Login to Access Video or Poster Abstract: MP08-19
Sources of Funding: None

Introduction

Pelvic MRI can be used in the setting of prostate cancer for pre-surgical evaluation of local disease extension although limitations are debated. Less often discussed, however, is the human component; specifically, the radiologist interpreting the study. Herein, we investigate the accuracy and variability of pelvic MRI interpretation among our body radiology team versus a senior faculty member.

Methods

A single institution retrospective study identified 233 consecutive individuals diagnosed with prostate cancer who ultimately had a prostatectomy. All patients had pre-surgical pelvic 3T surface body coil MRI read by a fellowship trained body radiologist provided with relevant clinical information. Thereafter, a senior radiologist was selected to re-read all pelvic MRIs blinded to the initial interpretation. Both MRI readings were compared to final pathology report. Kappa (K) scores as well as sensitivity, specificity, positive predictive values (PPV), negative predictive value (NPV), and accuracy were calculated.

Results

When considering extraprostatic extension (EPE), there was low concordance comparing the initial versus repeat MRI interpretation (K=0.22). Additionally, when the senior radiologist re-read his own initial interpretation (n=93, blinded to initial result), concordance for EPE was greater (K=0.36) albeit similarly low. Regarding EPE, a comparison of initial MRI interpretation versus re-read by senior radiologist noted universal improvements in diagnostic characteristics include sensitivity (30.3% vs 56.1%), specificity (80.2% vs 88.6%), PPV (37.7% vs 66.1%), NPV (74.4% vs 83.6%), and accuracy (66.1% vs 79.4%). In contrast, seminal vesicle (SV) invasion interpretation was more uniform whereby initial MRI interpretation vs. re-read yielded similar sensitivity (18.2% vs 27.3%), specificity (97.2% vs 93.8%), PPV (40.0% vs 31.6%), NPV (91.9% vs 92.5%), and accuracy (89.7% vs 87.6%) (Table).

Conclusions

Even at an academic medical center, interobserver agreement amongst radiologists to evaluate local extent of disease on prostate MRI is relatively low. We report, however, improved characteristics when a senior member of the body radiology team reads the MRI. These findings underscore the importance of uniformity when defining criteria for EPE and SV invasion to allow for appropriate surgical planning.

Funding

None