Did you attend the recent webinar “Strategies for the Identification of Unknown Extractables and Leachables by LC-HRMS,” which I presented as part of Thermo Fisher’s Networking in Science Webinar Series? If so, you learned a lot about effective ways to evaluate and identify extractables and leachables (E&L) compounds in pharmaceutical packaging, single-use systems, drug delivery and medical devices.
Many attendees had questions – a number of which were answered live during the Q&A session. However, since we didn’t get to them all, the team at SGS wanted to share answers to the remaining questions here in the AnalyteGuru community so all can benefit.
The selection of an effective technique for the extraction of analytes that allows for their separation from interfering substances and matrices is a crucial part of any E&L study. Feasibility studies should be performed to experiment on sample preparation techniques to determine the suitability of a specific extraction procedure. A typical workflow for a liquid-liquid extraction study consists of trying solvents of different polarities (e.g., DCM, hexane, etc.) and adding internal standards of varied chemical structures at a predetermined concentration to assess the applicability of the sample preparation and solvent.
The library we use (mzCloud) is publicly accessible through Compound Discoverer. However, the software also allows for in-house library building.
The identification of unknowns in extractables is a crucial step for a successful leachable study. Although the ultimate goal is to identify all contaminants in the leachables, it’s easier to identify and quantify the extractables because the matrix is very simple (i.e., clean solvents used for extractable studies). In leachable studies, we are dealing with very complex matrices and sometimes, leachable contaminants get masked under the formulation peaks and so we have the risk of missing the contaminants in the noise. In this context, the identification of extractables is very important to guide the leachable study.
At SGS, we perform quantifications for E&L studies routinely. It can be done by semi-quantitation by surrogate standards or validated methods by authentic standards. The LC’s detection limit for an extractable study can be as low as 0.01 ppm. However, we usually try to work around 0.1 ppm levels, and so we manipulate our sample (i.e., preconcentration) to achieve this threshold.
While products with very low AET are challenging, there are certain strategies we can take in order to get reliable E&L data. Play with your sample prep (e.g., extraction technique, concentration factor, etc.). In addition, there is some flexibility in terms of detection limit. It is possible to achieve low instrumentation detection limits with simple changes in the instrument method or even the solvent used for samples and standards.
While it’s correct that a true leachable study cannot be performed for certain medical devices, at SGS we have performed many simulation studies that produce reliable leachable profiles. By choosing the right solvents (e.g., perspiration analog, pH adjusted solutions) and temperature, it is possible to recreate to a certain extent, the conditions to which a medical device will be subjected during its clinical use. In the case that the extractables detected from the chemical characterization of the device pass the toxicological risk assessment, a leachable study may not be necessary to perform.
Here are a few references I find useful:
It is true that sample preparation is a very time-consuming and almost artisanal step. However, I disagree with the comment about not being possible for complex formulations. Over the past few years, there have been many advances in sample prep. There are many techniques out there for the preconcentration and purification of E&L compounds, from liquid-liquid extraction to solid phase extraction. Don’t get discouraged — there is always a way! I recommend reading our review paper where we describe the latest advances and future developments in sample preparation [Singh, Gagandeep, et al. “Analytical challenges and recent advances in the identification and quantitation of extractables and leachables in pharmaceutical and medical products.” TrAC Trends in Analytical Chemistry 141 (2021): 116286].
Toxicity of the compound depends on the detected concentration and the route of administration. Although 2-MBT is a potential human carcinogen, the level can be controlled to be under the safety concern level. For this reason, it is of paramount importance that we have the tools to identify and quantify this and other potentially harmful chemicals in case they show up in an E&L analysis.
Screening analyses are very challenging because the same set of parameters are applied to all molecules in a complex matrix, from sample prep to instrumentation settings. We know that each molecule has its own “ideal” settings, but we have to make compromises (e.g., in terms of sensitivity), in order to have a useful and robust method for screening analysis. Our analytical methods are based on years of experience analyzing complex samples. In addition, we adjust them on a case-by-case basis depending on matrix and expected/target compounds.
For most applications we use ESI. However, APCI is also very useful, especially for those naughty molecules that produce a signal in the UV-Vis but do not on ESI MS. APCI is particularly useful for the identification of some small non-polar molecules.
In our experience, the LC grade solvents are worth the extra cost. The less interferences you can introduce to the system, the better. No need to spend time trying to understand where an unknown molecule comes from. Good quality solvents produce cleaner, easier to interpret data.
I personally have never used an ASE. However, its use has increased over the last few years. An ASE offers advantages like speeding of extraction experiments, as well as the possibility to reduce the number of devices required for extractions.
Extractable studies should be performed with a range of solvents of varying solvating power to ensure a representative pool of organic and inorganic extractable species are generated. For an IV bag with large volume parenteral products, we would choose two or three solvents, for example: acidified water (pH 3), alkaline water (pH 9), and ethanol water.
If the molecule has a signal in the UV, we use that detector for quantification because it has a lower relative response factor. Quantification by MS carries a lot of uncertainty, and we try to use it for identification purposes only. The adduct ratios could vary in different runs, especially sodium adduct.
I’ve never worked on separation of stereoisomers myself. But have read interesting applications in which an enantioselective column is used to separate enantiomers. It is important to keep in mind that MS cannot discriminate between stereoisomers though. So, you must be very careful when building your LC method if you are intending to separate stereoisomers. For structural isomers, the fragmentation pattern could be different, which can be captured by MS/MS spectra. The retention index could be used to differentiate the structural isomers as well.
Interested in hearing the original E&L webinar? Listen to the recording today.