of creatinine by the Jaffe procedure is well documented. We describe a procedure for the elimination of interfering bilirubin by oxidation to biliverdin through the ...
CLIN. CHEM.
38/9, 1749-1751
(1992)
A Simplified Procedure for Eliminating the Negative Interference of Bilirubinin the Jaff#{233} Reaction for Creatinine N. O’Leary,’
A. Pembroke,2
and P. F. Duggan’
The negative interference by bilirubin in the measurement
of creatinine by the Jaffe procedure is well documented. We describe a procedure for the elimination of interfering bilirubin by oxidation to biliverdin through the addition of potassium femcyanide before the alkaline picrate reagent. We detected no negative interference with the addition of as much as 600 zmol/L bilirubin. The procedure shows good agreement with a recently described blank rate method for analyzing icteric samples (Clin Chem 1991 ;37:1460-1).
when sealed from the atmosphere; picric acid reagent, 25 mmol/L, and 30% Brij 35 surfactant, 1 mLIL, were stable for 3 months at room temperature. Combined picrate solution-5:1 (by vol) alkaline reagentlpicric acid reagents-was stable for at least 5 days at 4#{176}C. Potassium ferricyanide solutions were made up in 154 mmol/L sodium chloride and were stable for at least 3 days at 4#{176}C. indefinitely
Methods
Aliquots of 100 mmol/L bilirubin dissolved in 100 interference due to bilirubin in the Jaff#{233}mmol/L NaOH were added to a plasma pool. The pH of procedure for serum and (or) plasma creatinine is well the pool did not change by more than 0.1 pH units, and documented (1-4). This interference is evident with the addition of an equivalent amount of NaOH to the both conjugated and unconjugated biirubin, is indepenplasma pool had no detectable effect on the assay. The dent of creatinine concentration, and is directly related dilutional effect of the added bilirubin was compensated to bilirubin concentration (5). for by appropriate calculation. Many approaches have been adopted to minimize this Three assay procedures were used with the Hitachi interference. A commercial reagent has been shown to 704 (Tokyo, Japan). be effective for the measurement of icteric samples (5), Rate method. Twenty microliters of plasma was mixed but it requires initial blank monitoring at 10 a, which with 350 ML of alkaline reagent for 5 mm, after which limits its application. The reagent also contains sodium 70 MLof picrate reagent was added (creatinine kit, cat. dodecyl sulfate, which tends to precipitate when reno. 816426; Boehringer, Mannheim, FRG). The rate of agents are stored at lower temperatures. High alkaline bichromatic absorbance change (505 nm analytical picrate concentrations have also been shown to be effecwavelength, 570 nm blanking wavelength) was meative, but the reaction time is quite short and requires sured 60 s after picrate was added for a further 60 a. continuous measurement (5). Potassium ferricyanide Blank rate method. Twenty microliters of sample was also has been incorporated into the Jaff#{233} procedure (6), incubated with 350 L of alkaline reagent (9). After 4 but is unstable in the working reagent (7) and needs to min, the rate of absorbance decline (505 nm analytical be added to the reagent at the time of analysis. More wavelength, 570 nm blanking wavelength), which is recently, blank rate procedures involving subtraction of linear with time, was measured for 60s. Seventy microthe rate of decline in absorbance of the ictenc sample in liters of picrate reagent was added, and after 60 s the the alkaline reagent from the increasing absorbance in rate of absorbance change was measured for a further 60 the mixed picrate reagent have been used with Hitachi s. Concentration