers report in the August 26th BJU International online.

Some earlier studies reported endorectal MRI to have a sensitivity and specificity exceeding 90% for detecting T3 disease, but these included all patients who presented for prostatectomy and generally involved academic radiologists specialized in MRI or the genitourinary system.

Dr. Jonathan S. Brajtbord and colleagues from The Mount Sinai Medical Center, New York, evaluated the accuracy of endorectal MRI in predicting the presence of extracapsular extension and seminal vesicle invasion in 1161 men who underwent robotic-assisted laparoscopic prostatectomy. Endorectal MRI results were available for 179 of these patients.

MRIs were ordered on selected higher risk patients and performed at both private imaging centers and academic institutions, investigators say, thereby better reflecting the current practice environment in the USA than previous study designs did.

Eighty-one of 110 patients with histopathologically organ-confined disease (74%) were correctly diagnosed as such on endorectal MRI, whereas only 30 of 69 patients with pT3 disease (43%) were correctly predicted by endorectal MRI.

These figures result in an overall accuracy of 64%, with an overall sensitivity of 43% and specificity of 73% for diagnosing pT3 disease. Positive predictive value of endorectal MRI was 50% and negative predictive value was 67%.

The positive predictive value of 50% is "equivalent to a coin toss," the authors say. "This demonstrates that endorectal MRI has a limited ability to preoperatively predict seminal vesicle invasion and extracapsular extension, and a positive test does not reliably indicate the presence of extraprostatic disease."

Results were similar when locally advanced tumors were stratified by extracapsular extension and seminal vesicle invasion, and when the analysis was limited to high-risk patients.

Moreover, sensitivity and specificity rates were similar at academic and non-academic centers.

According to the researchers, "The Prostate-Specific Antigen Best Practice Statement of 2009 states that it is 'generally unnecessary' for patients with a PSA score of 25 or less to undergo radiological staging by CT or MRI."

"The data obtained in the present study support this recommendation," the authors conclude.

"It should be noted that improvements in MRI technology currently being investigated have shown promise over standard endorectal MRI," they add. "Such improvements may improve preoperative staging, although, at this time, the endorectal MRI does not appear to be of much clinical utility."

BJU Intl. Posted online August 26, 2010. Abstract