| Type: | Package |
| Title: | Two-Stage Adaptive Dose-Finding Clinical Trial Design |
| Version: | 2.0.1 |
| Maintainer: | Alyssa Vanderbeek <vanderbeekam@gmail.com> |
| Description: | Simulate and implement early phase two-stage adaptive dose-finding design for binary and quasi-continuous toxicity endpoints. See Chiuzan et al. (2018) for further reading <doi:10.1080/19466315.2018.1462727>. |
| Depends: | R (≥ 3.5.0), shiny, shinydashboard |
| Imports: | Matrix, grDevices, utils |
| License: | LGPL-3 |
| Encoding: | UTF-8 |
| LazyData: | true |
| RoxygenNote: | 7.1.1 |
| Suggests: | knitr, rmarkdown |
| VignetteBuilder: | knitr |
| NeedsCompilation: | no |
| Packaged: | 2021-08-05 15:29:32 UTC; Alyssa |
| Author: | Alyssa Vanderbeek [aut, cre], Laura Cosgrove [ctb], Elizabeth Garrett-Mayer [ctb], Cody Chiuzan [ctb] |
| Repository: | CRAN |
| Date/Publication: | 2021-08-06 04:50:09 UTC |
Calculates likelihood of safety for single dose
Description
Function LRtox() calculates the likelihood of safety for a single dose
and designates whether to escalate to the next dose (safe) or stop dose escalation and move onto stage 2 (unsafe).
Usage
LRtox(coh.size, ndlt, p1, p2, K = 2)
Arguments
coh.size |
cohort size (number of patients) per dose (Stage 1) |
ndlt |
number of observed DLTs |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
Value
List object that gives the likelihood ratio of safety and indicates whether to escalate to the next highest dose level, or stop dose escalation and move onto stage 2.
Examples
LRtox(coh.size = 3, ndlt = 2, p1 = 0.40, p2 = 0.15, K = 2)
LRtox(coh.size = 3, ndlt = 1, p1 = 0.40, p2 = 0.15, K = 2)
Calculates likelihood of safety for single dose, using nTTP
Description
(nTTP) Function LRtox.nTTP() calculates the likelihood of safety for a single dose
and designates whether to escalate to the next dose (safe) or stop dose escalation and move onto stage 2 (unsafe).
Usage
LRtox.nTTP(tox_grades, ntox, coh.size, W, p1, p2, K = 2, std.nTTP = 0.15)
Arguments
tox_grades |
data frame of observed AE grades for each patient (rows) across all toxicity types (columns). e.g. for one patient, grades for 3 toxicity types might be c(3, 2, 4), where they experienced a grade 3 AE for tox type 1, grade 2 AE for tox type 2, etc. |
ntox |
number (integer) of different toxicity types |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 5 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
std.nTTP |
the standard deviation of nTTP scores at each dose level (constant across doses) |
Value
List object that gives the likelihood ratio of safety and indicates whether to escalate to the next highest dose level, or stop dose escalation and move onto stage 2.
Examples
ntox = 3 # three different types of toxicity
coh.size = 3 # number of patients enrolled per dose
# Observed AE grades for each patient on tested dose
obs = data.frame(tox1 = c(3, 2, 4),
tox2 = c(1, 1, 2),
tox3 = c(2, 3, 3))
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
LRtox.nTTP(obs, ntox, coh.size, W, p1, p2, K = 2, std.nTTP = 0.15)
Sample array of toxicity probabilities for 6 doses. Taken from Du et al.
Description
This is a sample array of toxicity probabilities to be used for trial simulations with nTTP as the toxicity endpoint. In this example, we have 3 toxicity types, 6 test doses, and 5 AE grades (0-4). See the nTTP_simulation_example vignette for more details.
Usage
data("TOX")Format
The format is: num [1:6, 1:5, 1:3] 0.791 0.738 0.685 0.662 0.605 0.39 0.172 0.195 0.19 0.2 ...
Source
https://pubmed.ncbi.nlm.nih.gov/30403559/
Examples
data(TOX)
TOX
Generates parameters for the beta distribution # I don't think we need to show this as a separate function, but put together with gen.eff.stg1 or be called by gen.eff.stg1
Description
Function beta.ab() returns parameters alpha and beta for generating beta r.v. (per dose)
Usage
beta.ab(m, v)
Arguments
m |
mean efficacies for given dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
efficacy variance per dose. Values range from 0 - 1. (e.g., 0.01) |
Value
Vector of alpha and beta values for generating beta random variable for a dose.
Calculate DLT probability corresponding to average nTTP for each dose
Description
Calculate DLT probability corresponding to average nTTP for each dose
Usage
dlt.prob(dose, ntox, TOX, grade.thresh)
Arguments
dose |
number of doses to be tested (scalar) |
ntox |
number (integer) of different toxicity types (e.g, hematological, neurological, GI) |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4, since the probability of a grade 0 event may not be 0). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
grade.thresh |
grade (0-4) at which each toxicity type qualifies as a DLT |
Value
ptox - Vector of DLT probabilities per dose.
