Steven A. Zdravkovic is a research scientist at PPD Laboratories, 8551 Research Way, Suite 90, Middleton, WI 53562. 

Steven A. Zdravkovic

An extractable/leachable (E/L) study is routinely performed on the materials comprising the manufacturing, packaging, and/or delivery system(s) of a pharmaceutical product to ensure they do not have a negative impact on the safety or efficacy of a drug. At its core, an E/L study involves the determination of a profile of substances that may be forcibly extracted (extractables) from a packaging material(s) or which may leach (leachables) from the material(s) and into the drug product during its manufacture, storage, and/or administration. These studies typically require the analysis of an array of matrices, ranging from polar aqueous solutions (e.g., aqueous drug formulations, aqueous buffers, water) to low-polarity organic solvents (e.g., ethyl acetate, hexane, dichloromethane). While some of these solvents are directly compatible with the analytical instrumentation employed within an E/L study, others require further processing prior to analysis. One such incompatibility occurs for the analysis of semi-volatile organic compounds in aqueous sample matrices by gas chromatography (GC), which is an analytical technique commonly used for both extractable and leachable studies. This lack of compatibility is a consequence of the high polarity and boiling point of water, which can cause issues such as peak distortion, excessive back flash in the inlet, or damage to the analytical column.

	To analyze aqueous sample matrices with GC instrumentation, the compounds of interest must first be transferred into a more volatile, less polar solvent. For E/L studies, this has historically been performed using liquid/liquid extraction (LLE) as evident by this technique’s use in E/L-related scientific publications and industry recommendations (1-6), as well as the author’s own experience performing these studies. LLE operates by combining the aqueous sample with an immiscible organic solvent. Compounds of interest in the aqueous sample phase are transferred into the organic solvent phase via a mixing step, provided they have a sufficient affinity for that phase. After phase separation, the organic portion can be analyzed directly or processed further as needed. Although LLE performs well in many circumstances, there are several drawbacks associated with its use, for example:

		LLE frequently requires multiple extractions of a single sample, pooling of the extracts, and evaporation of solvent to produce a sample for analysis. This makes the procedure labor-intensive.

		Problematic emulsions can be encountered if an amphiphilic substance is present, such as a surfactant, formulation excipient, or an active ingredient. These emulsions prevent phase separation that can render the technique unusable if not resolved.
		Polar excipients that may be found in aqueous drug product formulations can partially transfer into the organic phase. These excipients may produce chromatographic interference in the GC analysis.

		Concentration enrichment for the compounds of interest can be difficult. This issue is a consequence of the need to mix, separate, and evaporate large volumes of the sample and extraction solvent.

		The large volume of organic solvent that is frequently required for LLE precludes it from being considered an environmentally friendly technique.

	An alternative to LLE that does not share these negative attributes is solid-phase extraction (SPE). This technique operates by extracting compounds of interest from the aqueous matrix onto a solid sorbent that is then rinsed to remove interferences, dried to remove water, and eluted as a concentrated fraction of GC compatible solvent. This technique provides several advantages over LLE for the preparation of aqueous samples for GC analysis, as highlighted in the following:

		SPE is less labor intensive than LLE, a result of the extraction occurring in a single loading step without the need for mixing, phase separation, and solvent evaporation. Furthermore, SPE is easily automated using a wide range of commercially available systems that provide the option to further reduce hands-on labor.

		Emulsion formation is not an issue for SPE, as there is no mixing of immiscible liquid phases.

		Polar excipients in drug formulations can be separated from the compounds of interest via a wash step resulting in a purified sample for analysis.
		The concentration of the target compounds in a sample can be easily enriched by loading a large volume of sample onto the stationary phase followed by elution with a small volume of organic solvent.

		The minimal volume of organic solvent used by SPE makes it a more environmentally friendly technique than LLE.

	Because of these advantages, SPE has seen widespread use in areas such as environmental and biological sample analysis. However, despite its advantages, and its establishment as a routine technique in these and other areas, SPE has seen minimal usage for the preparation of samples within the E/L field. To help realize the potential that this technique has for use in E/L studies, the article will discuss the operating principles of SPE, situations within E/L studies where SPE may be advantageous, and studies that have been published using SPE for E/L work.

