Influence of methodology on the detection and diagnosis of congenital analbuminemia.
The most commonly used serum albumin assays in clinical laboratories rely on dye-binding techniques (8). Characteristically, these methods accurately measure the concentration of serum albumin near the limits of the reference range. In conditions such as nephrotic syndrome, the serum albumin and the albumin:globulin ratio are low. Under these conditions, results of the bromcresol green (BCG) methods can be falsely inflated (9-15) and bromcresol purple-binding methods prone to falsely lowered results (16), although reagent manufacturers have adopted many variations of the dye-binding albumin methodology in an attempt to assure accuracy.
Dye-binding albumin methods would likely be used to initially assess albumin status in undiagnosed patients with analbuminemia. Dye-binding albumin methods have falsely indicated the presence of albumin in analbuminemia in a rat model (17) and in a case report (18); however, few analytical details were reported. Here we compare the ability of two BCG methods to detect the absence of serum albumin in two new cases of congenital analbuminemia and comment on the role of methodology in previous studies.
Patient DN was admitted as a newborn with hypoglycemia and apnea; investigations revealed apparently low serum albumin (15 g/L). His serum albumin concentration remained low despite 24-h urine protein excretion rates within reference values. He was readmitted on three additional occasions in his first 6 months of life with respiratory distress and was eventually started on treatment for chronic asthma. There was no identifiable cause for the patient's low serum albumin and no evidence of a protein-losing enteropathy, based on a [Tc.sup.99]-labeled albumin scan.
Patient KB was admitted at 2 days of age for treatment of cellulitis and was subsequently readmitted three times over the next 6 months for respiratory distress involving upper respiratory tract infections and pneumonia. The reported serum albumin remained low (10-15 g/L), and there was no identifiable source of protein loss.
Serum protein electrophoresis performed at 10 weeks (patient DN) and 12 weeks (patient KB) revealed an absence of an albumin band (0-3 g/L), and the diagnosis of congenital analbuminemia was considered. Both neonates had no symptoms directly attributable to the lack of albumin, and both presented with recurrent respiratory tract symptoms related to bronchial asthma or pneumonia, which may be a coincidental observation (19-21).
In each case serum albumin >10 g/L was reported from an Ektachem 700XR analyzer at an early stage of clinical investigation. Reported albumin concentrations were variable, with no corresponding clinical changes but with method-dependent variability (Table 1). By immunoassay, albumin was <0.01 g/L (assay detection limit) for patient DN. BCG assays are prone to positive bias (9-15); however, the two BCG assays produced different results with analbuminemic sera (Table 1). Moreover, the Ektachem 700XR analyzer did not detect that the concentration of albumin was substantially below the lower limit of assay linearity. These observations prompted an evaluation of the dynamic range for each albumin assay.
The manufacturer's lower limit of albumin assay linearity was confirmed to be 10 g/L with saline-diluted serum on the Ektachem 700XR. Saline-diluted serum retains a constant albumin:globulin ratio, and alteration of this ratio is known to affect the Ektachem 700XR method (9-12). Pure human albumin solutions (Bayer Corp.) revealed a nonlinear calibration curve below 20 g/L, with a positive intercept of 4.6 g/L estimated by polynomial regression. False albumin results (6-10 g/L) were also produced by 10-50 g/L purified human IgG (Bayer Corp.). Both positive biases would be anticipated in the analysis of sera from analbuminemic patients with the Ektachem 700XR analyzer.
The Hitachi 717 albumin assay demonstrated a linear response with saline-diluted serum to 4 g/L (lowest concentration tested), which is below the manufacturer's stated lower limit of 10 g/L. In the absence of globulins, pure human albumin solutions demonstrated a linear calibration curve with a near-zero intercept (0.03 g/L); the assay demonstrated no reaction with purified human IgG. A false albumin result (8 g/L) for analbuminemic patient DN from the Hitachi 717 analyzer was suppressed because it was below the dynamic range of the assay (Table 1).
The reaction times of these two BCG assays are 96 s (Hitachi 717) and 145 s (Ektachem 700XR); the shorter reaction time can confer greater albumin specificity (22). The lack of sample dilution in the multilayer dry film technology of the Ektachem 700XR may also promote the reaction of globulins with BCG.
The serum electrophoresis results indicated non-zero concentrations of albumin in each case (Table 1). This can be attributed to a small densitometry peak caused by the staining of non-albumin proteins that migrate near albumin-including transthyretin (prealbumin) and [[alpha].sub.1]-acid glycoprotein, (3). Although non-zero results were obtained with both agarose gel (patient DN) and cellulose acetate (patient KB) electrophoresis, other analbuminemia studies have occasionally reported albumin concentrations of 0 g/L with electrophoresis. Variability of albumin quantification by electrophoresis may be partially attributed to falsely low measurements caused by protein absorption to the matrix (e.g., paper).
