1
|
Hu C, Qiao X, Huang R, Hu C, Bao J, Wang X. Development and Validation of a Multimodality Model Based on Whole-Slide Imaging and Biparametric MRI for Predicting Postoperative Biochemical Recurrence in Prostate Cancer. Radiol Imaging Cancer 2024; 6:e230143. [PMID: 38758079 PMCID: PMC11148661 DOI: 10.1148/rycan.230143] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/23/2023] [Revised: 02/16/2024] [Accepted: 03/25/2024] [Indexed: 05/18/2024]
Abstract
Purpose To develop and validate a machine learning multimodality model based on preoperative MRI, surgical whole-slide imaging (WSI), and clinical variables for predicting prostate cancer (PCa) biochemical recurrence (BCR) following radical prostatectomy (RP). Materials and Methods In this retrospective study (September 2015 to April 2021), 363 male patients with PCa who underwent RP were divided into training (n = 254; median age, 69 years [IQR, 64-74 years]) and testing (n = 109; median age, 70 years [IQR, 65-75 years]) sets at a ratio of 7:3. The primary end point was biochemical recurrence-free survival. The least absolute shrinkage and selection operator Cox algorithm was applied to select independent clinical variables and construct the clinical signature. The radiomics signature and pathomics signature were constructed using preoperative MRI and surgical WSI data, respectively. A multimodality model was constructed by combining the radiomics signature, pathomics signature, and clinical signature. Using Harrell concordance index (C index), the predictive performance of the multimodality model for BCR was assessed and compared with all single-modality models, including the radiomics signature, pathomics signature, and clinical signature. Results Both radiomics and pathomics signatures achieved good performance for BCR prediction (C index: 0.742 and 0.730, respectively) on the testing cohort. The multimodality model exhibited the best predictive performance, with a C index of 0.860 on the testing set, which was significantly higher than all single-modality models (all P ≤ .01). Conclusion The multimodality model effectively predicted BCR following RP in patients with PCa and may therefore provide an emerging and accurate tool to assist postoperative individualized treatment. Keywords: MR Imaging, Urinary, Pelvis, Comparative Studies Supplemental material is available for this article. © RSNA, 2024.
Collapse
Affiliation(s)
| | | | - Renpeng Huang
- From the Departments of Radiology (Chenhan Hu, X.Q., Chunhong Hu,
J.B., X.W.) and Pathology (R.H.), the First Affiliated Hospital of Soochow
University, 188 Shizi Road, Suzhou 215006, China
| | - Chunhong Hu
- From the Departments of Radiology (Chenhan Hu, X.Q., Chunhong Hu,
J.B., X.W.) and Pathology (R.H.), the First Affiliated Hospital of Soochow
University, 188 Shizi Road, Suzhou 215006, China
| | - Jie Bao
- From the Departments of Radiology (Chenhan Hu, X.Q., Chunhong Hu,
J.B., X.W.) and Pathology (R.H.), the First Affiliated Hospital of Soochow
University, 188 Shizi Road, Suzhou 215006, China
| | - Ximing Wang
- From the Departments of Radiology (Chenhan Hu, X.Q., Chunhong Hu,
J.B., X.W.) and Pathology (R.H.), the First Affiliated Hospital of Soochow
University, 188 Shizi Road, Suzhou 215006, China
| |
Collapse
|
2
|
Boyer MJ, Carpenter DJ, Gingrich JR, Raman SR, Sirohi D, Tabriz AA, Rompre-Broduer A, Lunyera J, Basher F, Bitting RL, Kosinski A, Cantrell S, Gordon AM, Ear B, Gierisch JM, Jacobs M, Goldstein KM. Genomic classifiers and prognosis of localized prostate cancer: a systematic review. Prostate Cancer Prostatic Dis 2024:10.1038/s41391-023-00766-z. [PMID: 38200096 DOI: 10.1038/s41391-023-00766-z] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/25/2023] [Revised: 10/26/2023] [Accepted: 11/20/2023] [Indexed: 01/12/2024]
Abstract
BACKGROUND Refinement of the risk classification for localized prostate cancer is warranted to aid in clinical decision making. A systematic analysis was undertaken to evaluate the prognostic ability of three genomic classifiers, Decipher, GPS, and Prolaris, for biochemical recurrence, development of metastases and prostate cancer-specific mortality in patients with localized prostate cancer. METHODS Data sources: MEDLINE, Embase, and Web of Science were queried for reports published from January 2010 to April 2022. STUDY SELECTION prospective or retrospective studies reporting prognosis for patients with localized prostate cancer. DATA EXTRACTION relevant data were extracted into a customized database by one researcher with a second overreading. Risk of bias was assessed using a validated tool for prognostic studies, Quality in Prognosis Studies (QUIPS). Disagreements were resolved by consensus or by input from a third reviewer. We assessed the certainty of evidence by GRADE incorporating adaptation for prognostic studies. RESULTS Data synthesis: a total of 39 studies (37 retrospective) involving over 10,000 patients were identified. Twenty-two assessed Decipher, 5 GPS, and 14 Prolaris. Thirty-four studies included patients who underwent prostatectomy. Based on very low to low certainty of evidence, each of the three genomic classifiers modestly improved upon the prognostic ability for biochemical recurrence, development of metastases, and prostate cancer-specific mortality compared to standard clinical risk-classification schemes. LIMITATIONS downgrading of confidence in the evidence stemmed largely from bias due to the retrospective nature of the studies, heterogeneity in treatment received, and era in which patients were treated (i.e., prior to the 2000s). CONCLUSIONS Genomic classifiers provide a small but consistent improvement upon the prognostic ability of clinical classification schemes, which may be helpful when treatment decisions are uncertain. However, evidence from current management-era data and of the predictive ability of these tests is needed.