	SPE is a form of liquid chromatography and thus can be performed in normal phase, reverse phase, or ion-exchange retention mechanisms. However, unlike other chromatographic techniques that separate compounds into individual bands or responses, the goal of SPE is to completely retain the compounds of interest while other undesired substances are eluted and discarded. This on/off operation leads to the categorization of SPE as a form of “digital chromatography” (7). The retention, or lack thereof, of compounds within a sample is typically achieved via optimization of the separation mode. In the case of E/L studies, for which a vast majority of the compounds encountered are relatively hydrophobic in nature, reverse phase is typically used.

	An SPE method consists of four steps (see 

). The first step involves conditioning the stationary phase with a small volume (1-5 mL) of a water-miscible organic solvent, such as methanol, to wet the stationary phase, followed by an aqueous solution similar to the sample matrix. Once conditioned, the sample is flowed over the stationary phase. The volume of sample applied is determined based on the enrichment factor to be achieved, retention of the target compounds on the stationary phase, and complexity of the sample matrix. After loading, an optional wash step may be performed to separate interferences from the compounds of interest. To prevent water from entering the final sample, a drying step is typically performed by drawing air over the column. Once dried, a small volume (1-2 mL) of GC-compatible solvent can be applied to the stationary phase, collected, and used for analysis.

	Considerations for the use of SPE and performance characteristics

	The issue of retention must be considered when using SPE to ensure the compounds of interest are adequately recovered from the sample matrix. Retention is primarily a function of the mode of chromatography used, and as previously mentioned, reverse phase is typically employed for E/L studies. In reverse phase, a non-polar stationary phase adsorbs relatively non-polar compounds from a polar solvent. Thus, the sample matrix must be sufficiently polar so that it does not reduce the affinity of the compounds of interest for the stationary phase. In E/L studies, however, it is not uncommon to have an excipient or co-solvent in the sample matrix that reduces its polarity enough to potentially cause issues with recovery. In these circumstances, this incompatibility can be overcome by diluting the sample with water so the final solution loaded onto the stationary phase is sufficiently polar.

	Another issue related to retention is the desorption of a compound of interest from the stationary phase during sample loading. This issue, commonly known as breakthrough, is a function of the affinity of a compound for the stationary phase and the volume of sample being loaded. Effectively, breakthrough occurs when the capacity factor of a substance on the stationary phase for a given matrix is lower than the volume of sample being applied. In the case of a reverse-phase retention mechanism, polar compounds have a higher propensity for breakthrough than less polar compounds.

	A previously published study (8) assessed the use of SPE for the preparation of aqueous samples for analysis by gas chromatography in tandem with mass spectrometry (GC/MS), including an evaluation of recovery and breakthrough of 31 common E/L compounds covering a range of properties in these matrices. In this study, accuracy and precision were evaluated in water, 50/50 isopropanol/water, and 1% polysorbate 80 at concentrations of 20 ng/mL, 50 µg/mL, and 5 µg/mL, respectively. The results of this evaluation showed that this set of compounds produced adequate recovery (defined as being 70%-130% of the theoretical concentration) and adequate precision values (defined as having a percent relative standard deviation of ≤15%) for a majority of the compounds tested. This study also assessed breakthrough using water and 10/90 isopropanol/water as solvents. The results of this evaluation showed that most compounds showed no breakthrough for volumes up to, and likely exceeding, 100 mL of each solvent loaded onto the stationary phase.

	Use of SPE for the preparation of analytically challenging samples

	There are three characteristics of certain samples that can pose challenges for the execution of E/L studies. The first pertains to products that are administered in large volumes (≥ 0.1 L) over relatively short periods of time (≤ 1 dose/day). For these products, the large volume serves to dilute any extractable/leachable compounds that may be present without reducing the actual patient exposure to the compound. Thus, the levels at which these substances are required to be quantified to ensure adequate characterization can be analytically challenging. The second challenge pertains to products with complex formulations containing excipients, or the active ingredient itself, which need to be excluded from the final analytical sample. If not removed, these substances can produce interfering responses in the chromatographic data that may prevent the detection of any compounds of interest that may be present. The third challenge pertains to drug products that contain excipients, or amphiphilic active ingredients, which cause emulsion formation when prepared via liquid/liquid extraction. If the emulsion is not resolved, it is not possible to produce a sample for analysis.

