Validating the biological origin of in vivo fractional intracellular volume from VERDICT-MRI in the prostate

Validating the biological origin of in vivo fractional intracellular volume from VERDICT-MRI in the prostate | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Validating the biological origin of in vivo fractional intracellular volume from VERDICT-MRI in the prostate Marta Masramon, Manju Mathew, Saurabh Singh, Thomy Mertzanidou, and 21 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7723707/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Reliable imaging biomarkers are needed to characterize clinically significant prostate cancer (csPCa). The clinically established measure, apparent diffusion coefficient (ADC) from multiparametric (mp)MRI, lacks specificity to key histological features. The fractional intracellular volume (fIC) from in vivo Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors (VERDICT)-MRI has shown improved detection of csPCa. This prospective study (NCT04792138) quantifies the accuracy of fIC as a measure of prostate epithelial cell density. Fifty-two participants (mean age, 65 years ± 6; Gleason grades ≥ 3 + 4) with biopsy-confirmed PCa underwent mpMRI and VERDICT-MRI before radical prostatectomy. Personalized molds from preoperative mpMRI enabled histology to MRI correspondence. Comparisons between histological epithelial density measures and MR markers were made on a region of interest (ROI) basis using Pearson’s correlation coefficient. VERDICT fIC corresponded more strongly with epithelial fraction, cell density and epithelial density (r = 0.784, 0.711, 0.747) than both VERDICT derived ADC (r=-0.639, -0.587, -0.609) and mpMRI ADC (r-0.353, -0.273, -0.326). Differences between benign tissue and csPCa in VERDICT fIC (0.16 ± 0.10 vs 0.51 ± 0.13) were similar to epithelial fraction (0.21 ± 0.08 vs 0.43 ± 0.12). These findings support in vivo VERDICT fIC as a marker of epithelial cell density and show promise for wider clinical use. Health sciences/Biomarkers Biological sciences/Cancer Health sciences/Medical research Health sciences/Oncology Health sciences/Urology Prostate Cancer MRI VERDICT-MRI Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Multiparametric MRI (mpMRI) plays a crucial role in the diagnosis and management of prostate cancer (PCa) by aiding in detection, localization, risk stratification, and staging of tumors. Despite its advantages, mpMRI has limited specificity[ 1 ] and insufficient sensitivity, missing 10–20% of tumors[ 2 , 3 ]. Improving detection of clinically significant lesions requires an understanding of how MRI signals relate to histology and ensuring that these signals are specific to relevant histological features such as epithelial cell density, an increase of which is a key descriptor of adenocarcinoma – the most common type of prostate cancer[ 4 ]. The apparent diffusion coefficient (ADC), a metric derived from the diffusion-weighted MRI acquisition within mpMRI, is routinely used to examine tissue microstructure. However, ADC compounds all physiological changes affecting water diffusion into a single measure, leading to poor discrimination of clinically significant (cs) from clinically insignificant PCa/benign change[ 5 , 6 ]. To address this limitation, specialized diffusion-weighted MRI protocols combined with computational modelling offer a more precise characterization of tissue microstructure by examining specific features such as cell size and density[ 7 ]. The Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors (VERDICT) framework is based on a biophysical model of diffusion in tumors[ 8 ]. It considers three water compartments – intracellular, vascular, and extracellular-extravascular – and estimates the volume fraction of each along with their diffusion coefficients. The fractional intracellular volume (fIC) derived from VERDICT has been explored in several studies (NCT02721784, NCT02689271, NCT04792138, NCT05017181) and has shown promise as an imaging biomarker when used with mpMRI, achieving higher accuracy in detecting clinically significant PCa (area under the curve (AUC): 0.96) than ADC maps (AUC: 0.85). It has also demonstrated statistically significant differences between true-positive and false-positive lesions[ 9 – 12 ], with potential to reduce negative biopsies. This study aims to verify the biological interpretation of in vivo VERDICT fIC. We use personalized 3D-printed prostate molds to match whole-mount prostatectomy slices with in vivo VERDICT-MRI. We hypothesize that fIC correlates with epithelial cell density, since this is an assumption of the VERDICT model. We test our hypothesis by performing region of interest (ROI)-based comparisons of MRI and histology-derived measures of cell density and epithelial fraction. We also analyze differences between tissue composition and MR parameters in benign and csPCa tissue. Methods Study Participants Ethical approval for this prospective study was granted by London-Queen Square Research Ethics Committee (19/LO/1803) on 23 January 2020. This study was performed in accordance with the Declaration of Helsinki. Patients were involved in research and ethic committee meetings reviewing study design and protocol prior to recruitment. We analyze data from the Histo-MRI clinical trial (registered with ClinicalTrials.gov, NCT04792138, first posted 10/03/2021 and ongoing since 2020 with consecutive sampling; amendments were made to increase sample size, 15/09/2023), where men with histologically confirmed PCa awaiting prostatectomy were scanned using VERDICT-MRI (n=68). Patients attending a One-stop Prostate Cancer clinic at University College London Hospital were screened against the inclusion and exclusion criteria and eligible patients were invited. All screened participants were provided with Participant Information Leaflets detailing the objectives of the study and their involvement; those interested provided informed written consent. Inclusion criterium was men aged 18+ with no MRI contraindications. Exclusion criteria included inability to undergo MRI, inability to provide informed consent or prior/ongoing PCa treatment (surgery, radiotherapy, hormone treatment). MRI files were anonymized using DICOM cleaner software, and whole slide images saved under study identifications. Participants missing ADC on the mpMRI, or with large MRI artifacts, histology deformations, or lesions smaller than the diffusion-weighted imaging voxel size (1.5 mm²) were excluded (Figure 1). Tissue preparation To ensure slice-to-slice correspondence between histology and MR images, custom molds were created from manual prostate contours on pre-operative T2W axial images using Horos[13]. Reference slices, set as the plane with the largest lesion detected on the mpMRI, were chosen and used to place cutting guides within the mold. After removal of the prostates during radical prostatectomy, specimens were inked to denote laterality and silicon catheters inserted through the urethra. Prostates were sealed in their molds and placed in neutral buffered formalin for 24-hour fixation. The full specimen handling protocol is described by Bourne et al. [14]. For histological analysis, a histopathologist (MM – 5 years of experience) cut the reference slice following the cutting guides, spaced 5mm apart. The remainder of the prostate was progressively sliced at 5mm thickness. Tissue sections were processed per standard protocol. MRI acquisition The MRI acquisitions included mpMRI and VERDICT-MRI[15]. VERDICT scans were performed on a 3.0 T MRI scanner (Achieva or Ingenia; Philips, Best, the Netherlands). A pulse-gradient spin-echo sequence was acquired using a 32-channel cardiac coil (Achieva) or body coils (Ingenia). B-values included 90, 500, 1500, 2000 and 3000 s/mm 2 obtained in 3 orthogonal directions, with a b = 0 s/mm 2 image for every b-value. This was used to normalize T2 and T1 dependencies. Further sequence parameters were as follows: phase-encoding direction A/P, FOV = 220x220 mm 2 , reconstruction matrix = 176x176 and voxel size = 1.25x1.25x5.0 mm 3 . Each b-value had a corresponding TR and TE. VERDICT scans lasted 10mins 95s (Achieva) or 17mins 41s (Ingenia). For more details, see Singh et al. [15]. ADC maps were derived from clinical mpMRI (1.5 or 3T, see Supplemental Material) and from VERDICT diffusion-weighted images with b-values 0, 90, 500 and 1500 s/mm 2 (mpMRI ADC and VERDICT ADC, respectively). Radiologists (AR – ‘in-training’, NT – 5 years of experience, SP – 20 years of experience) drew benign and PCa ROIs on the MR images in consensus using MIM 7.2.9[16]. VERDICT model The VERDICT mathematical model assumes no exchange between water populations [8] . The normalized signal is the sum of three parametric models: Histological analysis A semi-automated pipeline was created to process the histology images and produce quantitative maps. The appearance of the Hematoxylin and Eosin staining on histology is often highly variable [18] . To normalize the images, the brightness, contrast, and saturation of stained whole-slide histopathology was standardized. Shrinkage was accounted for using a correction factor of 1.15, proposed by Schned et al. and confirmed by Jonmarker et al. for specimens processed with formalin fixation[19,20]. The following steps were done on a patch-by-patch basis (535x535mm) on QuPath[21]. Cell density was calculated from cell nuclei segmentations (Figure 2a). Patches were segmented into stroma, lumen and epithelial tissue using a pixel classifier trained