Share：
Share this content in WeChat
X
[Chinese] [PDF] 606 2
• Clinical Article •
Study on the clinical value of assessing prostate cancer aggressiveness and prognosis based on multi-parameter MRI radiomics cluster analysis
QI Dong  YANG Hongkai  QI Xuan  DUAN Shaofeng  HE Yongsheng 

Cite this article as: QI D, YANG H K, QI X, et al. Study on the clinical value of assessing prostate cancer aggressiveness and prognosis based on multi-parameter MRI radiomics cluster analysis[J]. Chin J Magn Reson Imaging, 2025, 16(9): 124-131. DOI:10.12015/issn.1674-8034.2025.09.019.


[Abstract] Objective To identify the intrinsic imaging phenotype based on multi parameter magnetic resonance imaging (mpMRI) radiomics clustering analysis of prostate cancer, in order to evaluate the invasiveness of prostate cancer and predict prognosis.Materials and Methods We retrospectively collected preoperative mpMRI and clinical pathological data of 185 patients with pathologically confirmed prostate cancer from January 2022 to January 2024. The mpMRI includes ZOOMit diffusion weighted imaging (ZOOMit-DWI), ZOOMit apparent diffusion coefficient (ZOOMit-ADC), T2WI, and T2WI fat suppression (T2WI-FS). Cluster analysis was performed based on extracting radiomics features from mpMRI to obtain clustering subtypes. Chi square test was used to analyze the categorical variables Gleason score, significant prostate cancer (sigPCA), P504S, lymph node metastasis (LNM), prostate-specific antigen (PSA), visible cancer thrombus in the vasculature, perineural invasion (PNI), and Ki-67 in clinical pathological variables. Independent sample t-test was used to evaluate the age and prostate volume (PV) of clinical continuous variable data, exploring the relationship between clinical pathological variables and subtypes. Is there a significant difference or association.Results Two cluster subtypes were obtained. Cluster 1 was associated with a higher incidence of clinically significant prostate cancer (92.857%) and lymph node metastasis (21.429%). There was a statistically significant difference between cluster 1 and cluster 2, with P values of 0.024 (sigPCA) and 0.028 (LNM), respectively. There was a statistically significant difference in Gleason scores between the two subtypes of clusters (P = 0.035). In cluster 1, the proportion of (4 + 3) scores was the highest, at 32.143%, followed by (3 + 4) and (5 + 4) scores, both at 15.476%. In Cluster 2, the proportion of (3 + 4) scores is the highest, at 26.733%. The proportions of (3 + 5) and (5 + 3) for the two subtypes of clusters are 2.381% and 1.980%, respectively. There was no statistically significant difference in the incidence of PNI between the two clusters (P = 0.754).Conclusions The potential imaging phenotype of multi parameter MRI in prostate cancer is associated with the incidence of sigPCA and LNM, which symbolize high invasiveness and poor prognosis. This can help evaluate the prognosis of PCA patients and provide individualized treatment after risk stratification.
[Keywords] prostate cancer；multi parameter magnetic resonance imaging；cluster analysis；invasiveness；prognosis

QI Dong   YANG Hongkai   QI Xuan   DUAN Shaofeng   HE Yongsheng*  

Imaging Department of Maanshan People's Hospital, Maanshan 243000, China

Corresponding author: HE Y S, E-mail: heyongsheng881@163.com

Conflicts of interest   None.

Received  2025-03-17
Accepted  2025-08-25
DOI: 10.12015/issn.1674-8034.2025.09.019
Cite this article as: QI D, YANG H K, QI X, et al. Study on the clinical value of assessing prostate cancer aggressiveness and prognosis based on multi-parameter MRI radiomics cluster analysis[J]. Chin J Magn Reson Imaging, 2025, 16(9): 124-131. DOI:10.12015/issn.1674-8034.2025.09.019.

