From: Sriram Krishnaswami <ksriram@ufl.edu>
Subject: Simulations for Bioequivalence testing
Date: Tue, 30 May 2000 18:36:56 -0400

Dear NMusers,

I was wondering if it is possible to use monte carlo simulations on NONMEM to test for various measures (Cmax, AUC, Cmax/AUClqc etc.) to differentiate test from the reference compound. From what I gather from a couple of papers,
1) El-Tahtawy et al Pharm Res 15&12& 11, (1998, 1995 and 1994),
2) Bois et al, Pharm Res , 11, 966 and 715 (1994)
3) Tozer et al, Pharm Res, 1995
a very large number of volunteers (something like a million) have been simulated based on various scenarios followed by PK analysis and statistics on them. All of them have used SAS for the simulations as well as follow up stats.

Has anybody used NONMEM plus Excel or something for this purpose?

Thanks in advance,
Sriram
--
Sriram Krishnaswami
Dept. of Pharmaceutics
College of Pharmacy
University of Florida
Gainesville, FL 32610-0494
Home: 352 846 5968
Lab: 352 846 3257
Fax: 352 392 4447

 

 

*****

 

 

From: "Jogarao Gobburu 301-594-5354 FAX 301-480-3212" <GOBBURUJ@cder.fda.gov>
Subject: Re: Simulations for Bioequivalence testing
Date: Wed, 31 May 2000 08:50:35 -0400 (EDT)

Dear Sriram,
1. Yes, it is possible to perform monte carlo simulations using NM for your need.
2. You need the following:
a. Routine to generate a input dataset for your trial design
b. NONMEM for simulations
c. Routine to calculate non-compartmental parameters (nca), preferrably that can be invoked from command line
d. SAS (or the equivalent) to conduct stat testing.
Dr.Nick Holford and I have been using a combination of NONMEM, AWK (nca), and SAS to perform simulations/power-calculations of BE trials.

Regards,
Joga Gobburu,
Pharmacometrics, CDER, FDA.

 

 

*****

 

 

From: Jeff Wald <jwald@pharsight.com>
Subject: Re: Simulations for Bioequivalence testing
Date: Wed, 31 May 2000 15:01:16 -0400

To interested readers - Pharsight's Trial Simulator version2.0 contains all of the elements listed by Dr. Gobburu.

The simulator allows one to build a mixed effects model and simulate it based upon a trial design. For analysis of the simulated results the mathematical engine of WinNonlin is available for noncompartmental computations as is WinNonlin's ANOVA module (supporting general linear models and bioequivalence computations). S-Plus is integrated into the product for further statistical and graphical comparisons. In addition the simulator allows one to build up multiple scenarios for simulation and then analyze/compare them together in the same set of results. Scenarios may address different study design issues (sampling schedules, number of subjects....) or different elements of the drug model (within, between, or residual variance, values of fixed effect model parameters, etc). Simulation of different analyses and analysis of the results is automated via the simulator.

More information can be obtained at www.pharsight.com.

Regards, Jeff