Past literature reviews have focused on the clinical attributes of analbuminemic patients (1, 2), with few comments on the analytical methods used to confirm the diagnoses. Case studies report albumin concentrations of 0-12 g/L among patients that by definition lack serum albumin, although most results were 0.03-3 g/L (1, 2, 1821, 23-41). We evaluated the influence of methodology on past case reports and our own cases (Fig. 1). Cases reports were included in this analysis if albumin methods were described (2, 3, 5, 18-41); others were excluded (42-44). Reports of albumin with modified mobility by electrophoresis were also excluded (45, 46). Fig. 1 shows that most variation in the apparent albumin concentration in analbuminemia reports is accounted for by the albumin methodology (one-way ANOVA, P <0.001). Both dye binding and the Howe (47) sulfate precipitation methods are particularly prone to overestimation of serum albumin. Many of the albumin immunoassays reported nonzero concentrations in analbuminemic patients. This may be accounted for by both the imprecision of assays and by weak reactions with serum albumin fragments produced by incomplete transcription and translation of the albumin gene (48).
In summary, some conventional BCG methods are prone to overestimations of albumin that may delay diagnosis of congenital analbuminemia. When dynamic ranges of BCG methods are assigned using saline-diluted serum (with a constant albumin:globulin ratio), there is no assurance that very low concentrations or the absence of albumin can be detected when the albumin:globulin ratio is abnormally low. The risk of reporting inaccurate albumin results is present not only in analbuminemia but also in hypoalbuminemic states such as nephrotic syndrome and burn injuries, using these methods (8, 9). The concentration of albumin reported by clinical laboratories for specimens from analbuminemic patients is method-dependent. Of the four categories of albumin methods, dye-binding methods demonstrate the greatest overestimation of albumin concentrations and can produce misleading reports that indicate the presence of albumin. Immunoassays of serum albumin are likely the most accurate (8). Electrophoretic quantification is widely available and often produces near-zero albumin results that are consistent with the diagnosis of analbuminemia. The challenge for clinicians and clinical laboratorians is to realize that initial dye-binding albumin test results can be falsely inflated and that a more definitive albumin method must be used to evaluate congenital analbuminemia.
[FIGURE 1 OMITTED]
(1.) Russi E, Weigand K. Analbuminemia. Klin Wochenschr 1983;61:541-5.
(2.) Dammacco F, Miglietta A, D'Addabbo A, Fratello A, Moschetta R, Bonomo L. Analbuminemia: report of a case and review of the literature. Vox Sang 1980;39:153-61.
(3.) Kallee E. Bennhold's analbuminemia. a follow-up study of the first two cases (1953-1992). J Lab Clin Med 1996;127:470-80.
(4.) Tarnoky AL. Genetic and drug-induced variation in serum albumin. Adv Clin Chem 1980;21:101-46.
(5.) Hollander CS, Bernstein G, Oppenheimer JH. Abnormalities of thyroxine binding in analbuminemia [Letter]. J Clin Endocrinol Metab 1968;28: 1064-6.
(6.) Wilkins TA, Midgley JE, Barron N. Comprehensive study of a thyroxin-analog-based assay for free thyroxin ("Amerlex [FT.sub.4]"). Clin Chem 1985;31:164453.
(7.) Stockigt JR, Stevens V, White EL, Barlow JW. "Unbound analog" radioimmunoassays for free thyroxin measure the albumin-bound fraction. Clin Chem 1983;29:1408-10.
(8.) Doumas BT, Peters T. Serum and urine albumin: a progress report on their measurement and clinical significance. Clin Chim Acta 1997;258:3-20.
(9.) Gorus FK, Van Blerk M. Ektachem slides overestimate serum albumin in patients with nephrotic syndrome. Eur J Clin Chem Clin Biochem 1993;31: 159-63.
(10.) Leerink CB, Winckers EKA. Multilayer-film bromocresol green method for albumin measurement significantly inaccurate when albumin/globulin ratio is <0.8 [Letter]. Clin Chem 1991;37:766-7.
(11.) Green RJ, Day BS, Powers DM. Multilayer-film bromocresol green method for albumin measurement significantly inaccurate when albumin/globulin ratio is <0.8 [Response to Letter]. Clin Chem 1991;37:767-8.
