SEARS {SEARS}

A randomized distributed phase I-II seamless dose escalation/expansion schema for dose optimization and selection in early oncology clinical development

Description

Implements seamless randomized phase I-II SEARS design for finding the optimal design. Practitioners can use a rich set of parameters to explore various real scenarios for their studies. The function can generate operating characteristics via simulation for practitioners to examine the design’s properties

Usage

SEARS(p.p, p.d, p.tox, k1, k2, pi_t, pi_e, pT, eff_a = 0.5, eff_b = 0.5,
      plac_a = 0.5, plac_b = 0.5, tox_a = 1, tox_b = 1, csize, csize2,
      p.star = 0.2, q.star = 0.6, f.star = 0.06, p.star2 = 0.2,
      q.star2 = 0.98, d.cs, p.cs, phase1_size, n_earlystop,
      extrasafe_BOIN = FALSE, offset_BOIN = 0.05, Nsim, n_catchup,
      control_arm = "", power_c = 0.5, lower_bound = 0.05, weight1,
      weight2, seed = 100)

Arguments

p.p

the true placebo response rate

p.d

the true dose response rate vector

p.tox

the true dose toxicity rate vector

k1

the safety rule cutoff value in phase I

k2

the safety rule cutoff value in phase II

pi_t

the physician-specified upper toxicity rate threshold

pi_e

the physician-specified lower response rate threshold

pT

the target toxicity rate

eff_a

the hyperparameter "a" for priors of the response rate for experimental doses. The default value is eff_a = 0.5

eff_b

the hyperparameter "b" for priors of the response rate for experimental doses. The default value is eff_b = 0.5

plac_a

the hyperparameter "a" for prior of the response rate for the control arm. The default value is plac_a = 0.5

plac_b

the hyperparameter "b" for prior of the response rate for the control arm. The default value is plac_b = 0.5

tox_a

the hyperparameter "a" for priors of the toxicity rates. The default value is tox_a = 1

tox_b

the hyperparameter "b" for priors of the toxicity rates. The default value is tox_b = 1

csize

the cohort size in phase I

csize2

the cohort size in phase II

p.star

the fixed cutoff probability for toxicity in phase I. The default value is p.star = 0.2

q.star

the fixed cutoff probability for efficacy in phase I. The default value is q.star = 0.6

f.star

the small probability cutoff. It will be used for futility dose exclusion. The default value is f.star = 0.06

p.star2

the fixed cutoff probability for toxicity in phase II. The default value is p.star2 = 0.2

q.star2

the fixed cutoff probability for efficacy in phase II. The default value is q.star2 = 0.98

d.cs

the prespecified maximum allowable number of patients enrolled for each dose

p.cs

the prespecified maximum allowable number of patients enrolled for the placebo

phase1_size

the prespecified maximum sample size of phase I trial

n_earlystop

the cutoff number in phase I. When the number of patients enrolled at a certain dose reaches this value in phase I, this dose will be graduated to phase II

extrasafe_BOIN

the logical value which indicates whether a more stringent stopping rule will be applied to phase I BOIN design. The default value is extrasafe_BOIN = FALSE

offset_BOIN

the small positive number (between 0 and 0.5) to control how strict the stopping rule is when extrasafe_BOIN = TRUE. The default value is offset_BOIN = 0.05

Nsim

the number of simulated trials

n_catchup

the catch-up cutoff when employing the adaptive randomization in phase II

control_arm

the argument is for phase II design only. If this argument is "fixed", then allocation probability of control arm (the first component of the allocation probability vector) will be fixed to \frac{1}{K}. The default of this argument will return unfixed results; K indicates total number of arms (including control arm)

power_c

the power correction parameter of the allocation probability. The default value is power_c = 0.5

lower_bound

the lower bound of the allocation probability in phase II design. It must be a value between 0 and \frac{1}{K}. The default value is lower_bound = 0.05; K indicates total number of arms (including the control arm)

weight1

the penalized weight in the utility function for the toxicity

weight2

the additional penalized weight in utility function for dose(s) which has (have) the toxicity probability greater than the pre-specified DLT rate

seed

the seed. The default value is seed = 100

Value

SEARS() returns a list with following elements (1) type I error (2) average sample size for the trial (3) average sample size for each dose (4) average sample size for placebo (5) selection percentage for each dose (6) average toxicity events for each dose

Author(s)

Chia-Wei Hsu, Haitao Pan

Examples

SEARS(p.p = 0.2, p.d = c(0.2, 0.2, 0.2, 0.2, 0.2), p.tox = c(0.03, 0.06, 0.17, 0.3, 0.5),
      k1 = 0.95, k2 = 0.8, pi_t = 0.17, pi_e = 0.2, pT = 0.17, eff_a = 0.5, eff_b = 0.5,
      plac_a = 0.5, plac_b = 0.5, tox_a = 1, tox_b = 1, csize = 3, csize2 = 3,
      p.star = 0.2, q.star = 0.6, f.star = 0.06, p.star2 = 0.2, q.star2 = 0.98,
      d.cs = 36, p.cs = 36, phase1_size = 30, n_earlystop = 100, extrasafe_BOIN = FALSE,
      offset_BOIN = 0.05, Nsim = 10, n_catchup = 3, control_arm = "fixed", power_c = 0.5,
      lower_bound = 0.05, weight1 = 0.5, weight2 = 0.5, seed = 100)