Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Machine-Learning and Stochastic Tumor Growth Models for Predicting Outcomes in Patients With Advanced Non-Small-Cell Lung Cancer

Machine-Learning and Stochastic Tumor Growth Models for Predicting Outcomes in Patients With... PURPOSE: The prediction of clinical outcomes for patients with cancer is central to precision medicine and the design of clinical trials. We developed and validated machine-learning models for three important clinical end points in patients with advanced non-small-cell lung cancer (NSCLC)-bjective response (OR), progression-free survival (PFS), and overall survival (OS)-using routinely collected patient and disease variables. METHODS: We aggregated patient-level data from 17 randomized clinical trials recently submitted to the US Food and Drug Administration evaluating molecularly targeted therapy and immunotherapy in patients with advanced NSCLC. To our knowledge, this is one of the largest studies of NSCLC to consider biomarker and inhibitor therapy as candidate predictive variables. We developed a stochastic tumor growth model to predict tumor response and explored the performance of a range of machine-learning algorithms and survival models. Models were evaluated on out-of-sample data using the standard area under the receiver operating characteristic curve and concordance index (C-index) performance metrics. RESULTS: Our models achieved promising out-of-sample predictive performances of 0.79 area under the receiver operating characteristic curve (95% CI, 0.77 to 0.81), 0.67 C-index (95% CI, 0.66 to 0.69), and 0.73 C-index (95% CI, 0.72 to 0.74) for OR, PFS, and OS, respectively. The calibration plots for PFS and OS suggested good agreement between actual and predicted survival probabilities. In addition, the Kaplan-Meier survival curves showed that the difference in survival between the low- and high-risk groups was significant (log-rank test P < .001) for both PFS and OS. CONCLUSION: Biomarker status was the strongest predictor of OR, PFS, and OS in patients with advanced NSCLC treated with immune checkpoint inhibitors and targeted therapies. However, single biomarkers have limited predictive value, especially for programmed death-ligand 1 immunotherapy. To advance beyond the results achieved in this study, more comprehensive data on composite multiomic signatures is required. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JCO Clinical Cancer Informatics Wolters Kluwer Health

Machine-Learning and Stochastic Tumor Growth Models for Predicting Outcomes in Patients With Advanced Non-Small-Cell Lung Cancer

Download PDF
11 pages

Loading next page...
 
/lp/wolters-kluwer-health/machine-learning-and-stochastic-tumor-growth-models-for-predicting-0RiCReK3PP
Publisher
Wolters Kluwer Health
Copyright
(C) 2019 by Lippincott Williams & Wilkins, Inc.
ISSN
2473-4276
DOI
10.1200/CCI.19.00046
Publisher site
See Article on Publisher Site

Abstract

PURPOSE: The prediction of clinical outcomes for patients with cancer is central to precision medicine and the design of clinical trials. We developed and validated machine-learning models for three important clinical end points in patients with advanced non-small-cell lung cancer (NSCLC)-bjective response (OR), progression-free survival (PFS), and overall survival (OS)-using routinely collected patient and disease variables. METHODS: We aggregated patient-level data from 17 randomized clinical trials recently submitted to the US Food and Drug Administration evaluating molecularly targeted therapy and immunotherapy in patients with advanced NSCLC. To our knowledge, this is one of the largest studies of NSCLC to consider biomarker and inhibitor therapy as candidate predictive variables. We developed a stochastic tumor growth model to predict tumor response and explored the performance of a range of machine-learning algorithms and survival models. Models were evaluated on out-of-sample data using the standard area under the receiver operating characteristic curve and concordance index (C-index) performance metrics. RESULTS: Our models achieved promising out-of-sample predictive performances of 0.79 area under the receiver operating characteristic curve (95% CI, 0.77 to 0.81), 0.67 C-index (95% CI, 0.66 to 0.69), and 0.73 C-index (95% CI, 0.72 to 0.74) for OR, PFS, and OS, respectively. The calibration plots for PFS and OS suggested good agreement between actual and predicted survival probabilities. In addition, the Kaplan-Meier survival curves showed that the difference in survival between the low- and high-risk groups was significant (log-rank test P < .001) for both PFS and OS. CONCLUSION: Biomarker status was the strongest predictor of OR, PFS, and OS in patients with advanced NSCLC treated with immune checkpoint inhibitors and targeted therapies. However, single biomarkers have limited predictive value, especially for programmed death-ligand 1 immunotherapy. To advance beyond the results achieved in this study, more comprehensive data on composite multiomic signatures is required.

