The Accuracy of Prostate MRI Interpretation: Impact of the Individual Radiologist and Clinical Factors - Beyond the Abstract

Prostate multi paramedic MRI (mpMRI) has become an extremely important tool in the diagnosis and management of prostate cancer. The use of mpMRI to identify patients who require a biopsy, as well as to target suspicious lesions during biopsy, has significantly improved the detection of clinically significant prostate cancer.1-3 Despite the standardization provided by PI-RADS, however, considerable variability in radiologists’ interpretations of concerning prostate lesions has been reported in the literature.4-6 Such variation in mpMRI interpretation presents challenges in clinical decision-making and prostate cancer management. With the increasing widespread implementation of prostate mpMRI beyond high volume research centers to routine clinical use, understanding the variability among individual radiologists’ accuracy and the clinical factors that affect prostate mpMRI is important.

We generated a study cohort of consecutive men who underwent prostate mpMRI from our prospective institutional registry. Measures of test performance (e.g. sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV)) were compared between the nine radiologists, who interpreted the studies as part of the clinical workflow of the abdominal imaging section at our institution. While most studies compare mpMRI to either radical prostatectomy or standard TRUS-guided biopsy histopathology, we utilized MRI-targeted biopsy core pathology as the reference standard. 

Sensitivity, specificity, PPV and NPV were 83.6%, 50.5%, 41.0% and 88.2%, respectively. Although we found variation in test performance among individual radiologists, no significant differences were observed. Interestingly, the additional prostate mpMRI experience was not beneficial in improving accuracy. We found that a number of non-modifiable patient variables, including prostate lesion location, prior biopsy history, PSA, and age, were predictive of prostate mpMRI test performance (Table 1).

Table 1. Multivariate logistic regression analysis of clinical factors predictive of measures of test performance.


PSA = prostate-specific antigen; MRI = magnetic resonance imaging; PPV = positive predictive value; NPV = negative predictive value; CI = confidence interval.

As our use and understanding of prostate mpMRI continue to grow, clinicians must remain aware of variability in the interpretation of this imaging modality, as well as the clinical factors that may affect its performance. The implementation of the PI-RADS scoring system has led to a significant reduction in such variability,7 which we expect will further improve with subsequent versions. Finally, while prostate mpMRI has emerged as the new de facto imaging test for the evaluation of localized prostate cancer, further study of newer technology is critical. Promising examples include contrast-enhanced ultrasound and shear wave elastography (for patients unable to undergo MRI), PSMA-PET/CT (for staging high-risk disease or detecting disease in patients with biochemical recurrence), diffusion tensor imaging, and high-resolution MRI.

Written by: Nicholas A. Pickersgill, BS and Eric H. Kim, MD, Division of Urology, Washington University School of Medicine, St. Louis, Missouri, United States

1. Kasivisvanathan V, Rannikko AS, Borghi M, et al. MRI-targeted or standard biopsy for prostate cancer diagnosis. N Engl J Med 2018;378(19):1767-1777.
2. Rastinehad AR, Turkbey B, Salami SS, et al.  Improving detection of clinically significant prostate cancer: magnetic resonance imaging/transrectal ultrasound fusion guided prostate biopsy.  J Urol 2014; 191: 1749-1754.
3. Tompson JE, Moses D, Shnier R, et al.  Multiparametric magnetic resonance imaging-guided diagnostic biopsy detects significant prostate cancer and could reduce unnecessary biopsies and overdetection: a prospective study. J Urol 2014; 192: 67-74.
4. Rosenkrantz AB, Ginocchio LA, Cornfeld D, et al.  Inter-observer reproducibility of the PI-RADS version 2 lexicon: a multi-center study of six experienced prostate radiologists.  Radiology 2016; 280: 793-804. 
5. Rosenkrantz AB, Lim RP, Haghighi M, et al.  Comparison of interreader reproducibility of the Prostate Imaging Reporting and Data system and Likert scales for evaluation of multiparametric prostate MRI.  Am J Roentgenol 2013; 201: W612-W618.
6. Sonn GA, Fan RE, Ghanouni P, et al. Prostate magnetic resonance imaging interpretation varies substantially across radiologists. Eur Urol Focus 2017;(17):30266-30263.
7. Weinreb JC, Barentsz JO, Choyke PL, et al.  Prostate Imaging—Reporting and Data System: 2015, version 2. Eur Urol 2016; 69: 16-40.

Read the Abstract