on expert annotations (histopathologist MM - 5 years of experience; Figure2b). Epithelial cell density was calculated by weighing cell density by epithelial fraction. All steps of this pipeline were visually checked by a single histopathologist (MM). Lesions in histology were contoured and graded by histopathologists (primary annotator MM and senior reviewers AF – 20 years of experience – and AH – 6 years of experience). Figures 3 & 4 show representative examples. Statistical analysis The primary outcome of this study was the correlation (with 95% confidence intervals) of features from histology and MR derived maps. In each sample, ROIs were drawn in cancerous and benign areas of tissue using consistent anatomical landmarks across MR and histology. Cancerous ROIs were based on radiologist annotations on mpMRI; benign ROIs were drawn directly on histology (to exclude any confounding pathology) and transferred to MRI. The average value of each ROI was used. Statistical significance was based on 95% confidence intervals. The study was not powered to detect differences between correlations. All analysis was performed in Python. The secondary outcome was the AUC between csPCa and benign tissue. Both analyses included the full patient cohort. Code availability The code used in this study is publicly available at https://github.com/Martamasramon/VERDICT-fIC-validation (commit 8e84b1eba228bc33c05143874db74ae9249da987). Results Participant Characteristics From the total of 68 participants, four were excluded due to lack of ADC maps on mpMRI, six due to image artifacts in MRI, two due to deformations in histology, three due to small lesions on histology and one because of small lesions on MRI (<0.5mm 2 , Figure 1). Participant and lesion characteristics of the remaining participants (n=52) are summarized in Table 1. The mean age of these men is 65 years ± 6. The mean prostate-specific antigen level is 8.4 ng/mL ± 6.1. The mean time between PSA and prostatectomy is 56 ± 39 days and the mean time between mpMRI and VERDICT-MRI is 108 ± 115 days. A single PCa and benign ROI was taken from each patient. Lesions have Gleason grades 3+4 (n=37), 3+4+5 (n=4), 4+3 (n=7), 4+3+5 (n=2), 4+4 (n=1) and 4+5 (n=1). Comparison of VERDICT fIC and mpMRI ADC against histological features in benign and clinically significant PCa ROIs When analyzing benign and cancerous ROIs in the full cohort, fIC correlates most strongly to the histological epithelial fraction (r=0.784 [0.70, 0.84]). It correlates with lower strength to epithelial cell density (r=0.747 [0.66, 0.81]) and cell density (r=0.711 [0.60, 0.79]). VERDICT ADC maps show weaker correlations to these histological parameter maps (r=-0.639 [-0.54, -0.71], -0.609 [-0.51, -0.69] and -0.587 [-0.45, -0.69], respectively) while mpMRI ADC exhibits the weakest correlations (r=-0.353 [-0.18, -0.50], -0.326 [-0.17, -0.47] and -0.273 [-0.09, -0.43], respectively). In all cases, the correlation with epithelial fraction is strongest. These findings are summarized in Table 2 and illustrated in Figure 5. Comparison of MR-derived parameters and tissue composition in benign and cancerous prostate regions Table 3 presents how fIC, ADC and tissue composition differ between benign and csPCa prostate regions. Tissue composition changes between benign and csPCa are characterized by an increase in epithelial fraction (0.21 ± 0.08 to 0.43 ± 0.12, AUC=0.94, p < 0.001) and a decrease in lumen fraction (0.24 ± 0.13 to 0.12 ± 0.04, AUC=0.86, p < 0.001) and stroma fraction (0.57 ± 0.16 to 0.49 ± 0.13, AUC=0.68, p < 0.001). Cell density and epithelial cell density are both higher in csPCa (3780 ± 790 and 1810 ± 750, respectively) compared to benign tissue (2360 ± 780 and 680 ± 380, respectively). As in the case of cell density, epithelial cell density and epithelial fraction, VERDICT fIC shows a statistically significant difference between benign and csPCa tissue (0.16 ± 0.10 to 0.51 ± 0.13, AUC=0.99, p < 0.001). Mean VERDICT ADC and mpMRI ADC values in benign and csPCa regions are 1.00 ± 0.19 to 0.68 ± 0.14 and 1.36 ± 0.32 to 1.05 ± 0.31, respectively. Discussion and Conclusions This study investigated the biological origin of the fractional intracellular volume from Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors, or VERDICT, MRI using 3D personalized molds to enable accurate histological comparison. Region of interest analysis showed strong correlation of fractional intracellular volume with epithelial fraction (r=0.784) and histological cell density (r=0.711). This supports our hypothesis that the fractional intracellular volume estimate is positively correlated with epithelial cell density. The VERDICT model assumes that the intracellular component captures the signal solely from epithelial cells while the signal from stromal cells and lumen is accounted for by the extra-cellular extra-vascular component. Importantly, results show that fIC correlates better with epithelial fraction than cell density, which has been shown to be a less effective Gleason Pattern predictor than tissue fractions[22]. Previous ROI-based studies that describe the relationship between histological cell density and ADC in PCa report correlations between r = -0.50 and r = -0.695[23–26]. Our findings for ADC from mpMRI show a correlation of r=-0.273. We speculate that the variability in these results is due to differences in the b-values, scanner manufacturers and diffusion sequences used to calculate ADC maps, as well as cell counting algorithms. Notably, the higher correlation of ADC with epithelial fraction presented here (r=-0.353) has been previously noted by Chatterjee et al., who reported correlations with cell count and epithelial fraction in fixed tissue of -0.598 and -0.647, respectively[27]. They also showed that both correlations are stronger in fresh tissue. Our results also show that VERDICT ADC is more strongly correlated to histological measures of cell density and epithelial fraction than mpMRI ADC (-0.587 and -0.639 vs -0.273 and -0.353). We believe that this is primarily due to using a consistent imaging sequence. In this study we derive the VERDICT ADC from diffusion-weighted images with b-values 0, 90, 500 and 1500 s/mm2, each with a given TR and TE, while each mpMRI ADC uses a different number of diffusion-weighted images, with different b-values (see Supplemental Material) and inconsistent TR and TE. The fact that VERDICT fIC outperforms VERDICT ADC highlights the importance of complex microstructural models to discriminate the relevant signal components. Further analysis investigated differences in tissue composition and MR-derived parameters between benign and csPCa ROIs. Consistent with previous findings[27], csPCa regions exhibit a higher epithelial fraction and lower luminal fraction compared to benign tissue (0.21 ± 0.08 and 0.24 ± 0.13 vs 0.43 ± 0.12 and 0.12 ± 0.04). VERDICT fIC demonstrates similar changes to the epithelium, with values increasing from 0.16 ± 0.10 in benign regions to 0.51 ± 0.13 in csPCa. The highest AUC values correspond to epithelial fraction and fIC (0.94 and 0.99, respectively), highlighting their ability to differentiate between tissue types, while both VERDICT ADC and mpMRI ADC exhibit lower discriminatory performance (AUC=0.92 and 0.76, respectively). We note certain limitations in our study. Resection of the prostate creates shape changes due to removal of mechanical tension and compression and hemodynamic pressure[28], dehydration and embedding of the tissue during processing causes tissue shrinkage and thin sectioning may cause further deformations[14]. This compromises the histology-MR mapping and affects the distribution of tissue fractions; future work could incorporate deformable registration or 3D histology reconstruction to mitigate these effects. Lesions smaller than 0.5 mm² were excluded from our analysis, as they fall below the resolution threshold of diffusion-weighted MRI; investigating whether VERDICT can detect such small lesions will require higher-resolution acquisitions or ultra-high-field imaging[29]. MpMRI was performed on various scanners while VERDICT-MRI was consistently obtained from Philips machines; assessing reproducibility across vendors and field strengths in future multi-center studies will be essential for clinical translation. Finally, the validation design of this study restricted our analysis to prostatectomy cases, and the limited cohort size represents an additional constraint. In conclusion, this study established the biological origin of the Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors’ fractional intracellular volume, showing strong correlation with epithelial fraction and histological cell density in region of interest-based analysis, and aiding interpretability of a biomarker that has already shown potential to improve prostate cancer characterization non-invasively[30] [9]. Future work will further investigate the relationship between the model’s parameters and histological metrics across different Gleason scores and growth patterns, with a focus on mpMRI-invisible lesions, in larger, multi-center cohorts using a range of scanner manufacturers. Abbreviations ADC apparent diffusion coefficient AUC area under the curve cs clinically significant fIC fractional intracellular volume mpMRI multiparametric MRI PCa prostate cancer ROI region of interest VERDICT Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors WSI Whole-Slide Imaging Declarations Data Availability The datasets analyzed within this study are not publicly available due to data protection laws but may be available from the corresponding author on reasonable request and subject to approval by the relevant institutional data protection office. Funding Declaration This work is supported by the EPSRC-funded UCL Centre for Doctoral Training in Intelligent, Integrated Imaging in Healthcare (i4health) [EP/S021930/1]; EPSRC grant numbers EP/N021967/1, EP/R006032/1; Prostate Cancer UK, Targeted Call 2014, Translational Research St.2, grant number PG14-018-TR2; the National Institute for Health and Care Research, University College London Hospitals Biomedical Research Centre; and Cancer Research UK National Cancer Imaging Translational Accelerator. Author Contributions M. Masramon conceived the study, carried out the computational analysis, and was responsible for drafting and writing the manuscript. M. Mathew contributed to patient recruitment and led the histological analysis. S.P. and E.P. contributed to the conceptualization of the study, funding acquisition, and supervision. A.R., N.T., S.P., A.H., A.F. and M. Mathew provided image annotations. S.S., T.M., J.C., M.