serum vs plasma - hormone ELISA - (Jan/04/2012 )

Hello,
Quick question - I have a set of plasma samples which we have access to to look at thyroid hormones using ELISA.
The kit specifies that serum should be used, not plasma.

I am not quite sure how / why plasma would interfere with the assay - any idea?
Is there a reasonable protocol out there to prepare plasma derived serum? I haven't come accross any and given that bloods were collected on anti-coagulant, this may not be feasible...

Hope someone can advise.

Many thanks
Ecatharina

who is the manufacturer and what is the product # of the kit that you intend to use?

Thanks for your reply John - the kits are competitive ELISA from Astra Biotech (for TSH, Tg, T3, T4, fT3, fT4 - sorry, I don't have #nbs right here).
Kits already purchased previously by tech, and plasmas are not from in-house study.

Ecatharina

Ben, Ben, sorry, not John!

I can't get the kit inserts on line for some reason, but it looks like these are licensed invitro diagnostic medical devices in Germany. I suspect that the matrix type used for the license applications was serum and the reason the inserts specify serum is because that's what the license calls for. I have seen procedures to make plasma clot by the addition of sufficient Ca++ or Mg++ to saturate the EDTA anticoagulant so that a clot forms but the matrix that you end up with will contain excess divalent cations and EDTA so is somewhat removed from actual serum, so this form of serum may give more problems than simply working with the matrix type you have. I suspect that the kits should work OK in plasma for research purposes. What you could do is obtain matched serum and plasma samples either commercially or from individuals in the lab (5 sets for example) and assess the matched pairs in the assays unspiked and spiked with exogenous analyte. If you don't see huge differences between the plasma and serum, then you are probably OK.

good luck!

Ben Ben

You may see offset between serum plasma matrices; curves may be in parallel with one another. Rather than draw 5 sets of individuals (and you would need samples in the low mid and high regions of the curves for all these products...other than at a hospital I don't know where you would find them)
you could prepare serial dilutions of the kit high calibrator in a normal plasma and with the kit 0 calibrator. If the curves are on top of one another then you are all set to go and analyze your samples. If the curves are different then you do have a matrix issue and will have to compensate for the difference.

Thanks for your replies - heparin was the anticoagulant, so the Ca++ flooding would not work (shame!)
I heard back from the company, who confirmed that they did not recommend plasma only because they did not validate the kit with that matrix, not due to actual interferences.
I have done pretty much what you suggested PAO - it looks good enough to go.

Otherwise - a quick thought - would the cryoprecipitation from thawing frozen plasma in the fridge pretty much remove any fibrinogen in the sample (thereby "correcting" the matrix effect once the precipitate is spun down and supernatant used)?

good question. I have no answer for you. If your samples were frozen then I suggest you thaw them and centrifuge to remove particulates before you assay.

FYI......the small molecule kits are competative the TSH and TBG are probably sandwich type assays.