Friday 9 December 2011

Help on: Acceptable Variation for HPLC Retention Time

MTS HELPDESK
Do you have any problems relating to analytical chemistry for pharmaceuticals or training? Send your questions to the MTS helpdesk using our contact form.

Question:
“I assess assay results for imported pharmaceutical products performed by a third party lab. The methods we use do not always specify an "allowed" retention time range. What would a reasonable guide be when assessing this – plus or minus 10%? Basically, what I should like to know is when should we ask the laboratory to investigate/explain a retention time shift from one assay to the next? Put another way, when should the laboratory make system adjustments like the temperature of the column, flow rate etc. in order for the retention time to conform? Often the method will say the active elutes at say 10 minutes. Does this mean that 9 to 11 minutes is also OK, or are there justifiable limits that we can impose?”

Answer:
“The reasons for variation of retention time from analysis to analysis includes small differences in the following: the composition of the mobile phase (in terms of the aqueous and organic solvent portions); the pH of the mobile phase (particularly important for ionisable molecules like acids and bases); HPLC columns (column to column variation), and HPLC systems (particularly the system volume). As a result of these variables it is expected that the retention time will vary and usually small variations do not cause any problems. The potential problems associated with retention time variation from run to run (not injection to injection in the same run, that is a more serious problem which may indicate a problem) includes:
  • The change in runtime means that a peak is not fully collected by the end of the run and thus the analysis has to be repeated.
  • For analysis of mixtures, such as impurity analysis, peaks may be difficult to assign if the retention time varies.
  • Automatic integration events may have problems relating to identifying peaks correctly, thus requiring extra time for reprocessing.
To answer your question, it is difficult to set an allowable retention time range without a good understanding of the individual method. Ideally the person who develops the method should investigate its robustness and thus gain information on the range of retention times which might be expected due to normal variation and where the method still performs as required. Of course, it is not an ideal world and often you will be presented with data from methods where this has not been done, or if it has, it is not written into the method. From experience I would say that about 10 to 20% of the time quoted in the method is a reasonable guide but be careful not to apply it too restrictively. If you have an analysis where the retention time is on the limits, the best way to assess if there is a problem is to look at the chromatograms for the analysis and compare to a previous satisfactory analysis (preferably an example chromatogram should be included in the method). If there aren’t any differences except for retention time, and the system suitability testing is all fine, then the results should be satisfactory to use. If the retention time is very different to that expected and well outside of this approximate range then I would suspect that something has been changed in the way the method was applied and would want the analyst to investigate further.

Although pharmacopoeia sometimes include adjustments to temperature, flow rate etc. to make sure that a peak is at a particular retention time these changes are actually deviations to the method and unless you are following a pharmacopoeia method or have validated your method to show that these alterations are valid then it is not a good idea to change the method parameters.”

1 comment:

  1. Dear Miss Oona McPolin,
    I appreciate very much your answers to the questions, they help me very much! Thanks in advance,

    Sincerely Yours
    Fatos Osmani
    Senior HPLC Analyst
    Profarma sh.a.
    Tirane Albania

    ReplyDelete