Shirin Sabouri1, Silvia Chang2,3,4, Emily Pang2, Rehab Mohammedeid2, Edward Jones5, Larry Goldenberg3,4, Peter Black3,4, and Piotr Kozlowski1,2,3,4
1UBC MRI Research Centre, Vancouver, BC, Canada, 2Department of Radiology, The University of British Columbia, Vancouver, BC, Canada, 3Department of Urologic Sciences, The University of British Columbia, Vancouver, BC, Canada, 4Vancouver Prostate Centre, Vancouver, BC, Canada, 5Department of Pathology and Laboratory Medicine, The University of British Columbia, Vancouver, BC, Canada
Synopsis
Luminal water
imaging (LWI) is an MRI technique that
detects regions of decreased lumen volume in prostate. Recent studies on LWI
have shown promising results regarding the accuracy of this technique in
diagnosis of prostate cancer. However, to the best of our knowledge no study
has yet compared the performance of
LWI against the current clinical assessment. In this
study, we perform a comparison between the diagnostic accuracy of LWI with the
PI-RADSv2.1 assessment. Our results show that LWI performs similar to the PI-RADSv2.1 in the entire prostate
and peripheral zone, and outperforms significantly in transition zone.
Purpose:
To compare the accuracy of detection of prostate adenocarcinoma (PCa)
achieved from luminal water imaging (LWI) and mp-MRI PI-RADSv2.1. Introduction:
Luminal Water Imaging (LWI), an MRI technique that detects regions of
decreased lumen volume in prostate, was first introduced1 with the
purpose of detection and grading of PCa. Since then, a number of follow up studies
2-5 have been conducted to evaluate the applicability and accuracy of
this technique; all with consistent and promising results. Although the results
of these primary studies reveal the potential benefits of LWI in clinical
prostate MRI, it is necessary to compare its performance against the current clinical
assessment before proposing it as a method for population-level utilization. In this
study, we perform a detailed comparison between the diagnostic accuracy of LWI and
the PI-RADSv2.16 assessment performed by an experienced radiologist.Methods:
40 patients with biopsy proven PCa were examined with LWI, high spatial
resolution T2-weighted (T2W), diffusion-weighted imaging (DWI), and dynamic
contrast-enhanced (DCE) MRI on a 3T scanner [Philips Medical Systems, Best, The
Netherlands] prior to undergoing radical prostatectomy. MRI signals were
acquired with a combined endorectal/pelvic phased-array coils. Sequence
parameters are presented in Table 1. From LWI data, maps of seven metrics: T2-short,
T2-long, gmT2,
LWF, Ashort, Along, and Ncomp were generated
by fitting the signal decay curve to a multi-exponential function, as previously
described1,7. Maps of these parameters were correlated to
whole-mount histology following image registration.
The T2W, DWI and DCE sequences were interpreted by a radiologist, with
20 years’ experience in interpreting prostate MRI, according to the current PI-RADSv2.1.
The images were interpreted on a picture archiving computer system (PACS)
station without knowledge of the prostate-specific antigen (PSA), PSA density
or biopsy results. Each lesion given a PI-RADS score of 3 or above has
localized into the sector diagram as per PI-RADSv2.1. An overall PI-RADS score
of 4 or 5 was considered positive. Scores of 3 or less were considered negative. The
sensitivity, specificity and area under the curve (AUC) were calculated based
on the receiver operating characteristics (ROC) analysis.
ROC analysis for LWI data was performed based on
logistic generalized linear mixed effect models (GLMMs), with the binary
dependent variable being the malignancy status, and the predictor variable
being a LWI parameter. For ROC analysis in each specific zone (peripheral zone (PZ), and transition zone (TZ), separately) the patient
identifier was considered as the random intercept in the model, while in the
entire prostate (including PZ, TZ, and
stroma), both the patient identifier and the location were
considered as the random intercepts.
ROC analysis
for PI-RADSv2.1 was performed using logistic regression, with the binary
dependent variable being the malignancy status, and the predictor variable
being the PI-RADS score. The significance of the difference between the AUC values was tested using two-tailed
t-test (MedCalc v. 19.1, Mariakerke, Belgium).
