Development and External Validation of a Machine Learning Model for Prediction of Lymph Node Metastasis in Patients with Prostate Cancer.

Lymph node metastasis Machine leaning Prostate cancer

Journal

European urology oncology
ISSN: 2588-9311
Titre abrégé: Eur Urol Oncol
Pays: Netherlands
ID NLM: 101724904

Informations de publication

Date de publication:
Oct 2023
Historique:
received: 13 09 2022
revised: 10 01 2023
accepted: 03 02 2023
pubmed: 4 3 2023
medline: 4 3 2023
entrez: 3 3 2023
Statut: ppublish

Résumé

Pelvic lymph node dissection (PLND) is the gold standard for diagnosis of lymph node involvement (LNI) in patients with prostate cancer. The Roach formula, Memorial Sloan Kettering Cancer Center (MSKCC) calculator, and Briganti 2012 nomogram are elegant and simple traditional tools used to estimate the risk of LNI and select patients for PLND. To determine whether machine learning (ML) can improve patient selection and outperform currently available tools for predicting LNI using similar readily available clinicopathologic variables. Retrospective data for patients treated with surgery and PLND between 1990 and 2020 in two academic institutions were used. We trained three models (two logistic regression models and one gradient-boosted trees-based model [XGBoost]) on data provided from one institution (n = 20267) with age, prostate-specific antigen (PSA) levels, clinical T stage, percentage positive cores, and Gleason scores as inputs. We externally validated these models using data from another institution (n = 1322) and compared their performance to that of the traditional models using the area under the receiver operating characteristic curve (AUC), calibration, and decision curve analysis (DCA). LNI was present in 2563 patients (11.9%) overall, and in 119 patients (9%) in the validation data set. XGBoost had the best performance among all the models. On external validation, its AUC outperformed that of the Roach formula by 0.08 (95% confidence interval [CI] 0.042-0.12), the MSKCC nomogram by 0.05 (95% CI 0.016-0.070), and the Briganti nomogram by 0.03 (95% CI 0.0092-0.051; all p < 0.05). It also had better calibration and clinical utility in terms of net benefit on DCA across relevant clinical thresholds. The main limitation of the study is its retrospective design. Taking all measures of performance together, ML using standard clinicopathologic variables outperforms traditional tools in predicting LNI. Determining the risk of cancer spread to the lymph nodes in patients with prostate cancer allows surgeons to perform lymph node dissection only in patients who need it and avoid the side effects of the procedure in those who do not. In this study, we used machine learning to develop a new calculator to predict the risk of lymph node involvement that outperformed traditional tools currently used by oncologists.

Sections du résumé

BACKGROUND BACKGROUND
Pelvic lymph node dissection (PLND) is the gold standard for diagnosis of lymph node involvement (LNI) in patients with prostate cancer. The Roach formula, Memorial Sloan Kettering Cancer Center (MSKCC) calculator, and Briganti 2012 nomogram are elegant and simple traditional tools used to estimate the risk of LNI and select patients for PLND.
OBJECTIVE OBJECTIVE
To determine whether machine learning (ML) can improve patient selection and outperform currently available tools for predicting LNI using similar readily available clinicopathologic variables.
DESIGN, SETTING, AND PARTICIPANTS METHODS
Retrospective data for patients treated with surgery and PLND between 1990 and 2020 in two academic institutions were used.
OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS METHODS
We trained three models (two logistic regression models and one gradient-boosted trees-based model [XGBoost]) on data provided from one institution (n = 20267) with age, prostate-specific antigen (PSA) levels, clinical T stage, percentage positive cores, and Gleason scores as inputs. We externally validated these models using data from another institution (n = 1322) and compared their performance to that of the traditional models using the area under the receiver operating characteristic curve (AUC), calibration, and decision curve analysis (DCA).
RESULTS AND LIMITATIONS CONCLUSIONS
LNI was present in 2563 patients (11.9%) overall, and in 119 patients (9%) in the validation data set. XGBoost had the best performance among all the models. On external validation, its AUC outperformed that of the Roach formula by 0.08 (95% confidence interval [CI] 0.042-0.12), the MSKCC nomogram by 0.05 (95% CI 0.016-0.070), and the Briganti nomogram by 0.03 (95% CI 0.0092-0.051; all p < 0.05). It also had better calibration and clinical utility in terms of net benefit on DCA across relevant clinical thresholds. The main limitation of the study is its retrospective design.
CONCLUSIONS CONCLUSIONS
Taking all measures of performance together, ML using standard clinicopathologic variables outperforms traditional tools in predicting LNI.
PATIENT SUMMARY RESULTS
Determining the risk of cancer spread to the lymph nodes in patients with prostate cancer allows surgeons to perform lymph node dissection only in patients who need it and avoid the side effects of the procedure in those who do not. In this study, we used machine learning to develop a new calculator to predict the risk of lymph node involvement that outperformed traditional tools currently used by oncologists.

Identifiants

pubmed: 36868922
pii: S2588-9311(23)00038-X
doi: 10.1016/j.euo.2023.02.006
pii:
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Pagination

501-507

Informations de copyright

Copyright © 2023 European Association of Urology. Published by Elsevier B.V. All rights reserved.

Auteurs

Ali Sabbagh (A)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA.

Samuel L Washington (SL)

Department of Urology, University of California-San Francisco, San Francisco, CA, USA; Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA.

Derya Tilki (D)

Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany; Department of Urology, University Hospital Hamburg-Eppendorf, Hamburg, Germany.

Julian C Hong (JC)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA.

Jean Feng (J)

Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA.

Gilmer Valdes (G)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA.

Ming-Hui Chen (MH)

Department of Statistics, University of Connecticut, Storrs, CT, USA.

Jing Wu (J)

Department of Computer Science and Statistics, University of Rhode Island, Kingston, RI, USA.

Hartwig Huland (H)

Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany.

Markus Graefen (M)

Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany.

Thomas Wiegel (T)

Department of Radio Oncology, University Hospital Ulm, Ulm, Germany.

Dirk Böhmer (D)

Department of Radiation Oncology, Charité University Hospital, Berlin, Germany.

Janet E Cowan (JE)

Department of Urology, University of California-San Francisco, San Francisco, CA, USA.

Matthew Cooperberg (M)

Department of Urology, University of California-San Francisco, San Francisco, CA, USA; Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA.

Felix Y Feng (FY)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA; Department of Urology, University of California-San Francisco, San Francisco, CA, USA.

Mack Roach (M)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA.

Bruce J Trock (BJ)

Division of Epidemiology, Brady Urological Institute, Johns Hopkins Medical Institution, Baltimore, MD, USA.

Alan W Partin (AW)

Department of Urology, Brady Urological Institute, Johns Hopkins Medical Institution, Baltimore, MD, USA.

Anthony V D'Amico (AV)

Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, MA, USA.

Peter R Carroll (PR)

Department of Urology, University of California-San Francisco, San Francisco, CA, USA.

Osama Mohamad (O)

Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA; Department of Urology, University of California-San Francisco, San Francisco, CA, USA. Electronic address: osama.mohamad@ucsf.edu.

Classifications MeSH