Examples
# Number of test doses
dose = 6
# Number of toxicity types
ntox <- 3
# Array of toxicity event probabilities
TOX = array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001, #prob of tox for dose 1 and tox type 1
0.791, 0.172, 0.032, 0.004, 0.001, #prob of tox for dose 2 and tox type 1
0.758, 0.180, 0.043, 0.010, 0.009, #prob of tox for dose 3 and tox type 1
0.685, 0.190, 0.068, 0.044, 0.013, #prob of tox for dose 4 and tox type 1
0.662, 0.200, 0.078, 0.046, 0.014, #prob of tox for dose 5 and tox type 1
0.605, 0.223, 0.082, 0.070, 0.020), #prob of tox for dose 6 and tox type 1
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000, #prob of tox for dose 1 and tox type 2
0.968, 0.029, 0.002, 0.001, 0.000, #prob of tox for dose 2 and tox type 2
0.813, 0.172, 0.006, 0.009, 0.000, #prob of tox for dose 3 and tox type 2
0.762, 0.183, 0.041, 0.010, 0.004, #prob of tox for dose 4 and tox type 2
0.671, 0.205, 0.108, 0.011, 0.005, #prob of tox for dose 5 and tox type 2
0.397, 0.258, 0.277, 0.060, 0.008), #prob of tox for dose 6 and tox type 2
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004, #prob of tox for dose 1 and tox type 3
0.917, 0.070, 0.007, 0.001, 0.005, #prob of tox for dose 2 and tox type 3
0.652, 0.280, 0.010, 0.021, 0.037, #prob of tox for dose 3 and tox type 3
0.536, 0.209, 0.031, 0.090, 0.134, #prob of tox for dose 4 and tox type 3
0.015, 0.134, 0.240, 0.335, 0.276, #prob of tox for dose 5 and tox type 3
0.005, 0.052, 0.224, 0.372, 0.347), #prob of tox for dose 6 and tox type 3
nrow = 6, byrow = TRUE)
# Grades at which each tox type qualifies as DLT
grade.thresh = c(3, 3, 4)
dlt.prob(dose = dose, ntox = ntox, TOX = TOX, grade.thresh = grade.thresh)
Generates efficacy outcomes for stage 1 when using binary toxicity
Description
Function eff.stg1() uses a beta-binomial distribution to generate
outcomes (Ys) corresponding to acceptable dose assignments from stage 1.
Usage
eff.stg1(dose, dose.tox, p1, p2, K, coh.size, m, v, nbb = 100)
Arguments
dose |
number of doses to be tested (scalar) |
dose.tox |
vector of true toxicities for each dose. Values range from 0 - 1. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
nbb |
binomial parameter (default = 100 cells per patient) |
Value
List of efficacy outcomes for subjects enrolled during stage 1 (dose-escalation)
Y.safe - vector of efficacy outcomes for each subject assigned to an acceptable safe dose
d.safe - vector of dose allocation for each subject assigned to an acceptable safe dose
tox.safe - number of dose-limiting toxicities for each safe dose level
Y.alloc - vector of efficacy outcomes for all subjects from stage 1 (acceptable and unsafe doses)
d.alloc - vector of dose allocation for all subjects from stage 1 (acceptable and unsafe doses)
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p2) and unacceptable (p1) DLT rates used for establishing safety
p1 <- 0.40
p2 <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Vector of true mean efficacies per dose (here mean percent persistence per dose)
m <- c(5, 15, 40, 65, 80) # MUST BE THE SAME LENGTH AS dose.tox
# Efficacy (equal) variance per dose
v <- rep(0.01, 5)
# Total sample size (stages 1&2)
N <- 25
# Stopping rule: if dose 1 is the only safe dose, allocate up to 9 pts.
stop.rule <- 9
eff.stg1(dose = dose, dose.tox = dose.tox, p1 = p1, p2 = p2, K = K,
coh.size = coh.size, m, v, nbb = 100)
Generates efficacy outcomes for stage 1 when using nTTP to measure toxicity
Description
Function eff.stg1.nTTP() uses a beta-binomial distribution to generate
outcomes (Ys) corresponding to acceptable dose assignments from stage 1.
