NONMEM software

The non-linear mixed effects model is the most widely used method and has proven to be very useful for continuous measures of drug effect, categorical response data, and survival-type data. The nonlinear mixed-effects modeling software (NONMEM) introduced by Sheiner and Beal is one of the most commonly used programs for population analysis. A detailed review of software for performing population PK/PD analysis is available. ... 

Base Model Choice. The choice was a steady-state one-compartment model with first-order absorption or a steady-state oral two-compartment model with first-order absorption. The disposition parameters were to be expressed in volume and clearance. Intersubject variability and residual error were also to be assessed. The best-fit model, using the software NONMEM, was to be the final base model. The criteria for accepting the NONMEM base model included (a) improved fitting of the diagnostic scatterplots (observed vs. predicted concentration, residual/weighted residual vs. predicted concentration... 

The software NONMEM (17) with first-order conditional estimation (FOCE + INTER) was used throughout the analysis. The proportional error model was used for intraindividual variability, and interindividual variability was assumed lognormally distributed. Interindividual variabilities were assumed independent that is, diagonal matrices for OMEGA were used throughout for model development. These details were not explicitly stated in the plan but were maintained as... 

Estimation of nonlinear mixed effects models has been implemented in a number of software packages and includes different estimation methods [12]. As NONMEM is the most commonly used software to estimate population parameters this program is base for the following description. 

The NONMEM (nonlinear mixed-effects modeling) software (Beal et al. 1992), mostly used in population pharmacokinetics, was developed at the University of California and is presently distributed by Globomax. For data management, post processing and diagnostic plots, the software S-plus (Mathsoft) is frequently used. 

The nonlinear mixed-effects method is depicted in Figure 10.4 and is described here using the conventions of the NONMEM software (2, 3) and the description by Vozeh ef a/. (3). It is based on the principle that the individual pharmacokinetic parameters of a patient population arise from a distribution that can be described by the population mean and the interindividual variance. Each individual pharmacokinetic parameter can be expressed as a population mean and a deviation, typical for an individual. The deviation is the difference between the population mean and the individual parameter and is assumed to be... 

First-Order (NONMEM) Method. The first nonlinear mixed-effects modeling program introduced for the analysis of large pharmacokinetic data was NONMEM, developed by Beal and Sheiner. In the NONMEM program, linearization of the model in the random effects is effected by using the first-order Taylor series expansion with respect to the random effect variables r], and Cy. This software is the only program in which this type of linearization is used. The jth measurement in the ith subject of the population can be obtained from a variant of Eq. (5) as follows ... 

Some pharmacometricians may be involved in complex software projects, such as the development of software for ADMET (absorption, distribution, metabolism, excretion, and toxicity) predictions or software tools that can be used by other scientists. Examples of such tools include Perl-speaks-NONMEM (14) or Xpose (15). Such tasks often require a diverse set of programming skills and strong programming practices. 

The validation process is outlined from writing user requirements specification to testing and validating specific analysis using estimation methods. This is followed with brief examples of validation approaches for some commonly encountered software, such as S-Plus , SAS , WinNonlin , and NONMEM . 

NONMEM For the operational qualification, a careful review of the parameters discussed in Section 2.9 of the NONMEM Users Guide—Part III (17) should be performed. These values should be identified and set during the IQ and tested properly during the QQ. The specific examples provided for NQNMEM s PREDPP, NM-TRAN, and associated library subroutines are highly recommended as a starting point for the QQ. The Phenobarbital and Theophylline data files provided with the software (18) offer even more extensive testing appropriate (with modification) for a PQ. The output is well documented and individuals may seek to modify or parameterize the examples for their needs. 

The graphs and examples are geared toward NONMEM simply because NONMEM is the most widely used computer program for population PK/PD analysis. The principles, on the other hand, are quite general and should be easily adoptable for use with other software employing the same methodological strategy as NONMEM does. 

Maitre et al. (15) proposed an improvement on the traditional approach. The approach consists of using individual Bayesian posthoc PK or PK/PD parameters from a population modeling software such as NONMEM and plotting these parameter estimates against covariates to look for any possible model parameter covariate relationship. The individual model parameter estimates are obtained using a base model—a model without covariates. The covariates are in turn tested to determine individual significant covariate predictors, which are in turn used to form a full model. The final irreducible model is obtained by backward elimination. The drawback for this approach is the same as that for the traditional approach. 

To obtain the empirical estimates of a, Kowalski and Hutmacher (33) simulated 300 chnical trials for each combination of sample size and p, where the proportional reduction in CUP (0) was fixed to zero. Covariate and base models were fitted to each of the trials and the likelihood ratio tests were performed at the 5% level of significance. The percentage of trials where a statistically significant difference in CUP was observed provided an empirical estimate of a (i.e, PIoi = 0 is rejected when i/o is true). The data were analyzed with the NONMEM population phar-macokinetics/pharmacodynamics analysis software. The results suggested that an approximate nine-point change in the objective function should be used to assess statistical significance at the 5% level rather than the commonly used critical value of 3.84 for one degree of freedom. 

This ability is available in many software programs. NONMEM (Iconus, EUicott City, MD) has been widely used to estimate population models arising from both sparse and intensely sampled data. Other programs include WinNonMix (Pharsight Corp., Palo Alto, CA), Kinetica 2000 (Innaphase Corp, Philadelphia, PA), and Pop-Kinetics (SAAM Institute, Seattle, WA). ADAPT II and WinNonlin have focused on PK/PD models and have been combined with Bayesian approaches to estimate population models. 

The theory and techniques described in this chapter focus on the application of logistic regression to binary outcome data and the development of models to describe the relationship between binary endpoints and one or more explanatory variables (covariates). While many software options are available for fitting fixed or mixed effects logistic regression models, this chapter endeavors to illustrate the use of nonlinear mixed effects modeling to analyze binary endpoint data as implemented in the NONMEM software. 

Utilization of the Poisson and ZIP in population PK/PD modeling requires coding the appropriate distribution into the software selected for analysis. Example code will be given as appropriate for NONMEM implementation however, the fundamentals are applicable to other software programs. 

Parameter estimation without an appropriate assessment of reliabihty of the estimates yields no conhdence in such estimates. Estimation of uncertainty enables the use of such parameter estimates in data synthesis. Embarking on data synthesis (e.g., clinical trial simulation) using model parameter estimates without associated uncertainty or poorly dehned uncertainty will produce unreliable outcomes. Sometimes it is impossible to obtain standard errors for population model parameter estimates when small sample sizes are used for population PK/PD modeling. The bootstrap with winsorization has been proposed for the estimation of inestimable uncertainty—standard errors—for population PK/PD parameters that are usually not obtainable using software such as NONMEM because of small sample size... 

FIGURE 33.4 Schematic illustrating how NONMEM and a more generalized software program, like SAS (SAS Institute, Cary, NC) or S-Plus (Insightful Corp., Seattle, WA), can be used to interact and simulate clinical trials. 

Data were analyzed using the nonlinear mixed effects model software program NONMEM (Version 5 level 1.1 double precision (32)). Nelfinavir and M8 were fitted simultaneously. The molecular weight of nelfinavir and M8 is comparable with a ratio of M8 to nelhnavir of 1.028. Therefore, the concentrations were not corrected and are expressed in nanogram per milliliter. 

Perhaps the area where PopPK has made the largest impact is in drug development. Prior to the introduction of NONMEM as a commercial software package, there was little one could do with pharmacokinetic data collected from Phase 3 clinical studies beyond, perhaps, summary statistics and correlations between some summary measure of the pharmacokinetic data, like the mean or median concentration in a subject, and subject... 

---