How was the VADS System built and were there any assumptions that had to be made during the process?
The VADS System combines the following data to develop suggested antimicrobial regimens.
Pharmacokinetics of the antimicrobial
Drug concentration profiles related to multiple dose regimens
Pharmacodynamics of the antimicrobial
The optimum concentration profile for exposure of the pathogen to the antimicrobial
Susceptibility profile of the pathogen
Minimal inhibitory concentration (MIC) for the specific isolate, or
Consideration of the MIC distribution of multiple isolates of this pathogen from available datasets
It is not reasonable to conclude that combining these factors will allow precise regimen calculations for a specific pathogen MIC. However, the above data categories may be combined to give reasonable regimen targets and to eliminate unreasonable regimens.
During system development, many decisions have been necessary where clear direction is not available. For example:
Use of free, unbound drug concentrations in serum or plasma as opposed to total drug concentrations (protein-bound and unbound fractions combined)
Methods for extrapolation of pharmacokinetic trials with small numbers of animals (4-6 animals) to populations of animals
Determination of pharmacodynamic targets with minimal available data
Available data is often derived from in-vitro or animal-model studies in other animal species
Available data is often based on pathogens, diseases, and animal species other than the situation for which the model is being developed
Pharmacodynamic data is often developed using single or multiple dose strategies that do not reflect the serum or plasma profiles of extended release drugs used in veterinary medicine
Conflicting evidence and/or opinion as to the best pharmacodynamic target (e.g., 50% vs. 100% of the dosing interval above the MIC for -lactam antibiotics in the therapy of Gram negative organisms)
Proportion of the animal population for which the suggested regimen is adequate
Pharmacokinetic data is typically reported as mean ± variance.
The VADS System is constructed such that suggested regimens are modeled to address 95% of the animal population.
Extrapolating from one or several pharmacokinetic studies with 5-10 healthy animals apiece to an entire population of diseased animals is a concern. The reality is that these are the data available to work with. Population pharmacokinetic studies are needed in diseased animals to advance our understanding of the relationship between pharmacokinetic parameters, pathogen MICs, and clinical efficacy. These relationships may then be used to guide decisions in other diseases until clinical trial data is available.
To address these decisions in the absence of clear guidance, the VADS System collaborators have based decisions on the best evidence available. The methods for developing the suggested regimens and the data sources are completely transparent with provisions made for feedback on methods and regimens. Using this approach, it is hoped that the VADS System will function as a central point for discussion of modeling approaches and for identification of research needed to move forward in this area.