-V.P., L.S., V.K., A.G., E.D., G.S., D.P., L.C., and C.M.M. contributed to the administration of the clinical trial and/or data acquisition. D.A. and A.P. supported methodology development. T.P. provided advice for the statistical analysis. D.C.A. contributed to funding acquisition. All authors contributed to the review and editing of the manuscript. Competing Interests Statement Dr Punwani serves as a consultant for Qubim, NVision, and Gold Standard Phantoms. mDr Shaw previously acted as a consultant for Angle plc. Dr Grey acts as a proctor for Sonablate, and as a consultant, proctor, and grant holder for Angiodynamics. He has also been a speaker for HC21/Aquilant. Dr Singh serves as a consultant for Docus AI. Dr Alexander is a director and shareholder of Queen Square Analytics. The remaining authors declare no competing financial interests. References Ahmed, H. U. et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study. Lancet 389 , 815–822 (2017). Zhao, Y. et al. Comparison of Multiparametric Magnetic Resonance Imaging with Prostate-Specific Membrane Antigen Positron-Emission Tomography Imaging in Primary Prostate Cancer Diagnosis: A Systematic Review and Meta-Analysis. Cancers (Basel) . 14 , 3497 (2022). Sathianathen, N. J. et al. Negative Predictive Value of Multiparametric Magnetic Resonance Imaging in the Detection of Clinically Significant Prostate Cancer in the Prostate Imaging Reporting and Data System Era: A Systematic Review and Meta-analysis. Eur. Urol. 78 , 402–414 (2020). Epstein, J. I., Cubilla, A. L. & Humphrey, P. A. Tumors of the prostate gland, seminal vesicles, penis, and scrotum. AFIP Atlas of Tumor Pathology (American Registry of Pathology, 2011). Rosenkrantz, A. B. et al. Interobserver reproducibility of the PI-RADS version 2 lexicon: A multicenter study of six experienced prostate radiologists. Radiology 280 , 793–804 (2016). Westphalen, A. C. et al. Variability of the positive predictive value of PI-RADS for prostate MRI across 26 centers: Experience of the society of abdominal radiology prostate cancer disease-focused panel. Radiology 296 , 76–84 (2020). Martinez-Heras, E., Grussu, F., Prados, F., Solana, E. & Llufriu, S. Diffusion-Weighted Imaging: Recent Advances and Applications. Seminars Ultrasound CT MRI . 42 , 490–506 (2021). Panagiotaki, E. et al. Noninvasive quantification of solid tumor microstructure using VERDICT MRI. Cancer Res. 74 , 1902–1912 (2014). Johnston, E. W. et al. VERDICT MRI for prostate cancer: Intracellular volume fraction versus apparent diffusion coefficient. Radiology 291 , 391–397 (2019). Singh, S. et al. Avoiding Unnecessary Biopsy after Multiparametric Prostate MRI with VERDICT Analysis: The INNOVATE Study. Radiology 10.1148/radiol.212536 (2022). Appayya, M. B. et al. The intracellular component of VERDICT (Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors) MRI distinguishes Gleason 4 pattern better than Apparent Diffusion Coefficient. (2018). Sen, S. et al. Differentiating False Positive Lesions from Clinically Significant Cancer and Normal Prostate Tissue Using VERDICT MRI and Other Diffusion Models. Diagnostics 12, (2022). Horos Project. Horos. Preprint at https://horosproject.org/. Bourne, R. M. et al. Apparatus for histological validation of in vivo and ex vivo magnetic resonance imaging of the human prostate. Front Oncol 7 , (2017). Singh, S. et al. Histo-MRI map study protocol: a prospective cohort study mapping MRI to histology for biomarker validation and prediction of prostate cancer. BMJ Open. 12 , e059847 (2022). MIM Software Inc. MIM 7.2.9. Preprint at https://www.mimsoftware.com/ Bonet-Carne, E. et al. VERDICT-AMICO: Ultrafast fitting algorithm for non-invasive prostate microstructure characterization. NMR Biomed 32 , (2019). Suvarna, K., Layton, C. & Bancroft, J. Bancroft’s Theory and Practice of Histological Techniques . (2018). Schned, A. R. et al. Tissue-shrinkage Correction Factor in the Calculation of Prostate Cancer Volume. Am. J. Surg. Pathol. 20 , 1501–1506 (1996). Jonmarker, S., Valdman, A., Lindberg, A., Hellström, M. & Egevad, L. Tissue shrinkage after fixation with formalin injection of prostatectomy specimens. Virchows Arch. 449 , 297–301 (2006). Bankhead, P. et al. QuPath: Open source software for digital pathology image analysis. Sci Rep 7 , (2017). Chatterjee, A. et al. Changes in Epithelium, Stroma, and Lumen Space Correlate More Strongly with Gleason Pattern and Are Stronger Predictors of Prostate ADC Changes than Cellularity Metrics. Radiology 277 , 751–762 (2015). Gibbs, P. et al. Correlation of ADC and T2 Measurements With Cell Density in Prostate Cancer at 3.0 Tesla. at (2009). http://journals.lww.com/investigativeradiology Zelhof, B. et al. Correlation of diffusion-weighted magnetic resonance data with cellularity in prostate cancer. BJU Int. 103 , 883–888 (2009). Glazer, D. I. et al. Diffusion-weighted endorectal MR imaging at 3T for prostate cancer: correlation with tumor cell density and percentage Gleason pattern on whole mount pathology. Abdom. Radiol. 42 , 918–925 (2017). Xi, Z. W. et al. Diffusion-weighted imaging of prostate cancer: Correlation between apparent diffusion coefficient values and tumor proliferation. J. Magn. Reson. Imaging . 29 , 1360–1366 (2009). Chatterjee, A. et al. Changes in epithelium, stroma, and lumen space correlate more strongly with gleason pattern and are stronger predictors of prostate ADC changes than cellularity metrics1. Radiology 277 , 751–762 (2015). Orczyk, C., Taneja, S. S., Rusinek, H. & Rosenkrantz, A. B. Assessment of change in prostate volume and shape following surgical resection through co-registration of in-vivo MRI and fresh specimen ex-vivo MRI. Clin. Radiol. 69 , e398–e403 (2014). Molendowska, M. et al. Diffusion MRI in prostate cancer with ultra-strong whole‐body gradients. NMR Biomed 37 , (2024). Johnston, E. et al. INNOVATE: A prospective cohort study combining serum and urinary biomarkers with novel diffusion-weighted magnetic resonance imaging for the prediction and characterization of prostate cancer. BMC Cancer 16 , (2016). Tables Tables 1 to 3 are available in the Supplementary Files section Additional Declarations Competing interest reported. Dr Punwani serves as a consultant for Qubim, NVision, and Gold Standard Phantoms. mDr Shaw previously acted as a consultant for Angle plc. Dr Grey acts as a proctor for Sonablate, and as a consultant, proctor, and grant holder for Angiodynamics. He has also been a speaker for HC21/Aquilant. Dr Singh serves as a consultant for Docus AI. Dr Alexander is a director and shareholder of Queen Square Analytics. The remaining authors declare no competing financial interests. Supplementary Files SupplementalMaterial.docx Tables.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7723707","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":531401476,"identity":"f846925f-da75-4bc4-b270-5f2c6410295a","order_by":0,"name":"Marta Masramon","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIiWNgGAWjYNCDDxU2DAzsELYBHnWMDXDWjDNpDAzMpGhh5m07TFiLbvvZ5w9+MGxL3M7e/oCBh+18Yn8zA+OHHwyHjXFpMTuTbtjYw3A7cWfPgQQGCZ7biTMOMzBL9jAcNsOp5UAaYwMPUMuGGwkHGAwkgAygw6QZGA7b4NRy/hlj4x+wlsQGhgSDcyAtzL/xarmRxtgMsSWZgeFAwgGQFjaQLbgdduMZ42wZg9vGG84cYzjYcCDZeMZhxjbLHoN03N4/n8bw8U3FbdkNx9sfPv77z062v7358I0fFdaGDbj0gAE0Dg5AKFBE4YvIUTAKRsEoGAUEAQCwilnE4rd/TgAAAABJRU5ErkJggg==","orcid":"","institution":"University College London","correspondingAuthor":true,"prefix":"","firstName":"Marta","middleName":"","lastName":"Masramon","suffix":""},{"id":531401477,"identity":"e26de4ec-c4a2-4bb8-a79a-bcbe5882b8de","order_by":1,"name":"Manju Mathew","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Manju","middleName":"","lastName":"Mathew","suffix":""},{"id":531401478,"identity":"f82fe185-fcd4-48b2-a26a-569498932276","order_by":2,"name":"Saurabh Singh","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Saurabh","middleName":"","lastName":"Singh","suffix":""},{"id":531401479,"identity":"e301b870-687c-4046-b421-57ae149e5bec","order_by":3,"name":"Thomy Mertzanidou","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Thomy","middleName":"","lastName":"Mertzanidou","suffix":""},{"id":531401480,"identity":"1dcb93d2-85df-4d3a-9e2b-a4b40364b3f2","order_by":4,"name":"Joey Clemente","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Joey","middleName":"","lastName":"Clemente","suffix":""},{"id":531401481,"identity":"a2bb97b3-5298-473a-92a6-e2d63bcd397b","order_by":5,"name":"Adam