Stability Flashcards
What statistics are used in stability testing?
Regression analysis and 95% confidence limit
How would you set up a stability protocol for a new tablet formulation, what conditions?
- ICHQ1 – Stability
- New product every 3 months first year , 6 months second year and annually thereafter
- If new product may want to do some accelerated studies for shelf life studies
- 30 degrees C 65RH and 40 75RH
You have miscellaneous peaks on a stability batch – what would you do?
Phase 1a – determination of obvious lab error analyst
Phase 1 b - determination of obvious lab error analyst and manager - Hypothesis Testing
Phase 2 – Manufacturing Investigation
Phase 3 – Review of both investigations and batch disposition decision.
Investigate any changes associated with the product
Review previous stability data / batch data to determine when the peak started to become visible – this should give you an idea of timeline for direct cause.
Notify DMRC that you are investigating a miscellaneous peak
Identify peak – what is it? Is it harmful to patients ? – Toxicologist / Medical
Quarantine batches in the warehouse, notify customer to hold stock if its not already with patients
involve the following in your risk assessment risk assessment:
Medical Department, Toxicologist , Pharmacovigilance, QC or Contract Lab to ID peak
You have a stability sample for your DP which has an extra peak in the chromatogram which is OOT (not yet OOS). What are your concerns?….
Talk through lab investigation
Talk about where else, other temperature conditions, accelerated batches initially. Was it there at time zero at a lower concentration (just a blip)
Talk through processing conditions being different (looking for deg products) -
Talk about contamination
API, Any changes? check technical / QAA, supplier management risk assessment
If there was a overcharge deviation they had not told us about the level of concern depends on whether it is solvent from our process being reused or solvent from someone elses product
If it is a single product facility, at the same stage then we might get over-reacted material and that would likely be in our deg products and we would have some tox data on that so could make a judgement. If it is from a different stage then more of a concern
If it was from someone else’s process in a multi product facility then that’s a whole different story as that could be cross contamination where there could be interactions and we will have no idea how toxic it is.
Would defiantly look into for cause audit looking at their PQS, QAA, deviation management, management of control of solvents, change over and cleaning processes etc
Potential OOS at 24 months stability for related substances. OSD for angina/blood pressure. What investigation would I do ?
Ask anything known about the related substances and spec upfront and if it’s a true result.
OOS Question Considerations:-
Raise Lab Investigation PQS
Place any batches on hold / keep samples in laboratory
Use MHRA OOS / OOT guidance and initiate phase 1a
Objective of phase 1a is to determine clear and obvious errors
If nothing identified move to phase 1b
Phase 1b is conducted with analyst and supervisor and is a more detailed review of analytical steps to determine what could have gone wrong.
Hypothesis testing is initiated here and moves to phase 2
Phase 2 manufacturing investigation initiated
If this is also inconclusive and there is scientific justification – retesting may be considered
Number of retests must be scientifically and statistically valid
Phase 3 – batch disposition decision.
How would /which statistical analysis would have potentially detected this / how should stability results be monitored?
Control chart looking for and special cause variation, shifts or trends (nelson rules)
How could I test for related substances, what else could I check?
The main purpose of a test for related substances is to control degradation impurities. An assay can be used to compare the sample to the standard. The limit test is can also be used to identify the impurities in the substance and compare it with standard. In general, limit test is defined as quantitative or semi quantitative test designed to identify and control small quantities of impurity which is likely to be present in the substance
What would you do if you had an OOT?
If it was out of trend as a QP I would want to understand why it was out of trend and what the implications were. I would look at previous timepoints to look for any signals and look to see if the data could be extrapolated to see when the batch is likely to become OOS.
What are required timelines for an OOT versus any investigation? Why?
Depends on findings of initial investigation, this is still only an OOT and all released batches have passed release testing however if initial investigations highlight a batch specific problem vs. potential problem that could be systemic / could impact all batches on the market.
If confirmed OOS and label claim is 3 years would I recall and what level of recall would it be? What if manufacturing manager said not to recall and/or no patient risk?
If systemic then what is the product, what is the supply picture like, is it a critical medicine, when is the batch predicted to go oos vs how long the product stays in the supply chain - all will feed into my decision. Lets say for example this is paracetamol and the market is flooded in generics this is a no brainer, no matter what the manufacturing manager says.
