get.boundary.kb {Keyboard}

Dose Escalation or De-escalation Boundaries for Single-agent Trials

Description

This function generates the optimal dose escalation or de-escalation boundaries when conducting a single-agent trial with the Keyboard design.

Usage

get.boundary.kb(
  target,
  ncohort,
  cohortsize,
  marginL = 0.05,
  marginR = 0.05,
  cutoff.eli = 0.95,
  n.earlystop = 100,
  extrasafe = FALSE,
  offset = 0.05
)

Arguments

target

The target dose-limiting toxicity (DLT) rate.

ncohort

The total number of cohorts.

cohortsize

The number of patients in the cohort.

marginL

The difference between the target and the lower bound of the "target key" (proper dosing interval) to be defined.
The default is 0.05.

marginR

The difference between the target and the upper bound of the "target key" (proper dosing interval) to be defined.
The default is 0.05.

cutoff.eli

The cutoff value to eliminate an overly toxic dose and all higher doses for safety.
The recommended value is 0.95.

n.earlystop

The early stopping parameter. If the number of patients treated at the current dose reaches n.earlystop, then stop the trial and select the MTD based on the observed data. The default value is 100.

extrasafe

Set extrasafe=TRUE to impose a stricter stopping rule for extra safety, expressed as the stopping boundary value in the result.

offset

A small positive number (between 0 and 0.5) to control how strict the stopping rule is when extrasafe=TRUE. A larger value leads to a stricter stopping rule. The default value is 0.05.

Details

The Keyboard design relies on the posterior distribution of the toxicity probability to guide dosage. To make a decision of dose escalation or de-escalation, given the observed data at the current dose, we first identify an interval that has the highest posterior probability, referred to as the "strongest key". This key represents where the true dose-limiting toxicity (DLT) rate of the current dose is most likely located. If the strongest key is to the left of the "target key", then we escalate the dose because the data suggest that the current dose is most likely too low; if the strongest key is to the right of the target key, then we de-escalate the dose because the observed data suggest that the current dose is likely too high; and if the strongest key is the target key, then we retain the current dose because the observed data support the notion that the current dose is most likely to be in the proper dosing interval. Graphically, the strongest key is the one with the largest area under the posterior distribution curve of the DLT rate of the current dose.

An attractive feature of the Keyboard design is that its dose escalation and de-escalation rules can be tabulated before the onset of the trial. Thus, when conducting the trial, no calculation or model fitting is needed, and we need to count only the number of DLTs observed at the current dose; the decision to escalate or de-escalate the dose is based on the pre-tabulated decision rules.

Given all observed data, the Keyboard design uses isotonic regression to obtain an efficient statistical estimate of the maximum tolerated dose (MTD) by utilizing the fact that toxicity presumably increases with the dose.

For patient safety, we apply the following Bayesian overdose control rule after each cohort: if at least 3 patients have been treated at the given dose and the observed data indicate that the probability of the current dose's toxicity rate being above the target toxicity rate is more than 95%, then we stop the trial to avoid exposing future patients to these overly toxic doses. The probability threshold can be specified with cutoff.eli. When a dose is eliminated, the design recommends the next lower dose for treating the next cohort of patients. If the lowest dose is overly toxic, then the trial terminates early and no dose is selected as the MTD.

Value

The function returns a matrix, which includes the dose escalation, de-escalation, and elimination boundaries.

Note

In most clinical applications, the target DLT rate is often a rough guess, but finding a dose level with a DLT rate reasonably close to the target rate (which ideally would be the MTD) is of interest.

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

See Also

Other single-agent functions: get.oc.kb(), select.mtd.kb()

Examples

### Single-agent trial ###

bound <- get.boundary.kb(target=0.3, ncohort=10, cohortsize=3)
print(bound)