This is a post submitted by Paul Ambrose (with a few comments by yours truly).
Ambrose et. al. suggested to the FDA an others that there was an alternative pharmacometric-based method to using historical data to estimate the magnitude of antibiotic treatment effect. The pharmacometric-based method overcomes the numerous limitations inherent in the interpretation of historical data and is based upon contemporary data and clinical endpoints. Better yet, the methods would utilize authenticated clinical data that was part of recent NDAs. The method relies upon the construct exposure-response or pharmacokinetic-pharmacodynamic relationships, which are typically evaluated as part of modern-day drug development.
The upper and lower regions of exposure-response relationships for efficacy provide estimates of drug effect as exposure increases toward maximal-treatment effect or as exposure decreases toward no-treatment effect (Figure 1). The difference between the patient response rates at the upper and lower asymptotes closely approximates the treatment effect (i.e., the maximal benefit of therapy in a given patient population). Such an estimate provides the critical missing data needed to design non-inferiority studies with appropriate statistical power, which in turn obviates the need for superiority study designs. Moreover, these data make it possible to estimate the no-treatment response rate without exposing patients to any risk (mortality or mortality) incurred by conducting superiority studies that are placebo-controlled or which utilize suboptimal dose ranges or comparator regimens).
The upper and lower regions of exposure-response relationships for efficacy provide estimates of drug effect as exposure increases toward maximal-treatment effect or as exposure decreases toward no-treatment effect (Figure 1). The difference between the patient response rates at the upper and lower asymptotes closely approximates the treatment effect (i.e., the maximal benefit of therapy in a given patient population). Such an estimate provides the critical missing data needed to design non-inferiority studies with appropriate statistical power, which in turn obviates the need for superiority study designs. Moreover, these data make it possible to estimate the no-treatment response rate without exposing patients to any risk (mortality or mortality) incurred by conducting superiority studies that are placebo-controlled or which utilize suboptimal dose ranges or comparator regimens).
Comments from Antibiotics – The Perfect Storm - For those of you not in the know – Paul is simply proposing to take the various drug exposures that might occur across a large clinical trial and extrapolate them to zero as a way of estimating the “no treatment”effect. The treatment effect is then defined by the difference between that and, I suppose the mean or maximal drug exposure for the group of treated patients.
Paul and his collaborators have a number of other comments and critiques of the FDA’s stance. I refer you to their FDA Docket submission for more details.
I must say, I thought this was a pretty interesting read when it comes to this topic. Liked the material. . . . John
ReplyDelete.