Usage
eff.stg1.nTTP(
dose,
p1,
p2,
K,
coh.size,
m,
v,
nbb = 100,
W,
TOX,
ntox,
std.nTTP
)
Arguments
dose |
number of doses to be tested (scalar) |
p1 |
toxicity under null (unsafe nTTP). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe nTTP). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
nbb |
binomial parameter (default = 100 cells per patient) |
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 4 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
ntox |
number (integer) of different toxicity types |
std.nTTP |
the standard deviation of nTTP scores at each dose level (assumed constant across doses) |
Value
List of efficacy outcomes for subjects enrolled during stage 1 (dose-escalation)
Y.safe - vector of efficacy outcomes for each subject assigned to an acceptable safe dose
d.safe - vector of dose allocation for each subject assigned to an acceptable safe dose
tox.safe - number of dose-limiting toxicities for each safe dose level
Y.alloc - vector of efficacy outcomes for all subjects from stage 1 (acceptable and unsafe doses)
d.alloc - vector of dose allocation for all subjects from stage 1 (acceptable and unsafe doses)
all_nttp - all observed nTTP values
Examples
# Number of pre-specified dose levels
dose <- 6
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Efficacy (equal) variance per dose
v <- rep(0.01, 6)
# Dose-efficacy curve
m = c(10, 20, 30, 40, 70, 90)
# Number of toxicity types
ntox <- 3
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# standard deviation of nTTP values
std.nTTP = 0.15
# Array of toxicity event probabilities
TOX <- array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001,
0.791, 0.172, 0.032, 0.004, 0.001,
0.758, 0.180, 0.043, 0.010, 0.009,
0.685, 0.190, 0.068, 0.044, 0.013,
0.662, 0.200, 0.078, 0.046, 0.014,
0.605, 0.223, 0.082, 0.070, 0.020),
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000,
0.968, 0.029, 0.002, 0.001, 0.000,
0.813, 0.172, 0.006, 0.009, 0.000,
0.762, 0.183, 0.041, 0.010, 0.004,
0.671, 0.205, 0.108, 0.011, 0.005,
0.397, 0.258, 0.277, 0.060, 0.008),
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004,
0.917, 0.070, 0.007, 0.001, 0.005,
0.652, 0.280, 0.010, 0.021, 0.037,
0.536, 0.209, 0.031, 0.090, 0.134,
0.015, 0.134, 0.240, 0.335, 0.276,
0.005, 0.052, 0.224, 0.372, 0.347),
nrow = 6, byrow = TRUE)
eff.stg1.nTTP(dose = dose, p1 = p1, p2 = p2, K = K, coh.size = coh.size,
m = m, v = v, nbb = 100, W = W, TOX = TOX, ntox = ntox, std.nTTP = std.nTTP)
Obtain average nTTP at each dose level
Description
Obtain average nTTP at each dose level
Usage
get.thresh(dose, ntox, W, TOX)
Arguments
dose |
number of doses to be tested (scalar) |
ntox |
number (integer) of different toxicity types (e.g, hematological, neurological, GI) |
W |
matrix defines burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 4 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
Value
Vector of average nTTP for each dose level.
Examples
# Number of test doses
dose = 6
# Number of toxicity types
ntox <- 3
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Array of toxicity event probabilities
TOX = array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001, #prob of tox for dose 1 and tox type 1
0.791, 0.172, 0.032, 0.004, 0.001, #prob of tox for dose 2 and tox type 1
0.758, 0.180, 0.043, 0.010, 0.009, #prob of tox for dose 3 and tox type 1
0.685, 0.190, 0.068, 0.044, 0.013, #prob of tox for dose 4 and tox type 1
0.662, 0.200, 0.078, 0.046, 0.014, #prob of tox for dose 5 and tox type 1
0.605, 0.223, 0.082, 0.070, 0.020), #prob of tox for dose 6 and tox type 1
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000, #prob of tox for dose 1 and tox type 2
0.968, 0.029, 0.002, 0.001, 0.000, #prob of tox for dose 2 and tox type 2
0.813, 0.172, 0.006, 0.009, 0.000, #prob of tox for dose 3 and tox type 2
0.762, 0.183, 0.041, 0.010, 0.004, #prob of tox for dose 4 and tox type 2
0.671, 0.205, 0.108, 0.011, 0.005, #prob of tox for dose 5 and tox type 2
0.397, 0.258, 0.277, 0.060, 0.008), #prob of tox for dose 6 and tox type 2
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004, #prob of tox for dose 1 and tox type 3
0.917, 0.070, 0.007, 0.001, 0.005, #prob of tox for dose 2 and tox type 3
0.652, 0.280, 0.010, 0.021, 0.037, #prob of tox for dose 3 and tox type 3
0.536, 0.209, 0.031, 0.090, 0.134, #prob of tox for dose 4 and tox type 3
0.015, 0.134, 0.240, 0.335, 0.276, #prob of tox for dose 5 and tox type 3
0.005, 0.052, 0.224, 0.372, 0.347), #prob of tox for dose 6 and tox type 3
nrow = 6, byrow = TRUE)
get.thresh(dose = dose, ntox = ntox, W = W, TOX = TOX)
Simulate full trial (both stages) x times when using nTTP to measure toxicity
Description
Results are displayed in a matrix format, where each row represents one trial simulation
Usage
nTTP.indiv.sim(W, TOX, ntox, dose)
Arguments
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 5 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
ntox |
number (integer) of different toxicity types |
dose |
number of doses to be tested (scalar) |
Value
List of the following objects:
sim.Y - estimated efficacy per each dose assignment
sim.d - dose assignment for each patient in the trial
Calculates randomization probabilities and dose allocation for next patient
Description
Function rand.prob() calculates the updated randomization probabilities based on observed efficacies up to that point.
