Artifactually elevated BUN values on the Boehringer Mannheim-Hitachi 737 and 705 automated chemistry analysers and the development of a kinetic BUN method

The measurement of blood urea nitrogen (BUN) as serum or plasma urea is one of the most frequently ordered laboratory tests for evaluating renal functions. Conventional analysis of reactions of urea with diacetyl monoxime or urease with the Berthelot reaction [1] is rapidly giving way to the totally enzymatic reactions that are more convenient to perform and more easily adapted to modern instrumentation. However, the automated analysis ofthese reactions may yield falsely elevated BUN levels for some patients.


Introduction
The measurement of blood urea nitrogen (BUN) as serum or plasma urea is one of the most frequently ordered laboratory tests for evaluating renal functions. Conventional analysis of reactions of urea with diacetyl monoxime or urease with the Berthelot reaction [1] is rapidly giving way to the totally enzymatic reactions that are more convenient to perform and more easily adapted to modern instrumentation. However, the automated analysis ofthese reactions may yield falsely elevated BUN levels for some patients.   [2]. R2 contains urease, GLDH, and 2-oxogluarate in addition to an unspecified buffer. On both instruments, serum or calibrator is mixed with R1 and incubated for several minutes. At the end of this incubation period an initial bichromatic absorbance is measured as the difference between 376 nm and 660 nm or between 376 nm and 415 nm on the Hitachi 737 and 705, respectively. R2 is then added, mixed, and incubated for several additional minutes before a second bichro-matic absorbance measurement is taken. The BUN concentration is proportional to the difference between the first and second absorbance measurements after a reagent blank is subtracted. The method is based on a two-stage reaction in which urea is hydrolysed by urease to form ammonia and CO2. urease urea+ H20 + 2H +2NH4 / + CO.
In the presence of GLDH and NADH, NH4 + reacts with 2-oxoglutarate to form L-glutamate and NAD + GLDH NH4 + + 2-oxoglutarate + NADH L-glutamate + NAD + + HO.   specified an endpoint (ENDP), the manufacturer uses this nomenclature for two-point kinetic reactions. This code is used to take two separate absorbance readings after the addition of the second reagent, R2. Tables 3 and  4 show the new instrument settings for the Hitachi 737 and 705, respectively. Both R1 and R2 are stored at 4 C and are stable for at least 10 days.

Analyses by diffirent chemical methods
The BUN was measured on the DuPont aca, Technicon SMAC, Kodak Ektachem 700, and Beckman Astra. The results are summarized in table 5. The DuPont aca method is a two-step kinetic enzymatic assay employing urease and GLDH in which the sample is preincubated with all reagents except 2-oxoglutarate, which is added as a start reagent. On the SMAC, a diacetyl monoxime method employing a dialysis step is used. The Ektachem 700 uses an enzymatic method with urease in a dry chemistry slide. The Astra uses an enzymatic reaction with urease and measures the rate of change in conductance. The BUN values from the aca, SMAC, Ektachem, and Astra were similar and compatible with the patient's history, physical condition, and creatinine values.
Since both the Hitachi instruments and aca use similar enzymatic reactions, the main differences between the two systems are (1) the preincubation conditions; and (2) the end point versus kinetic method for measuring absorbance changes. Any type ofprecipitation that might occur during the preincubation period and then dissolve upon the addition of the second reagent would produce a falsely elevated BUN by the Hitachi method but not by the kinetic method or diacetyl monoxime method on the SMAC. To test this possibility, the patient's sample was mixed in a test-tube with appropriate proportions of the BMD reagent 1. A large amount of fine precipitation was observed forming as the serum contacted the reagent.
When BMD reagent 2 was added and mixed the precipitate completely disappeared. The exact cause of the precipitation in BMD reagent is unknown, since the composition of the R1 reagent is not given in the manufacturer's literature. It is known, however, that patients with abnormally high levels of monoclonal immunoglobulins are prone to precipitation [3]. A serum protein electrophoresis confirmed our suspicion: a monoclonal peak was observed in the gamma globulin region. Immunochemical analysis revealed in IgM level of 21 g/1.
In addition, the patient was found to have an abundant amount oflambda light chains in his urine. The patient's physician was notified and he is currently undergoing a full medical evaluation to determine the exact aetiology of his abnormal protein.

Linearity
Human serum pools containing very low and extremely high urea concentrations were mixed in known proportions and assayed in triplicate to determine the linearity of our method. It proved to be linear to at least 1000 mg/1 on both instruments.
Correlation with a kinetic enzymatic method--randomly selected serum samples were analysed in parallel on the Hitachi 737 and on the COBAS FARA (Roche Analytical Instruments Inc., Nutley, New Jersey 07110, USA) with enzymatic reagents from Reagent Applications, Inc. The correlation statistics are N 80, y 0"983 x + 10, r 0"999, and Sy.x 14. At a later time, parallel analyses were performed on the Hitachi 737 and 705 with randomly selected serum samples using the proposed enzymatic rate method on both instruments. For N 31, y 0"964 x 58, r 0"999, and Sy.x 39 where units are in mg/1.

Discussion
Two additional patients tiave been found who produce falsely elevated BUN values due to precipitation in R1 when analysed on the Hitachi 737 and 705 with BMD reagents. It was interesting to note that each patient had a different form of multiple myeloma; one patient had an elevated IgG and the other an elevated IgA. The marked discrepancies between BUN values on the Hitachi 737 and 705 were puzzling at first because both instruments use similar reagents. It was discovered, however, that this discrepancy was due to a difference in bichromatic absorbance measurements made on the Hitachi 737 (376 nm and 660 nm) and on the Hitachi 705 (376 nm and 415 nm). Because a fine precipitate produces significantly less absorbance at 660 nm than at 415 nm, the difference in absorbance between 376 and 660 nm was much greater than that observed between 376 and 415 nm.
In all three cases studied, normal BUN values were obtained on both instruments by using the proposed method. The method has been in use for over six months without any problems. An additional advantage to the new method is the improved day-to-day precision compared to the original method using BMD reagents.