Apr 2, 1985 - Walter Slavin,2and A. Ralph Henderso&. We compared results by ..... Gardiner PE, Ottaway JM, Fell GB,Hall DJ. Determination of aluminum.
CLIN. CHEM. 31/11, 1882-1884 (1985)
Direct-CalibrationMethodfor Determinationof Aluminumin Serum Is Comparablewiththe Protein-Precipitation Technique Cohn Bradley,1 Fred
V. Leung,’3
Walter Slavin,2and A. Ralph Henderso&
We compared results by two methods for serum Al determination: matrix modification with direct calibration in a stabilized-temperature platform furnace (C/in Chem 28, 2139, 1982) and a technique involving extraction with nitric acid before atomic absorption spectrometry (Cm Chem 30, 1216, 1984). The two methods gave similar results with use of either a deuterium or a Zeeman system of background correction, but gave different slopes for standard additions (4. s per gIL), depending on the Al content of the serum, an effect not seen with aqueous solutions. These differences do not affect the accuracy of the Al determination up to 150
Additional Keyphrases: variation,source of atomicabsorption spectrometry deuterium and Zeeman corrections compared renaldialysispatients Aluminum (Al) concentrations in serum are determined by electrothermal atomic absorption spectrometry, to monitor increased Al concentrations in patients who are being treated by regular hemodialysis (1). Serum (matrix) interactions with Al that cause analytical error have been circumvented by various methods (2,3). We use a stabilized-temperature platform furnace method with matrix modification to determine Al in serum and urine (4). Our results compare favorably with those by electrothermal atomic absorption spectrometry methods (5) or the method of standard additions (6). Brown et al. (7) recently proposed an alternative technique requiring precipitation of serum protein with nitric acid before determination of Al. They developed this technique because, using our matrix-modification technique (4), they found differences in the slopes of calibration lines (by standard additions) between uremic patients’ sera and aqueous standards. They therefore argued that when serum is analyzed directly (diluting with water or matrix modifier), the results are in error. However, they did not compare results by their proposed technique with those by the deprecated directcalibration method (4). Here we directly compare our method with theirs, using serum specimens from patients being treated with renal dialysis. We also compare results by both techniques, after background correction with a deuterium arc background detector and with a Zeeman magnetic system. Our technique and the acid-precipitation method give similar resuIts, both for concentration and slopes of the standard additions lines. In both methods, however, there are deviations from linearity for serum with high A! concentrations.
Materials and Methods Instrumentation We used a Model 5000 atomic absorption spectrophotometar equipped with a deuterium-arc background corrector, an HGA 500 graphite furnace, and an AS4O autosampler. A Zeeman magnet background corrector was fitted for subsequent studies. All equipment, including a Data System 10 used to record and process absorbance signal proffles, was from Perkin-Elmer Corp., Norwalk, CT 06852. The program used in the Data System 10 automatically initiates the baseline correction on the spectrophotometer 1 s before atomization. Procedure We determined the concentration of Al in aqueous standards, quality-control sera (8), and patients’ sera, using a stabilized-temperatureplatform furnace as previously described (4). The graphite furnace settings were modified as follows: dry-temperature 100#{176}C, ramp 5 s, hold 10 5; 120 #{176}C, ramp 10 s, hold 10 s; 250 #{176}C, ramp 1 s, hold 10 s; char-temperature 1500 #{176}C, ramp 1 s, hold 40 s. All other settings were as previously listed (4). The protein precipitation method we used was that of Brown et al. (7). We added equal volumes of aqueous Al standards (standard-additions method, ref. 9) with concentrations of 50, 100, and 150 pg/L to aliquots of pooled serum, and assayed. After the absorbance of the original serum sample was determined, the instrument was re-set to zero with the autozero function. The supplemented standards were then used to determine linearity and slope of the line. With a specimen initially containing 50 pg of Al per liter, the final Al of the highest standard would not exceed 100 pgfL on the platform, a concentration below the linear range of the method, which is up to 150 pg/L. Specimens with higher Al concentrations were analyzed after two- to fourfold dilution with matrix modifier. Al concentrations determined by standard additions were compared with those by the direct calibration method in which the calibration line was prepared from data obtained after the addition of Al standards to a low-Al serum pool blanked against this serum pool. Large molecular mass compounds (Mr> 1000), including proteins, were removed by ultrafiltration of serum samples. We used the Centrifree Micropartition Unit with YMT membranes supplied by Ainicon Corp., Danvers, MA 01923, as previously described (10).
Statistical Analysis Student’s t-test, variance ratios (F) test, Tukey-Kramer test, and linear regressions and correlations were by standard methods (11, 12). 1Depanent ofClinical Biochemistry,University Hospital,London, Ontario, Canada N6A 5A5. 2 Perkin-Elmer Corporation,Ridgefleld,CT 06877. 3Addreas correspondenceto this author. Received April 2, 1985; accepted August 5, 1985. 1882 CLINICALCHEMISTRY, Vol. 31, No. 11, 1985
Results
and DIscussion
We compared our direct calibration method (4) with the acid-precipitation technique (7), using deuterium back-
ground correction. Sera from 32 patients gave a mean of 72.66 pg/L (range 0-225) and 69.91 p.g/L (range 0-200) for the direct and acid-precipitation procedures, respectively (p
