Study: mpMRI predicts lymph node involvement

A combination of findings on preoperative multi-parametric magnetic resonance imaging showed good accuracy for predicting lymph node involvement on final pathology in men who underwent robot-assisted radical prostatectomy for prostate cancer, reported researchers from the National Institutes of Health, Bethesda, MD.

Bethesda, MD-A combination of findings on preoperative multi-parametric magnetic resonance imaging (mpMRI) showed good accuracy for predicting lymph node involvement on final pathology in men who underwent robot-assisted radical prostatectomy (RARP) for prostate cancer, reported researchers from the National Institutes of Health, Bethesda, MD.

“Preoperative evaluation of lymph node status is an essential component of staging prior to definitive therapy for prostate cancer,” said Steven Abboud, NIH Medical Research Scholar.

“Our study suggests that for men who are part of a fusion biopsy protocol and undergoing mpMRI, the finding of high-risk features on that imaging may help urologists assess the risk of nodal involvement and decide whether or not to perform extended lymph node dissection during radical prostatectomy,” added Abboud, who worked on the study with Peter Pinto, MD, and co-authors. The findings were presented at the 2015 AUA annual meeting in New Orleans.

mpMRI characteristics associated with having positive lymph nodes were investigated in a retrospective study that reviewed clinical information, mpMRI, and surgical pathology reports for 367 patients who underwent RARP with lymph node dissection between 2007 and 2014. Nineteen men (5%) had positive lymph nodes on postoperative surgical pathology.

Next: Factors linked with LN invasion identified

Factors linked with LN invasion identified

On univariate analysis, preoperative PSA, higher clinical stage disease, mpMRI suspicion score, total number of prostatic lesions seen on mpMRI, presence of extracapsular extension (ECE) on mpMRI, and seminal vesical invasion (SVI) on mpMRI were all significantly associated with lymph node invasion.

Multivariate logistic regression analysis controlling for confounding variables found preoperative PSA, mpMRI suspicion score, presence of ECE on mpMRI, and presence of SVI on mpMRI were significantly associated with cancer-positive lymph nodes. In a receiver operating curve analysis, the combination of the three mpMRI characteristics predicted lymph node involvement on final pathology with an area under the curve of 0.88, which was higher than PSA alone.

“Our experience shows that some men with lymph node involvement on final pathology have microdisease that may be missed on preoperative CT scans and standard pelvic MRI,” Abboud told Urology Times.

He suggested that it would be interesting in the future to compare the performance of the MRI characteristics for predicting lymph node involvement against currently used nomograms.