	For each of these challenges, SPE can be used to overcome the issue and produce an acceptably pure and/or enriched sample for analysis by GC. The problem of quantifying compounds of interest in dilute samples can be solved by SPE’s ability to achieve high levels of sample enrichment. This benefit is a product of the theoretically limitless volume of sample that can be loaded into the stationary phase and the small volume of solvent needed to elute the compounds of interest for analysis. For example, SPE has been demonstrated to adequately handle volumes of 100 mL of aqueous sample with minimal loss of recovery. For such a loading scenario with a 2 mL-eluent volume, a fiftyfold concentration of the compounds of interest would be afforded.

The second problem can be resolved in cases where the interference in the sample has different properties than the compounds of interest. Polar interferences in an aqueous sample are not retained well in reverse-phase mode, and any residual amount can be eluted from the stationary phase during the wash step. In the case of many drug products, these interferences typically originate as excipients or other additives in the formulation, and include sugars, glycols, alcohols, and other compounds that are very polar and thus, would not be retained. This principal was demonstrated by processing a sample consisting of 20% glycerol and 20% sucrose using both SPE and LLE. In each case, 10 mL of sample was extracted and processed to produce a final 2 mL-aliquot for analysis. 

 presents an overlay of the chromatograms obtained from the GC analysis of each sample. The chromatograms clearly show the SPE procedure was able to exclude a majority of the glycerol and sucrose from the final sample, resulting in less chromatographic interference as compared with the LLE result. Additionally, the SPE procedure isolated several aromatic hydrocarbon impurities from the sample, which were not captured by the LLE procedure.

	SPE can also be used to overcome the problem of emulsion formation. This issue is typically encountered when a surfactant, such as polysorbate 80, is present in the drug product vehicle, or the active ingredient itself is amphiphilic enough to produce an emulsion. Because SPE uses a solid sorbent, it is not possible for emulsions to be formed. Thus, simply switching from LLE to SPE allows this concern to be remedied. Nevertheless, avoiding emulsion formation is pointless if the method cannot recover the peaks of interest in a sample with an appropriate level of accuracy and precision. To that end, work has been published demonstrating the ability of SPE to recover a wide range of compounds from polysorbate 80 solutions. As a first example, and as previously discussed, SPE was shown to adequately recover a range of 31 common extractables from a 1% polysorbate 80 solution spiked at 5 µg/mL. Not only was adequate recovery obtained, but a twofold enrichment factor was achieved. As a second example, SPE was used to recover a series of aliphatic ketones from a 5% polysorbate 80 solution (9). This work reported adequate accuracy and precision values down to a level of 10 ng/mL.

	In the E/L field, LLE is the preferred technique for the preparation of aqueous samples for analysis by GC. SPE, however, is another technique widely used in other fields of chemical analysis which can provide several advantages over LLE for E/L studies. These advantages include a reduction in time required to prepare a sample, elimination of emulsions encountered using LLE, purification of the analytical sample, enrichment of the concentration of target compounds, and a reduction in organic solvent consumption. Although SPE has not been used extensively in E/L, work has been performed to demonstrate its suitability and performance for this purpose, which in turn supports its immediate use in these studies.

. 67 (5), 448-511 (2013).
	2. J. Story et al., 

. 64 (2), 101-112 (2010).
	3. H. Kiminami et al., 

. 69 (6), 713-724 (2015).
	4. W. Ding et al., 

. 34 (8), 1-11 (2014).
	5. D. Jenke et al, 

. doi:10.5731/pdajpst.2016.007229 (2016).
	6. USP General Chapter <1664> Assessment of Drug Product Leachables Associated with Pharmaceutical Packaging/Delivery Systems (US Pharmacopeial Convention, Rockville, MD). 
	7. E.M. Thurman, M.S. Mills, Solid Phase Extraction: Principals and Practice (John Wiley and Sons, Hoboken, NJ, 1st ed., 1998).
	8. S.A. Zdravkovic, 

. 112, 126-138 (2015).
	9. S.A. Zdravkovic, 

	When referring to this article, please cite it as S. Zdravkovic, “Solid-Phase Extraction for the Preparation of Aqueous Sample Matrices for Gas Chromatographic Analysis in Extractable/Leachable Studies," 

Articles

Solid-phase extraction has several advantages over liquid/liquid extraction for extractables and leachables studies.
