Machine learning models for predicting post-cystectomy recurrence and survival in bladder cancer patients

Zaki Hasnain; Jeremy Mason; Karanvir Gill; Gus Miranda; Inderbir S Gill; Peter Kuhn; Paul K Newton

doi:10.1371/journal.pone.0210976

Machine learning models for predicting post-cystectomy recurrence and survival in bladder cancer patients

PLoS One. 2019 Feb 20;14(2):e0210976. doi: 10.1371/journal.pone.0210976. eCollection 2019.

Authors

Zaki Hasnain¹, Jeremy Mason², Karanvir Gill², Gus Miranda³, Inderbir S Gill³, Peter Kuhn^{1

2

3

4

5}, Paul K Newton^{1

5

6}

Affiliations

¹ Department of Aerospace and Mechanical Engineering, Viterbi School of Engineering, University of Southern California, University Park Campus, Los Angeles, CA, United States of America.
² Department of Biological Sciences, Dornsife College of Letters, Arts, and Sciences, University of Southern California, University Park Campus, Los Angeles, CA, United States of America.
³ USC Institute of Urology, Keck School of Medicine, University of Southern California, Health Sciences Campus, Los Angeles, CA, United States of America.
⁴ Department of Biomedical Engineering, Viterbi School of Engineering, University of Southern California, University Park Campus, Los Angeles, CA, United States of America.
⁵ Norris Comprehensive Cancer Center, Keck School of Medicine, University of Southern California, Health Sciences Campus, Los Angeles, CA, United States of America.
⁶ Department of Mathematics, University of Southern California, University Park Campus, Los Angeles, CA, United States of America.

Abstract

Currently in patients with bladder cancer, various clinical evaluations (imaging, operative findings at transurethral resection and radical cystectomy, pathology) are collectively used to determine disease status and prognosis, and recommend neoadjuvant, definitive and adjuvant treatments. We analyze the predictive power of these measurements in forecasting two key long-term outcomes following radical cystectomy, i.e., cancer recurrence and survival. Information theory and machine learning algorithms are employed to create predictive models using a large prospective, continuously collected, temporally resolved, primary bladder cancer dataset comprised of 3503 patients (1971-2016). Patient recurrence and survival one, three, and five years after cystectomy can be predicted with greater than 70% sensitivity and specificity. Such predictions may inform patient monitoring schedules and post-cystectomy treatments. The machine learning models provide a benchmark for predicting oncologic outcomes in patients undergoing radical cystectomy and highlight opportunities for improving care using optimal preoperative and operative data collection.

Publication types

Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't

MeSH terms

Aged
Aged, 80 and over
Cystectomy*
Databases, Factual*
Disease-Free Survival
Female
Humans
Machine Learning*
Male
Middle Aged
Models, Biological*
Survival Rate
Urinary Bladder Neoplasms* / mortality
Urinary Bladder Neoplasms* / surgery

Grants and funding

This work was supported by Federal funds from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, under Contract No. HHSN261201500003B; Seven Bridges Genomics, Inc.; as well as institutional support from the University of Southern California Institute of Urology and the University of Southern California Michelson Center for Convergent Biosciences.