Collapse
Affiliation(s)
- Matthew J Boyer
- Durham VA Health Care System, Durham, NC, USA.
- Department of Radiation Oncology, Duke University School of Medicine, Durham, NC, USA.
| | | | - Jeffrey R Gingrich
- Durham VA Health Care System, Durham, NC, USA
- Department of Urology, Duke University School of Medicine, Durham, NC, USA
| | - Sudha R Raman
- Department of Population Health Sciences, Duke University School of Medicine, Durham, NC, USA
| | - Deepika Sirohi
- Department of Pathology, University of Utah School of Medicine, Salt Lake City, UT, USA
| | - Amir Alishahi Tabriz
- Department of Health Outcomes and Behavior, Moffitt Cancer Center, Tampa, FL, USA
| | | | - Joseph Lunyera
- Division of General Internal Medicine, Department of Medicine, Duke University School of Medicine, Durham, NC, USA
| | - Fahmin Basher
- Division of Medical Oncology, Department of Medicine, Duke University School of Medicine, Durham, NC, USA
| | - Rhonda L Bitting
- Durham VA Health Care System, Durham, NC, USA
- Division of Medical Oncology, Department of Medicine, Duke University School of Medicine, Durham, NC, USA
| | - Andrzej Kosinski
- Department of Biostatistics & Bioinformatics, Duke University School of Medicine, Durham, NC, USA
| | - Sarah Cantrell
- Duke University Medical Center Library & Archives, Duke University School of Medicine, Durham, NC, USA
| | | | - Belinda Ear
- Durham VA Health Care System, Durham, NC, USA
| | - Jennifer M Gierisch
- Durham VA Health Care System, Durham, NC, USA
- Division of General Internal Medicine, Department of Medicine, Duke University School of Medicine, Durham, NC, USA
- Department of Population Health, Duke University School of Medicine, Durham, NC, USA
| | | | - Karen M Goldstein
- Durham VA Health Care System, Durham, NC, USA
- Division of General Internal Medicine, Department of Medicine, Duke University School of Medicine, Durham, NC, USA
| |
Collapse
|
3
|
Molecular Characterization of Clear Cell Renal Cell Carcinoma Reveals Prognostic Significance of Epithelial-mesenchymal Transition Gene Expression Signature. Eur Urol Oncol 2021; 5:92-99. [PMID: 34840106 DOI: 10.1016/j.euo.2021.10.007] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/29/2021] [Revised: 10/17/2021] [Accepted: 10/31/2021] [Indexed: 11/23/2022]
Abstract
BACKGROUND There is an ongoing need to develop prognostic biomarkers to improve the management of clear cell renal cell carcinoma (ccRCC). OBJECTIVE To leverage enriched pathways in ccRCC to improve risk-stratification. DESIGN, SETTING, AND PARTICIPANTS We retrospectively identified two complementary discovery cohorts of patients with ccRCC who underwent (1) radical nephrectomy (RNx) with inferior vena cava tumor thrombectomy (patients = 5, samples = 24) and (2) RNx for localized disease and developed recurrence versus no recurrence (n = 36). Patients with localized ccRCC (M0) in The Cancer Genome Atlas (TCGA, n = 386) were used for validation. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS A differential expression gene (DEG) analysis was performed on targeted RNA next-generation sequencing data from both discovery cohorts. Using TCGA for validation, Kaplan-Meier survival analysis and multivariable Cox proportional hazard testing were utilized to investigate the prognostic impact of DEGs, cell cycle proliferation (CCP), and a novel epithelial-mesenchymal transition (EMT) score on progression-free (PFS) and disease-specific (DSS) survival. RESULTS AND LIMITATIONS In the discovery cohorts, we observed overexpression of WT1 and CCP genes in the tumor thrombus versus the primary tumor, as well as in patients with recurrence versus those without recurrence. A hallmark pathway analysis demonstrated enrichment of the EMT- and CCP-related pathways in patients with high WT1 expression in the TCGA (validation) ccRCC cohort. CCP and EMT scores were derived in the validation cohort, which was stratified into four risk groups using Youden Index cut points: CCPlow/EMTlow, CCPlow/EMThigh, CCPhigh/EMTlow, and CCPhigh/EMThigh. The CCPhigh/EMThigh risk group was associated with the worst PFS and DSS (both p < 0.001). In a multivariable analysis, CCPhigh/EMThigh was independently associated with poor PFS and DSS (hazard ratio = 4.6 and 10.3, respectively; p < 0.001). CONCLUSIONS We demonstrate the synergistic prognostic impact of EMT in tumors with a high CCP score. Our novel EMT score has the potential to improve risk stratification and provide potential novel therapeutic targets. PATIENT SUMMARY Genes involved in epithelial-mesenchymal transition provides important prognostic information for patients with clear cell renal cell carcinoma.
Collapse
|