get.boundary: Generate the optimal dose escalation and deescalation...

In BOIN: Bayesian Optimal INterval (BOIN) Design for Single-Agent and Drug- Combination Phase I Clinical Trials

Description

Use this function to generate the optimal dose escalation and deescalation boundaries for conducting the trial.

Usage

get.boundary(
  target,
  ncohort,
  cohortsize,
  n.earlystop = 100,
  p.saf = 0.6 * target,
  p.tox = 1.4 * target,
  cutoff.eli = 0.95,
  extrasafe = FALSE,
  offset = 0.05
)

Arguments

target	the target DLT rate
ncohort	the total number of cohorts
cohortsize	the cohort size
n.earlystop	the early stopping parameter. If the number of patients treated at the current dose reaches n.earlystop, stop the trial and select the MTD based on the observed data. The default value n.earlystop=100 essentially turns off the type of early stopping.
p.saf	the highest toxicity probability that is deemed subtherapeutic (i.e., below the MTD) such that dose escalation should be made. The default value is p.saf = 0.6 * target.
p.tox	the lowest toxicity probability that is deemed overly toxic such that deescalation is required. The default value is p.tox=1.4*target.
cutoff.eli	the cutoff to eliminate an overly toxic dose for safety. We recommend the default value (cutoff.eli=0.95) for general use.
extrasafe	set extrasafe=TRUE to impose a more strict 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 more strict stopping rule. The default value (offset=0.05) generally works well.

Details

The dose escalation and deescalation boundaries are all we need to run a phase I trial when using the BOIN design. The decision of which dose to administer to the next cohort of patients does not require complicated computations, but only a simple comparison of the observed DLT rate at the current dose with the dose escalation and deescalation boundaries. If the observed DLT rate at the current dose is smaller than or equal to the escalation boundary, we escalate the dose; if the observed toxicity rate at the current dose is greater than or equal to the deescalation boundary, we deescalate the dose; otherwise, we retain the current dose. The dose escalation and deescalation boundaries are chosen to minimize the probability of assigning patients to subtherapeutic or overly toxic doses, thereby optimizing patient ethics. get.boundary() also outputs the elimination boundary, which is used to avoid treating patients at overly toxic doses based on the following Bayesian safety rule: if Pr(p_j > φ | m_j , n_j ) > 0.95 and n_j ≥ 3, dose levels j and higher are eliminated from the trial, where p_j is the toxicity probability of dose level j, φ is the target DLT rate, and m_j and n_j are the number of toxicities and patients treated at dose level j. The trial is terminated if the lowest dose is eliminated.

The BOIN design has two built-in stopping rules: (1) stop the trial if the lowest dose is eliminated due to toxicity, and no dose should be selected as the MTD; and (2) stop the trial and select the MTD if the number of patients treated at the current dose reaches n.earlystop. The first stopping rule is a safety rule to protect patients from the case in which all doses are overly toxic. The rationale for the second stopping rule is that when there is a large number (i.e., n.earlystop) of patients assigned to a dose, it means that the dose-finding algorithm has approximately converged. Thus, we can stop the trial early and select the MTD to save the sample size and reduce the trial duration. For some applications, investigators may prefer a more strict safety stopping rule than rule (1) for extra safety when the lowest dose is overly toxic. This can be achieved by setting extrasafe=TRUE, which imposes the following more strict safety stopping rule: stop the trial if (i) the number of patients treated at the lowest dose >=3, and (ii) Pr(toxicity\ rate\ of\ the\ lowest\ dose > \code{target} | data) > \code{cutoff.eli}-\code{offset}. As a tradeoff, the strong stopping rule will decrease the MTD selection percentage when the lowest dose actually is the MTD.

Value

get.boundary() returns a list object, including the dose escalation and de-escalation boundaries $lambda_e and $lambda_d and the corresponding decision tables $boundary_tab and $full_boundary_tab. If extrasafe=TRUE, the function also returns a (more strict) safety stopping boundary $stop_boundary.

Note

We should avoid setting the values of p.saf and p.tox very close to the target. This is because the small sample sizes of typical phase I trials prevent us from differentiating the target DLT rate from the rates close to it. In addition, in most clinical applications, the target DLT rate is often a rough guess, and finding a dose level with a DLT rate reasonably close to the target rate will still be of interest to the investigator. The default values provided by get.boundary() are generally reasonable for most clinical applications.

Author(s)

Suyu Liu and Ying Yuan

References

Liu S. and Yuan, Y. (2015). Bayesian Optimal Interval Designs for Phase I Clinical Trials, Journal of the Royal Statistical Society: Series C, 64, 507-523.

Yan, F., Zhang, L., Zhou, Y., Pan, H., Liu, S. and Yuan, Y. (2020).BOIN: An R Package for Designing Single-Agent and Drug-Combination Dose-Finding Trials Using Bayesian Optimal Interval Designs. Journal of Statistical Software, 94(13),1-32.<doi:10.18637/jss.v094.i13>.

Yuan Y., Hess K.R., Hilsenbeck S.G. and Gilbert M.R. (2016). Bayesian Optimal Interval Design: A Simple and Well-performing Design for Phase I Oncology Trials, Clinical Cancer Research, 22, 4291-4301.

See Also

Tutorial: http://odin.mdacc.tmc.edu/~yyuan/Software/BOIN/BOIN2.6_tutorial.pdf

Paper: http://odin.mdacc.tmc.edu/~yyuan/Software/BOIN/paper.pdf

Examples

## get the dose escalation and deescalation boundaries for BOIN design with
## the target DLT rate of 0.3, maximum sample size of 30, and cohort size of 3
bound <- get.boundary(target=0.3, ncohort=10, cohortsize=3)
summary(bound) # get the descriptive summary of the boundary
plot(bound)    # plot the flowchart of the design with boundaries

BOIN documentation built on Jan. 20, 2021, 1:06 a.m.