It also gives the dose allocation for the next enrolled patient based on these probabilities.
Usage
rand.prob(y.eff, d.safe)
Arguments
y.eff |
vector of all efficacy outcomes for each dose allocation |
d.safe |
vector of dose assignment |
Value
List object giving
Rand.Prob - randomization probability for each safe dose (from stage 1)
Next.Dose - the dose to enroll the next patient on
Examples
y.eff <- c(9, 1, 0, 34, 10, 27, 38, 42, 60, 75, 48, 62)
d.safe <- c(1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4)
rand.prob(y.eff, d.safe)
Stage 2 Adaptive Randomization
Description
Function rand.stg2() fits a linear regression for the continuous
efficacy outcomes,
computes the randomization probabilities/dose and allocates the next patient to a dose that
is considered acceptably safe and has the most promising efficacy. Dose safety is still
monitored using LR and doses
that become unacceptable are discarded (never re-visited).
Usage
rand.stg2(
dose,
dose.tox,
p1,
p2,
K,
coh.size,
m,
v,
N,
stop.rule = 9,
cohort = 1,
samedose = TRUE,
nbb = 100
)
Arguments
dose |
number of doses to be tested (scalar) |
dose.tox |
vector of true toxicities for each dose. Values range from 0 - 1. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
N |
total sample size for stages 1&2 |
stop.rule |
if only dose 1 safe, allocate up to 9 (default) patients at dose 1 to collect more info |
cohort |
cohort size (number of patients) per dose (Stage 2). Default is 1. |
samedose |
designates whether the next patient is allocated to the same dose as the previous patient. Default is TRUE. Function adjusts accordingly. |
nbb |
binomial parameter (default = 100 cells per patient) |
Value
List of the following objects:
Y.final - vector of all efficacy outcomes (Ys) corresponding to dose assignments (Stages 1&2)
d.final - vector of all dose assignments(Stages 1&2)
If dose allocation stops early, put NAs in d.final and y.final until it reaches the total sample size.
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p_yes) and unacceptable (p_no) DLT rates used for establishing safety
p_no <- 0.40
p_yes <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Vector of true mean efficacies per dose (here mean percent persistence per dose)
m <- c(5, 15, 40, 65, 80) # MUST BE THE SAME LENGTH AS dose.tox
# Efficacy(equal) variance per dose
v <- rep(0.01, 5)
# Total sample size (stages 1&2)
N <- 25
# Stopping rule: if dose 1 is the only safe dose, allocate up to 9 pts.
stop.rule <- 9
rand.stg2(dose, dose.tox, p_no, p_yes, K, coh.size, m, v, N, stop.rule = stop.rule,
cohort = 1, samedose = TRUE, nbb = 100)
Stage 2 Adaptive Randomization with nTTP to monitor toxicity
Description
Function rand.stg2.nTTP() fits a linear regression for the continuous
efficacy outcomes, computes the randomization probabilities/dose and allocates the next patient
to a dose that is considered acceptably safe and has the highest efficacy. Dose safety
(with nTTP) is still monitored using LR and doses that become unacceptable are discarded
(never re-visited).
Usage
rand.stg2.nTTP(
dose,
p1,
p2,
K,
coh.size,
m,
v,
N,
stop.rule = 9,
cohort = 1,
samedose = TRUE,
nbb = 100,
W,
TOX,
ntox,
std.nTTP = 0.15
)
Arguments
dose |
number of doses to be tested (scalar) |
p1 |
toxicity under null (unsafe nTTP). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe nTTP). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
N |
total sample size for stages 1&2 |
stop.rule |
if only dose 1 safe, allocate up to 9 (default) patients at dose 1 to collect more info |
cohort |
cohort size (number of patients) per dose (Stage 2). Default is 1. |
samedose |
designates whether the next patient is allocated to the same dose as the previous patient. Default is TRUE. Function adjusts accordingly. |
nbb |
binomial parameter (default = 100 cells per patient) |
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 4 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
ntox |
number (integer) of different toxicity types |
std.nTTP |
the standard deviation of nTTP scores at each dose level (constant across doses) |
Value
List of the following objects:
Y.final - vector of all efficacy outcomes (Ys) corresponding to dose assignments (Stages 1&2)
d.final - vector of all dose assignments(Stage 1&2)
n1 - Stage 1 sample size
If dose allocation stops early, put NAs in d.final and y.final until it reaches the total sample size.
