Chromatography's Role in Spotting False Leachables

False positives are an underappreciated hazard in leachables testing, and they can be costly. Flag the wrong compound as a container-derived contaminant, and you risk derailing a regulatory submission over an artifact that was never actually a safety concern. This challenge becomes especially acute with protein-based biologics, where excipients like sugars and surfactants sit at concentrations dwarfing the trace-level leachables analysts are trying to detect, and where even how control solutions are stored can quietly seed false leads.

A recent case study examining the anti-T cell immunoreceptor with Ig and Immunoreceptor tyrosine-based inhibitory motif (ITIM) domains (TIGIT) antibody tiragolumab puts this problem into sharp relief: two unknown compounds turned up above the analytical evaluation threshold during stressed-sample testing, only to be traced back to excipient-related artifacts rather than true leachables. LCGC International spoke with Jie Du of Cardinal Health, author of the paper resulting from this study,¹ about what this investigation reveals about distinguishing genuine leachables from look-alikes, and what it means for how analytical scientists approach leachables studies in complex biologic formulations going forward.

Gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS) are both used for organic leachables screening. Walk me through how you would decide which technique to deploy first for an unknown leachable detected above the analytical evaluation threshold (AET), and what physicochemical properties of the analyte would drive that decision.

The choice between GC-MS and LC-MS is driven by the volatility and polarity of the analyte. LC-MS is primarily used to screen for non-volatile (polar) and some semi-volatile (moderately polar) extractables and leachables (E&L) compounds, while GC-MS is used for volatiles (non-polar) or semi-volatiles.

Typically, an unknown detected by one technique is first characterized by that same technique. In our study, two unknowns were detected during LC-MS screening, so structure elucidation was initially performed using LC-high resolution mass spectrometry (HRMS). Based on its retention behavior, we determined one unknown was moderately polar and suitable for further analysis by GC-MS, which ultimately aided in its identification.

The AET often demands trace or ultra-trace level sensitivity. How would you optimize an LC-MS method to achieve the required sensitivity in a protein-based drug matrix, and what figures of merit would you report to demonstrate that sensitivity is adequate?

Leachables screening falls into the category of nontargeted analysis (NTA), which makes analyte-specific optimization challenging. A non-discriminative chromatographic method (e.g., a linear gradient elution) and generic MS parameters (e.g., source and acquisition settings) for small molecules are often used to maximize compound coverage. While optimizing chromatography or MS parameters on a set of representative leachables may help, the most significant sensitivity gains come from sample preparation.

The key figure of merit for sensitivity is the limit of quantitation (LOQ). For an NTA method, defining the LOQ is challenging because there are no set target analytes. A common practice is to establish the LOQ using one or more "representative" leachable compounds. The LOQ must be at or below the AET, but it is important to recognize that the choice of representative compounds will significantly influence the result.

Drug excipients like amino acids, sugars, and surfactants can be orders of magnitude more concentrated than leachables. Describe the sample preparation strategies you would employ — chromatographically or otherwise — to reduce matrix interference before MS detection.

To manage complex biological matrices, we use techniques such as simple dilution, protein precipitation, liquid-liquid extraction (LLE), and solid phase extraction (SPE). Our paper describes an alternative extraction technique—salting-out assisted liquid-liquid extraction (SALLE)—which is particularly effective. These methods can reduce matrix interference and concentrate the analytes (e.g., via solvent evaporation), boosting sensitivity.

While chromatography for a general screening method cannot be over-optimized to avoid reducing compound coverage, we can employ various chromatographic strategies during the targeted identification of an unknown. For example, changing chromatographic conditions can separate an unknown from interfering excipients, facilitating accurate identification.

The study uses LC-HRMS for leachables identification. Explain the advantage of HRMS over unit-resolution MS in this context. How does accurate mass data help distinguish a true leachable from an excipient-related artifact?

The primary advantage of HRMS is its ability to provide accurate mass measurements for both the parent molecule and its fragments. This is crucial for determining a compound's molecular formula, which is the first step in structure elucidation. With unit-resolution MS, a single nominal mass can correspond to many possible elemental compositions, making definitive identification difficult.

Moreover, HRMS data are usually processed using a molecular feature extraction algorithm for the thorough extraction of even the smallest features (“peaks”). Screening at trace levels requires such data analysis tools.

Accurate mass data is helpful for distinguishing true leachables from excipient-related artifacts, as exemplified by Unknown 1 in our paper. The molecular formula determined by HRMS for Unknown 1 was found to be the sum of methionine (an excipient) and hexose (a logical degradant of another excipient). This was a strong clue that Unknown 1 was an artifact generated by the excipients rather than a true leachable. One can reach this conclusion even without performing a formal stress study of the excipients.

When two compounds co-elute in an LC-MS run and one is an excipient degradant masking a potential leachable, what chromatographic strategies would you use to resolve them? Consider both stationary phase chemistry and mobile phase optimization.

It should be noted that optimizing chromatography to resolve one compound might cause you to lose others in NTA. However, there are several techniques to resolve a targeted unknown leachable from an excipient degradant:

1. Change in stationary phase chemistry: This is often the most effective approach. As exemplified by the strategies used for Unknown 1 in our paper, switching from a reversed-phase (RP) C8 column to hydrophilic interaction liquid chromatography (HILIC) (zwitterionic and diol) columns increased the retention of polar compounds and allowed the separation of methionine and hexose, which had co-eluted on the RP column.
2. Mobile phase and gradient optimization: Adjusting mobile phase parameters like pH, organic modifier, and buffer concentration can alter selectivity. Additionally, using a shallower gradient increases the analyte's interaction with the stationary phase and leads to better resolution. The comparison of shallow and steep gradients in HILIC was demonstrated in our paper for Unknown 1.

