Pay now or Later? Early Method Development and Phase-Appropriate Validation-Considerations and Issues for Pharmaceutical Drug Development

By Harley Everett Wilcox, MBA

Senior Scientific Advisor

ABC Laboratories
www.abclabs.com

I do not know an explicit percentage, but I would guess roughly 30% - 50% of early purity methods for drug products have issues prior to phase II clinical.  The issue often observed include:

• New chromatographic peaks showing up in the drug product during stability

• Peaks migrating, or moving around over time

• System Suitability issues over time

Often, minimal efforts are put into early HPLC methods as the drug product may change and extra efforts would appear to be not useful and down-right costly.  In addition, time-lines are often tight. Additional method development efforts suggested for methods supporting phase one studies should be in concert with the established specifications and thus the issue.  Drug product specifications are required for release and stability monitoring and typical determined as the IND nears. If little is known about a new formulation, degradation profiles, and purity of lab/engineering, batches how does one determine purity specifications? Often, ICH guidelines or drug substance data are referenced for total impurities of NMT 2% and no unknown impurity greater than 0.2% represent typical specifications. OK, now we have specifications and we validate the early development methods as phase appropriate conducting minimal accuracy, linearity, precisions, stability of solution, LOD and LOQ.  We stay away from robustness, intermediate precisions, and complete specificity which would could be completed probably in a few weeks. Why do phase appropriate? Again, as things may change the cost would be sunk.

Often and unfortunately, either at release or during stability the analyst will encounter an investigation as the product does not meet specification because of:

• New peaks > unknowns specification

• New peaks, or higher than expected impurities from early manufacturing

Drug developers will exclaim ‘those unknowns are very likely excipient or excipients related’ and thus not a real OOS, or ‘I thought the method was validated’.  However, as the unknown peaks move around from run to run, and often have insufficient resolution from known active impurities, more work is necessary to confirm the unknowns as excipients or not. Typically, this requires evaluating placebo, placebo on stability, stressed excipients and possible some excipient drug compatibility studies. Gains made from the savings for early development/phase appropriateness are lost and unless more efforts are applied to understand the method, we may revisit costly investigations in the future.  The message—it is most cost effective to insure that methods and specifications are compatible and or conduct appropriate studies during method development to understand degradents and representative impurity levels.