Nitrosamine false positives: when the method makes the nitrosamine

Most impurity results answer one question: how much of this compound is in the sample?

A nitrosamine result has a second question in front of it. Was the compound in the sample, or did the analysis make it?

That sounds like an unusual worry. For nitrosamines it is a normal one, and the regulators say so. The European Medicines Agency’s nitrosamine Q&A lists “in situ formation of nitrosamines during analysis” as a known source of false results, next to contamination from gloves, membranes and solvents (EMA/409815/2020 Rev. 23, Q&A 8, p. 14).

The chemistry is the reason. A nitrosamine forms when an amine meets a nitrosating agent, usually nitrite. A drug product can carry both: the amine is often the drug itself or an impurity of it, and nitrite comes in with excipients and water. A sample preparation brings them together in solution, sometimes with acid or heat. That is close to the recipe for making the compound you are trying to measure.

This piece is about the measurement. Compliance strategy belongs to your regulatory team. It covers four ways the method makes the result, and the experiments that show whether a positive is real.

1. Nitrite in the diluent: the melatonin case

Yalamanchili and co-workers (2026) were developing an LC-MS/MS method for the nitrosamine of melatonin, N-nitroso-melatonin, in a melatonin drug product. Recoveries in the samples were inconsistent, while the standards behaved.

They suspected the nitrosamine was forming during sample preparation. So they tested it directly: they added a known amount of sodium nitrite during the preparation. In the normal preparation there was no nitrosamine peak. With nitrite added, the peak appeared, at 50 ppm (pp. 4–5).

Their conclusion was that “even trace levels of nitrite present in the diluent could induce in situ NDSRI formation.” The fix was to add a nitrite scavenger, 3,4-diaminotoluene, before the diluent, so the nitrite reacts with the scavenger and not with the drug (p. 5).

Two points from this case are worth keeping.

The first is the experiment. Spiking nitrite into the preparation is a simple test, and it gives a clear answer. If the result grows with added nitrite, your preparation can make the analyte.

The second is the symptom. The first sign was poor recovery and inconsistent peak areas, the kind of result that usually gets blamed on the instrument or the column. Melatonin is sold as a natural health product in Canada, so the case matters for NHP labs as well as pharma.

2. Heat: ranitidine and the GC-MS inlet

Ranitidine was withdrawn in 2020 after NDMA was found in it. The Saudi FDA’s review of the case describes how it differed from the sartan recalls (Aldawsari et al. 2021).

In the sartans, NDMA came from the synthesis of the drug substance. In ranitidine, NDMA was absent in recently manufactured batches and increased with time (p. 7). Higher storage temperatures made it worse.

And for the analyst, the key finding: ranitidine heated under accelerated conditions in a GC-MS instrument released NDMA (p. 7). A GC inlet runs hot. A method that heats the sample can create the NDMA it is measuring.

This is one reason FDA’s published methods for nitrosamines in drug products are LC with high-resolution MS. FDA’s own method for sartans explains that its earlier GC-MS methods could not see one of the nitrosamines, NMBA, which is why an LC-HRMS method was written (FDA, LC-HRMS method for six nitrosamines in ARB drugs, 2019, p. 1).

If your lab has a GC-MS method for a volatile nitrosamine, the question to ask is whether the matrix can produce the nitrosamine at inlet temperature. For some products the answer is no. Ranitidine showed that for some products it is yes.

3. DMF: the compound next to NDMA

Dimethylformamide is a common pharmaceutical solvent. It is also small, polar and close to NDMA in mass.

EMA is direct about the result: “false positives have been observed from DMF co-eluting with NDMA”, and the Q&A requires accurate mass techniques, MS/MS or high-resolution accurate mass, to identify a nitrosamine peak (Q&A 8, p. 14).

A 2025 study shows how common DMF is (Sibhat et al. 2025). The authors tested five commercial 500 mg metformin tablets. NDMA was at most 12 ng per tablet. DMF was found in every product, at 50 to 653.5 ng per tablet (sections 3.4 and abstract). On their single-quadrupole MS they read NDMA at m/z 75 and DMF at m/z 74, and they needed a pentafluorophenyl (PFP) column and a water and methanol gradient to separate the two peaks (sections 2.2.2, 3.1).

Two practical points. With a nominal-mass detector, a separation between DMF and NDMA is part of the method, and you need to show it. And their extraction solvent, dichloromethane, also interfered with the NDMA peak and had to be evaporated before injection (section 3.3). The solvent you extract with is part of the separation problem too.