Examples
# Number of pre-specified dose levels
dose <- 6
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Total sample size (stages 1&2)
N <- 25
# Efficacy (equal) variance per dose
v <- rep(0.01, 6)
# Dose-efficacy curve
m = c(10, 20, 30, 40, 70, 90)
# Number of toxicity types
ntox <- 3
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Standard deviation of nTTP value
std.nTTP = 0.15
# Array of toxicity event probabilities
TOX <- array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001,
0.791, 0.172, 0.032, 0.004, 0.001,
0.758, 0.180, 0.043, 0.010, 0.009,
0.685, 0.190, 0.068, 0.044, 0.013,
0.662, 0.200, 0.078, 0.046, 0.014,
0.605, 0.223, 0.082, 0.070, 0.020),
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000,
0.968, 0.029, 0.002, 0.001, 0.000,
0.813, 0.172, 0.006, 0.009, 0.000,
0.762, 0.183, 0.041, 0.010, 0.004,
0.671, 0.205, 0.108, 0.011, 0.005,
0.397, 0.258, 0.277, 0.060, 0.008),
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004,
0.917, 0.070, 0.007, 0.001, 0.005,
0.652, 0.280, 0.010, 0.021, 0.037,
0.536, 0.209, 0.031, 0.090, 0.134,
0.015, 0.134, 0.240, 0.335, 0.276,
0.005, 0.052, 0.224, 0.372, 0.347),
nrow = 6, byrow = TRUE)
rand.stg2.nTTP(dose = dose, p1 = p1, p2 = p2, K = K, coh.size = coh.size,
m = m, v = v, N = N, stop.rule = 9,
cohort = 1, samedose = TRUE, nbb = 100, W = W, TOX = TOX, ntox = ntox, std.nTTP = std.nTTP)
Launch Shiny app
Description
Launch Shiny app
Usage
runExample()
Identify safe/acceptable doses from stage 1 based on observed binary toxicity
Description
Function safe.dose() distinguishes acceptable from unacceptable doses
Usage
safe.dose(dose, dose.tox, p1, p2, K, coh.size)
Arguments
dose |
number of doses to be tested (scalar) |
dose.tox |
vector of true toxicities for each dose. Values range from 0 - 1. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
Value
List of the following objects:
alloc.safe - matrix of assignments only for acceptable doses (to be used in stage 2) and their corresponding toxicities
alloc.total - vector of all dose assignments from stage 1
n1 - total number of subjects allocated in stage 1
Examples
dose = 5 # Dose levels
dose.tox <- c(0.05, 0.10, 0.15, 0.20, 0.30) # True toxicity per dose
p1 = 0.40 # Unacceptable DLT rate
p2 = 0.15 # Acceptable DLT rate
K = 2 # Likelihood-ratio (LR) threshold
coh.size = 3 # Assign 3 pts per dose in stage 1
safe.dose(dose = dose, dose.tox = dose.tox, p1 = p1, p2 = p2, K = K, coh.size = coh.size)
Identify safe/acceptable doses from stage 1 based on nTTP scores.
Description
Function safe.dose.nTTP() distinguishes acceptable from unacceptable doses
Usage
safe.dose.nTTP(dose, p1, p2, K, coh.size, W, TOX, ntox, std.nTTP = 0.15)
Arguments
dose |
number of doses to be tested (scalar) |
p1 |
toxicity under null (unsafe nTTP). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe nTTP). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 4 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
ntox |
number (integer) of different toxicity types |
std.nTTP |
the standard deviation of nTTP scores at each dose level (constant across doses) |
Value
List of the following objects:
alloc.safe - matrix of assignments only for acceptable doses (to be used in stage 2) and their corresponding toxicities
alloc.total - vector of all dose assignments from stage 1
n1 - total number of subjects allocated in stage 1
all_nttp - all observed nTTP values
Examples
# Number of pre-specified dose levels
dose <- 6
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Number of toxicity types
ntox <- 3
# Standard deviation of nTTP values
std.nTTP = 0.15
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Array of toxicity event probabilities
TOX <- array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001,
0.791, 0.172, 0.032, 0.004, 0.001,
0.758, 0.180, 0.043, 0.010, 0.009,
0.685, 0.190, 0.068, 0.044, 0.013,
0.662, 0.200, 0.078, 0.046, 0.014,
0.605, 0.223, 0.082, 0.070, 0.020),
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000,
0.968, 0.029, 0.002, 0.001, 0.000,
0.813, 0.172, 0.006, 0.009, 0.000,
0.762, 0.183, 0.041, 0.010, 0.004,
0.671, 0.205, 0.108, 0.011, 0.005,
0.397, 0.258, 0.277, 0.060, 0.008),
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004,
0.917, 0.070, 0.007, 0.001, 0.005,
0.652, 0.280, 0.010, 0.021, 0.037,
0.536, 0.209, 0.031, 0.090, 0.134,
0.015, 0.134, 0.240, 0.335, 0.276,
0.005, 0.052, 0.224, 0.372, 0.347),
nrow = 6, byrow = TRUE)
safe.dose.nTTP(dose = dose, p1 = p1, p2 = p2, K = K, coh.size = coh.size,
W = W, TOX = TOX, ntox = ntox, std.nTTP = std.nTTP)
Generate plots for estimated percent allocation and response per dose.