[1]
BRAY F, LAVERSANNE M, SUNG H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2024, 74(3): 229-263. DOI: 10.3322/caac.21834.
[2]
SIEGEL R L, MILLER K D, FUCHS H E, et al. Cancer statistics, 2022[J]. CA A Cancer J Clinicians, 2022, 72(1): 7-33. DOI: 10.3322/caac.21708.
[3]
BRENDLE C, MARTIROSIAN P, SCHWENZER N F, et al. Diffusion-weighted imaging in the assessment of prostate cancer: Comparison of zoomed imaging and conventional technique[J]. Eur J Radiol, 2016, 85(5): 893-900. DOI: 10.1016/j.ejrad.2016.02.020.
[4]
HU L, FU C X, SONG X Y, et al. Automated deep-learning system in the assessment of MRI-visible prostate cancer: comparison of advanced zoomed diffusion-weighted imaging and conventional technique[J/OL]. Cancer Imaging, 2023, 23(1): 6 [2024-05-06]. https://pubmed.ncbi.nlm.nih.gov/36647150/. DOI: 10.1186/s40644-023-00527-0.
[5]
ZHU X H, SHAO L Z, LIU Z Y, et al. MRI-derived radiomics models for diagnosis, aggressiveness, and prognosis evaluation in prostate cancer[J]. J Zhejiang Univ Sci B, 2023, 24(8): 663-681. DOI: 10.1631/jzus.B2200619.
[6]
SIGORSKI D, GULCZYŃSKI J, SEJDA A, et al. Investigation of neural microenvironment in prostate cancer in context of neural density, perineural invasion, and neuroendocrine profile of tumors[J/OL]. Front Oncol, 2021, 11: 710899 [2024-05-06]. https://pmc.ncbi.nlm.nih.gov/articles/PMC8281889/. DOI: 10.3389/fonc.2021.710899.
[7]
HE Y S, DUAN S F, WANG W L, et al. Integrative radiomics clustering analysis to decipher breast cancer heterogeneity and prognostic indicators through multiparametric MRI[J/OL]. NPJ Breast Cancer, 2024, 10: 72 [2025-03-14]. https://pmc.ncbi.nlm.nih.gov/articles/PMC9252170/. DOI: 10.1038/s41523-024-00678-8.
[8]
WELS M G, LADES F, MUEHLBERG A, et al. General purpose radiomics for multi-modal clinical research[C]//Medical Imaging 2019: Computer-Aided Diagnosis. San Diego, USA. SPIE, 2019: 150. DOI: 10.1117/12.2511856.
[9]
ZWANENBURG A, VALLIÈRES M, ABDALAH M A, et al. The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping[J]. Radiology, 2020, 295(2): 328-338. DOI: 10.1148/radiol.2020191145.
[10]
LU X F, MENG J L, ZHOU Y J, et al. MOVICS: an R package for multi-omics integration and visualization in cancer subtyping[J]. Bioinformatics, 2021, 36(22/23): 5539-5541. DOI: 10.1093/bioinformatics/btaa1018.
[11]
OYEWOLE G J, THOPIL G A. Data clustering: application and trends[J]. Artif Intell Rev, 2023, 56(7): 6439-6475. DOI: 10.1007/s10462-022-10325-y.
[12]
CHEN R, REN S C, YIU M K, et al. Prostate cancer in Asia: a collaborative report[J]. Asian J Urol, 2014, 1(1): 15-29. DOI: 10.1016/j.ajur.2014.08.007.
[13]
MOHLER J L, ANTONARAKIS E S, ARMSTRONG A J, et al. Prostate cancer, version 2.2019, NCCN clinical practice guidelines in oncology[J]. J Natl Compr Canc Netw, 2019, 17(5): 479-505. DOI: 10.6004/jnccn.2019.0023.
[14]
MOTTET N, VAN DEN BERGH R C N, BRIERS E, et al. EAU-EANM-ESTRO-ESUR-SIOG guidelines on prostate cancer-2020 update. part 1: screening, diagnosis, and local treatment with curative intent[J]. Eur Urol, 2021, 79(2): 243-262. DOI: 10.1016/j.eururo.2020.09.042.
[15]