Retter","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"Retter","suffix":""},{"id":531401482,"identity":"56851a65-62f7-42e9-bd89-9fd2e7e7dcb1","order_by":6,"name":"Natasha Thorley","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Natasha","middleName":"","lastName":"Thorley","suffix":""},{"id":531401483,"identity":"0281836c-e074-47c4-9b39-219f2812f68e","order_by":7,"name":"Adam Phipps","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"Phipps","suffix":""},{"id":531401484,"identity":"48130a11-9cf7-445b-a2dc-d3be4534df72","order_by":8,"name":"Thomas Parry","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Parry","suffix":""},{"id":531401485,"identity":"0c53e708-9ed6-468a-8ad9-da4453a4fdf1","order_by":9,"name":"Marianthi-Vasiliki Papoutsaki","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Marianthi-Vasiliki","middleName":"","lastName":"Papoutsaki","suffix":""},{"id":531401486,"identity":"81c9d28a-5868-4ecc-a088-79a2840a947e","order_by":10,"name":"Lorna Smith","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Lorna","middleName":"","lastName":"Smith","suffix":""},{"id":531401487,"identity":"625c9282-0d42-415b-abe4-57c36008dd9d","order_by":11,"name":"Francesco Grussu","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Francesco","middleName":"","lastName":"Grussu","suffix":""},{"id":531401488,"identity":"562a9432-79f7-4532-82c9-eb0404adbe9b","order_by":12,"name":"Veeru Kasivisvanathan","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Veeru","middleName":"","lastName":"Kasivisvanathan","suffix":""},{"id":531401489,"identity":"551028e5-a963-4851-8d07-17d8fcef163a","order_by":13,"name":"Alistair Grey","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Alistair","middleName":"","lastName":"Grey","suffix":""},{"id":531401490,"identity":"375296db-0588-4cea-b22b-99e079f49be2","order_by":14,"name":"Eoin Dinneen","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Eoin","middleName":"","lastName":"Dinneen","suffix":""},{"id":531401492,"identity":"c44c11c8-fc21-4b7a-9ff1-93b178aa2b3f","order_by":15,"name":"Greg Shaw","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Greg","middleName":"","lastName":"Shaw","suffix":""},{"id":531401494,"identity":"7ae8f6ca-2da0-4a37-9efe-a444e8289ed9","order_by":16,"name":"Dominic Patel","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Dominic","middleName":"","lastName":"Patel","suffix":""},{"id":531401495,"identity":"b5824922-26fc-48cd-aafb-f4e3b5caedf7","order_by":17,"name":"Lucy Caselton","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Lucy","middleName":"","lastName":"Caselton","suffix":""},{"id":531401498,"identity":"6a7f9a21-81cf-4e9e-bddb-5adc719c2821","order_by":18,"name":"Caroline M. Moore","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Caroline","middleName":"M.","lastName":"Moore","suffix":""},{"id":531401499,"identity":"cdd753c8-3649-4c36-b313-112e92d434a9","order_by":19,"name":"David Atkinson","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Atkinson","suffix":""},{"id":531401500,"identity":"b61ce433-5f58-4f39-b607-37600e05f5a9","order_by":20,"name":"Aiman Haider","email":"","orcid":"","institution":"University College London Hospitals NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Aiman","middleName":"","lastName":"Haider","suffix":""},{"id":531401501,"identity":"eec13ac7-e53d-4f4a-a236-498210246bc1","order_by":21,"name":"Alex Freeman","email":"","orcid":"","institution":"University College London Hospitals NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Alex","middleName":"","lastName":"Freeman","suffix":""},{"id":531401502,"identity":"c4105310-46ab-4f15-a371-315b1990754e","order_by":22,"name":"Daniel C. Alexander","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"C.","lastName":"Alexander","suffix":""},{"id":531401503,"identity":"c7e5a23f-4290-438d-89c3-c4c7b7d1fc8a","order_by":23,"name":"Shonit Punwani","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Shonit","middleName":"","lastName":"Punwani","suffix":""},{"id":531401504,"identity":"7454ee00-7bb7-43c7-b576-a8e5bd528fe9","order_by":24,"name":"Eleftheria Panagiotaki","email":"","orcid":"","institution":"University College London","correspondingAuthor":false,"prefix":"","firstName":"Eleftheria","middleName":"","lastName":"Panagiotaki","suffix":""}],"badges":[],"createdAt":"2025-09-26 16:38:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7723707/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7723707/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94210298,"identity":"f14bff68-4577-4c50-9044-b73a101ea7e9","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"png","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":97400,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/c21377ce838905253fa31ee0.png"},{"id":94210304,"identity":"c0b8f749-71ac-446e-a45b-8bbdd1d24062","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":319973,"visible":true,"origin":"","legend":"","description":"","filename":"MainBody.docx","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/f1465ed0e00341ef9b1871f0.docx"},{"id":94210307,"identity":"d8a1162b-4465-41ec-a064-2332aba60e63","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"json","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":23762,"visible":true,"origin":"","legend":"","description":"","filename":"28e14931f0a041eeb29fcabb7818462b.json","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/ec657de7e1d6dda0b1e98890.json"},{"id":94210310,"identity":"50378e73-b40a-4b3e-a229-b4c72c76a5f2","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35555,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/a342871d70ae0db7f348162e.docx"},{"id":94210311,"identity":"688a6fc6-3a1c-4a3d-97b7-fba8c1b0152d","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":100006,"visible":true,"origin":"","legend":"","description":"","filename":"28e14931f0a041eeb29fcabb7818462b1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/aec266c6c777e359c34f52ad.xml"},{"id":94211196,"identity":"67b2f5bf-c1bc-4d9c-8558-49002d503007","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":97400,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/bf44ce72d6c5e9a91a61b075.png"},{"id":94210314,"identity":"ddd718f8-03b4-4b24-85f0-178148cfd668","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":572292,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/888c6d493552f3feead1a95a.pdf"},{"id":94210317,"identity":"98eab87c-dddb-45bf-b1b6-b433a8c5860a","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"pdf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":307205,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/76873f2f4fbaac99082482c8.pdf"},{"id":94211200,"identity":"61cc23d9-c839-4c4d-9032-52ca64aa53f2","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"pdf","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":246086,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/1d77e2408d80bce3a2960139.pdf"},{"id":94210316,"identity":"f5dce304-d415-405c-a68b-3e5f8a97c3e7","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"pdf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":216499,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/98cbcfa04dc3354cee31a220.pdf"},{"id":94210312,"identity":"bb84090a-b626-452c-8dbe-62793441c8f6","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50381,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/27039e6432a3ee91e1eff09b.png"},{"id":94210308,"identity":"9ce78049-d049-46b3-ba17-bfd8962562a4","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":96716,"visible":true,"origin":"","legend":"","description":"","filename":"28e14931f0a041eeb29fcabb7818462b1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/048256645596bc3ebc40919f.xml"},{"id":94211199,"identity":"294d12b4-9872-48d2-8c0a-f0a22d466b6d","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":116472,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/ea77edfdc9b1bd75ad162fda.html"},{"id":94210302,"identity":"a10ad629-e2ea-4f6e-b740-57a7c70b490d","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":286572,"visible":true,"origin":"","legend":"\u003cp\u003eParticipant selection flowchart. ADC: Apparent diffusion coefficient. VERDICT: Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors. WSI: Whole-Slide Imaging.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/a5af0d54fc01cfdf4c50042a.png"},{"id":94211193,"identity":"da4a5f54-0b1a-45a0-9649-8d214726170e","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":799444,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA.\u003c/strong\u003e Pipeline to obtain cell density maps from histology. High-resolution images were divided into 512-pixel-wide patches, nuclei were segmented (outlined in red), counted and normalized by the total area of the patch, each patch value constituted a pixel in the resulting cell density map. \u003cstrong\u003eB. \u003c/strong\u003eTissue type segmentation: original image (left) and segmentation result (right). Green corresponds to epithelium, blue to lumen and yellow to stroma.\u003c/p\u003e","description":"","filename":"Binder11.