(12.) Buffone GJ, DeArmas W, Frawley V, Ou CN, Schimbor C. Multilayer-film bromocresol green method for albumin measurement compared to three others. Clin Chem 1984;30:581-2.
(13.) Hill PG. The measurement of albumin in serum and plasma. Ann Clin Biochem 1985;22:565-78.
(14.) Speicher CE, Widish JR, Gaudot FJ, Hepler BR. An evaluation of the overestimation of serum albumin by bromocresol green. Am J Clin Pathol 1978;69:347-50.
(15.) Slater L, Carter PM, Hobbs JR. Measurement of albumin in the sera of patients [Technical Bulletin No. 34]. Ann Clin Biochem 1975;12:33-40.
(16.) Maguire GA, Price CP. Bromocresol purple method for serum albumin gives falsely low values in patients with renal insufficiency. Clin Chim Acta 1986;155:83-8.
(17.) Joles JA, Kortlandt W. Massive overestimation of albumin in analbuminemic rat serum by bromcresol green. Clin Chem 1987;33:2121.
(18.) Berger GMB, Stephen CR, Finestone A, Beatty DW. Analbuminaemia. Clinical and laboratory features in a South African patient. S Afr Med J 1985;67:418-22.
(19.) Bowan H, Hermodson M, Hammond CA, Motulsky AG. Analbuminemia in an American Indian girl. Clin Genet 1976;9:513-26.
(20.) Weigand K, Russi E, von Schulthess G, Bavaud C. Turnover of autologous alpha 1-antitrypsin in a patient with congenital analbuminemia. Klin Wochenschr 1983;61:547-52.
(21.) Montgomery DAD, Neill DW, Dowdle EBD. Idiopathic hypoalbuminaemia. Clin Sci 1962;22:141-54.
(22.) Webster D. A study of the interaction of bromocresol green with isolated serum globulin fractions. Clin Chim Acta 1974;53:109-15.
(23.) Baldo G, Fellin R, Manzato E, Baiocchi MR, Ongaro G, Baggio G, et al. Characterization of hyperlipidemia in two patients with analbuminemia. Clin Chim Acta 1983;128:307-19.
(24.) Frohlich J, Pudek MR, Cormode EJ, Sellers EM, Abel JG. Further studies on plasma proteins, lipids, and dye- and drug-binding in a child with analbuminemia. Clin Chem 1987;27:1213-6.
(25.) Weinstock JV, Kawanishi H, Sisson J. Morphologic, biochemical and physiologic alterations in a case of idiopathic hypoalbuminemia (analbuminemia). Am J Med 1979;67:132-9.
(26.) Goulle JP, Laine G, Sauger F, Maitrot B, Bouillerot A, Gray H, et al. Modifications biochimiques dans un cas d'analbuminemie. Clin Chim Acta 1976;71:143-7.
(27.) Benhold H, Kallee E. Comparative studies on the half-life of 1131-labeled albumins and nonradioactive human serum albumin in a case of analbuminemia. J Clin Investig 1959;38:863-72.
(28.) Cormode EJ, Lyster DM, Israels S. Analbuminemia in a neonate. J Pediatr 1975;86:862-7.
(29.) Waldmann TA, Gordon RS, Rosse W. Studies on the metabolism of the serum proteins and lipids in a patient with analbuminemia. Am J Med 1964;37:960-8.
(30.) Banter FC, Steinfeld JL, Waldmann T, Delea CS. Metabolism of infused serum albumin in the hypoproteinemia of gastrointestinal protein loss and in analbuminemia. Trans Assoc Am Physicians 1961;74:180-94.
(31.) Gordon RS, Banter FC, Waldmann T. Idiopathic hypoalbuminemias: clinical staff conference at the National Institutes of Health. Ann Intern Med 1959;51:553-76.
(32.) Christen H, Franglen G. Analbuminamie oder albuminmaskierung? Diskussionsbeitrag zur analbuminemie. Schweiz Med Wochenschr 1972;102: 1900-6.
(33.) Keller H, Noseda G, Morell A, Riva G. Analbuminemia. Minerva-Med 1972; 63:1296-306.
(34.) Di Guardo G, Monticelli C, Leoni G, Calabrese M, Pozzoli E, Porta F, Fraccaro M. Un caso di analbuminemia. Lab (Milan) 1977;4:113-7.
(35.) Reynaud JP, Sagniez M, Rousseau A, Nicolas G. Etude biologique d'un cas d'analbuminemie. Semin Hop Paris 1983;59:124-8.
