Naproxen Interference with the Ion-Selective Electrode

electrochemical: Yellow Springs Instrument Co., Yellow Springs, OH 45387) for assaying galactose. For both fluorometric procedures, all reagents and standards were prepared as previously described (1). The manual assay requires 400 jiL of Tris buffer (10 mmolJL); 600 ,iiL of a solution containing p-hydroxyphenylacetic acid (50 mg/ L) and horseradish peroxidase (EC 1.11.1.7) (10 mgIL); 160 1zL of galactose oxidase solution (4.1 U/mL); and 400 tL of blank, standard, or sample filtrate. This mixture is incubated at 40#{176}C for 40 mm, then mixed with 160 iL of 0.4 molJL NaOH solution and the fluorescence is measured (317 nm excitation, 414 nm emission). The YSI was fitted with a galactose oxidase membrane and lactose buffer system (2). We used a 100 mg/L aqueous calibrator to define the assay range of 10-200 mgIL, a 10-fold increase in the manufacturer’s claimed detection limit. No electronic modification of this model was necessary. The interand intra-assay precision of all three methods was good, with respective CVs of 2.5-6.3% and 2.4-8.6%, the CVs being largest at the lowest concentrations. Analytical recoveries ranged from 95 to 105%. In comparison studies with patients’ samples, MAN = 0.6 mg/L + 1.00 CF and YSI = 8.02 mg/L + 1.102 CF. There was an excellent intraclass correlation (3), 0.987, between the CF and MAN results; between the YSI and CF, however, this correlation was poor: 0.524. Clearance of low concentrations of galactose from plasma is a measure of hepatic blood flow. To estimate clearance within ±5%, the accuracy of plasma steady-state during continuous infusion must be defined to within 2 mg/L. We found that overestimation of galactose by the YSI, which directly affected the intraclass correlation, resulted from a combination of factors: (a) less specificity of the galactose oxidase membranes, which was not totally correctable by subtraction of baseline values; (b) direct oxidation of interfering compounds at the platinum electrode, notably heparim flush and glycerol (plasma samples consistently gave positive readings across a blank membrane); and (c) operation at an inappropriately high sensitivity, thereby exaggerating the previous factors. Evidently, either of the fluorogenic assays may be used for measuring low concentration of galactose, but the electrochemical detection system should not be used to measure values well beyond the manufacturer’s specifications.