The Impact of Physico-chemical Properties of Extractable Compounds on Their Analytical Responses: An Evaluation of GC/MS Response Data
Presented by Jan Baeten.
In the last few years, the Analytical Chemistry approaches to characterize extracts of Medical Devices - In line with the new ISO 10993-18:2020 Standard - has grown considerably. Biocompatibility assessments, based on (experimental) Chemical Characterization of Medical Devices is now one of the standard approaches for a number of toxicological end points. However, one of the aspects that is under constant debate is how to make a non-targeted analysis - to detect and identify a broad set of compounds - sufficiently protective or quantitative allowing a subsequent reliable toxicological risk analysis.
While authorities indicate that a number of surrogate standards (eg 3 for GC/MS, 5 for LC/MS) should be used with varying retention times or functional groups to assist in the quantification of the detected extractables, no little or no guidance is given on the selection criteria on which surrogate to select to quantify a detected compound.
In a first step to address this issue, Nelson Labs has evaluated over 2700 compounds of which the analysis of authentic standards provided relative response values in GC/MS. GC/MS was selected as a first action as it is often a more standardized chromatographic and MS- method and it is considered as the “work horse” in E/L testing. Conclusions in GC/MS could also lead to guidance on how to optimize the orthogonal and complementary non-targeted analysis methodology (combining different techniques and detectors) it low responding GC/MS compounds could be well defined.
The database of over 2700 compounds was evaluated against a number of physico-chemical properties, such as boiling point, molecular weight, retention time, LogKow… While a lot of these parameters show some impact on the response of a compound, there is one parameter that stands out and would allow either a differentiated approach in the use of responses for compounds detected in GC/MS or it would allow selecting surrogate standards in a more scientific way.