Description
Generate plots for estimated percent allocation and response per dose.
Usage
sim.plot(sims)
Arguments
sims |
output from sim.trials |
Value
Error plots of estimated (1) percent allocation per dose, and (2) estimated response per dose.
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p_yes) and unacceptable (p_no) DLT rates used for establishing safety
p_no <- 0.40
p_yes <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Vector of true mean efficacies per dose (here mean T-cell persistence per dose (%))
m <- c(5, 15, 40, 65, 80) # MUST BE THE SAME LENGTH AS dose.tox
# Efficacy (equal) variance per dose
v <- rep(0.01, 5)
# Total sample size (stages 1&2)
N <- 25
# Stopping rule
stop.rule <- 9
numsims = 100
set.seed(1)
simulations = sim.trials(numsims = numsims, dose, dose.tox, p1 = p_no, p2 = p_yes,
K, coh.size, m, v, N, stop.rule = stop.rule, cohort = 1,
samedose = TRUE, nbb = 100)
# sim.plot(simulations)
Visualize simulation results (Stages 1&2)
Description
Results from simulated trials (using sim.trials() function)
displayed in tabular and/or graphical format
Usage
sim.summary(sims, print = TRUE)
Arguments
sims |
output from sim.trials |
print |
logical specifying whether to print tables in console |
Value
Printed tables and a list of the following objects:
pct.treated - IQR (25th percentile, median, 75th percentile) of percent of subjects treated at each dose level
efficacy - IQR of efficacy observed at each dose level
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p_yes) and unacceptable (p_no) DLT rates used for establishing safety
p_no <- 0.40
p_yes <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Vector of true mean efficacies per dose (here mean percent persistence per dose)
m <- c(5, 15, 40, 65, 80) # MUST BE THE SAME LENGTH AS dose.tox
# Efficacy(equal) variance per dose
v <- rep(0.01, 5)
# Total sample size (stages 1&2)
N <- 25
# Stopping rule: if dose 1 is the only safe dose, allocate up to 9 pts.
stop.rule <- 9
simulations = sim.trials(numsims = 100, dose, dose.tox, p1 = p_no, p2 = p_yes, K,
coh.size, m, v, N, stop.rule = stop.rule, cohort = 1, samedose = TRUE, nbb = 100)
summary = sim.summary(simulations)
Simulate full trial (both stages) x times
Description
Results are displayed in a matrix format, where each row represents one trial simulation.
Usage
sim.trials(
numsims,
dose,
dose.tox,
p1,
p2,
K,
coh.size,
m,
v,
N,
stop.rule = 9,
cohort = 1,
samedose = TRUE,
nbb = 100
)
Arguments
numsims |
number of simulated trials |
dose |
number of doses to be tested (scalar) |
dose.tox |
vector of true toxicities for each dose. Values range from 0 - 1. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
N |
total sample size for stages 1&2 |
stop.rule |
if only dose 1 safe, allocate up to 9 (default) patients at dose 1 to collect more info. |
cohort |
cohort size (number of patients) per dose (Stage 2). Default is 1. |
samedose |
designates whether the next patient is allocated to the same dose as the previous patient. Default is TRUE. Function adjusts accordingly. |
nbb |
binomial parameter (default = 100 cells per patient) |
Value
List of the following objects:
sim.Y - estimated efficacy per each dose assignment
sim.d - dose assignment for each patient in the trial
safe.d - indicator of whether dose was declared safe
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p_yes) and unacceptable (p_no) DLT rates used for establishing safety
p_no <- 0.40
p_yes <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Vector of true mean efficacies per dose (here mean percent persistence per dose)
m <- c(5, 15, 40, 65, 80) # MUST BE THE SAME LENGTH AS dose.tox
# Efficacy(equal) variance per dose
v <- rep(0.01, 5)
