The 'decider' package: DECIsion making in oncology Dose Escalation trials with logistic Regression
decider-package.RdPerforms Bayesian logistic regression for binary dose-toxicity
data from monotherapy and/or (two-component) combination therapy dose-finding trials.
The underlying model is referred to as Bayesian logistic regression model (BLRM)
and specified according to Neuenschwander et al. (2008, 2014, 2016).
All Bayesian models were implemented in the
in the Stan modeling language (Stan Development Team (2020))
using the rstan-package
and the rstantools-package.
Currently, only the so-called joint BLRM is included in the package. Different
methods for the derivation of dosing recommendations are supported, among others the
escalation with overdose control (EWOC) criterion that goes back to Babb et al. (1998).
The main functions are sim_jointBLRM() and
scenario_jointBLRM().
Refer to their documentation entries for a detailed description of the underlying methods.
The methods implemented by this package were mainly developed in the context of oncology dose-finding trials, but can also be applied to different therapeutic areas provided that the methodology of estimating a monotonically increasing dose-toxicity relationship based on binary safety data applies. For the documentation and the argument names used by the package, the typical terminology of oncology trials is used. That is, the dose to be determined by the trials is referred to as the maximum tolerated dose (MTD), which is determined based on binary safety data given by so-called dose-limiting toxicities (DLT). These DLTs are a pre-specified set of adverse events that are considered to be dose limiting. In this context, the goal of a trial is to determine the MTD so that it has a true DLT rate in some pre-specified dosing interval. Note that other therapeutic areas often use slightly different notions instead.
References
Stan Development Team (2020). RStan: the R interface to Stan. R package version 2.21.2. https://mc-stan.org.
Neuenschwander, B., Branson, M., & Gsponer, T. (2008). Critical aspects of the Bayesian approach to phase I cancer trials. Statistics in medicine, 27(13), 2420-2439, doi:10.1002/sim.3230.
Schmidli, H., Gsteiger, S., Roychoudhury, S., O'Hagan, A., Spiegelhalter, D., & Neuenschwander B. (2014). Robust meta-analytic-predictive priors in clinical trials with historical control information.
Neuenschwander, B., Matano, A., Tang, Z., Roychoudhury, S., Wandel, S., & Bailey, S. (2014). A Bayesian Industry Approach to Phase I Combination Trials in Oncology. In: Zhao. W & Yang, H. (editors). Statistical methods in drug combination studies. Chapman and Hall/CRC, 95-135, doi:10.1201/b17965.
Neuenschwander, B., Roychoudhury, S., & Schmidli, H. (2016). On the use of co-data in clinical trials. Statistics in Biopharmaceutical Research, 8(3), 345-354, doi:10.1080/19466315.2016.1174149.
Babb, J., Rogatko, A., & Zacks, S. (1998). Cancer phase I clinical trials: Efficient dose escalation with overdose control. Statistics in medicine 17(10), 1103-1120.
Zhou, H., Yuan, Y., & Nie, L. (2018). Accuracy, safety, and reliability of novel phase I designs. Clinical Cancer Research, 24(21), 5483-5484 <doi: 10.1158/1078-0432.ccr-18-0168>.