A Bayesian method to estimate the optimal threshold of a marker used to select patients' treatment

Abstract

The use of a quantitative treatment selection marker to choose between two treatment options requires the estimate of an optimal threshold above which one of these two treatments is preferred. Herein, the optimal threshold expression is based on the definition of a utility function which aims to quantify the expected utility of the population (e.g. life expectancy, quality of life) by taking into account both efficacy (success or failure) and toxicity of each treatment option. Therefore, the optimal threshold is the marker value that maximizes the expected utility of the population. A method modelling the marker distribution in patient subgroups defined by the received treatment and the outcome is proposed to calculate the parameters of the utility function so as to estimate the optimal threshold and its 95% credible interval using the Bayesian inference. The simulation study found that the method had low bias and coverage probability close to 95% in multiple settings, but also the need of large sample size to estimate the optimal threshold in some settings. The method is then applied to the PETACC-8 trial that compares the efficacy of chemotherapy with a combined chemotherapy + anti-epidermal growth factor receptor in stage III colorectal cancer.

Keywords

Biomarker threshold expected utility treatment selection Bayesian predictive marker

Get full access to this article

View all access options for this article.

References

Ballman

. Biomarker: predictive or prognostic?. J Clin Oncol 2015; 33: 3968–3971.

Italiano

. Prognostic or predictive? It’s time to get back to definitions!. J Clin Oncol 2011; 29: 4718.

Lièvre

Bachet

Boige

, et al. KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol 2008; 26: 374–379.

Byar

. Assessing apparent treatment-covariate interactions in randomized clinical trials. Stat Med 1985; 4: 255–263.

Vickers

Kattan

Sargent

. Method for evaluating prediction models that apply the results of randomized trials to individual patients. Trials 2007; 8: 14, .

Janes

Pepe

Bossuyt

, et al. Measuring the performance of markers for guiding treatment decisions. Ann Intern Med 2011; 154: 253–259.

Janes

Brown

Huang

, et al. An approach to evaluating and comparing biomarkers for patient treatment selection. Int J Biostat 2014; 10: 99–121.

Huang

Gilbert

Janes

. Assessing treatment-selection markers using a potential outcomes framework. Biometrics 2012; 68: 687–696.

Zhang

Nie

Soon

, et al. The use of covariates and random effects in evaluating predictive biomarkers under a potential outcome framework. Ann Appl Stat 2014; 8: 2336–2355.

10.

Sox

Higgins

Owens

. Medical decision making, 2nd ed. Chichester: John Wiley & Sons, 2013.

11.

Janes

Pepe

Huang

. A framework for evaluating markers used to select patient treatment. Med Decis Making 2014; 34: 159–167.

12.

Collins

Moons

. Comparing risk prediction models. BMJ 2012; 344: e3186.

13.

Van Calster

Vickers

. Calibration of risk prediction models: impact on decision-analytic performance. Med Decis Making 2015; 35: 162–169.

14.

Cullen

Frey

. Probabilistic techniques in exposure assessment, New York: Plenum Press, 1999.

15.

Forbes

Evans

Hastings

, et al. Statistical distributions, 4th ed. Hoboken: John Wiley & Sons, 2000.

16.

Delignette-Muller

Dutang

. fitdistrplus: an R package for fitting distributions. J Stat Softw 2015; 64: 1–34.

17.

Vickers

Elkin

. Decision curve analysis: a novel method for evaluating prediction models. Med Decis Making 2006; 26: 565–74.

18.

Huang

Laber

Janes

. Characterizing expected benefits of biomarkers in treatment selection. Biostatistics 2015; 16: 383–399.

19.

Baker

Cook

Vickers

, et al. Using relative utility curves to evaluate risk prediction. J R Stat Soc Ser A Stat Soc 2009; 172: 729–748.

20.

Efron

Tibshirani

. Bootstrap methods for standard errors, confidence intervals, and other measures of statistical accuracy. Stat Sci 1986; 1: 54–75.

21.

Schisterman

Perkins

. Confidence intervals of the Youden index and corresponding optimal cut-point. Commun Stat Simul Comput 2007; 36: 549–563.

22.

Subtil

Rabilloud

. A Bayesian method to estimate the optimal threshold of a longitudinal biomarker. Biom J 2010; 52: 333–347.

23.

Subtil

Rabilloud

. Estimating the optimal threshold for a diagnostic biomarker in case of complex biomarker distributions. BMC Med Inform Decis Making 2014; 14: 53.

24.

Gelman

Carlin

Stern

, et al. Bayesian data analysis, 3rd ed. Boca Raton: CRC Press, 2014.

25.

R Core Team. R: a language and environment for statistical computing, Vienna, Austria: R Foundation for Statistical Computing, 2017.

26.

Taieb

Tabernero

Mini

, et al. Oxaliplatin, fluorouracil, and leucovorin with or without cetuximab in patients with resected stage III colon cancer (PETACC-8): an open-label, randomised phase 3 trial. Lancet Oncol 2014; 15: 862–873.

27.

Dmitrienko

Molenberghs

Chuang-Stein

, et al. Analysis of clinical trials using SAS: a practical guide, Cary: SAS Institute, 2005.

28.

Janes

Brown

Pepe

. Designing a study to evaluate the benefit of a biomarker for selecting patient treatment. Stat Med 2015; 34: 3503–3515.

29.

Escobar

. Estimating normal means with a Dirichlet process prior. J Am Stat Assoc 1994; 89: 268–277.

30.

Ohlssen

Sharples

Spiegelhalter

. Flexible random-effects models using Bayesian semi-parametric models: applications to institutional comparisons. Stat Med 2007; 26: 2088–2112.

31.

Blanche

Dartigues

Jacqmin-Gadda

. Review and comparison of ROC curve estimators for a time-dependent outcome with marker-dependent censoring. Biom J 2013; 55: 687–704.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.19 MB