New England Section of the American Urological Association

Back to 2019 Abstracts


Clinically-Insignificant Prostate Cancer (Gleason Grade Group 1) or Benign Pathology in PI-RADS 5 Lesions with Extraprostatic Extension on Multi-Parametric MRI
Kamyar Ghabili Amirkhiz, MD, Jason Hao, MD, Sarah Amalraj, BS, Michael Leapman, MD, Jeffrey Weinreb, MD, Preston Sprenkle, MD
Yale School of Medicine, New Haven, CT

Clinically-Insignificant Prostate Cancer (Gleason Grade Group 1) or Benign Pathology in PI-RADS 5 Lesions with Extraprostatic Extension on Multi-Parametric MRI
Background
The clinical utility of extraprostatic extension (EPE) on multi-parametric magnetic resonance imaging (mpMRI) is unknown. We sought to investigate the rate of benign or clinically-insignificant prostate cancer (ciPCa) in biopsy of PI-RADS 5 lesions with EPE, and to identify clinical and imaging parameters associated with these findings.
Methods
We retrospectively queried our institutional mpMRI-ultrasound fusion (targeted) biopsy database to identify patients with EPE detected on mpMRI along with a PI-RADS 5 lesion who underwent targeted biopsy between October 2014 and April 2018. mpMRI findings were assessed, including prostate and lesion volumes, and zonal location of the lesion (peripheral or transition). We measured the rate of benign or clinically-insignificant PCa, defined as Gleason grade group (GG) 1, detected on the targeted biopsy of the lesion with EPE. Logistic regression and receiver operating characteristics curves with an area under the curve (AUC) were used to assess the ability of clinical and mpMRI characteristics to predict GG≥2 prostate cancer on the targeted biopsy of those lesions.
Results
Of 300 PI-RADS 5 lesions that underwent targeted biopsy during the study period, 117 (39%) were associated with EPE on mpMRI. On targeted biopsy of those 117 lesions, 5 (4.3%), 14 (12%), and 98 (83.7%) lesions harbored benign pathology, GG1, and GG≥2 prostate cancer, respectively. Benign or ciPCa was detected in 32% of lesions in the first quartile of prostate-specific antigen (PSA) density (<0.13), 16.7% of lesions in the interquartile range of PSA density (0.13-0.30), and 3.1% of lesions with PSA density >0.30 (p=0.003). Using a threshold of 0.13, PSA density was 82.6% sensitive and 42.1% specific for detecting GG≥2 prostate cancer on PI-RADS 5 lesions with EPE. On multivariable analysis, PSA density (OR 2.5 per 0.1 decrease in unit, 95%CI 1.14-5.26, p=0.02) was associated with an increased likelihood of benign or ciPCa in those lesions. Compared with lesion volume and PSA, PSA density had the highest discriminative ability for GG≥2 prostate cancer in those lesions (AUC 0.71).
Conclusions
Clinically-insignificant findings (benign or ciPCa) were identified in a minority of PI-RADS 5 lesions with EPE. In this setting, patients with PSA density <0.13 could be more frequently detected with ciPCa on the targeted biopsy.


Back to 2019 Abstracts