Unknown Impurity of Sertraline - Formamide?

Confirmatory Testing & Analytical Challenges
1

Dear Nitrosamine-Community,

I am currently investigating an analytical method for the NDSRI N-Nitroso Sertraline (monoisotopic mass: 334) using UHPLC-TQ-MS. This method was developed by an external laboratory. However, we observed a significant discrepancy in ion ratios between matrix spike samples and standards prepared in pure solvent, indicating the presence of an interference. All three measured transitions are shared by both the interference and N-Nitroso Sertraline, and no internal standard is being used.

By adjusting the LC parameters, we were able to partially separate the analyte peak from the interference and obtain full-scan spectra for both N-Nitroso Sertraline and the interfering compound (see figure below).

grafik
grafik1538×375 93.2 KB

Full-scan analysis revealed that the monoisotopic mass of the interfering compound is 333. Therefore, the observed peaks likely originate from the [M+1] isotope of the interference. Additionally, both N-Nitroso Sertraline and the interfering compound exhibit highly similar isotopic distribution patterns, further supporting the hypothesis that they share the same Sertraline core structure.

It has recently come to my attention that formamide-related impurities can occasionally form from the API. In such cases, the nitroso group (NO, mass: 30) could be replaced by a formamide group (CHO, mass: 29) on the secondary amine of Sertraline. This would be consistent with the observation that the interference has a monoisotopic mass one unit lower than the analyte.

To form a formamide impurity from the API, formic acid is required. While formic acid is absent from the sample preparation process, it is present in the LC solvents (Mobile Phase A: water + 0.1% FA; Mobile Phase B: methanol + 0.1% FA). Given the simplicity of the sample preparation, residual API may remain in the injected sample. My current theory is that, post-injection, residual API could react with formic acid in the LC solvent, generating a formamide impurity that appears as an interference in the chromatogram.

I would be very interested in hearing your thoughts on this hypothesis. Your insights would be greatly appreciated, especially since I will not have access to the TQ for further confirmatory measurements in the near future. My next steps include replacing formic acid with acetic acid and recording an MRM transition using a precursor mass of 334 while keeping the fragment masses unchanged, as well as testing alternative transitions.

Thank you very much, and best regards,
Andreas

2

Based on the data provided it seems that the impurity is probably N-formyl Sertraline.
This is a common degradation impurity in many formulations due to presence of formic acid and formaldehyde at trace levels in many excipients.
A casual google search will also provide you with this impurity.
I am not sure your problem will be solved by replacing formic acid. I believe you will have to chromatographically separate the peaks in the absence of LC HRMS

1 Like
3

Agree, Formyl are very commonly formed and usually are higher orders of magnitude than nitroso, so always they show the isotopic overlap . We have commonly seen in many product, but it can be challenging but these can be chromatographically separated.

3 Likes
4

Absolutely. Anirban has mentioned that the impurity forms at much greater concentration than the nitrosamine.
The level of this impurity will increase during the shelf life and this has to be accounted during the method development to avoid issues during the method life cycle.

1 Like
5

Hello, Andreas!
It is known that amines can interact with excipients (Maillard reaction) or some of the impurities contained in the excipients to generate API derived formyl impurities.
If by spiking the matrix you meant adding N-nitroso sertraline Std solution to a portion of the dosage form, it is probably about the formyl derivative formed during manufacturing and storage.
If this is formed post injection, the API alone will generate the same - was it also tested?
Also, apart changing the eluent, I would try to use a reconstituted composition of the dosage form: Excipients + API std soln, and then spiked with Nitroso-sertraline. It is not expected that formamide impurity is formed during the solution preparation.

1 Like
6

Dear all,

thank you very much for your replies and suggestions. Chromatographic separation seems to be most viable approach for us to deal with this impurity.

@mflorea Yes, by spiking I meant adding the analyte to the dosage form. The API was also tested by the external lab, however, we do not have access to any raw data and I do not know if they accounted for the impurity. I appreciate your input—I’ll be testing your suggestions by measuring different samples to identify the source of the impurity and determine at which stage it forms.

Best regards
Andreas

7

Probably it can be further confirmed by synthesizing N-Formyl Sertraline and spiking it in the sample solution. It is easy to synthesize N-Formyl Sertraline or it is easily available with many impurity vendors.
N-Nitroso Sertraline CAS: 3006789-98-3
N-Formyl Sertraline CAS: 147250-57-5.
We can verify the result by injecting Sertraline API as such in the same method to rule out the possibilty of interaction with execipient to form N-Formyl Sertraline.
Regards
Sachin

2 Likes
8

I had developed and validated the method of N-Nitroso Sertraline , but you can separate N-Nitroso Sertraline by changing in mobile phase compositions. I had separated base to base peak of N-Nitroso Sertraline with other impurities.

You can properly separate and Quantify N-Nitroso Sertraline in Drug substance and Drug products easily.

1 Like
9

Dear Viraj,

thank you for your input. Are you by chance able to share some specifics regarding the LC parameters?

Best regards,
Andreas

1 Like
10

Yes can you please tell me the make of your Instrument and its Model Number. Becaus it’s MS Specific molecule, which will give different ionization patterns with different makes of instruments. Some where we can see In-Source Fragmentation also.

1 Like
11

There are many manufacturers available which will provide N-Nitroso Sertraline. But based on Root of Synthesis you will observ 3-4 Peaks co-eluting with N-Nitroso Sertraline. As per my Thinking there might be the StereoIsomer also present of N-Nitroso Sertraline. We can say it’s a Racemic Mixture.

If we can see the three dimentional structure of the N-Nitroso Sertraline there is N-Methyl Nitroso Group and 1,2 Dichloro phenyl is placed in Dash-Away position.

IMG_0765
IMG_0765300×300 1.63 KB

1 Like
12

Unless using a chiral column, the enantiomers are not separated in Nitrosamine LC-MS analysis. However, asymmetrical nitrosamines can exist in two different stable forms, rotamers, distinguishable by several analytical methods - due to the partial double bond character of the N–N bond. Therefore, it is not unusual to see two chromatographic peaks for N-Nitrosamines, regardless of the existence or not of optical isomers

3 Likes
13

Hi Viraj,

sure! I am working with a UHPLC 1290 Infinity II (Agilent) coupled to a 6495D TQ (Agilent). Samples are ionized by APCI.
Thank you also for your input regarding isomers!

Best regards
Andreas

14

Hi mflorea,

thank you for your input! We did indeed observe double-peaks for a different nitrosamine standard, which are most likely related to your answer.

Best regards
Andreas