Results:
The values of AUC, sensitivity, and specificity
calculated from ROC analyses are summarized in Tables 2. Among the single LWI
parameters, the highest AUC was obtained for: gmT2, LWF and Along (0.93) in the entire prostate, gmT2, and LWF (0.91) in the PZ, and T2-long (0.94)
in the TZ. The AUC values obtained from multi-parametric ROC analysis
of LWI were equal to 0.94, 0.92 and 0.97, in the entire prostate, PZ, and TZ,
respectively. The AUC values obtained from ROC analysis of PI-RADSv2.1 were all
smaller, equal to 0.88, 0.91 and 0.78, in the entire prostate, PZ, and TZ,
respectively, with the significant differences observed in the entire prostate,
and TZ. Specificity values obtained from PI-RADSv2.1 were always higher than
multi-parametric LWI (0.99, 0.98, and 1.00 vs. 0.95, 0.90, and 0.98 in the
entire prostate, PZ, and TZ, respectively). Sensitivity values obtained from PI-RADSv2.1
were lower than multi-parametric LWI in the PZ and TZ (0.84 vs 0.86, and 0.57
vs. 0.78, respectively), but higher in the entire prostate (0.76 vs. 0.73). Discussion:
Comparing the diagnostic
accuracies obtained from LWI and PI-RADSv2.1 assessment shows that LWI provides
higher AUC values than those obtained from PI-RADSv2.1. In comparison between
the specificities, PI-RADSv2.1 performs slightly better compared to LWI. However,
in comparison between the sensitivities LWI outperforms considerably in TZ, and
performs relatively similar in the entire prostate and PZ.Conclusion:
This study was
conducted to compare the performance of LWI against the current clinical
assessment in detection of PCa. Results of this study show that LWI performs
equally well or better than PI-RADSv2.1 in diagnosis of PCa. The significant
improvement in the AUC and sensitivity provided by LWI in TZ is of high
importance, considering that TZ tumors usually pose a clinical challenge in detection
and are often missed during biopsy sessions8.This technique neither
requires the administration of a contrast agent, which is an inseparable
component of current PI-RADSv2.1, nor suffers from the artifacts and spatial
distortions commonly seen in DWI scans. Inclusion of this technique in clinical
prostate MRI can potentially increase the accuracy of diagnosis of PCa, and
reduce the concerns regarding the repetitive use of contrast agents. Acknowledgements
This study was
supported by the Canadian Institutes of Health Research. References
1.
Sabouri S, Chang S.D, Savdie R, et al. Luminal Water Imaging: A New MR Imaging T2
Mapping Technique for Prostate Cancer Diagnosis. Radiology.
2017;284(2):451-459.
2.
Sabouri S, Chang S.D, Goldenberg S.L, et al. Comparing
diagnostic accuracy of luminal water imaging with diffusion-weighted and
dynamic contrast-enhanced MRI in prostate cancer: A quantitative MRI study. NMR
Biomed. 2019;32(2):e4048.
3.
Chan R.W, Lau A.Z, Detzler G, Thayalasuthan V, Nam R.K,
Haider M.A. Evaluating the accuracy of multicomponent T2 parameters for luminal
water imaging of the prostate with acceleration using inner-volume 3D GRASE.
MRM. 2019;81(1):466-476.
4.
Devine W, Giganti F, Johnston EW, et al. Simplified
Luminal Water Imaging for the Detection of Prostate Cancer From Multiecho T2 MR
Images. JMRI. 2019;50(3):910-917.
5.
Carlin D, Orton M.R, Collins D, deSouza N.M. Probing
structure of normal and malignant prostate tissue before and after radiation
therapy with luminal water fraction and diffusion-weighted MRI. JMRI.
2019;50(2):619-627.
6.
Turkbey B, Rosenkrantz A.B, Haider M.A, et al. Prostate
Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging
Reporting and Data System Version 2. European Urology. 2019;76(3):340-351.
7.
Sabouri S, Fazli L, Chang S.D, et al. MR measurement of
luminal water in prostate gland: Quantitative correlation between MRI and
histology. JMRI. 2017;46(3):861-869.
8.
Lawrentschuk N, Haider M.A, Daljeet N, et al. Prostatic
evasive anterior tumours: the role of magnetic resonance imaging. BJU Int
2010;105:1231–1236.