EURURO Vol. 72 No. 4

E100 656

Letter to the Editor
Reply to Thorsten Derlin, Christoph-A. von Klot, and
Katja Hueper’s Letter to the Editor re: Sungmin Woo,
Chong Hyun Suh, Sang Youn Kim, Jeong Yeon Cho,
Seung Hyup Kim. Diagnostic Performance of Magnetic
Resonance Imaging for the Detection of Bone
Metastasis in Prostate Cancer: A Systematic Review
and Meta-analysis. Eur Urol. In press.
We thank Derlin et al for their interest in our meta-analysis
Althoughmost of the issuesmentioned in their letter have
already been dealt with in our meta-analysis, we acknowl-
edge that they are important and warrant further discussion.
First, the main endpoint of our meta-analysis was based
on a per-patient analysis rather than a per-lesion analysis.
Indeed, as Derlin et al commented, per-patient analysis will
result in significantly better diagnostic performance than
per-lesion analysis, as shown in the study by Mosavi et al
in which the former yielded sensitivity of 1.0 compared
with 0.56 from the latter. Therefore, it is clinically important
to know both the patient- and lesion-based performance.
However, per-lesion analysis was not feasible, as most
studies performing a per-lesion analysis regarding magnet-
ic resonance imaging (MRI) for bone metastases in prostate
cancer were not diagnostic test accuracy studies and
therefore did not provide sufficient information for the
generation of 2 2 contingency tables to calculate sensi-
tivity and specificity . In addition, various MRI
sequences were used by the investigators in each study
included. Three of the studies used only a single sequence,
while others used a multiparametric protocol. To account
for this heterogeneity, we performed a meta-regression
analysis for studies using diffusion-weighted imaging and
those using only conventional MRI sequences (ie, T1-
weighted imaging, short tau inversion recovery) and found
that this was not a potential source of heterogeneity in our
meta-analysis. However, we do advise that when interpret-
ing a bony lesion in the clinical setting of prostate cancer, a
multiparametric approach using all possible sequences
should be applied. Despite the fact that we could not
derive per-lesion diagnostic performance, we believe that
the per-patient performance presented in our meta-analysis
provides clinically meaningful value inmaking management
and therapeutic decisions.
Second, all ten studies included used the ‘‘best value
comparator’’ (combination of imaging/clinical/biological
data and follow-up) as the reference standard. Although
the gold standard of histopathological proof would be the
most robust method for identifying whether a bony lesion is
metastasis or not, this was not feasible, as biopsy or surgery
to determine pathology is not routinely performed, and
would not be ethically justifiable, for suspicious bony
lesions in prostate cancer patients. An alternative, as stated
by Derlin et al, would be to use the highly specific prostate-
specific membrane antigen (PSMA)-based positron emis-
sion tomography. In fact, there are recent efforts to analyze
PSMA and MRI collaboratively
Finally, there may be some limitations to direct
application of our results to routine practice owing to
possible biases. As mentioned by Derlin et al, all but two
studies were performed at single centers, four were
retrospective, and several had small study populations,
possibly giving rise to various types of bias. However,
despite such issues, these studies are currently the only
studies providing diagnostic test accuracy results for MRI in
detecting bone metastasis from prostate cancer. In order to
determine the diagnostic performance of MRI free from
such biases, future large-scale studies with a prospective,
multi-center design will be required.
Conflicts of interest: The authors have nothing to disclose.
References
[1] Woo S, Suh CH, Kim SY, Cho JY, Kim SH. Diagnostic performance of
magnetic resonance imaging for the detection of bone metastasis in
prostate cancer: a systematic review and meta-analysis. Eur Urol. in
press. .
[2]
[3]
[4]
.
available at
journal homepage:
DOIs of original articles: .
0302-2838/# 2017 European Association of Urology. Published by Elsevier B.V. All rights reserved.