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/29514d9b5049f277f0d4f1a6.png"},{"id":94211198,"identity":"4d1aa57e-c315-489d-8081-9480b43550ad","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":576618,"visible":true,"origin":"","legend":"\u003cp\u003eRepresentative example: 67-year-old man with PSA levels of 10.89ng/mL at referral, prostate volume of 36ml and Likert Score 3. WSI (left) and corresponding masked MR and histologically derived maps (T2W and ADC from mpMRI and fIC from VERDICT-MRI; cell density, epithelial cell density and epithelial fraction from histology). Contours on the WSI include benign prostatic hyperplasia (green), inflammation (blue), Gleason 3+3 (yellow) and Gleason 3+4 (red). A benign (white) and Gleason 3+4 (black) ROI is contoured on all images. WSI: Whole-Slide Imaging. ADC: Apparent diffusion coefficient. fIC: Fractional intracellular volume. VERDICT: Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors. ROI: Region of interest.\u003c/p\u003e","description":"","filename":"Binder12.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/36ef909a6595409055b16f3b.png"},{"id":94211195,"identity":"aab8480a-79ed-49e1-982d-45136f999e41","added_by":"auto","created_at":"2025-10-23 15:41:05","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":465950,"visible":true,"origin":"","legend":"\u003cp\u003eRepresentative example: 62-year-old man with PSA levels of 7.75ng/mL at referral, prostate volume of 37ml and Likert Score 3. WSI (left) and corresponding masked MR and histologically derived maps (T2W and ADC from mpMRI and fIC from VERDICT-MRI; cell density, epithelial cell density and epithelial fraction from histology). Contours on the WSI include cystic atrophy (cyan), inflammation (blue) and Gleason 3+4 (red). A benign (white) and Gleason 3+4 (black) ROI is contoured on all images. WSI: Whole-Slide Imaging. ADC: Apparent diffusion coefficient. fIC: Fractional intracellular volume. VERDICT: Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors. ROI: Region of interest.\u003c/p\u003e","description":"","filename":"Binder13.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/91a00bf2fa18dc33b290e8aa.png"},{"id":94211868,"identity":"e1dd744e-be35-4279-b273-a6a8164f88ba","added_by":"auto","created_at":"2025-10-23 15:49:05","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":548478,"visible":true,"origin":"","legend":"\u003cp\u003eLinear correlation between MRI maps (VERDICT fIC and ADC, mpMRI ADC) and histological parameters (cell density, epithelial cell density and epithelial fraction). r: Pearson’s correlation coefficient with 95% confidence interval from bootstrapping in brackets. *: p-value\u0026lt;0.05. VERDICT: Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors. fIC: Fractional intracellular volume. ADC: Apparent diffusion coefficient. mpMRI: multiparametric MRI.\u003c/p\u003e","description":"","filename":"Binder14.png","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/29f8b1f79b8087640da15329.png"},{"id":101881870,"identity":"509d761b-7a77-4924-9d1a-eefb614ff86a","added_by":"auto","created_at":"2026-02-04 15:17:22","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3414148,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/e1fae4e8-fa99-4768-a03e-db9b2ad06c2b.pdf"},{"id":94211867,"identity":"9b2821a3-4a3b-44f5-8a65-cf4c5777bfc4","added_by":"auto","created_at":"2025-10-23 15:49:05","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":35555,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/d9d9946e0bb444de33eaf5fc.docx"},{"id":94210299,"identity":"35d00b05-5b0a-4841-a3b6-ace57f400ffa","added_by":"auto","created_at":"2025-10-23 15:33:05","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":22783,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7723707/v1/56acb43872d470a845f2da5f.docx"}],"financialInterests":"Competing interest reported. Dr Punwani serves as a consultant for Qubim, NVision, and Gold Standard Phantoms. mDr Shaw previously acted as a consultant for Angle plc. Dr Grey acts as a proctor for Sonablate, and as a consultant, proctor, and grant holder for Angiodynamics. He has also been a speaker for HC21/Aquilant. Dr Singh serves as a consultant for Docus AI. Dr Alexander is a director and shareholder of Queen Square Analytics. The remaining authors declare no competing financial interests.","formattedTitle":"Validating the biological origin of in vivo fractional intracellular volume from VERDICT-MRI in the prostate","fulltext":[{"header":"Introduction","content":"\u003cp\u003e Multiparametric MRI (mpMRI) plays a crucial role in the diagnosis and management of prostate cancer (PCa) by aiding in detection, localization, risk stratification, and staging of tumors. Despite its advantages, mpMRI has limited specificity[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] and insufficient sensitivity, missing 10\u0026ndash;20% of tumors[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Improving detection of clinically significant lesions requires an understanding of how MRI signals relate to histology and ensuring that these signals are specific to relevant histological features such as epithelial cell density, an increase of which is a key descriptor of adenocarcinoma \u0026ndash; the most common type of prostate cancer[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The apparent diffusion coefficient (ADC), a metric derived from the diffusion-weighted MRI acquisition within mpMRI, is routinely used to examine tissue microstructure. However, ADC compounds all physiological changes affecting water diffusion into a single measure, leading to poor discrimination of clinically significant (cs) from clinically insignificant PCa/benign change[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTo address this limitation, specialized diffusion-weighted MRI protocols combined with computational modelling offer a more precise characterization of tissue microstructure by examining specific features such as cell size and density[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors (VERDICT) framework is based on a biophysical model of diffusion in tumors[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. It considers three water compartments \u0026ndash; intracellular, vascular, and extracellular-extravascular \u0026ndash; and estimates the volume fraction of each along with their diffusion coefficients. The fractional intracellular volume (fIC) derived from VERDICT has been explored in several studies (NCT02721784, NCT02689271, NCT04792138, NCT05017181) and has shown promise as an imaging biomarker when used with mpMRI, achieving higher accuracy in detecting clinically significant PCa (area under the curve (AUC): 0.96) than ADC maps (AUC: 0.85). It has also demonstrated statistically significant differences between true-positive and false-positive lesions[\u003cspan additionalcitationids=\"CR10 CR11\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], with potential to reduce negative biopsies.\u003c/p\u003e\u003cp\u003eThis study aims to verify the biological interpretation of in vivo VERDICT fIC. We use personalized 3D-printed prostate molds to match whole-mount prostatectomy slices with in vivo VERDICT-MRI. We hypothesize that fIC correlates with epithelial cell density, since this is an assumption of the VERDICT model. We test our hypothesis by performing region of interest (ROI)-based comparisons of MRI and histology-derived measures of cell density and epithelial fraction. We also analyze differences between tissue composition and MR parameters in benign and csPCa tissue.\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eStudy Participants\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eEthical approval for this prospective study was granted by London-Queen Square Research Ethics Committee (19/LO/1803) on 23 January 2020. This study was performed in accordance with the Declaration of Helsinki. Patients were involved in research and ethic committee meetings reviewing study design and protocol prior to recruitment. We analyze data from the Histo-MRI clinical trial (registered with ClinicalTrials.gov, NCT04792138, first posted 10/03/2021 and ongoing since 2020 with consecutive sampling; amendments were made to increase sample size, 15/09/2023), where men with histologically confirmed PCa awaiting prostatectomy were scanned using VERDICT-MRI (n=68). Patients attending a One-stop Prostate Cancer clinic at University College London Hospital were screened against the inclusion and exclusion criteria and eligible patients were invited. All screened participants were provided with Participant Information Leaflets detailing the objectives of the study and their involvement; those interested provided informed written consent. \u0026nbsp;Inclusion criterium was men aged 18+ with no MRI contraindications. Exclusion criteria included inability to undergo MRI, inability to provide informed consent or prior/ongoing PCa treatment (surgery, radiotherapy, hormone treatment). MRI files were anonymized using DICOM cleaner software, and whole slide images saved under study identifications. Participants missing ADC on the mpMRI, or with large MRI artifacts, histology deformations, or lesions smaller than the diffusion-weighted imaging voxel size (1.5 mm\u0026sup2;) were excluded (Figure 1).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eTissue preparation\u003c/h2\u003e\n\u003cp\u003eTo ensure slice-to-slice correspondence between histology and MR images, custom molds were created from manual prostate contours on pre-operative T2W axial images using Horos[13]. Reference slices, set as the plane with the largest lesion detected on the mpMRI, were chosen and used to place cutting guides within the mold. After removal of the prostates during radical prostatectomy, specimens were inked to denote laterality and silicon catheters inserted through the urethra. Prostates were sealed in their molds and placed in neutral buffered formalin for 24-hour fixation. The full specimen handling protocol is described by Bourne et al. [14].