# Total sample size (stages 1&2)
N <- 25
# Stopping rule: if dose 1 is the only safe dose, allocate up to 9 pts.
stop.rule <- 9
sim.trials(numsims = 10, dose, dose.tox, p1 = p_no, p2 = p_yes, K,
coh.size, m, v, N, stop.rule = stop.rule, cohort = 1, samedose = TRUE, nbb = 100)
Simulate full trial (both stages) x times when using nTTP to measure toxicity
Description
Results are displayed in a matrix format, where each row represents one trial simulation
Usage
sim.trials.nTTP(
numsims,
dose,
p1,
p2,
K,
coh.size,
m,
v,
N,
stop.rule = 9,
cohort = 1,
samedose = TRUE,
nbb = 100,
W,
TOX,
ntox,
std.nTTP = 0.15
)
Arguments
numsims |
number of simulated trials |
dose |
number of doses to be tested (scalar) |
p1 |
toxicity under null (unsafe nTTP). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe nTTP). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
m |
vector of mean efficacies per dose. Values range from 0 - 100. (e.g, T cell persistence - values b/w 5 and 80 per cent) |
v |
vector of efficacy variances per dose. Values range from 0 - 1. (e.g., 0.01) |
N |
total sample size for stages 1&2 |
stop.rule |
if only dose 1 safe, allocate up to 9 (default) patients at dose 1 to collect more info |
cohort |
cohort size (number of patients) per dose (Stage 2). Default is 1. |
samedose |
designates whether the next patient is allocated to the same dose as the previous patient. Default is TRUE. Function adjusts accordingly. |
nbb |
binomial parameter (default = 100 cells per patient) |
W |
matrix defining burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 4 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
ntox |
number (integer) of different toxicity types |
std.nTTP |
the standard deviation of nTTP scores at each dose level (constant across doses) |
Value
List of the following objects:
sim.Y - estimated efficacy per each dose assignment
sim.d - dose assignment for each patient in the trial
safe.d - indicator of whether dose was declared safe
Examples
# Number of pre-specified dose levels
dose <- 6
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Total sample size (stages 1&2)
N <- 25
# Efficacy (equal) variance per dose
v <- rep(0.01, 6)
# Dose-efficacy curve
m = c(10, 20, 30, 40, 70, 90)
# Number of toxicity types
ntox = 3
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Standard deviation of nTTP values
std.nTTP = 0.15
# Array of toxicity event probabilities
TOX = array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001,
0.791, 0.172, 0.032, 0.004, 0.001,
0.758, 0.180, 0.043, 0.010, 0.009,
0.685, 0.190, 0.068, 0.044, 0.013,
0.662, 0.200, 0.078, 0.046, 0.014,
0.605, 0.223, 0.082, 0.070, 0.020),
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000,
0.968, 0.029, 0.002, 0.001, 0.000,
0.813, 0.172, 0.006, 0.009, 0.000,
0.762, 0.183, 0.041, 0.010, 0.004,
0.671, 0.205, 0.108, 0.011, 0.005,
0.397, 0.258, 0.277, 0.060, 0.008),
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004,
0.917, 0.070, 0.007, 0.001, 0.005,
0.652, 0.280, 0.010, 0.021, 0.037,
0.536, 0.209, 0.031, 0.090, 0.134,
0.015, 0.134, 0.240, 0.335, 0.276,
0.005, 0.052, 0.224, 0.372, 0.347),
nrow = 6, byrow = TRUE)
sim.trials.nTTP(numsims = 10, dose = dose, p1 = p1, p2 = p2, K = K,
coh.size = coh.size, m = m, v = v, N = N, stop.rule = 9, cohort = 1,
samedose = TRUE, nbb = 100, W = W, TOX = TOX, ntox = ntox, std.nTTP = std.nTTP)
Generates DLTs and calculates the likelihood-ratio (LR) for each dose
Description
Gives toxicity profile (number of dose-limiting toxicities) and likelihood ratio per dose based on binary toxicity.
Usage
tox.profile(dose, dose.tox, p1, p2, K, coh.size)
Arguments
dose |
number of doses to be tested (scalar) |
dose.tox |
vector of true toxicities for each dose. Values range from 0 - 1. |
p1 |
toxicity under null (unsafe DLT rate). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe DLT rate). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
Value
4-column matrix containing dose assignment, dose-limiting toxicities at each dose, cohort number, and likelihood ratio.
Examples
# Number of pre-specified dose levels
dose <- 5
# Vector of true toxicities associated with each dose
dose.tox <- c(0.05, 0.10, 0.20, 0.35, 0.45)
# Acceptable (p2) and unacceptable (p1) DLT rates used for establishing safety
p1 <- 0.40
p2 <- 0.15
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Stopping rule: if dose 1 is the only safe dose, allocate up to 9 pts.
stop.rule <- 9
tox.profile(dose = dose, dose.tox = dose.tox, p1 = p1, p2 = p2, K = K, coh.size = coh.size)
Generate nTTPs toxicity scores and the likelihhood-ratio (LR) per dose
Description
The normalized total toxicity profiles (nTTP) are calculated by combining multiple toxicity grades and their weights. The nTTPs are considered a quasi-continuous toxicity measure that follows a normal distribution truncated to [0, 1]. The likelihood ratio per dose are based on nTTP toxicity.
