Ultra-trace Quantitative Analysis of NDMA in Empagliflozin-Metformin ER tablets with Effective Resolution of DMF Interference Using LC/TQ

Ultra-trace Quantitative Analysis of NDMA in Empagliflozin–Metformin ER Tablets with Effective Resolution of DMF Interference Using LC/TQ — Summary

Importance of the topic

The detection of N-nitrosodimethylamine (NDMA), a probable human carcinogen, in metformin-containing products has triggered product recalls and regulatory scrutiny. Accurate quantification of ultra-trace NDMA in combination drug formulations is essential for patient safety and product release decisions. Analytical challenges include interference from residual solvents such as N,N-dimethylformamide (DMF), and sample matrix effects from extended‑release (ER) tablet excipients and high metformin content. This study addresses these challenges by developing a targeted LC–tandem MS method that resolves NDMA from DMF and provides sensitive, reproducible quantitation suitable for QC environments.

Objectives and study overview

The main objectives were to:

• Develop and validate a sensitive LC–MS/MS method to quantify NDMA in empagliflozin–metformin ER tablets at ultra‑trace levels.
• Achieve chromatographic and mass-spectrometric separation of NDMA from DMF to avoid false positives.
• Demonstrate method performance across multiple empagliflozin strengths formulated with 1 g metformin (5, 10, 12.5 and 25 mg empagliflozin).

Study approach combined targeted sample preparation to avoid gel formation, UHPLC separation, and dynamic MRM detection on a triple quadrupole mass spectrometer with APCI source.

Methodology

Sample preparation and overall strategy:

• ER tablets (empagliflozin + 1 g metformin) were dissolved in water and a controlled addition of methanol prior to final dilution prevented gelation commonly encountered with high metformin loads and enabled reproducible extraction of analytes.
• NDMA was evaluated by spiking at trace levels (notably at 0.048 ppm relative to tablet API) to assess recovery and precision.

Chromatography and detection (summary):

• UHPLC system: Agilent 1290 Infinity III UHPLC.
• Column: pentafluorophenyl (PFP) type, 3.0 mm × 150 mm, 2.7 µm, operated at 40 °C.
• Mobile phases: 0.1% formic acid in water (A) and 0.1% formic acid in methanol (B) with a multi-step gradient that begins highly aqueous and ramps to high organic (programmed gradient noted in experimental table).
• Injection: 5 µL at 10 °C sample cooler.
• Detection: Agilent 6495D triple quadrupole (LC/TQ) with APCI positive ionization; dynamic MRM acquisition to maximize sensitivity and selectivity.
• UV detection at 220 nm was used for API elution monitoring and a diverter valve routed flow to MS between 3.5 and 21 min.

MRM optimization and source conditions (high-level): APCI+ was selected (vapourizer 350 °C; drying gas 290 °C at 11 L/min; nebulizer 25 psi), and compound-specific precursor/product transitions, collision energies and dwell timing were optimized to separate NDMA (retention ~5.7 min) from DMF (retention ~6.0 min).

Used instrumentation

• UHPLC: Agilent 1290 Infinity III.
• Column: PFP, 3.0 × 150 mm, 2.7 µm, 40 °C.
• MS: Agilent 6495D triple quadrupole LC/TQ.
• Ion source: APCI, positive mode (vapourizer 350 °C, drying gas 290 °C, nebulizer 25 psi).
• Detection mode: Dynamic MRM; UV monitoring at 220 nm with flow diversion to MS between 3.5 and 21 min.

Main results and discussion

Chromatographic separation and selectivity:

• Effective chromatographic separation was achieved: APIs eluted early (~3.1 min by UV), NDMA at ~5.7 min (MRM), and DMF at ~6.0 min (MRM). This resolution prevents DMF from inflating NDMA signals and reduces false positives.

Analytical performance:

• Calibration: linear over 0.024–0.250 ppm for NDMA with excellent linearity (r2 = 0.9998).
• Sensitivity: method sensitivity to 0.024 ppm reported; reliable quantitation at 0.048 ppm (used as QC spike level).
• Precision: repeatability at 0.048 ppm showed %RSD ~4.67% (below 5%).
• Recovery: spiked recoveries for NDMA at 0.048 ppm in 10 mg metformin sample equivalents were 99–105% across four different empagliflozin strengths (S31–S34), demonstrating robust extraction and quantitation in the ER matrix.

Representative data summarized textually:

• Unspiked tablet samples were below LOQ for NDMA under the conditions used.
• Standard spiked sample area (single standard) ~66,247 counts; spiked sample average areas ranged ~65,643–69,429 across formulations, underpinning the reported recoveries near 100%.

Discussion points:

• Methanol addition prior to final dilution is a pragmatic sample-prep modification that prevents gelation, ensuring consistent analyte recovery from high-metformin ER matrices.
• Use of APCI ionization and targeted dynamic MRM transitions enhances specificity for volatile nitrosamines while tolerating residual organic solvent presence.
• Baseline chromatographic separation between NDMA and DMF is critical to avoid misattribution of DMF-derived signal to NDMA; the method design successfully addresses this.

Benefits and practical applications

This method offers several practical advantages for pharmaceutical quality control and safety testing:

• High sensitivity and low LOQ allow detection of NDMA at regulatory-relevant levels in complex ER tablet matrices.
• Robust separation of NDMA from DMF minimizes false positives that have previously driven unwarranted product recalls.
• Satisfactory accuracy and precision (recoveries ~99–105%, RSD <5%) make the workflow suitable for routine batch testing, stability assessments, and root-cause investigations.
• Use of a targeted LC/TQ approach with dynamic MRM provides rapid, instrument-efficient analysis compatible with high-throughput QC labs.

Future trends and potential uses

Potential developments and broader applications include:

• Implementation of isotopically labeled internal standards for NDMA to further improve quantitative accuracy and compensate for matrix effects.
• Automation of sample preparation to increase throughput and reproducibility for large-scale release testing.
• Extension of the workflow to detect and quantify other nitrosamine impurities (NDSRIs) and volatile contaminants across diverse dosage forms.
• Complementary use of high-resolution MS (HRMS) for confirmatory analysis and non-targeted screening in cases of ambiguous results.
• Method harmonization across regulatory bodies and industry to standardize LOQ/acceptance criteria for nitrosamines in pharmaceuticals.

Conclusions

The study presents a sensitive, selective and reproducible UHPLC–LC/TQ method for quantifying NDMA in empagliflozin–metformin ER tablets while resolving DMF interference. The method achieves linear response from 0.024–0.250 ppm, demonstrates reliable quantitation at 0.048 ppm with RSD <5%, and delivers recoveries between 99% and 105% in a 10 mg metformin equivalent. These performance characteristics indicate suitability for routine QC testing, helping prevent false-positive NDMA findings resulting from DMF co‑elution.

Reference

1. FDA Prescribing Information for metformin combination product, label document accessed 2016 (FDA drug label reference).
2. Article reporting DMF interference with nitrosamine analysis: PubMed entry, 2020; DOI: 10.1208/s12248-020-00473-w.
3. Study on sample-prep/internal standard approaches for metformin ER tablets: Organic Process Research & Development; DOI: 10.1021/acs.oprd.5c00308.