Journal

JCO Clinical Cancer Informatics – Wolters Kluwer Health

Published: Sep 20, 2019

Keywords: CD274, ALK, EGFR, MTUS1, SLC17A5, KRAS, TENM1

References

  • An FDA perspective on the regulatory implications of complex signatures to predict response to targeted therapies
    Beaver JA; Tzou A; Blumenthal GM
  • Trends in stage distribution for patients with non-small cell lung cancer: A National Cancer Database survey
    Morgensztern D; Ng SH; Gao F
  • Modeling the development of metastases from primary and locally recurrent tumors: Comparison with a clinical data base for prostatic cancer
    Yorke ED; Fuks Z; Norton L
  • Stochastic models of progression of cancer and their use in controlling cancer-related mortality
    Kimmel M; Gorlova O
  • Simulation-based parameter estimation for complex models: A breast cancer natural history modelling illustration
    Chia YL; Salzman P; Plevritis SK
  • Estimating tumor growth rates in vivo
    Talkington A; Durrett R
  • An introduction to ROC analysis
    Fawcett T
  • Regression models and life-tables
    Cox DR
  • Random survival forests
    Ishwaran H; Kogalur UB; Blackstone EH
  • A neural network model for survival data
    Faraggi D; Simon R
  • Evaluating the yield of medical tests
    Harrell FE Jr; Califf RM; Pryor DB
  • Lung cancer prognostic index: A risk score to predict overall survival after the diagnosis of non-small-cell lung cancer
    Alexander M; Wolfe R; Ball D
  • Combining clinical, pathological, and demographic factors refines prognosis of lung cancer: A population-based study
    Putila J; Remick SC; Guo NL
  • A prognostic model to predict mortality among non-small-cell lung cancer patients in the U
    Lin J; Carter CA; McGlynn KA
  • 4-year mortality in patients with non-small-cell lung cancer: Development and validation of a prognostic index
    Blanchon F; Grivaux M; Asselain B
  • Treatment of advanced non-small-cell lung cancer patients with ECOG performance status 2: Results of an European Experts Panel
    Gridelli C; Ardizzoni A; Le Chevalier T
  • Older age itself rarely a restriction to treatment of lung cancer
    Langer CJ
  • A prognostic model for advanced stage nonsmall cell lung cancer
    Mandrekar SJ; Schild SE; Hillman SL
  • Safety and efficacy of anti-PD-1 in patients with baseline cardiac, renal, or hepatic dysfunction
    Kanz BA; Pollack MH; Johnpulle R
  • Anti-PD-1 in patients with advanced malignancies and baseline organ dysfunction
    Kanz BA; Pollack MLH; Eroglu Z
  • Prognostic models to predict survival in non-small-cell lung cancer patients treated with first-line paclitaxel and carboplatin with or without bevacizumab
    Hoang T; Dahlberg SE; Sandler AB
  • Treatment of advanced squamous cell carcinoma of the lung: A review
    Derman BA; Mileham KF; Bonomi PD
  • Yarchoan M, Hopkins A, Jaffee EM
    Yarchoan M; Hopkins A; Jaffee EM
  • Real-world data for clinical evidence generation in oncology
    Khozin S; Blumenthal GM; Pazdur R
  • A radiomics approach to assess tumour-infiltrating CD8 cells and response to anti-PD-1 or anti-PD-L1 immunotherapy: An imaging biomarker, retrospective multicohort study
    Sun R; Limkin EJ; Vakalopoulou M
  • Genomics of non-small cell lung cancer (NSCLC): Association between CT-based imaging features and EGFR and K-RAS mutations in 122 patients-An external validation
    Rizzo S; Raimondi S; de Jong EEC
  • Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach
    Aerts HJWL; Velazquez ER; Leijenaar RTH
  • Deep learning for lung cancer prognostication: A retrospective multi-cohort radiomics study
    Hosny A; Parmar C; Coroller TP
  • Combination of peri- and intratumoral radiomic features on baseline CT scans predicts response to chemotherapy in lung adenocarcinoma
    Khorrami M; Khunger M; Zagouras A
  • Deep learning predicts lung cancer treatment response from serial medical imaging
    Xu Y; Hosny A; Zeleznik R
  • Radiomics: The bridge between medical imaging and personalized medicine
    Lambin P; Leijenaar RTH; Deist TM
  • Integrative approach in the era of failing drug discovery and development
    Earm K; Earm YE
  • Scikit-learn: Machine learning in Python
    Pedregosa F; Varoquaux G; Gramfort A
  • A Package for Survival Analysis in S
    Therneau TM