4. Reagents and consumables

The fourth route is plain contamination. EMA lists nitrosamines in water, air, plastics and rubber, and cross-contamination from gloves, membranes and solvents (Q&A 8, p. 14).

FDA’s LC-HRMS method for sartans has a note that is easy to miss: avoid commercially available pre-made 0.1% formic acid in water and in methanol, “which may interfere with the detection of the analytes” (2019 method, p. 3). The method prepares its own. The same method uses glass centrifuge tubes and PVDF filters, and discards the first 1 mL through the filter.

None of this is unusual for trace analysis. What makes nitrosamines harder is the level.

Why the level makes all of this matter

Nitrosamine limits are set in nanograms per day. You turn them into a concentration by dividing by the maximum daily dose of the drug, so the target changes with every product.

For NDMA, the acceptable intake is 96 ng/day. FDA’s metformin method converts that to 0.038 ppm for the immediate-release product, at a maximum daily dose of 2,550 mg (FDA, LC-ESI-HRMS method for nitrosamines in metformin, 2020, p. 1).

The LOQ then decides what the result can be used for. EMA and Health Canada both use the same three steps (EMA Q&A 9, p. 15; Health Canada nitrosamine guidance, Q33):

• LOQ at or below the limit, for routine testing
• LOQ at or below 30% of the limit, to justify skip testing
• LOQ at or below 10% of the limit, to justify omitting the specification

FDA’s metformin method has an NDMA LOQ of 0.01 ppm (p. 1). By my arithmetic that is about 26% of the 0.038 ppm limit: enough for the 30% step, not for the 10% step.

At these levels, a small amount of nitrosamine made in the vial, or carried in on a reagent, is not a small error. It can be the whole result.

How to show a positive is real

This is the list I would work through before reporting a nitrosamine above its limit, or before trusting a method that has never shown one.

1. Spike nitrite into the sample preparation. If the result grows, the preparation can form the nitrosamine. This is the Yalamanchili experiment, and it is cheap.
2. Add a nitrite scavenger before the diluent, and confirm that the result becomes stable and recovery becomes consistent.
3. Look at heat and acid in the method. Sonication, evaporation, a hot GC inlet, an acidic diluent. Each one is a place where formation can happen.
4. Show the separation from DMF and any other known interferent, or use MS/MS or high-resolution accurate mass, as EMA requires.
5. Run reagent and procedural blanks with every batch, prepare acidified mobile phases in the lab, and watch for gloves and membranes as sources.
6. Confirm a positive with an orthogonal method. FDA’s metformin method was written for exactly this purpose: to confirm positives from its primary screening method (p. 1).

A false positive can trigger a recall investigation for a product that was fine. A method that forms the nitrosamine in some vials and not others can also hide a real one. The checks above protect against both.

Sources

Regulatory documents were checked on 2 October 2026. Limits and guidance change, so check the current version before you rely on a number.

• European Medicines Agency. Questions and answers for marketing authorisation holders/applicants on the CHMP Opinion for the Article 5(3) referral on nitrosamine impurities in human medicinal products. EMA/409815/2020 Rev. 23.
• Health Canada. Guidance on nitrosamine impurities in medications, 2026.
• US FDA, CDER Office of Testing and Research. LC-HRMS method for the determination of six nitrosamine impurities in ARB drugs, 21 May 2019.
• US FDA, CDER Office of Testing and Research. LC-ESI-HRMS method for the determination of nitrosamine impurities in metformin drug substance and drug product, 3 June 2020.
• Yalamanchili J, Himavathi G, Suresh S. A rapid and sensitive LC–MS/MS method for quantification of NDSRI in melatonin drug product via controlled in situ formation. Biomed Chromatogr 2026; 40(10):e70591. doi:10.1002/bmc.70591.
• Aldawsari FS, Alshehry YM, Alghamdi TS. N-nitrosodimethylamine (NDMA) contamination of ranitidine products: a review of recent findings. J Food Drug Anal 2021; 29(1):39–45. doi:10.38212/2224-6614.1133.
• Sibhat G et al. A comparative study in metformin tablet quality assessment: LC-MS and LC-MS/MS method quantification of N-nitroso-dimethylamine in the presence of dimethyl formamide. Int J Anal Chem 2025; 5625153. doi:10.1155/ianc/5625153. A small academic study; I use it for what it found in the tablets, not as a model method.

If your lab has a nitrosamine result it does not trust, send me the analyte, the matrix and the technique in one line on LinkedIn, and I will tell you which of these checks I would run first. The LC-MS troubleshooting cheatsheet is the short version for the instrument side.