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor histological analysis, a histopathologist (MM \u0026ndash; \u0026nbsp;5 years of experience) cut the reference slice following the cutting guides, spaced 5mm apart. The remainder of the prostate was progressively sliced at 5mm thickness. Tissue sections were processed per standard protocol.\u003c/p\u003e\n\u003ch2\u003eMRI acquisition\u003c/h2\u003e\n\u003cp\u003eThe MRI acquisitions included mpMRI and VERDICT-MRI[15]. \u0026nbsp; VERDICT scans were performed on a 3.0 T MRI scanner (Achieva or Ingenia; Philips, Best, the Netherlands). A pulse-gradient spin-echo sequence was acquired using a 32-channel cardiac coil (Achieva) or body coils (Ingenia). B-values included 90, 500, 1500, 2000 and 3000 s/mm\u003csup\u003e2\u003c/sup\u003e obtained in 3 orthogonal directions, with a b = 0 s/mm\u003csup\u003e2\u003c/sup\u003e image for every b-value. This was used to normalize T2 and T1 dependencies. Further sequence parameters were as follows: phase-encoding direction A/P, FOV = 220x220 mm\u003csup\u003e2\u003c/sup\u003e, reconstruction matrix = 176x176 and voxel size = 1.25x1.25x5.0 mm\u003csup\u003e3\u003c/sup\u003e. Each b-value had a corresponding TR and TE. VERDICT scans lasted 10mins 95s (Achieva) or 17mins 41s (Ingenia). For more details, see Singh et al. [15]. ADC maps were derived from clinical mpMRI (1.5 or 3T, see Supplemental Material) and from VERDICT diffusion-weighted images with b-values 0, 90, 500 and 1500 s/mm\u003csup\u003e2\u003c/sup\u003e (mpMRI ADC and VERDICT ADC, respectively).\u003c/p\u003e\n\u003cp\u003eRadiologists (AR \u0026ndash; \u0026lsquo;in-training\u0026rsquo;, NT \u0026ndash; 5 years of experience, SP \u0026ndash; 20 years of experience) drew benign and PCa ROIs on the MR images in consensus using MIM 7.2.9[16].\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eVERDICT model\u003c/h2\u003e\n\u003cp\u003eThe VERDICT mathematical model assumes no exchange between water populations\u003cs\u003e[8]\u003c/s\u003e. The normalized signal\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cimg width=\"8\" height=\"22\" src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAgAAAAWCAMAAADKDS1SAAAAAXNSR0IArs4c6QAAAE5QTFRFAAAAAAAAAAA6AABmADpmAGa2OgAAOgA6OjqQOpC2OpDbZgAAZrb/kDoAkGY6kNv/tmYAtmY6tpA6ttv/tv//25A62////9uQ//+2///b1I25hwAAAAF0Uk5TAEDm2GYAAAAJcEhZcwAADsQAAA7EAZUrDhsAAAAZdEVYdFNvZnR3YXJlAE1pY3Jvc29mdCBPZmZpY2V/7TVxAAAASElEQVQYV2NgIApI8jEyMnEyMEjysooxiHIxMIiz80M0SnADRcBAnI1FGMKS5GUWgcnyIDEEuBgkhUBSgmyMjBxQxUTZT0ARABbxAl83a9L2AAAAAElFTkSuQmCC\" alt=\"image\"\u003e\u0026nbsp;is the sum of three parametric models:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ch2\u003eHistological analysis\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eA semi-automated pipeline was created to process the histology images and produce quantitative maps. The appearance of the Hematoxylin and Eosin staining on histology is often highly variable\u003cs\u003e[18]\u003c/s\u003e. To normalize the images, the brightness, contrast, and saturation of stained whole-slide histopathology was standardized. Shrinkage was accounted for using a correction factor of 1.15, proposed by Schned et al. and confirmed by Jonmarker et al. for specimens processed with formalin fixation[19,20].\u003c/p\u003e\n\u003cp\u003eThe following steps were done on a patch-by-patch basis (535x535mm) on QuPath[21]. Cell density was calculated from cell nuclei segmentations (Figure 2a). Patches were segmented into stroma, lumen and epithelial tissue using a pixel classifier trained on expert annotations (histopathologist MM - 5 years of experience; Figure2b). Epithelial cell density was calculated by weighing cell density by epithelial fraction. All steps of this pipeline were visually checked by a single histopathologist (MM). Lesions in histology were contoured and graded by histopathologists (primary annotator MM and senior reviewers AF \u0026ndash; 20 years of experience \u0026ndash; and \u0026nbsp;AH \u0026ndash; 6 years of experience). Figures 3 \u0026amp; 4 show representative examples.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eStatistical analysis\u003c/h2\u003e\n\u003cp\u003eThe primary outcome of this study was the correlation (with 95% confidence intervals) of features from histology and MR derived maps. In each sample, ROIs were drawn in cancerous and benign areas of tissue using consistent anatomical landmarks across MR and histology. Cancerous ROIs were based on radiologist annotations on mpMRI; benign ROIs were drawn directly on histology (to exclude any confounding pathology) and transferred to MRI. The average value of each ROI was used. Statistical significance was based on 95% confidence intervals. The study was not powered to detect differences between correlations. All analysis was performed in Python. The secondary outcome was the AUC between csPCa and benign tissue. Both analyses included the full patient cohort.\u003c/p\u003e\n\u003ch2\u003eCode availability\u003c/h2\u003e\n\u003cp\u003eThe code used in this study is publicly available at https://github.com/Martamasramon/VERDICT-fIC-validation (commit 8e84b1eba228bc33c05143874db74ae9249da987).\u003cbr\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003eParticipant Characteristics\u003c/h2\u003e\n\u003cp\u003eFrom the total of 68 participants, four were excluded due to lack of ADC maps on mpMRI, six due to image artifacts in MRI, two due to deformations in histology, three due to small lesions on histology and one because of small lesions on MRI (\u0026lt;0.5mm\u003csup\u003e2\u003c/sup\u003e, Figure 1). Participant and lesion characteristics of the remaining participants (n=52) are summarized in Table 1. The mean age of these men is 65 years ± 6. The mean prostate-specific antigen level is 8.4 ng/mL ± 6.1. The mean time between PSA and prostatectomy is 56 ± 39 days and the mean time between mpMRI and VERDICT-MRI is 108 ± 115 days. A single PCa and benign ROI was taken from each patient. Lesions have Gleason grades 3+4 (n=37), 3+4+5 (n=4), 4+3 (n=7), 4+3+5 (n=2), 4+4 (n=1) and 4+5 (n=1).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eComparison of VERDICT fIC and mpMRI ADC against histological features in benign and clinically significant PCa ROIs\u003c/h2\u003e\n\u003cp\u003eWhen analyzing benign and cancerous ROIs in the full cohort, fIC correlates most strongly to the histological epithelial fraction (r=0.784 [0.70, 0.84]). It correlates with lower strength to epithelial cell density (r=0.747 [0.66, 0.81]) and cell density (r=0.711 [0.60, 0.79]). VERDICT ADC maps show weaker correlations to these histological parameter maps (r=-0.639 [-0.54, -0.71], -0.609 [-0.51, -0.69] and -0.587 [-0.45, -0.69], respectively) while mpMRI ADC exhibits the weakest correlations (r=-0.353 [-0.18, -0.50], -0.326 [-0.17, -0.47] and -0.273 [-0.09, -0.43], respectively). In all cases, the correlation with epithelial fraction is strongest. These findings are summarized in Table 2 and illustrated in Figure 5.\u003c/p\u003e\n\u003ch2\u003eComparison of MR-derived parameters and tissue composition in benign and cancerous prostate regions\u003c/h2\u003e\n\u003cp\u003eTable 3 presents how fIC, ADC and tissue composition differ between benign and csPCa prostate regions. Tissue composition changes between benign and csPCa are characterized by an increase in epithelial fraction (0.21 ± 0.08 to 0.43 ± 0.12, AUC=0.94, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) and a decrease in lumen fraction (0.24 ± 0.13 to 0.12 ± 0.04, AUC=0.86, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001) and stroma fraction (0.57 ± 0.16 to 0.49 ± 0.13, AUC=0.68, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001).\u0026nbsp;Cell density and epithelial cell density are both higher in csPCa (3780 ± 790 and 1810 ± 750, respectively) compared to benign tissue (2360 ± 780 and 680 ± 380, respectively).\u003c/p\u003e\n\u003cp\u003eAs in the case of cell density, epithelial cell density and epithelial fraction, VERDICT fIC shows a statistically significant difference between benign and csPCa tissue (0.16 ± 0.10 to 0.51 ± 0.13, AUC=0.99, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Mean VERDICT ADC and mpMRI ADC values in benign and csPCa regions are 1.00 ± 0.19 to 0.68 ± 0.14 and 1.36 ± 0.32 to 1.05 ± 0.31, respectively.\u003c/p\u003e"},{"header":"Discussion and Conclusions","content":"\u003cp\u003eThis study investigated the biological origin of the fractional intracellular volume from Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors, or VERDICT, MRI using 3D personalized molds to enable accurate histological comparison. Region of interest analysis showed strong correlation of fractional intracellular volume with epithelial fraction (r=0.784) and histological cell density (r=0.711). This supports our hypothesis that the fractional intracellular volume estimate is positively correlated with epithelial cell density.\u003c/p\u003e\n\u003cp\u003eThe VERDICT model assumes that the intracellular component captures the signal solely from epithelial cells while the signal from stromal cells and lumen is accounted for by the extra-cellular extra-vascular component. Importantly, results show that fIC correlates better with epithelial fraction than cell density, which has been shown to be a less effective Gleason Pattern predictor than tissue fractions[22].