Like the canaries used in coal mines to detect dangerous gases, FDA warning letters can be used to detect the latest focus of the agency. But why should discovery units care about FDA focus?
“Warning Letters” belong to the kind of mail most pharma companies could easily do without. Written by the FDA, addressed directly to the CEO, listing lots of critical issues, usually threatening dire consequences and on top being downloadable for everybody on the FDA homepage. On the positive side, like the canaries used in coal mines to detect dangerous gases, FDA warning letters can be used to detect the latest focus of the agency.
Currently, one of the hotter topics of the FDA is “data integrity”. What is behind this buzz word and why should the topic be at least considered by drug discovery units?
In the old times lab equipment created either some signals on a display or, luxury version, a plot on a slip of paper. All other information like experimental and equipment setup were written in a paper lab journal – end of data generation.
Thanks to computerised systems, this has changed dramatically. E-lab journals, electronic sample tracking or the multitude of instrument settings that may or may not have an effect on the experimental results and that may or may not be recorded make things a lot trickier.
“Record enough to justify results” is a minimalistic but common standard which has the drawback of frequently falling short once somebody has a closer look. Exactly that is what FDA did and that is how they developed their data integrity syndrome.
Applying a trace backward approach, starting from the recorded results and working towards the starting point of the process, they found that some drug manufactures did not have a complete trail of data available.
The end result was available but various bit and pieces on the way there were missing, opening at least the chance for unrecorded manipulations. Since the FDA considers something like this simply “adulteration”, data integrity became a big issue for them and for the industry as well.
Of course, FDA is unlikely to visit drug discovery, so why bother? Well, first of all, others may show a comparable interest in complete data trails, for instance a court of law dealing with patents. In any law suite focused on priority dates, a complete data trail can be extremely important.
Second, different data sets generated during a discovery process do not necessarily match in terms of evidence and conclusion. While this is to some degree the name of the beast, sooner or later the level of discrepancies may become hard to accept. When searching for the rational of discrepancies, experimental setup and over time changes thereof is always a good starting point.
If enough data is available of course. In case just the bare bones have been stored, the minimum data set to justify the results generated, it may be extremely hard if not impossible to dig out more information. Especially matching observed discrepancies with changes in experimental setup may be beyond the information content of the recorded data.
Take home message: recording more information than strictly necessary during a discovery process has a distinct upside potential and should be considered. Especially since cost and effort are usually affordable.
Take home message 2: ideas and concepts from the development and production part of pharma business may be worthwhile to consider for discovery as well.