was calculated by subtracting the rate analyzers (8,9). of absorbance decline in the alkaline reagent from the In this study we demonstrate the effectiveness of rate of absorbance increase after picrate was added. direct addition of potassium ferricyanide to patients’ Fern cyanide (combined reagent) method. Twenty miplasma samples to effect rapid oxidation of bilirubin to croliters of plasma was mixed with 20 zL of ferricya.. biliverdin, thereby removing interference in the Jaffe nide, Reagent 1 (2 mmol/L). After 5 miii, 400 1zL of procedure. combined picrate reagent, Reagent 2, was added, reducing the ferricyanide concentration to 91 pmolJL. Sixty MaterIals and Methods seconds after the addition of mixed picrate, the rate of Materials absorbance change (505 nm analytical wavelength, 570 Bilirubin from gallstones was obtained from Sigma nm blanking wavelength) was measured for a further 60s. Chemical Co. Ltd., Poole, UK. All other chemicals were Resufts of Analar grade from British Drug House, Poole, UK. Alkaline reagent, NaOH 200 mmol/L, was stable To evaluate the effect of ferricyanide addition on bilirubin, a nomcteric plasma pool was supplemented with 400 mol/L bilirubin and added to different con‘Biochemistry Laboratory, Regional Hospital, Cork,Ireland. centrations of ferricyanide. Figure 1 represents the 2 Biochemistry Department, University of Limerick, Plassy, absorbance change of the supplemented pool at 450 nm Limerick, Ireland. Received December 16, 1991; accepted March 4, 1992. (sample volume proportion = 6%). As shown, at ferricy-
Negative
CLINICAL CHEMISTRY, Vol.38, No. 9, 1992
1749
1.0
E a
#{149}
0.25 mmott
#{149}
0.50 flwflol’t.
#{163}1.OmmolS.
0.5 -O-
crs.tlnlns, imol/L.
2.OmmollL
o.e
BHIrubln, pmst&
V U
a
0.4
2
Table 1. Creatlnine Values of Six Icteric Patient Samples as Measured by Three Methods Rats method
as 0
20
40
60
50
100
Time/s Fig. 1. Effect of potassium ferricyanide on the absorbance of a plasmapool supplemented wIth 400 zmoI/L bilirubin #{149} Rate Method #{149} Blank Rate Method
0
E a V
a U 0
200
400
soo
Bilirubin umol/L FIg. 2. Effect of added blllrubln to a plasma pool on the estimationof creatinineby the rate,blankrate,andferricyanide procedures
snide concentrations of 1.0 mmol/L, the absorbance approaches zero after 100 s. The unsupplemented and supplemented pools were then assayed for creatinine by both the rate and ferncyanide procedures. Assay by the rate method demon-
strated
different
rates
of absorbance change, yielding
creatinine concentrations of 59 and 100 jmol/L, respectively. With the fernicyanide procedure, the unsupplemented and supplemented pools showed similar increased rates of absorbance change, particularly during the measuring period of 60 to 120 s, and gave values of 91 and 96 unol/L, respectively. The effect of added bilirubin was subsequently investigated by the three procedures and is illustrated in Figure
2.
There is a marked reduction in the apparent creatinine concentration with increasing bilirubin by the more widely used rate method. Both the fernicyanide and the blank rate procedures demonstrated very slight increases with added bilirubin. The creatinine concentrations for the unsupplemented pool by the rate and ferricyanide procedures were 101 and 91 pmol/L, respectively. This difference was also apparent when 100 nonictenic samples were compared by both procedures, i.e., y (ferricyanide) = -10.6 + 0.998 x (rate method), = 4.3 (creatinine range =441740 imol/L). This difference was due to starting the reaction with combined picrate reagent, as opposed to initial preincubation with alkaline reagent, and was independent of 1750
CLINICALCHEMISTRY,Vol.38, No. 9, 1992
Fsrrlcyanld.
