Bayesian dose finding in oncology for drug combinations by copula regression, by G. Yin and Y. Yuan.

Readme file

SERIES C
Applied Statistics

Bayesian dose finding in oncology for drug combinations by copula regression, by G. Yin and Y. Yuan, pages 211–224;

We demonstrate how to use our software below. The scenario is for a drug combination trial with 3 by 2 dose combinations.

The .exe file can be dowloaded from

http://odin.mdacc.tmc.edu/~gyin/software.htm

please enter the number of dose for drug 1: 3
please enter the number of dose for drug 2: 2

Please enter the true toxicity probability for the combination (1, 1): .05
Please enter the true toxicity probability for the combination (1, 2): .1
Please enter the true toxicity probability for the combination (2, 1): .15
Please enter the true toxicity probability for the combination (2, 2): .3
Please enter the true toxicity probability for the combination (3, 1): .45
Please enter the true toxicity probability for the combination (3, 2): .5

The true toxicity probability entered is :

Drug 2	2	0.1	0.3	0.5
	1	0.05	0.15	0.45
	-------------------------------------
		1	2	3
	Drug 1

Are they correct? (y/n): y

Please enter the target toxicity : .3

Please enter the prior estimate of toxicity probability for drug 1 at the dose level 1: .1
Please enter the prior estimate of toxicity probability for drug 1 at the dose level 2: .2
Please enter the prior estimate of toxicity probability for drug 1 at the dose level 3: .3

The prior estimate of toxicity for drug 1 is: 0.1, 0.2, 0.3
Are they correct? (y/n): y

Please enter the prior estimate of toxicity probability for drug 2 at the dose level 1: .1
Please enter the prior estimate of toxicity probability for drug 2 at the dose level 2: .2

The prior estimate of toxicity for drug 2 is : 0.1, 0.2
Are they correct? (y/n): y

Enter the total number of cohorts: 10

Enter the cohort size: 3

Enter the number of simulated trials: 100
Are you sure to continue? (y/n): y

// The simulation results can be found in the file results.txt //
// under the current directory when the simulation finishes. //

Simulating trial 0......
Simulating trial 1......
.
.
.
.
.
.
Simulating trial 96......
Simulating trial 97......
Simulating trial 98......
Simulating trial 99......

----------------------------------------------------------------
Wed Aug 20 12:18:11 2008
CPU time (hour)= 0.226991 # of trials = 100

The number of cohort = 10; cohort size = 3
Escalate if Pr(toxicity<0.3) > 0.8
De-escalate if Pr(toxicity<0.3) < 0.45

True toxicity probabilities:
0.10 0.30 0.50
0.05 0.15 0.45

Selection probabilities (%):
3.0 35.0 8.0
0.0 23.0 30.0

Number of patients treated at each dose:
3.4 5.8 2.8
7.3 5.8 4.7

Number of toxicity observed at each dose:
0.5 1.8 1.5
0.4 1.0 2.0

Total number of observed toxicity: 7.1

Percentage of inconclusive trials: 1.0%

Guosheng Yin
Department of Biostatistics
Unit 1411
M. D. Anderson Cancer Center
University of Texas
PO BOX 301402
Houston
TX 77230-1402
USA

E-mail: gsyin@mdanderson.org