Fiona McGuinness 29 April 2015 Harmonising Formulation Analysis Procedures

Introduction  What is best practice?  Why harmonise?  What will be harmonised?  Harmonisation process  Discussion

Best Practice?  3 sites doing 3 separate procedures  White paper?  What parameters and acceptance criteria examined?

Why Harmonise?  One company, one process  Opportunity to look at processes and make changes or improvements  Discuss experiences to decide which way is the best way

What will be harmonised?  Parameters examined: Validation Sampling of diet formulations for stability assessment Homogeneity assessment Stability trial formulation size Standard and extract stability Pre-dose analysis Number of samples for routine analysis Genetic toxicology support

Method Accuracy and Precision  Site 1: 6 replicates of lowest and highest concentrations  Site 2: 3 replicates at lowest, intermediate and highest concentrations  Site 3: 2 replicates at lowest, intermediate and highest concentrations

Harmonised Approach  5 replicates at each lowest and highest levels

Diet Stability Trials  Site 1 Samples taken immediately after preparation into bags and stored ambient and frozen for specified time-points  Site 2 In drum, at ambient storage. Samples taken from drum at specified time-points  Site 3 Sample taken in glass jar and stored at ambient for 21 days

Proposed Change  Samples from drum at each time-point  Frozen sample taken on Day 0 and stored for 22 days (or more if time allows)

Not yet… working towards  Pre-dose analysis

Standard and Extract Stability  Site 1: Assessed, at least 4 days to allow re-analysis if necessary  Site 2: Not routinely done  Site 3: Not routinely done

Other validation parameters ParameterSite 1Site 2Site 3 LOD/LOQEstimated 3x BN and 10x BN Not calculated SpecificityControl vehicle extracted and examined for peaks LinearityAt least 5 standards over a 10-fold range R 2 ≥0.999 Back-fits within 10% At least 5 standards no fixed range R 2 ≥0.999 Back-fits within 20% At least 5 standards no fixed range R 2 ≥no criteria Back-fits no criteria System Precision6 injections of lowest and highest standards CV≤2% 6 injections of mid- concentration standard CV≤2% Not routinely done

How do you decide?  Highlight similarities?

Other validation parameters ParameterSite 1Site 2Site 3 LOD/LOQEstimated 3x BN and 10x BN Not calculated SpecificityControl vehicle extracted and examined for peaks LinearityAt least 5 standards over a 10-fold range R 2 ≥0.995 Back-fits within 10% At least 5 standards no fixed range R 2 ≥0.995 Back-fits within 20% At least 5 standards no fixed range R 2 ≥no criteria Back-fits no criteria System Precision6 injections of lowest and highest standards CV≤2% 6 injections of mid- concentration standard CV≤2% Not routinely done

Calibration Standard Acceptance  Over to you…..

 Any questions or further discussion points??