Hypothesis to acceptance

February 1, 2009
Mike Butler
Pharmaceutical Technology Europe

Volume 21, Issue 2

During the last 20 years, pharmaceutical R&D spending has increased 15-fold while new product approvals have increased only 0.70-fold.

During the last 20 years, pharmaceutical R&D spending has increased 15-fold while new product approvals have increased only 0.70-fold. The number of new chemical entities making it to market remains low, with only 17 new approvals by FDA in 2007. While there are more opportunities to discover potentially drugable compounds, only approximately 20% of drugs that enter Phase I make it to market. The prospect of a continued rise in R&D expenditure suggests that new approaches are needed to reduce the current attrition rate.

Mike Butler

Obtaining data about a drug candidate before it enters Phase I can reduce the risk of failure by filtering out unsuitable candidates early on. Of particular importance are data on the pharmacokinetics of a drug because pharmacokinetics can often be related to pharmacodynamics. Although there have been some reports that drug failure caused by pharmacokinetics has been virtually eliminated, this seems prematurely optimistic in a number of respects. Drugs rarely fail for one reason and classifying the cause of attrition against a single component can be misleading. Knowledge of a drug's pharmacokinetics sits at the heart of effective selection: too little drug at the target will lead to lack of efficacy, while too much is likely to lead to toxicity.1 For these reasons, efforts are made to predict a drug's pharmacokinetic profile prior to Phase I clinical trials.

Currently, traditional in silico and in vitro modelling and allometric scaling tools are used to predict what the pharmacokinetic profile might look like in humans. Unfortunately, preclinical data obtained by allometric scaling from animals or in vitro or in silico modelling studies do not always reliably predict what will happen in humans.

Where preclinical studies may be unreliable, a Phase 0 trial in humans using a 'microdose' can be a solution. A microdose study uses a dose approximately 1/100th of that required to produce a pharmacological effect. Such microdose trials can inform decisions about drug candidates in humans before they enter a Phase I study. In particular, a drug candidate can be rejected if its Phase 0 pharmacokinetic profile is unfavourable. As these studies use very small doses, they are considered inherently safer than studies using the therapeutic dose. They are, therefore, less heavily regulated, quicker and cheaper to run than a Phase I study.

Microdosing studies are increasing in popularity as the evidence that results obtained at low drug doses predict those at the therapeutic dose has grown during the past 5 years. A recent review of related peer-reviewed literature shows that, of 18 drugs reported, 15 demonstrated pharmacokinetic parameters at a microdose within a factor of 2 compared with those observed at therapeutic doses.2 This difference in magnitude between a preliminary and therapeutic study is acceptable for allometric scaling studies and, therefore, the limited data currently published supports the use of microdosing studies.

The 18 drugs in this review had wide ranging pharmacokinetic and physiochemical properties with half-lives ranging 1–>40 h, bioavailability between 0 and 100%, and solubility from virtually insoluble to 170 g/L.2 Because low doses of drug are administered to human volunteers in microdose studies, the resulting plasma drug concentrations are also low, and very sensitive analytical techniques are required to reliably measure the presence of a drug and calculate its pharmacokinetic parameters. The most sensitive analytical technology is accelerator mass spectrometry (AMS), which can measure zeptomolar (10–21 ) drug concentrations.3

Microdose studies are moving from hypothesis to acceptance, but how can we be more confident of results produced by this technique? The answer: we need more data and, unfortunately, much of it is unpublished because of commercial sensitivities. However, more results are on the way, including those from the seven-drug European Microdosing AMS Partnership Programme (EUMAPP) completed in 2008 and published data expected later this year.4 Part of the EUMAPP programme was to incorporate human microdose data into physiologically-based pharmacokinetic models. Microdosing is moving onto the next level, with data being interpreted and modelled more intelligently, thereby helping drug discovery scientists.

Mike Butler is CEO at Xceleron (US/UK).


1. G. Lappin and L. Stevens, Expert Opin. Drug Metab. Toxicol., 4(8), 1021–1033 (2008).

2. G. Lappin and C. Garner, Expert Opin. Drug Metab. Toxicol., 4(12), 1499–1506 (2008).

3. M. Salehpour, G. Possnert and H. Bryhni, Anal. Chem., 80(10), 3515–3521 (2008).

4. European Microdosing AMS Partnership Programme. www.EUMAPP.com