NKETIAH G, ELSCHOT M, KIM E, et al. T2-weighted MRI-derived textural features reflect prostate cancer aggressiveness: preliminary results[J]. Eur Radiol, 2017, 27(7): 3050-3059. DOI: 10.1007/s00330-016-4663-1.
[16]
JI G H, LIU F, CHEN Z Q, et al. Correlation of magnetic resonance imaging and computed tomography with biological factor expression and lymph node metastasis in aggressive prostate cancer[J]. Urol Int, 2025, 109(2): 175-181. DOI: 10.1159/000541953.
[17]
SONG Z, ZHOU Q, ZHANG J L, et al. Marker Ki-67 is a potential biomarker for the diagnosis and prognosis of prostate cancer based on two cohorts[J]. World J Clin Cases, 2024, 12(1): 32-41. DOI: 10.12998/wjcc.v12.i1.32.
[18]
STEPHEN N, BADHE B A. Diagnostic utility of immunohistochemical markers alpha methyl acyl coA racemase (AMACR) and Ets related gene (ERG) in prostate cancer[J]. Int J Clin Exp Pathol, 2022, 15(9): 364-372.
[19]
YANG D J, SHI X, LEI Y, et al. The auxiliary diagnostic value of prostate-specific antigen and α-methylacyl-CoA racemase in prostate cancer[J]. Oncol Lett, 2020, 20(2): 1418-1422. DOI: 10.3892/ol.2020.11658.
[20]
OSÓRIO C F E M, COSTA W S, GALLO C B M, et al. Comparative study of gleason 7 (3+4) and (4+3) prostatic adenocarcinomas with prognostic criteria and immunohistochemical profiles of AMACR, PSA and ki-67[J/OL]. Int Braz J Urol, 2025, 51(1): e202409922 [2025-03-14]. https://pubmed.ncbi.nlm.nih.gov/39388614/. DOI: 10.1590/s1677-5538.ibju.2024.9922.
[21]
LI J B, KANG R, TANG D L. Cellular and molecular mechanisms of perineural invasion of pancreatic ductal adenocarcinoma[J]. Cancer Commun (Lond), 2021, 41(8): 642-660. DOI: 10.1002/cac2.12188.
[22]
ZHANG L J, WU B, ZHA Z L, et al. Perineural invasion as an independent predictor of biochemical recurrence in prostate cancer following radical prostatectomy or radiotherapy: a systematic review and meta-analysis[J/OL]. BMC Urol, 2018, 18(1): 5 [2025-03-14]. https://pubmed.ncbi.nlm.nih.gov/29390991/. DOI: 10.1186/s12894-018-0319-6.
[23]
WU S L, LIN X M, LIN S X, et al. Impact of biopsy perineural invasion on the outcomes of patients who underwent radical prostatectomy: a systematic review and meta-analysis[J]. Scand J Urol, 2019, 53(5): 287-294. DOI: 10.1080/21681805.2019.1643913.
[24]
KARSIYAKALI N, OZGEN M B, OZVEREN B, et al. The association between perineural invasion in mpMRI-targeted and/or systematic prostate biopsy and adverse pathological outcomes in robot-assisted radical prostatectomy[J]. Actas Urol Esp (Engl Ed), 2022, 46(6): 377-384. DOI: 10.1016/j.acuroe.2022.02.004.
[25]
ZHANG W, ZHANG W T, LI X, et al. Predicting tumor perineural invasion status in high-grade prostate cancer based on a clinical-radiomics model incorporating T2-weighted and diffusion-weighted magnetic resonance images[J/OL]. Cancers (Basel), 2022, 15(1): 86 [2025-03-14]. https://pmc.ncbi.nlm.nih.gov/articles/PMC9817925/. DOI: 10.3390/cancers15010086.
[26]
KARTASALO K, STRÖM P, RUUSUVUORI P, et al. Detection of perineural invasion in prostate needle biopsies with deep neural networks[J]. Virchows Arch, 2022, 481(1): 73-82. DOI: 10.1007/s00428-022-03326-3.
[27]
