EURURO Vol. 73 No. 1

139 148

Words of Wisdom
Re: Magnetic Resonance Imaging Underestimation of
Prostate Cancer Geometry: Use of Patient-specific Molds
to Correlate Images with Whole-mount Pathology
Priester A, Natarajan S, Khoshnoodi P, et al
J Urol 2017;197:320–6
Expert's summary:
Previous findings on size estimation for prostate cancer
via magnetic resonance imaging (MRI) were somewhat
contradictory. Priester et al used individually printed three-
dimensional (3D) prostate cancer moulds based on in vivo
imaging for optimal alignment of the histopathology with
multiparametric MRI (mpMRI). This allowed them to
correlate 3D tumour contours obtained via imaging with
those in the final pathology specimen. Actual tumour size
as measured in the pathology specimen for 118 matched
tumours was on average three times larger than as
measured by mpMRI, with a median extent of 13.5 mm
beyond the imaging contour. The tumour diameter at
imaging correlated with the actual tumour diameter for
Gleason score 7 cancer, but did not correlate with the
volume of Gleason score 6 (3 + 3) cancers. Factors
influencing the degree of underestimation were prostate
zonal localisation and Gleason score. The axis of measure-
ment (eg, base–apex or anterior–posterior) differed in the
influence on the accuracy of tumour length measurement
at mpMRI. The authors conclude that their findings impact
the clinical assessment of prostate cancer by mpMRI,
particularly its ability to delineate accurately the cancer
area for focal therapy planning.
Expert's comments:
The degree of underestimation of mpMRI tumour
volume as reported in this study might be more than
expected. From a pathology perspective, mpMRI underesti-
mation can be attributed to the cellular density of the
prostatic adenocarcinoma. A previous study reported on the
phenomenon of mpMRI-invisible “sparse” prostate cancer
areas, defined as benign prostatic glandular tissue inter-
mixed with carcinoma glands Gleason score 6 prostate
cancers probably contain a larger component of sparse areas
when compared to carcinomas with higher Gleason grades.
This may explain the observation by Priester et al that
mpMRI tumour volume does correlate with actual tumour
volume for Gleason score 7 but not for Gleason score
6 prostate cancers. The next question is whether mpMRI
tumour volume is more prognostic than tumour volume as
measured by the pathologist. Prostate cancer volume
measured in prostatectomy specimens loses its prognostic
significance after adjusting for Gleason score and patholog-
ical stage . Intriguingly, Priester et al found that mpMRI
tumour size was both larger and more accurate for higher-
grade prostate cancers. Rosenkrantz et al demonstrated
in a prostatectomy setting that mpMRI-measured tumour
volume was a prognostic factor for biochemical recurrence.
Further studies are warranted to investigate the preopera-
tive prognostic power of mpMRI-measured volume after
taking into account zonal location in addition to the
apparent diffusion coefficient and other clinical param-
eters.
1. Conflicts of interest: The author has nothing to disclose.
References
[1] .
[2] .
[3] .
[4] .
Theo van der Kwas
Pathology Department, University Health Network, Toronto, Canada
*Pathology Department[1_TD$DIFF], Laboratory Medicine Program, University
Health Network, 200 Elizabeth Street, Toronto, ON M5G2C4, Canada.
E-mail address: .
Crown Copyright © 2017 Published by Elsevier
B.V. on behalf of European Association of
Urology. All rights reserved.
ava ilable at
journal homepage:
0302-2838/