method
method
61
96
99
104
134 180
59
69
75
75
94
93
403
32
79
87
518
235
296
522
29
90
295 88
I-
0.2
Blank rats
ferricyanide presence (data not shown). In an external quality-assessment program the overall mean values for creatinine concentrations of two samples were 117 and 437 p.mol/L (n = 1322). These samples yielded values of 129 and 439 Mmol/L by the rate method, and values of 115 and 432 imol/L by the combined reagent procedure (the presence or absence of ferricyanide showed no detectable difference with these latter values). The closer agreement of the combined reagent procedure to the overall mean value may reflect the number of laboratories using a combined reagent approach. Six samples with different bilirubin concentrations were measured for creatinune by the three procedures (the limited range of creatinine concentrations was dictated by the bilirubin content of samples). Table 1 demonstrates close agreement of both the fernicyanide and blank rate methods; the rate method shows a reduction in the apparent creatinine concentrations similar to that seen in Figure 2. Discussion
A recent approach to eliminate bilirubin interference in the Jaffe reaction for serum creatinine was an off-line preincubation of icteric samples with bilirubin oxidase (10). This approach is tedious and involves the expense of the enzyme. The approach presented here could be used both on- and off-line, and does not involve additional cost. The fennicyanide and blank rate procedures demonstrated assay values for icteric samples that were -10 mol/L per 100 tmol/L of bilirubin higher than the rate
These findings are consistent with the observations of Osselaer and Lievens (9). Although newer Hitachi analyzers have the capability of carrying out blank rate protocols, older Hitachi models and many other instruments lack this capability. Ferricyanide was first used at a concentration of 15 1mol/L (6). This was increased by Knapp and Mayne (7) to 50 mol/L with no adverse effects on the assay. We used a final concentration of 91 pmol/L and observed no interference or effect on analytical sensitivity. The procedure can be applied to a wide range of analyzers, and could be particularly useflil when longer reaction times and lower picrate concentrations are used and when the interference by bilirubin is more substantial (5). procedure.
In conclusion, tion of potassium
we demonstrated that the simple addiferricyanide to plasma samples as part
of the analytical procedure is very effective in elirninating the negative interference of bilirubin in the Jaffe reaction for plasma creatinine estimation. References 1. Watkins RE, Feldcsmp CS, Thibert BJ, Zak B. Interesting interferences in direct serum creatinine reaction. Microchem J 197621:370-.84. 2. Soldin 8.1, Henderson L, Hill JG. The effect ofbilirubin on reaction rate methods forthe measurement of creatunune [Abstract]. Cliii Chem 197723:1141. 3. Jacks Eli, Halley NH. The effect of bilirubun on the direct kinetic determination of serum creatinine [Abstract]. Clin Chem 1978;24:1022. 4. Guy JM, Legg EF. Bilirubin interference in determinations of creatunine with the Hitachi 737 analyzer [Tech Briefi. Clin Chem 1990;36:1851-2.
5. Knapp ML, Hadid 0. Investigation
into negative
interference
by jaundiced plasma in kinetic Jaff#{233} methods forplasma creatinine determination. Ann Clin Biochem 1987;24:85.-97. 6. Van der Heiden DA, Neervoort WH. Eon kinetische bepaling van kreatinine via de Jaff#{233} reactie, welke met gesteroord worth door bilirirubine. Tojschr NVKC 1984;9:81-5. 7. Knapp ML, Mayne MD. Development of an automated kinetic Jaff#{233} method designed to minimiae bilirubin interference in plasma creatinine assays. Cliii Chim Acta 1987;168:239-46. 8. Hoffman RJ, Fold RD. A modified procedure forcreatinine (CR which eliminates bilirubin (BILT) interference on theHitaclii 737 [Abstract]. Clin Chem 1988;34:1313. 9. Osaelaer JC, Lievens MM. Moreon interference ofbilirubin in creatinine determination by Hitachi analyzers [Letter]. Cliii Chem
1991;37:1460-1. 10. Haesaig LK, Feldbruegge DH, Urban EA, Koch DD. A utilitarian procedure to avoid bilirubun interference on creatinine assays [Abstract].
Clin Chem 1991;37:932.
CUNICAL CHEMISTRY, Vol.38, No. 9, 1992
1751
