Applying non-targeted analysis workflows to the detection and identification of extractables and leachables in medical device polymers

Significance of the topic

The chemical characterization of extractables and leachables (E&L) from medical-device polymers is essential to assess potential toxicological risk to patients and to satisfy regulatory frameworks (e.g., ISO 10993 series). Non-targeted analytical workflows that combine broad chromatographic separation, high-resolution mass spectrometry and automated data processing are increasingly important to detect a wide chemical space of potential E&L components and to reduce time and variability in data interpretation.

Objectives and study overview

This study evaluated an integrated approach to E&L screening in medical-device polymers by combining: (i) complementary LC-MS/MS ionization methods (electrospray ionization, ESI, and soft ionization by chemical reaction in transfer, SICRIT), (ii) data-independent acquisition (DIA) high-resolution mass spectrometry, and (iii) automated non-targeted data processing using Insight Profiler software. The aim was to expand detectable chemical space, simplify data handling, and identify E&L species in extracts of four polymer materials used in sleep-apnea device components.

Materials and methods

Samples and extraction:

• Four polymer materials: mattress extract, silicone foam, polyesterurethane and polyetherurethane.
• Extractions performed in water or isopropanol (IPA), with five technical replicates per material/solvent.
• Comparative analysis against representative blanks; features required consistent detection in 4/5 replicates for downstream statistical consideration.

Analytical method:

• LC separation: Shim-pack XR-ODS III (2.0 × 150 mm, 2.2 μm), 40 °C, flow 0.4 mL/min, binary gradient (water + 0.1% formic acid and methanol), ~20 min cycle.
• HRMS acquisition (LCMS-9050 Q-TOF): TOF MS survey (m/z 70–1250, 100 ms); DIA-MS/MS with 41 windows (m/z 40–1250, 25 ms per scan) using 20–35 Da precursor isolation ranges and collision energy spread 5–55 V.
• Ionization: conventional ESI (300 °C; voltages +4.5 kV / −3.0 kV) and SICRIT soft ionization (plasma transfer ionization; temperature ~500 °C; voltage 1.8 kV; frequency 45 kHz) to broaden ionizable compound classes.

Used instrumentation

• Shimadzu LCMS-9050 Q-TOF mass spectrometer (TOF MS and DIA-MS/MS acquisition).
• Shim-pack XR-ODS III analytical column (2.0 × 150 mm, 2.2 μm).
• Electrospray ionization source (ESI) and SICRIT plasma-based soft ionization device coupled to LC inlet.

Data processing and software workflow

The LabSolutions Insight Profiler application was used to create a single automated processing cascade covering feature detection, chromatographic alignment, statistical filtering and compound identification across batches. Key elements:

• Feature detection and alignment to find ion signals behaving as chromatographic peaks across replicates.
• Statistical filters (including volcano-plot comparison of samples vs blanks) to highlight significant, reproducible features and remove high-variance ions.
• Compound identification against large screening lists and MS/MS libraries (ELSIE list via EPA CompTox, FDA-CLAP, NIST 2023), enabling batch annotation and interactive review without reprocessing.

Main results and discussion

• Combining ESI and SICRIT expanded the effective chemical space for LC-MS/MS. SICRIT improved detection of several siloxanes and other neutrals that were weak or absent in ESI data (e.g., decamethylcyclopentasiloxane, tetradecamethylcycloheptasiloxane).
• Multiple class representatives of E&L were detected across extracts, including antioxidants (Irganox series), surfactants (N-lauryldiethanolamine), neutralizing agents (N-butyldiethanolamine), plasticizers (phthalates, adipates), slip/anti-block agents (erucamide, stearamide), processing aids and oligomeric polyethers.
• Automated filtering criteria (consistent detection in ≥4/5 replicates, volcano-plot significance versus blanks) reduced candidate lists and prioritized reproducible, sample-specific features for library matching and manual review.
• MS/MS library matching in Insight Profiler identified many dominant E&L candidates; chromatograms and linked spectral matches facilitated interpretation and comparison between ionization modes without reprocessing.
• Use of SICRIT enabled simpler acquisition of additional compound classes by LC-MS/MS, potentially reducing reliance on GC-MS for certain semi-volatile neutrals, though orthogonal GC–MS remains relevant for volatile/SVOC classes.

Benefits and practical applications

• Integrated ESI + SICRIT acquisition increases detection coverage of E&L in polymer extracts while maintaining LC-MS/MS workflow simplicity.
• Insight Profiler’s single-method automated pipeline reduces analyst workload, improves reproducibility and accelerates screening across batches and experimental comparisons.
• Statistical filtering and library-driven identification streamline prioritization for toxicological assessment and reporting under ISO 10993 frameworks.
• The approach supports QA/QC and research laboratories performing routine or exploratory E&L surveillance for medical-device components.

Future trends and potential applications

• Further integration of alternate soft-ionization techniques and orthogonal separations (e.g., different LC chemistries, ion mobility) can continue to broaden detectable chemical classes and structural resolution.
• Improved, curated spectral libraries and in silico fragmentation prediction will strengthen confident identification of unknowns detected by non-targeted workflows.
• Standardized automated processing pipelines with built-in validation and reporting will be important for wider regulatory acceptance and more routine use in E&L risk assessment.
• Machine-learning assisted prioritization of features based on toxicological relevance and exposure likelihood could optimize resource allocation for follow-up identification and quantitation.

Conclusions

An automated non-targeted LC-DIA-MS/MS workflow combining ESI and SICRIT ionization modalities and processed with Insight Profiler provides an efficient, reproducible screening strategy for E&L in medical-device polymers. SICRIT extends LC-MS/MS detectability for certain neutral and semi-volatile constituents, while the single-method data-processing cascade simplifies feature detection, statistical filtering and library-based identification. The workflow is suited for research and screening applications but is currently provided as research-use-only and not validated for regulatory decision-making without further verification.

References

• ISO 10993-18: Chemical characterization of materials used in medical devices (part of the ISO 10993 series for biological evaluation).
• EPA CompTox Chemicals Dashboard — ELSIE chemical list (screening list used for E&L candidate matching).
• FDA CLAP list (Controlled List of Additives and Probable candidates for polymeric materials screening).
• NIST Mass Spectral Library (2023) used for MS/MS spectral matching.
• Shimadzu Corporation. LCMS-9050 Q-TOF mass spectrometer; Shim-pack XR-ODS III column specifications and method parameters as reported in the study.