Chromatography Forum

Hi there,

I'm working on an LC/MS/MS method to quantitate a drug and two metabolites in urine. I've worked out a separation method and optimal MS/MS parameters but without using urine so far. I have also done some tests for matrix effects using diluted Surine (synthetic urine) as a matrix blank, and I've determined that my first analyte elutes more than a full minute after the last matrix effect vanishes. So far, so good.

Because of this 1-minute gap I can set up a divert method, so that I can shunt the unretained matrix components to waste. Also good.

So here's the question - suppose I'm doing dilute-and-shoot of these urine samples, 1 part in 100, 10 µL injections onto a C8 column. Is the column going to get "gummed up" in some way after a certain number of injections? Is there any routine column cleanup procedure I should run periodically to keep it healthy?

Alright, so the detection limit we can get from a 1:100 dilution is about 100 times too high for our needs. So suppose I reduce the dilution... 1 in 10, injecting 10 µL. Or even 10 µL straight urine. I can still protect the MS using the divert valve... but am I going to ruin the column eventually?

Of course an alternative would be to do sample prep - and if it's the only feasible option that's what I'll do.

Human urine is not that big of a mess, and you may get good column life without any extra work. I would try that first. If column lifetime is too short, I would go for a simple SPE sample cleanup. I can advise you on suitable methods.

Human urine is not that big of a mess, and you may get good column life without any extra work. I would try that first. If column lifetime is too short, I would go for a simple SPE sample cleanup. I can advise you on suitable methods.

Thanks Uwe, I would appreciate any input. There is a published SPE method for the parent compound and I was thinking I would start there if needed. However the paper in question did not analyze the metabolites, so I will have to verify that the metabolites survive too.

In the literature method, the cartridge is preconditioned with 3 mL methanol and then a mixture of 5 mL urine + 2 mL PBS buffer (pH=7) is added. After a 1 mL PBS wash, the cartridge is dried with 20 min air via vacuum pump, and eluted with ethyl acetate. The eluate is dried under nitrogen and reconstituted in 0.2 mL mobile phase.

Just to add, the study will involve urine from 100-200 subjects, so that is to be considered when thinking of column lifetime.

I can't see the rational for this SPE procedure. At best, it is the equivalent of an extraction with ethylacetate, which opens the same question about the survival of the metabolites.

SPE methods need to be designed for what you want to accomplish. Either you give me some information about the nature of the analyte(s) that you are after, and what you want to get rid of, and we can work out a procedure together. If necessary, you can contact me at my e-mail below. Or you can look up procedures in the Oasis Applications booklet (available online from Waters), and select a proven procedure that looks OK for your purpose.

I bet it depends on the nature of the patients. I imagine patients with kidney issues will clog up a column faster than healthy patients.

A better option is to use Cadenza HS-C18 for direct inj. of urine. This column will exclude proteins and retain small molecules.

Below is LC-MS/MS non-targeted metabolomics application using Cadenza HS-C18:
http://www.imtaktusa.com/site_media/fil ... TI493E.pdf

It's beneficial to avoid SPE if you're trying to find that needle in the haystack.

I bet it depends on the nature of the patients. I imagine patients with kidney issues will clog up a column faster than healthy patients.

A better option is to use Cadenza HS-C18 for direct inj. of urine. This column will exclude proteins and retain small molecules.

Below is LC-MS/MS non-targeted metabolomics application using Cadenza HS-C18:
http://www.imtaktusa.com/site_media/fil ... TI493E.pdf

It's beneficial to avoid SPE if you're trying to find that needle in the haystack.

Hi Bryan - we know exactly the compounds we're looking for, we have standards, and we have MRM transitions for each. So it's not exactly a needle in the haystack.

To clarify, this is not a metabolomics project, nor is it non-targeted. We are screening urine samples of health care workers who have occupational exposure to a specific drug, as part of a wider effort to determine safe exposure limits (if any). There may of course be some subjects in poorer health than others but these are volunteers, not patients.

We're also on a somewhat limited budget so if we can accomplish what we need on the C8 or C18 column we have already... it would be much preferred by my higher-ups

Yes, I understand. Just wanted to share data where non-diluted urine has been injected into LC-MS.

Assuming dilute and shoot is not an option (due to sensitivity), you have 3 options:
a). inj. urine onto Cadenza HS-C18
b). sample pretreatment
c). inj. urine onto your current C8

a & b will result in long column lifetime. No clue how many
injections you'll get out of your column if you go with option c.

Hello ctroster,

I am developing a "dilute and shoot" method for benzodiazepines in urine and am trying to choose the best column. I was just wondering if you tried your method on the C8 column and if so, how is it holding up?

Thanks.

I don't see why one column would hold up better than another. If one column gets contaminated, any other column will get contaminated as well. If the column gets contaminated too early, you need to either remove some contaminants from the sample, or use a guard column.

Guard columns are great! Most are dirt cheap. I'm using Phenomenex, where (in the UK) I pay £19 per guard.

At this price it is cheaper to change a guard column every 20 runs than to use a good quality microcentrifuge-tube based spin filter! Not that I recommend blocking guards that often.

Hello ctroster,

I am developing a "dilute and shoot" method for benzodiazepines in urine and am trying to choose the best column. I was just wondering if you tried your method on the C8 column and if so, how is it holding up?

Thanks.

I haven't even run urine samples yet. Turns out in the end, the method I tried worked well enough to give me more than enough sensitivity needed for 1 of 3 compounds. But for the other two - metabolites of the first, expected to be present in lower concentrations - there was little improvement. Therefore I will need to do sample prep if I'm going to see all 3.

I don't see why one column would hold up better than another. If one column gets contaminated, any other column will get contaminated as well. If the column gets contaminated too early, you need to either remove some contaminants from the sample, or use a guard column.

Because proteins elute at the beginning of the run when using this column.
Proteins stick to (or coagulate inside of) conventional ODS.

So if:
- proteins stick to conventional ODS (and presumably C8 too)
- my guard column always sees the sample first
- urine samples are largely from healthy subjects (hospital workers with presumably an average rate of sickness, rather than hospital patients) thus there shouldn't be that many samples with large protein concentrations
- I'm using SPE too (turns out I will have to anyway)

Then I can be reasonably assured my main column won't get gummed up with proteins?

Yes of course. You should achieve long column lifetime with this cleanup procedure.