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePrevious ROI-based studies that describe the relationship between histological cell density and ADC in PCa report correlations between r = -0.50 and r = -0.695[23–26]. Our findings for ADC from mpMRI show a correlation of r=-0.273. We speculate that the variability in these results is due to differences in the b-values, scanner manufacturers and diffusion sequences used to calculate ADC maps, as well as cell counting algorithms. Notably, the higher correlation of ADC with epithelial fraction presented here (r=-0.353) has been previously noted by Chatterjee et al., who reported correlations with cell count and epithelial fraction in fixed tissue of -0.598 and -0.647, respectively[27]. They also showed that both correlations are stronger in fresh tissue.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur results also show that VERDICT ADC is more strongly correlated to histological measures of cell density and epithelial fraction than mpMRI ADC (-0.587 and -0.639 vs -0.273 and -0.353). We believe that this is primarily due to using a consistent imaging sequence. In this study we derive the VERDICT ADC from diffusion-weighted images with b-values 0, 90, 500 and 1500 s/mm2, each with a given TR and TE, while each mpMRI ADC uses a different number of diffusion-weighted images, with different b-values (see Supplemental Material) and inconsistent TR and TE. The fact that VERDICT fIC outperforms VERDICT ADC highlights the importance of complex microstructural models to discriminate the relevant signal components.\u003c/p\u003e\n\u003cp\u003eFurther analysis investigated differences in tissue composition and MR-derived parameters between benign and csPCa ROIs. Consistent with previous findings[27], csPCa regions exhibit a higher epithelial fraction and lower luminal fraction compared to benign tissue (0.21 ± 0.08 and 0.24 ± 0.13 vs 0.43 ± 0.12 and 0.12 ± 0.04). VERDICT fIC demonstrates similar changes to the epithelium, with values increasing from 0.16 ± 0.10 in benign regions to 0.51 ± 0.13 in csPCa. \u0026nbsp;The highest AUC values correspond to epithelial fraction and fIC (0.94 and 0.99, respectively), highlighting their ability to differentiate between tissue types, while both VERDICT ADC and mpMRI ADC exhibit lower discriminatory performance (AUC=0.92 and 0.76, respectively).\u003c/p\u003e\n\u003cp\u003eWe note certain limitations in our study. Resection of the prostate creates shape changes due to removal of mechanical tension and compression and hemodynamic pressure[28], dehydration and embedding of the tissue during processing causes tissue shrinkage and thin sectioning may cause further deformations[14]. This compromises the histology-MR mapping and affects the distribution of tissue fractions; future work could incorporate deformable registration or 3D histology reconstruction to mitigate these effects. Lesions smaller than 0.5 mm² were excluded from our analysis, as they fall below the resolution threshold of diffusion-weighted MRI; investigating whether VERDICT can detect such small lesions will require higher-resolution acquisitions or ultra-high-field imaging[29]. MpMRI was performed on various scanners while VERDICT-MRI was consistently obtained from Philips machines;\u0026nbsp;assessing reproducibility across vendors and field strengths in future multi-center studies will be essential for clinical translation. Finally, the validation design of this study restricted our analysis to prostatectomy cases, and the limited cohort size represents an additional constraint.\u003c/p\u003e\n\u003cp\u003eIn conclusion, this study established the biological origin of the Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors’ fractional intracellular volume, showing strong correlation with epithelial fraction and histological cell density in region of interest-based analysis, and aiding interpretability of a biomarker that has already shown potential to improve prostate cancer characterization non-invasively[30] [9]. Future work will further investigate the relationship between the model’s parameters and histological metrics across different Gleason scores and growth patterns, with a focus on mpMRI-invisible lesions, in larger, multi-center cohorts using a range of scanner manufacturers.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eADC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eapparent diffusion coefficient\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003earea under the curve\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ecs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eclinically significant\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003efIC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003efractional intracellular volume\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003empMRI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003emultiparametric MRI\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePCa\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eprostate cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eROI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eregion of interest\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eVERDICT\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eVascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eWSI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eWhole-Slide Imaging\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003eData Availability\u003c/p\u003e\n\u003cp\u003eThe datasets analyzed within this study are not publicly available due to data protection laws but may be available from the corresponding author on reasonable request and subject to approval by the relevant institutional data protection office.\u003c/p\u003e\n\u003cp\u003eFunding Declaration\u003c/p\u003e\n\u003cp\u003eThis work is supported by the EPSRC-funded UCL Centre for Doctoral Training in Intelligent, Integrated Imaging in Healthcare (i4health) [EP/S021930/1]; EPSRC grant numbers EP/N021967/1, EP/R006032/1; Prostate Cancer UK, Targeted Call 2014, Translational Research St.2, grant number PG14-018-TR2; the National Institute for Health and Care Research, University College London Hospitals Biomedical Research Centre; and Cancer Research UK National Cancer Imaging Translational Accelerator.\u003c/p\u003e\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\u003cp\u003eM. Masramon conceived the study, carried out the computational analysis, and was responsible for drafting and writing the manuscript. M. Mathew contributed to patient recruitment and led the histological analysis. S.P. and E.P. contributed to the conceptualization of the study, funding acquisition, and supervision. A.R., N.T., S.P., A.H., A.F. and M. Mathew provided image annotations. S.S., T.M., J.C., M.-V.P., L.S., V.K., A.G., E.D., G.S., D.P., L.C., and C.M.M. contributed to the administration of the clinical trial and/or data acquisition. D.A. and A.P. supported methodology development. T.P. provided advice for the statistical analysis. D.C.A. contributed to funding acquisition. All authors contributed to the review and editing of the manuscript.\u003c/p\u003e\n\u003cp\u003eCompeting Interests Statement\u003c/p\u003e\n\u003cp\u003eDr Punwani serves as a consultant for Qubim, NVision, and Gold Standard Phantoms. mDr Shaw previously acted as a consultant for Angle plc. Dr Grey acts as a proctor for Sonablate, and as a consultant, proctor, and grant holder for Angiodynamics. He has also been a speaker for HC21/Aquilant. Dr Singh serves as a consultant for Docus AI. Dr Alexander is a director and shareholder of Queen Square Analytics. The remaining authors declare no competing financial interests. \u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAhmed, H. U. et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study. \u003cem\u003eLancet\u003c/em\u003e \u003cb\u003e389\u003c/b\u003e, 815\u0026ndash;822 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao, Y. et al. Comparison of Multiparametric Magnetic Resonance Imaging with Prostate-Specific Membrane Antigen Positron-Emission Tomography Imaging in Primary Prostate Cancer Diagnosis: A Systematic Review and Meta-Analysis. \u003cem\u003eCancers (Basel)\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e, 3497 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSathianathen, N. J. et al. Negative Predictive Value of Multiparametric Magnetic Resonance Imaging in the Detection of Clinically Significant Prostate Cancer in the Prostate Imaging Reporting and Data System Era: A Systematic Review and Meta-analysis. \u003cem\u003eEur. Urol.\u003c/em\u003e \u003cb\u003e78\u003c/b\u003e, 402\u0026ndash;414 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEpstein, J. I., Cubilla, A. L. \u0026amp; Humphrey, P. A. \u003cem\u003eTumors of the prostate gland, seminal vesicles, penis, and scrotum. AFIP Atlas of Tumor Pathology\u003c/em\u003e (American Registry of Pathology, 2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRosenkrantz, A. B. et al. Interobserver reproducibility of the PI-RADS version 2 lexicon: A multicenter study of six experienced prostate radiologists. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e280\u003c/b\u003e, 793\u0026ndash;804 (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWestphalen, A. C. et al. Variability of the positive predictive value of PI-RADS for prostate MRI across 26 centers: Experience of the society of abdominal radiology prostate cancer disease-focused panel. