Chromatography Forum

In our tox lab we analyze blood and urine for morphine, codeine, hydrocodone, hydromorphone, 6-MAM, oxycodone, and oxymorphone as TMS derivatives with nalorphine as an internal standard (qualitative detection only, no quants). We use SPE and GCMS. We recently cut the nalorphine concentration in urine down due to trace morphine showing up in the negative controls, and now occasionally the nalorphine drops out almost completely or at least to failing level. It is currently 500 ng/mL (50 uL spiking volume) using a 1 mL sample of urine.

The problem happens randomly regardless of analyst, PPM, turbovap, etc. and happens to controls as well as samples. We adjust the pH to 6 prior to adding to the SPE column. We have collected the sample load, methanol wash, and water wash and see no breakthrough. An extra aliquot of elution solvent (MeCl/IPA/NH4OH 78/20/2) shows no extra being retained after elution. We have been unable to force a failure by playing with flow rates. In controls we will see a large drop in nalorphine with little to no drop in other drugs, so it seems to be specific to nalorphine. It is very sporadic and the drop amount is inconsistent. We are being extra careful about spiking the IS. Can anyone shed some light on what might be happening? We are working on a new method with deuterated IS but need to fix this in the meantime. I can share more specifics about the method if needed.

If you lose nalorphine from a sample, and you run a replicate of the sample, do you see it ? This will at least show whether it is the samples or the analysis that causes the problem ? Losing it from some controls and not others suggests that the problem is in the analysis, unless your controls are variable in some way.

Peter

When we re-extract a sample the nalorphine will show up, so it is not sample related. We see the problem in positive and negative controls, so we are fairly certain it is method related. It also does not suddenly appear when re-injected, so it is not instrument related.

Having narrowed it down that far, can you take a double size sample and then split it at each step of the sample prep into two replicates - the step that gives poor replicability is where the trouble lies. The drawback with this approach is that if you have a low frequency of the problem you will have to do a lot of troubleshooting before it turns up again.

Maybe someone will ne able to offer more specific advice.

Thanks for the suggestion! I suppose anything is worth trying at this point, even if there is a chance we won't catch the problem.

Did you ever figure out the cause of your nalorphine loss?

Not yet, unfortunately. Our next troubleshooting run will spike nalorphine at different stages of the sample prep prior to SPE to see if we are losing it before even going onto the column. I had a theory that the strong 1M pH 5.0 acetate buffer used during hydrolysis was somehow interfereing, as we do not see this problem in the blood extraction which does not use this strong buffer. Not sure why this would only impact nalorphine, however. I'm hoping in the next week or two we will have more information (some other priorities came up).

For anyone following this, a thorough review of our extractions going back to the beginning of the year suggests the problem started when we switched from buying small bottles of NH4OH to large bottles (for elution solvent). The problem seems to have gone away since we started aliquoting the large bottles into small vials to limit the headspace. Not 100% sure this was the problem, but so far all the evidence suggests it. Again, only the urine method was affected, not blood.