I recently used a serum creatinine assay kit from abcam (link below) and obtained some unexpected results. I was wondering if anyone here could share their experience with measurement of serum creatinine that could possibly shed light on what's going wrong:

http://www.abcam.com/Creatinine-Assay-Kit-ab65340.html

Serum samples taken from 2010 to 2011 from elderly patients with prostate cancer were stored at -80 degrees Celsius, previously aliquoted, hence at the time of assay at most two or three freeze thaw cycles would've been performed.

This assay is based on conversion of creatinine to sarcosine, which is then oxidised and reacts with a probe to produce red colour measured with a spectrophotometer set at 570nm. The reaction is performed using a reaction mix containing all supplied enzymes and then again without creatininase in the reaction mix to account for background. The concentration is determined from the difference between the sample reading and its background.

I ended up with measurements of serum creatinine around 10pg/microlitre whereas the normal range is 45 to 110pg/microlitre. I tested serum samples from three subjects, and all of them had around 10pg/uL. If I don't subtract background, the serum creatinine concentration would calculate to be at the lower end of the normal range, however the absorbance reading of the background is high enough that it can't be ignored. The serum samples had varying amounts of haemolysis, but the background reading should account for this. If anything, haemolysis should overestimate the concentration, not underestimate it. The concentration of 10pg/uL appears to be way below the lower limit of the normal range.

I did not dilute the serum samples.

The standard curve was linear and covered 0 to 200pg/uL, with absorbances from 0.0052 (water + reaction mix in the well) to 1.778, which suggests I didn't dilute the standards incorrectly either (given that our spectrophotometer measures from 0.000 to 4.000).

So far I can only think of three explanations, the first that the serum samples were diluted (I have confirmed that this is not the case), secondly that the 2 or 3 freeze-thaw cycles have caused significant degradation of creatinine (unlikely), and thirdly that perhaps there's something in the serum that is inhibiting the enzymes (unlikely, given that this assay is optimised for serum samples). What could possibly explain these unexpected results, and how may I be able to fix this problem?

If anyone could help, it would be much appreciated.