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e296\u003c/b\u003e, 76\u0026ndash;84 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMartinez-Heras, E., Grussu, F., Prados, F., Solana, E. \u0026amp; Llufriu, S. Diffusion-Weighted Imaging: Recent Advances and Applications. \u003cem\u003eSeminars Ultrasound CT MRI\u003c/em\u003e. \u003cb\u003e42\u003c/b\u003e, 490\u0026ndash;506 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePanagiotaki, E. et al. Noninvasive quantification of solid tumor microstructure using VERDICT MRI. \u003cem\u003eCancer Res.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 1902\u0026ndash;1912 (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJohnston, E. W. et al. VERDICT MRI for prostate cancer: Intracellular volume fraction versus apparent diffusion coefficient. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e291\u003c/b\u003e, 391\u0026ndash;397 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSingh, S. et al. Avoiding Unnecessary Biopsy after Multiparametric Prostate MRI with VERDICT Analysis: The INNOVATE Study. \u003cem\u003eRadiology\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1148/radiol.212536\u003c/span\u003e\u003cspan address=\"10.1148/radiol.212536\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAppayya, M. B. et al. The intracellular component of VERDICT (Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors) MRI distinguishes Gleason 4 pattern better than Apparent Diffusion Coefficient. (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSen, S. et al. Differentiating False Positive Lesions from Clinically Significant Cancer and Normal Prostate Tissue Using VERDICT MRI and Other Diffusion Models. \u003cem\u003eDiagnostics\u003c/em\u003e 12, (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHoros Project. Horos. Preprint at https://horosproject.org/.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBourne, R. M. et al. Apparatus for histological validation of in vivo and ex vivo magnetic resonance imaging of the human prostate. \u003cem\u003eFront Oncol\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSingh, S. et al. Histo-MRI map study protocol: a prospective cohort study mapping MRI to histology for biomarker validation and prediction of prostate cancer. \u003cem\u003eBMJ Open.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, e059847 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMIM Software Inc. MIM 7.2.9. Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mimsoftware.com/\u003c/span\u003e\u003cspan address=\"https://www.mimsoftware.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBonet-Carne, E. et al. VERDICT-AMICO: Ultrafast fitting algorithm for non-invasive prostate microstructure characterization. \u003cem\u003eNMR Biomed\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSuvarna, K., Layton, C. \u0026amp; Bancroft, J. \u003cem\u003eBancroft\u0026rsquo;s Theory and Practice of Histological Techniques\u003c/em\u003e. (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchned, A. R. et al. Tissue-shrinkage Correction Factor in the Calculation of Prostate Cancer Volume. \u003cem\u003eAm. J. Surg. Pathol.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 1501\u0026ndash;1506 (1996).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJonmarker, S., Valdman, A., Lindberg, A., Hellstr\u0026ouml;m, M. \u0026amp; Egevad, L. Tissue shrinkage after fixation with formalin injection of prostatectomy specimens. \u003cem\u003eVirchows Arch.\u003c/em\u003e \u003cb\u003e449\u003c/b\u003e, 297\u0026ndash;301 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBankhead, P. et al. QuPath: Open source software for digital pathology image analysis. \u003cem\u003eSci Rep\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChatterjee, A. et al. Changes in Epithelium, Stroma, and Lumen Space Correlate More Strongly with Gleason Pattern and Are Stronger Predictors of Prostate ADC Changes than Cellularity Metrics. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e277\u003c/b\u003e, 751\u0026ndash;762 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGibbs, P. et al. Correlation of ADC and T2 Measurements With Cell Density in Prostate Cancer at 3.0 Tesla. at (2009). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u0026thinsp;http://journals.lww.com/investigativeradiology\u003c/span\u003e\u003cspan address=\"http://\u0026thinsp;http://journals.lww.com/investigativeradiology\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZelhof, B. et al. Correlation of diffusion-weighted magnetic resonance data with cellularity in prostate cancer. \u003cem\u003eBJU Int.\u003c/em\u003e \u003cb\u003e103\u003c/b\u003e, 883\u0026ndash;888 (2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGlazer, D. I. et al. Diffusion-weighted endorectal MR imaging at 3T for prostate cancer: correlation with tumor cell density and percentage Gleason pattern on whole mount pathology. \u003cem\u003eAbdom. Radiol.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e, 918\u0026ndash;925 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXi, Z. W. et al. Diffusion-weighted imaging of prostate cancer: Correlation between apparent diffusion coefficient values and tumor proliferation. \u003cem\u003eJ. Magn. Reson. Imaging\u003c/em\u003e. \u003cb\u003e29\u003c/b\u003e, 1360\u0026ndash;1366 (2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChatterjee, A. et al. Changes in epithelium, stroma, and lumen space correlate more strongly with gleason pattern and are stronger predictors of prostate ADC changes than cellularity metrics1. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e277\u003c/b\u003e, 751\u0026ndash;762 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOrczyk, C., Taneja, S. S., Rusinek, H. \u0026amp; Rosenkrantz, A. B. Assessment of change in prostate volume and shape following surgical resection through co-registration of in-vivo MRI and fresh specimen ex-vivo MRI. \u003cem\u003eClin. Radiol.\u003c/em\u003e \u003cb\u003e69\u003c/b\u003e, e398\u0026ndash;e403 (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMolendowska, M. et al. Diffusion MRI in prostate cancer with ultra-strong whole‐body gradients. \u003cem\u003eNMR Biomed\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJohnston, E. et al. INNOVATE: A prospective cohort study combining serum and urinary biomarkers with novel diffusion-weighted magnetic resonance imaging for the prediction and characterization of prostate cancer. \u003cem\u003eBMC Cancer\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, (2016).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 3 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Prostate, Cancer, MRI, VERDICT-MRI","lastPublishedDoi":"10.21203/rs.3.rs-7723707/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7723707/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eReliable imaging biomarkers are needed to characterize clinically significant prostate cancer (csPCa). The clinically established measure, apparent diffusion coefficient (ADC) from multiparametric (mp)MRI, lacks specificity to key histological features. The fractional intracellular volume (fIC) from in vivo Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumors (VERDICT)-MRI has shown improved detection of csPCa. This prospective study (NCT04792138) quantifies the accuracy of fIC as a measure of prostate epithelial cell density. Fifty-two participants (mean age, 65 years\u0026thinsp;\u0026plusmn;\u0026thinsp;6; Gleason grades\u0026thinsp;\u0026ge;\u0026thinsp;3\u0026thinsp;+\u0026thinsp;4) with biopsy-confirmed PCa underwent mpMRI and VERDICT-MRI before radical prostatectomy. Personalized molds from preoperative mpMRI enabled histology to MRI correspondence. Comparisons between histological epithelial density measures and MR markers were made on a region of interest (ROI) basis using Pearson\u0026rsquo;s correlation coefficient. VERDICT fIC corresponded more strongly with epithelial fraction, cell density and epithelial density (r\u0026thinsp;=\u0026thinsp;0.784, 0.711, 0.747) than both VERDICT derived ADC (r=-0.639, -0.587, -0.609) and mpMRI ADC (r-0.353, -0.273, -0.326). Differences between benign tissue and csPCa in VERDICT fIC (0.16\u0026thinsp;\u0026plusmn;\u0026thinsp;0.10 vs 0.51\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13) were similar to epithelial fraction (0.21\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08 vs 0.43\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12). These findings support in vivo VERDICT fIC as a marker of epithelial cell density and show promise for wider clinical use.\u003c/p\u003e","manuscriptTitle":"Validating the biological origin of in vivo fractional intracellular volume from VERDICT-MRI in the prostate","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-23 15:33:00","doi":"10.21203/rs.3.rs-7723707/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8465b1b0-d3ab-4fe2-9b04-516a3c9f336d","owner":[],"postedDate":"October 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":56497112,"name":"Health sciences/Biomarkers"},{"id":56497113,"name":"Biological sciences/Cancer"},{"id":56497114,"name":"Health sciences/Medical research"},{"id":56497115,"name":"Health sciences/Oncology"},{"id":56497116,"name":"Health sciences/Urology"}],"tags":[],"updatedAt":"2026-02-04T11:26:30+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-23 15:33:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7723707","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7723707","identity":"rs-7723707","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}