Keyboard (version 0.1.3)

example.kb: Illustration of the Keyboard Design (Single-agent)

Description

This function serves solely as an example to illustrate how the Keyboard design determines dosing decisions when conducting a single-agent trial.

Usage

example.kb(center, half_width, s1, s2, a, b)

Value

A dose decision plot with the posterior distribution of the DLT rate of the current dose and the positions of the target and the strongest keys.

Arguments

center

The center of the target key (between 0 and 1).

half_width

The width, or tolerable deviation, to each side of the center (half of the target key's width).

s1

The lower boundary of the strongest key.

s2

The upper boundary of the strongest key.

a

The alpha parameter of the beta distribution.

b

The beta parameter of the beta distribution.

Author

Xiaomeng Yuan, Chen Li, Hongying Sun, Li Tang and Haitao Pan

Details

It is straightforward to visualize how the Keyboard design determines whether to stay at the current dose, escalate, or de-escalate. The "target key" (shown in blue) is the proper dosing interval determined for the trial by the investigators. The "strongest key" (in red) is the interval with the highest posterior probability (given the dosage data observed thus far) of the current dose's true DLT rate. If the strongest key is to the left of the target key, then we escalate the dose; if the strongest key is to the right of the target key, then we de-escalate the dose; and if the strongest key is the target key, then we retain the current dose. Thus, posterior probabilities are what directs dosage. Graphically, the strongest key is the one with the largest area under the posterior distribution curve of the DLT rate of the current dose.

References

Yan F, Mandrekar SJ, Yuan Y. Keyboard: A Novel Bayesian Toxicity Probability Interval Design for Phase I Clinical Trials. Clinical Cancer Research. 2017; 23:3994-4003. http://clincancerres.aacrjournals.org/content/23/15/3994.full-text.pdf

Pan H, Lin R, Yuan Y. Keyboard design for phase I drug-combination trials. Contemporary Clinical Trials. 2020. https://doi.org/10.1016/j.cct.2020.105972

Examples

## Clear all plots before switching between graphical parameters
opar <- par(no.readonly = TRUE)
on.exit(par(opar))
par(mfrow = c(3, 1)) # for many plots in the same screen
par(mar = c(5, 5, 2, 2)) # for only one plot per page

example.kb(center = 0.19, half_width = 0.03, s1 = 0.4, s2 = 0.46, a = 3, b = 4) # de-escalation
example.kb(center = 0.19, half_width = 0.03, s1 = 0.04, s2 = 0.1, a = 2, b = 12) # escalation
example.kb(center = 0.19, half_width = 0.03, s1 = 0.16, s2 = 0.22, a = 2, b = 5) # stay