TY - JOUR T1 - Is Retention Time and/or Structure Matching a Solution to the Challenge of Providing Quantitative Data when Screening for Extractables and Leachables by GC/MS? JF - PDA Journal of Pharmaceutical Science and Technology JO - PDA J Pharm Sci Technol DO - 10.5731/pdajpst.2021.012673 SP - pdajpst.2021.012673 AU - Dennis Jenke AU - Piet Christiaens AU - Philippe Verlinde AU - Jan Baeten AU - Jean-Marie Beusen Y1 - 2021/01/01 UR - http://journal.pda.org/content/early/2021/11/15/pdajpst.2021.012673.abstract N2 - Leachables in drug products or from medical devices could potentially and adversely affect patient safety. Thus, drug products and medical devices are chromatographically screened for organic leachables (and extractables), establishing these compounds′ identity and quantity. Accurate quantitation of extractables and leachables is challenging given their compound-to-compound variation in response factors. One proposed means for managing variation, and improving quantitation accuracy, is the use of retention time (RT) and structure to match analytes with their most relevant quantitation surrogate. Although the scientific basis for relationships between RT and structure versus response is unclear, the use of matching was investigated using databases of response factors (RF) or relative response factors (RRF), RT and structures for extractables/leachables. Gas chromatography with mass spectrometry detection (GC/MS) was investigated as response variation in this technique is less than with other screening methods such as liquid chromatography with MS detection (LC/MS). The overall RF variation across RT and structure makes it difficult to establish whether RT and response or structure and response can be correlated. Rigorous statistical analysis of the data concludes that there are no discernible relationships between these quantities; however, casual visual examination suggests that subtle relationships might exist. The effect that RT or structure matching could have on quantitation accuracy if the visual trends were real was considered. If the visual trends were real, it was estimated that RT matching could at most improve quantitation accuracy by 25% and that structure matching could improve accuracy by at most 50%. However, these improvements do not address the response variation that is independent of RT or structure and thus it is concluded that RT or structure matching are not viable solutions to RF variation. Rather, it is recommended that databases of authentic RFs be aggressively populated to provide accurate quantitation. Compounds for which authentic RFs cannot be secured are most effectively quantified using the median RF. ER -