ZHAN P C, LIU K Y, QIU Q Y, et al. Radiomics based on CT for the noninvasive prediction of perineural invasion of cholangiocarcinoma[J]. Radiol Pract, 2023, 38(7): 910-915. DOI: 10.13609/j.cnki.1000-0313.2023.07.017.
[28]
YANG M M, YANG L. Value of contrast-enhanced CT radiomics and nomogram in predicting perineural invasion in gastric cancer[J]. Chin J CT MRI, 2023, 21(11): 121-124.
[29]
YOU H J, SHANG W T, MIN X D, et al. Sight and switch off: Nerve density visualization for interventions targeting nerves in prostate cancer[J/OL]. Sci Adv, 2020, 6(6): eaax6040 [2025-03-14]. https://pubmed.ncbi.nlm.nih.gov/32076639/. DOI: 10.1126/sciadv.aax6040.
[30]
QI D, QI X, YANG H K, et al. Current status of potential magnetic resonance imaging markers in the neural microenvironment in prostate cancer patients[J]. Chin J Magn Reson Imag, 2024, 15(8): 194-200. DOI: 10.12015/issn.1674-8034.2024.08.031.
[31]
ZHAI C F, HE Y S, QI X, et al. Prediction of Ki67 expression in prostate cancer by radiomics models based on multiparameter magnetic resonance imaging[J]. J Mol Imag, 2024, 47(8): 793-799. DOI: 10.12122/j.issn.1674-4500.2024.08.03.
[32]
CHEEWADHANARAKS S, SEREEBORWORNTHANASAK K, SIRIPONGSATIAN D, et al. Exploring the correlation between multiparameters detected in primary prostate cancer using F-18 PSMA-1007 PET/MRI and their potential for predicting metastasis[J]. Nucl Med Mol Imaging, 2025, 59(1): 79-90. DOI: 10.1007/s13139-024-00884-z.
[33]
DU B, QI X, YANG H K, et al. To analyze the value of radiomics based on different diffusion model parameter maps in the early diagnosis of clinically significant prostate cancer by magnetic resonance imaging[J]. Chin J Magn Reson Imag, 2024, 15(2): 83-89. DOI: 10.12015/issn.1674-8034.2024.02.012.
[34]
CASTIGLIONI I, RUNDO L, CODARI M, et al. AI applications to medical images: From machine learning to deep learning[J]. Phys Med, 2021, 83: 9-24. DOI: 10.1016/j.ejmp.2021.02.006.
[35]
CROCKETT M J, BAI X, KAPOOR S, et al. The limitations of machine learning models for predicting scientific replicability[J/OL]. Proc Natl Acad Sci USA, 2023, 120(33): e2307596120 [2025-03-14]. https://pubmed.ncbi.nlm.nih.gov/37549293/. DOI: 10.1073/pnas.2307596120.
[36]
GAO S W, QIU Z T, SONG Y Y, et al. Unsupervised clustering reveals new prostate cancer subtypes[J]. Transl Cancer Res, 2017, 6(3): 561-572. DOI: 10.21037/tcr.2017.05.15.
[37]
PEREZ-JOHNSTON R, ARAUJO-FILHO J A, CONNOLLY J G, et al. CT-based radiogenomic analysis of clinical stage I lung adenocarcinoma with histopathologic features and oncologic outcomes[J]. Radiology, 2022, 303(3): 664-672. DOI: 10.1148/radiol.211582.
[38]
ASHRAF A B, DAYE D, GAVENONIS S, et al. Identification of intrinsic imaging phenotypes for breast cancer tumors: preliminary associations with gene expression profiles[J]. Radiology, 2014, 272(2): 374-384. DOI: 10.1148/radiol.14131375.

PREV The value of intravoxel incoherent motion diffusion weighted imaging in the evaluation of bladder cancer grade and myometrial invasion
NEXT The value of ADC histogram in preoperative prediction of lymphvascular space invasion in early cervical cancer
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn