Upcoming event

Predicting pathologic tumor size in prostate cancer based on multiparametric prostate MRI and preoperative findings

  • Aydin Pooli,
  • David C. Johnson,
  • Joseph Shirk,
  • Daniela Markovic,
  • Taylor Y. Sadun,
  • Anthony E. JR. Sisk,
  • Amirhossein Mohammadian Bajgiran,
  • Sohrab Afshari Mirak,
  • Ely R. Felker,
  • Alexa K Hughes,
  • Steven S. Raman,
  • Robert E. Reiter

Publication: Journal of Urology, October 2020


Oncologic efficacy of focal therapies in prostate cancer (PCa) depends heavily on accurate tumor size estimation. We aim to evaluate the agreement between radiologic tumor size (RTS) and pathologic tumor size (PTS) and identify predictors of PTS.

Materials and Methods

This single-arm study cohort included all consecutive patients with biopsy-proven PCa and a corresponding PIRADSv2 ≥3 index tumor on mpMRI who subsequently underwent radical prostatectomy. RTS was defined as maximum tumor diameter on mpMRI and compared to whole mount histopathology tumor correlates. The difference between RTS and PTS was assessed and clinical, pathologic, and radiographic predictors of PTS were examined.


Four hundred sixty-one consecutive lesions in 441 men were included for statistical analysis. The mean RTS and PTS were 1.57 cm and 2.37 cm, respectively (p <0.001). RTS consistently underestimated PTS regardless of the preoperative covariates and the degree of underestimation increased with smaller RTS and lower PIRADSv2 scores. PTS was significantly larger for biopsy Gleason Grade Group (GG) 5 compared to GG 1 (mean change=0.37 cm, p=0.014), PI-RADSv2 5 lesions compared to PI-RADSv2 4 (mean change=0.26, p=0.006) and higher PSA density. The correlations between RTS vs PTS according to biopsy GG and radiologic covariates were generally low with correlation coefficients ranging between 0.13-0.65.


mpMRI frequently underestimates PTS and the degree of underestimation increases with smaller RTS and lower PIRADSv2 scores. Therefore, a larger ablation margin may be required for smaller tumors and lesions with lower PIRADSv2 scores. These variables must be considered when estimating treatment margins in focal therapy.