B-277 Validation of an Enzymatic Method for Quantification of Ethylene Glycol in Human Serum

Background Ethylene glycol (EG) is a clear and odorless, liquid used in automotive and household products. EG is toxic when ingested requiring rapid and accurate detection for effective patient management. Having a method to measure EG quickly that exhibits good specificity and sensitivity, is key for clinical decision making. In this study we validated the Catachem enzymatic method for ethylene glycol quantification in human serum. Methods The assay was validated on Roche Cobas c502 analyzer. Accuracy and precision were performed by running EG spiked patient samples at three different levels (LV1 = 25 mg/dL, LV 2 = 100 mg/dL, LV 3 = 200 mg/dL) in replicates of twenty for intra-assay accuracy/precision and in duplicates over 10 days for Inter-assay accuracy/precision. Lower limit of quantification (LLOQ) and upper limit of quantification (ULOQ) was assigned by evaluating samples that passed at a cut-off of 20% coefficient of variation (CV). The assay was assessed for: analytical measuring range (AMR) verification, carry over studies, sample stability, freeze-thaw stability, interferences and specificity with various chemicals. A method comparison was performed by using thirteen patient samples from ARUP reference laboratory that utilizes the same enzymatic assay, in addition these specimens were also measured by gas chromatography-flame ionization detection (GC-FID) method. Results Assay accuracy and precision results are summarized in table 1 below. Intra-assay accuracy had %bias and %CV <10%. Inter-assay accuracy had %bias and %CV <20%. The LLOQ and ULOQ passed at 10 mg/dL and 300 mg/dL respectively. The assay was linear over the AMR (10 – 300 mg/dL), there was no carry over observed. Samples stability studies passed (i.e. <20% bias from initial concentration) for up to 7 days room temperature and 7 days refrigerated. Freeze-thaw stability studies passed with <10% bias observed for up to 6 freeze-thaw cycles. Specificity studies for the following substances passed with less than 10% bias observed at the following cut-offs: 2,3 butanediol (up to 1000 mg/L), 2-propanol (up to 1000 mg/dL), acetone (up to 1000 mg/dL), ethanol (up to 1000 mg/dL), glycerol (up to 5000 mg/dL), methanol (up to 1000 mg/dL), propylene glycol (up to 1250 mg/dL). To test the assay against random false positive results, we run 100 sera from randomly selected patients and all results yielded ethylene glycol concentrations of <10 mg/dL. Interference studies passed with < 10% bias observed for the following substances up to the specified cut-offs: lactate (up to 227 mg/dL), lactate dehydrogenase (up to 6266 U/L), bilirubin (up to 56 mg/dL), hemoglobin (up to 25 mg/dL), triglycerides (up to 484 mg/dL). A method comparison study using 15 patient samples between UCLA vs. ARUP laboratory that uses the same enzymatic methodology revealed a slope (m) of 1.1 with a regression coefficient (R2) of 0.996 and an average bias of 14%. Seven of the same were also measured against GC-FID method and that comparison results were: m = 0.917, R2 = 0.998 and an average bias of 24%. Conclusion We conclude that the catachem enzymatic method for ethylene glycol determination is suitable for patient testing.