HIL interferences on Acetaminophen Assays: Evaluation of Syva EMIT® and Microgenics DRI® and Roche Assays
Yan Zhang, University of Rochester Medical Center
Acetaminophen is one of the most popular non-prescription pain relievers with a therapeutic range between 10 and 20ug/dL. However, its toxic side effects for patients on chronic acetaminophen therapy and large dose injection on liver damage have been well documented. Accurate measurement of serum acetaminophen levels plays a vital role in the proper management of acetaminophen overdoses and, more importantly, prognosis and treatment with antidotes to avoid hepatic necrosis. We performed a study to evaluate the Syva EMIT and Microgenics DRI acetaminophen assays on Roches cobas c501 analyzer in comparison to the Roche acetaminophen assay as COBAS INTEGRA 800 system with a focus on the hemolytic, icteric, and lipemia interference effects on the assay performance.
Both Syva EMIT and Microgetic DRI acetaminophen assays are homogenous enzyme immunoassay techniques. The assays are based on the competition between the acetaminophen in the patients’ sera and the enzyme glucose-6-phosphate dehydrogenase (G6PDH) labeled acetaminophen with the acetaminophen-specific antibodies (at a fixed amount) in the ready-to-use reagents. The G6PDH enzyme activity is determined by spectrophotometric measurement at wavelength of 340nm by measuring the increased level of NADH converted from NAD in the reagents. The complex of acetaminophen-G6PDH conjugated with the drug specific antibodies inhibits the enzyme activity, which results in decreased absorbance at 340nm when no acetaminophen is present in the serum.
The user-defined parameters were added to cobas c501 analyzer for both assays. Imprecision was evaluated using three levels of controls purchased from BioRad ten times a day (within-day precision) or once a day for 25 days (between-day precision). Accuracy was examined by comparing acetaminophen levels from EMIT and DRI assays from 30 patient sera (acetaminophen level up to 360 ug/dL) to the results from COBAS INTEGRA 800 system. Linearity was calculated based on average results from 5 serum samples at acetaminophen levels of 0, 50, 100, 150, and 200ug/dL measured in duplicate and the assay performance at 240ug/dL was also tested. Limit of detection (LOD) and limit of quantification (LOQ) was performed by measuring a blank serum sample five times and serum sample at 1, 2, and 2.5ug/dL. LOD was determined by the mean plus three times standard deviation for the blank serum and LOQ was determined by the lowest acetaminophen concentration which gave no more than 10% coefficient of variation (CV). The investigation of hemolytic (H), icteric (I), and lipemia (L) interferences was carried out by spiking (10% total volume) concentrated hemolyzed sample, bilirubin, and intralipids into serum of acetaminophen levels at 5, 10, and 30ug/dL to achieve H/I/L indices up to 10000, 60, and 10000, respectively.
Our study indicated that the within-day precision for EMIT and DRI at low, medium, and high levels of controls were 5.3%, 2.5%, 4.5%, and 7.3%, 4.9%,5.4% CV, respectively. The between-day precision for both assays at the three levels of controls were 3.3%, 4.2%, 3.9%, and 7.5%, 5.9%, 6.8% CV, respectively. The correlation between EMIT (y) and INTEGRA (x) was y = 1.08 x + 0.94 (R2 = 0.9976) and the correlation between DRI (y) and INTEGRA (x) was y = 1.05 x + 2.51 (R2 = 0.9990) with the highest concentration at 360ug/dL measured on INTEGRA (Figure 1).
LODs for EMIT and DRI were 0.42 and 3.04ug/dL, while LOQs for both assays was 1.14 and 3.09ug/dL. The linear range for EMIT was tested up to 240ug/dL with the same regress parameters as up to 200ug/dL with slope of 0.99 and intercept of 2.71 (R2 = 0.9993). DRI assay was linear up to 200ug/dL with slope of 1.05 and intercept of 1.15 at (R2 = 0.9998).
No hemolytic or icteric interference was observed for EMIT and DRI assays at all levels of acetaminophen for H index up to 10000 and I index up to 60, although acetaminophen assay on INTERGRA was affected significantly at nearly all acetaminophen concentrations for H index as low as 100 (except acetaminophen = 5ug/dL at H = 100 wasn’t affected) (Figure 2). It didn’t appear to us that the Roche assay was affected by icteric condition until it reached to I = 20 at acetaminophen = 5ug/dL or I reached to about 30 and 50 at acetaminophen = 10 and 30ug/dL. The lipemia level slightly interfered with Roche acetaminophen assay with the higher L index producing higher assay results. To our surprise, however, both EMIT and DRI assays were affected by high L index in a negative fashion with more effect at higher acetaminophen levels.
In conclusion, the user-defined EMIT and DRI acetaminophen assays were evaluated on cobas c501 system and imprecision, accuracy, LOD, LOQ were established. The hemolytic, icteric, and lipemia interference analysis at acetaminophen 5, 10 and 30ug/dL indicated that these two assays were not influenced by H/I indices up to 10000 and 60 respectively, although they were more prone to lipemia interference especially at acetaminophen 10ug/dL and more so at acetaminophen 30ug/dL. In comparison, Roche INTEGRA acetaminophen assay was dramatically affected by hemolytic interference, and highly affected by bilirubin levels at I index beyond 20, 30, and 50 for acetaminophen levels at 5, 10, and 30 ug/dL, respectively. It, however, didn’t appear to be affected by L index up to 10000 on tested acetaminophen levels.