(36.) Goule JP, Laine G, Sauger F, Maitrot B, Bouillerot A, Gray H, et al. Etude biochimique du premier cas d'analbuminie en France. Ann Biol Clin 1976; 34:403-9.
(37.) Irunberry J, Abbadi M, Khati B, Benabadji M, Rocha E. Trois cas d'analbuminemie dans une fratrie. Rev Eur Etud Clin Biol 1971;16:371-9.
(38.) Beck GE, Dorta T. Un cas d'analbuminemie. Helv Med Acta 1959;26:76471.
(39.) Baldo-Enzi G, Baiocchi MR, Vigna G, Andrian C, Mosconi C, Fellin R. Analbuminaemia: a natural model of metabolic compensatory systems. J Inherit Dis 1987;10:317-29.
(40.) Fujiwara M. Clinical studies on human serum albumin. Report 3. Analytical studies of serum protein of a patient with analbuminemia. Acta Haematol Jpn 1971;34:569-75.
(41.) Mahkota S, Dolar P, Pusenjak T. Primer analbuminemije. Zdrav Vestn 1965;34:95-9.
(42.) Freeman T. Analbuminaemia-a study of albumin and transferrin metabolism. In: Birke G, Norbery R, Plantin LO, eds. Physiology and pathophysiology of plasma protein metabolism. New York: Pergamon Press, 1969:75-85.
(43.) Klaber M, Baker H. Analbuminaemia with cutaneous angiomata. J R Soc Med 1978;71:67-9.
(44.) Belokrinitskaia OA, Sitnikova LA, Aleksandrovskaia TN. A case of congenital analbuminemia. Klin Med (Mosc) 1985;63:119-21.
(45.) Gras J, Padros R, Marti MI, Gomez Acha JA. Pseudo-analbuminemia due to the presence of a slow albumin variant moving in the alpha 1 zone. Clin Chim Acta 1980;104:125-8.
(46.) Carretero JP, Marcet MIM, Gomez Acha JA. Analbuminemia congenita. Sangre (Barc) 1978;23:876-81.
(47.) Howe PE. The use of sodium sulfate as the globulin precipitant in the determination of protein in blood. J Biol Chem 1921;49:93-108.
(48.) Watkins S, Madison J, Galliano M, Minchiotti L, Putnam FW. Analbuminemia: three cases resulting from different point mutations in the albumin gene. Proc Natl Acad Sci U S A 1994;91:9417-21.
Andrew W. Lyon, ,  * Paul Meinert,   Garth A. Bruce,   Victor A. Laxdal, ,  and Mark L. Salkie ,  (Departments of  Pathology and  Pediatrics, College of Medicine, University of Saskatchewan, and Departments of  Laboratory Medicine and  Pediatrics, Saskatoon District Health, Saskatoon, Saskatchewan, Canada 57N OW8; * address correspondence to this author at: Department of Laboratory Medicine, Royal University Hospital, 103 Hospital Dr., Saskatoon, SK, Canada 57N OW8; fax 306-655-2223, e-mail firstname.lastname@example.org; poster presented at the 42nd Annual Conference of the Canadian Society of Clinical Chemists and the Canadian Association of Medical Biochemists, June 14-18, 1998, Ottawa, Canada)
Table 1. Analysis of serum albumin and total protein on single specimens from the analbuminemia patients using different methods. Method Albumin, Total g/L protein, g/L Patient DN Ektachem 700XR (a) 15 45 Hitachi 717 (b) <10 (c) 42 Serum electrophoresis (d) 3.40 Radial immunodiffusion <0.01 (f) albumin immunoassay (e) Patient KB Ektachem 700XR (a) 11 36 Serum electrophoresis (g) 1 (a) Reagents: Vitros [R], Ortho Clinical Diagnostics. (b) Reagents: Boehringer Mannheim Canada. (c) The measured value of 8 g/L was below the assay linearity limit and was suppressed by the instrument. (d) SP-17 Titan agarose gel electrophoresis system, Helena Laboratories. (e) Kallestad [TM] Endoplate [TM] single radial immunodiffusion method, Sanofi Pasteur Diagnostics. (f) Estimated by linear regression and interpolation as 0.003 g/L. (g) Microzone system cellulose acetate electrophoresis, Beckman.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Technical Briefs|
|Author:||Lyon, Andrew W.; Meinert, Paul; Bruce, Garth A.; Laxdal, Victor A.; Salkie, Mark L.|
|Date:||Nov 1, 1998|
|Previous Article:||Is prostate-specific antigen a marker for pregnancies affected by down syndrome?|
|Next Article:||Lactate dehydrogenase-immunoglobulin G complex in the serum of the postburn patient.|