The text highlights that frozen controls can lead to misassignment of excipient degradants as leachables. How would you design a chromatographic control strategy — including choice of reference standards, blank subtraction, and retention time locking — to distinguish true leachables from freeze-thaw artifacts?

The core of our control strategy was to prepare and analyze a placebo control. By comparing the compound profiles of the drug product to this placebo, we can effectively "subtract" any peaks originating from the excipients.

While reference standards and retention time locking are mainstays for targeted analysis, they are less common for broad NTA screening. However, once a specific leachable becomes a target for monitoring, those techniques can and should be applied in a similar fashion to how one would track drug impurities.

Chemical reactions between primary leachables and drug components can generate secondary leachables. How would you use chromatographic retention behavior and MS fragmentation patterns together to hypothesize the structure of a secondary leachable you have never seen before?

You start by linking the unknown to a drug component, primarily through its MS fragmentation patterns. You need a thorough understanding of the drug components and their potential degradants. Stable substructures, like the imidazole ring of histidine in our paper's Unknown 2, can appear as fragments in both the drug component and the secondary leachable.

You then piece together other clues: the compositional difference between the parent ion and its fragments, the assigned structures of new fragments, and logical chemical reaction mechanisms. Finally, the chromatographic behavior must be justifiable. For example, if a secondary leachable forms via esterification, it will likely be more non-polar and have a longer retention time on a reversed-phase column than its parent compounds.

Describe how you would validate an LC-MS leachables method for a monoclonal antibody formulation. Which validation parameters are most critical in this matrix, and how does protein precipitation or other sample prep affect your assessment of recovery and matrix effects?

A leachables study on the drug is preceded by an extractables study on the container. The extractables study and toxicological assessment (if applicable) determine if any compounds of interest require accurate quantitation. A leachables study involves two general tasks: the full quantitation of known target leachables and the non-targeted screening for unknowns. Each requires a different validation approach.

1. Quantitation of Targeted Leachables: For specific target leachables identified from an extractables study, a full method validation is performed according to ICH Q2 guidelines for quantitative impurity methods. This includes parameters such as specificity, linearity, precision, accuracy, and LOQ.
2. NTA for Unknowns: The primary goal here is identification, with semi-quantitation against surrogate standard(s). For NTA, we perform a method qualification, which typically evaluates specificity, sensitivity (LOQ), precision, and accuracy, but often with less stringent acceptance criteria.

In a complex biological matrix, accuracy and sensitivity (LOQ) are the most critical parameters. Sample preparation techniques like protein precipitation can inadvertently remove leachables, so recovery must be carefully assessed by spiking a cohort of representative leachables into the drug matrix. Achieving adequate recovery at the LOQ level can be particularly challenging.

Your paper notes that GC-MS was used alongside LC-HRMS for further identification. For what classes of leachable compounds would GC-MS provide superior characterization compared to LC-MS, and what derivatization approaches might you consider broadening GC-MS coverage of polar or thermally labile species?

GC-MS is superior for characterizing volatile and semi-volatile compounds, which are often less polar than compounds analyzed by LC-MS. These compounds may not ionize well in an ESI source, resulting in poor LC-MS signals. The electron ionization (EI) used in GC-MS also produces rich, reproducible fragmentation patterns that can be matched against large, established libraries like NIST, which greatly facilitates identification.

Derivatization methods are commonly used in GC-MS for extractables analysis to broaden GC-MS coverage of polar and thermally labile species. The most common method is silylation using BSTFA (with TMCS), in which a trimethylsilyl (TMS) replaces an active hydrogen (in -OH, -NH₂, -COOH groups). This makes the molecule less polar, more volatile, and more thermally stable. Other derivatization techniques include acylation and alkylation/esterification. However, derivatization is not feasible for leachables in aqueous biological samples, as water interferes with the reaction. An extraction step must be performed first, meaning extremely polar compounds that cannot be extracted are not candidates for this approach.

Two compounds detected above the AET were ultimately identified as excipient-related artifacts, a key finding of this study. Walk me through the chromatographic and spectral evidence you would systematically gather to build a defensible case that a compound flagged above the AET is a false positive rather than a true leachable from the container closure system.

Identification of a compound flagged above the AET generally starts with the collection of its HRMS and MS/MS spectra, followed by a careful review of the determined molecular formula and MS² fragments. Linkage to excipients or other drug components may be made at this step. For example, the determined molecular formula for Unknown 1 in our paper was C₁₁H₂₃NO₈S. The occurrence of both N and S in the molecular formulae of E&L compounds is less common. The source of S was linked to methionine, which was confirmed by the matching fragment m/z 148.0438.

When you suspect that drug components are involved in the formation of the unknown, the most critical step is to analyze a placebo control that has been incubated under the same conditions as the sample but has not contacted the container. If the unknown appears in both, it is not a leachable.

It should be noted that not all MS peaks represent real compounds in the sample. As we demonstrated with Unknown 1, adducts can complicate data interpretation. We proved it was a gas-phase adduct by using alternative chromatography to separate its constituent components, which made the artifact disappear.

In sum, every leachable result must be carefully reviewed with a healthy dose of skepticism to avoid reporting false positives.

Reference

1. Du, J.; Pearson, R.; Zhou, Y. et al. Investigation of Excipient-Related Artifacts Observed in the Leachables Study of Tiragolumab by LC-HRMS and GC-MS. J Pharm Biomed Anal. 2026, 278, 117545. DOI: 10.1016/j.jpba.2026.117545