Metformin Interference in LC-MS/MS Analysis of Plasma Methoxycatecholamines Marianne Bergmann (Presenter) Lillebaelt Hospital, Vejle >> POSTER (PDF) Presenter Bio: Biochemist at Lillebaelt Hospital, Vejle, Denmark. 4 years’ experience in developing and validating LC-MS/MS methods in a clinical laboratory. Master in pharmaceutical Sciences from Copenhagen University in 2006.

Authors: Marianne L. Bergmann and Anne Schmedes
Biochemistry and Immunology, Lillebaelt Hospital, Vejle, Denmark

The LC-MS/MS analysis for plasma methoxycatecholamines has been in routine use at Lillebaelt Hospital since October 2015. During the spring of 2017 we experienced an increasing problem with sample chromatograms showing lower peak height for metanephrine and d3-metanephrine, but not for normetanephrine and d3-normetanephrine. The problem only affects a few samples in each run, but these samples also show poor peak shape. We discovered that the problem was caused by metformin – a drug used to treat type 2 diabetes. Metformin co-elutes with metanephrine and causes major ion suppression. Because metformin is widely used and in high doses (up to 2000 mg/day), we set out to solve this problem.

Problem

Problem: The LC-MS/MS analysis for plasma methoxycatecholamines has been in routine use at Lillebaelt Hospital since October 2015. During the spring of 2017 we experienced an increasing problem with sample chromatograms showing lower peak height for metanephrine and d3-metanephrine, but not for normetanephrine and d3-normetanephrine. The problem only affects a few samples in each run, but these samples show poor peak shape and often double peaks. We suspected that it could be caused by an interfering compound in these samples that co-elutes with metanephrine, causing ions suppression. A full-scan of one of these samples displayed that indeed a massive peak of 130 m/z co-eluted with metanephrine.

Further analysis confirmed that the interferent is metformin – a drug used to treat type 2 diabetes.

Method Information

Method Information: Sample extraction is performed on a Hamilton STARlet workstation, using solid phase extraction on Biotage EVOLUTE EXPRESS WCX 30 mg 96-well plate.

The analysis is conducted on a Waters Acquity ultra performance liquid chromatograph (UPLC) with Xevo TQ-S tandem mass spectrometer operated in electrospray positive mode. Chromatographic separation is achieved with a Waters BEH amide (100 x 2.1 mm, 1.7 µm) column. The mobile phases consist of 2 mM ammonium acetate, 0.1% formic acid in water (mobile phase A) and acetonitrile (mobile phase B). 4 µL is injected at a flow rate of 0.7 mL/min. Gradient elution is as follows: 0-1 min. 10% A, 90% B ; 1- 1.5 min. linear gradient to 20% A, 80% B; 1.5-2.5 min. 90% A, 10% B; 2.5-5 min. equilibrate with 10% A, 90% B. The multiple reaction monitoring (MRM) transitions for normetanephrine, metanephrine and their respective internal standards are: metanephrine 180.12>148.12; d3-metanephrine 183.12>151.12; normetanephrine 166.08>134.09; d3-normetanephrine 169.08 >137.09.

Troubleshooting Steps

Troubleshooting Steps: At a glance metanephrine and metformin are not that similar molecules, and they have different pka values and therefore we thought that it should be possible to chromatographically separate them.

First we tried to get a better separation between metanephrine and metformin by changing the gradient, flow rate and mobile phases (pH and ion strength) in the LC-MS/MS method. However the peaks moved together and the best separation we were able to get was 0.04 min.

Then we turned to the extraction procedure. The Biotage EVOLUTE EXPRESS WCX 30 mg plate and a Phenomenex strata WCX extraction plate was tried with different wash steps (water, 10 mM ammonium acetate, methanol) and elution solutions (acidic, basic, different compositions of methanol, acetonitrile and water). Still, when metformin is washed away, metanephrine is also lost, and the other way around.

Next we have tried to use the fact that at very basic pH, metanephrine should be neutral while metformin should be positively charged. Thus on a strong cation exchange plate like Phenomenex Strata® SCX we should be able to elute metanephrine with a basic solution leaving metformin bound to the stationary phase.

Finally, we tried different HILIC and Amide columns, and were able to find a couple that separated metanephrine and metformin.

Outcome

Outcome:

The LC-MS/MS method has been changed so that metanephrine and metformin no longer co-elutes.