Usage
tox.profile.nTTP(dose, p1, p2, K, coh.size, ntox, W, TOX, std.nTTP = 0.15)
Arguments
dose |
number of doses to be tested (scalar) |
p1 |
toxicity under null (unsafe nTTP). Values range from 0 - 1. |
p2 |
toxicity under alternative (safe nTTP). Values range from 0 - 1; p1 > p2 |
K |
threshold for LR. Takes integer values: 1,2,...(recommended K=2) |
coh.size |
cohort size (number of patients) per dose (Stage 1) |
ntox |
number (integer) of different toxicity types (e.g, hematological, neurological, GI) |
W |
matrix defines burden weight of each grade level for all toxicity types. The dimensions are ntox rows by 5 columns (for grades 0-4). See Ezzalfani et al. (2013) for details. |
TOX |
matrix array of toxicity probabilities. There should be ntox matrices. Each matrix represents one toxicity type, where probabilities of each toxicity grade are specified across each dose. Each matrix has the same dimensions: n rows, representing number of doses, and 5 columns (for grades 0-4). Probabilities across each dose (rows) must sum to 1. See Ezzalfani et al. (2013) for details. |
std.nTTP |
the standard deviation of nTTP scores at each dose level (constant across doses) |
Value
mnTTP - 4-column matrix containing dose assignment, mean nTTP at each dose, cohort number, and likelihood ratio.
all_nttp - all observed nTTP values
Examples
# Number of pre-specified dose levels
dose <- 6
# Acceptable (p2) and unacceptable nTTP values
p1 <- 0.35
p2 <- 0.10
# Likelihood-ratio (LR) threshold
K <- 2
# Cohort size used in stage 1
coh.size <- 3
# Number of toxicity types
ntox <- 3
# Standard deviation of nTTP values
std.nTTP = 0.15
# Toxicity burden weight matrix
W = matrix(c(0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 1
0, 0.5, 0.75, 1.0, 1.5, # Burden weight for grades 0-4 for toxicity 2
0, 0.00, 0.00, 0.5, 1), # Burden weight for grades 0-4 for toxicity 3
nrow = ntox, byrow = TRUE)
# Array of toxicity event probabilities
TOX = array(NA, c(dose, 5, ntox))
TOX[, , 1] = matrix(c(0.823, 0.152, 0.022, 0.002, 0.001, #prob of tox for dose 1 and tox type 1
0.791, 0.172, 0.032, 0.004, 0.001, #prob of tox for dose 2 and tox type 1
0.758, 0.180, 0.043, 0.010, 0.009, #prob of tox for dose 3 and tox type 1
0.685, 0.190, 0.068, 0.044, 0.013, #prob of tox for dose 4 and tox type 1
0.662, 0.200, 0.078, 0.046, 0.014, #prob of tox for dose 5 and tox type 1
0.605, 0.223, 0.082, 0.070, 0.020), #prob of tox for dose 6 and tox type 1
nrow = 6, byrow = TRUE)
TOX[, , 2] = matrix(c(0.970, 0.027, 0.002, 0.001, 0.000, #prob of tox for dose 1 and tox type 2
0.968, 0.029, 0.002, 0.001, 0.000, #prob of tox for dose 2 and tox type 2
0.813, 0.172, 0.006, 0.009, 0.000, #prob of tox for dose 3 and tox type 2
0.762, 0.183, 0.041, 0.010, 0.004, #prob of tox for dose 4 and tox type 2
0.671, 0.205, 0.108, 0.011, 0.005, #prob of tox for dose 5 and tox type 2
0.397, 0.258, 0.277, 0.060, 0.008), #prob of tox for dose 6 and tox type 2
nrow = 6, byrow = TRUE)
TOX[, , 3] = matrix(c(0.930, 0.060, 0.005, 0.001, 0.004, #prob of tox for dose 1 and tox type 3
0.917, 0.070, 0.007, 0.001, 0.005, #prob of tox for dose 2 and tox type 3
0.652, 0.280, 0.010, 0.021, 0.037, #prob of tox for dose 3 and tox type 3
0.536, 0.209, 0.031, 0.090, 0.134, #prob of tox for dose 4 and tox type 3
0.015, 0.134, 0.240, 0.335, 0.276, #prob of tox for dose 5 and tox type 3
0.005, 0.052, 0.224, 0.372, 0.347), #prob of tox for dose 6 and tox type 3
nrow = 6, byrow = TRUE)
tox.profile.nTTP(dose = dose,
p1 = p1,
p2 = p2,
K = K,
coh.size = coh.size,
ntox = ntox,
W = W,
TOX = TOX,
std.nTTP = std.nTTP)