Misleading acute hypercalcemia due to hyperlipidemia: a method-dependent error

OBJECTIVE

To report a case of artifactual hypercalcemia in a patient with hyperlipidemia.

METHODS

We present clinical data and laboratory findings in a 33-year-old woman with generalized fatigue, a recent 5-kg weight loss, and a papular rash on the extremities as well as a history of diabetes and hypertension.

RESULTS

Physical examination revealed an obese patient with eruptive xanthomas and lipemia retinalis. Laboratory tests showed hyperlipidemia, hypercalcemia (serum calcium measured by spectrophotometry), anemia, hyperproteinemia, hyperuricemia, and hyperbilirubinemia. After 4 days of a low-fat, low-cholesterol diet and gemfibrozil therapy, the serum triglyceride level decreased, and the serum calcium concentration returned to normal.

CONCLUSION

In patients with hypercalcemia without an obvious cause, a spurious measurement should be considered.

Problems with measurements caused by high concentrations of serum solids

There have been numerous reports of spectrophotometric and volume problems caused by elevated levels of lipids in blood. The offending lipids, primarily triglycerides, not only cause turbidity leading to optical aberrations when added to analytical reagents, but also result in short-sampling errors leading to the measurement of inaccurate volumes of sample. Numerous methods have been developed to clear the lipemia, including ultracentrifugation organic solvent extraction, chemical precipitation and, most recently, enzymic hydrolysis. Although the latter procedures eliminate the optical problems, they do not deal with the volume dilution error created by the triglycerides. In turn, corrective mathematics have been developed to compensate for the inaccurate pipetting caused by the elevated lipids in a sample; however, these empirical calculations are not truly accurate at high concentrations of total lipids. This monograph will describe the problems caused by the presence of elevated lipids and the means available for treating them.

Nonenzymatically glycosylated proteins

Nonenzymatic glycosylation takes place in all proteins with a free-reacting lysine or valine in the presence of glucose. The formation of glycosylated plasma albumin, hemoglobin (Hb A1c), and skin collagen provides a diagnostic index of short- to long-term time-concentration of glucose in vivo. A wide range of assay methods are available, with affinity chromatographic, isoelectric focusing, and spectrophotometric methods providing the best accuracy and versatility. Glycosylated hemoglobin assays indicate glucose pressure over the previous 2 to 3 months and are of diagnostic value in general diabetic control, while glycosylated plasma albumin determinations are preferable in acute episodes in the life of a diabetic (e.g., pregnancy, infection, stress, trauma, surgery), since they provide an overview of changing blood glucose values of the previous 2 to 4 weeks. Glycosylated collagen estimations reflect tissue aging and are relevant in healing processes. Glycosylation alters the biologic activity of proteins, and these may relate to the manifold complications concomitant on the lifelong elevation of blood and tissue glucose in the diabetic (C6a). Assays for glycosylated hemoglobin have been routinely performed in clinical chemistry laboratories for a decade, and convenient determination for other nonenzymatically glycosylated proteins is proceeding apace.

A comparison of the determination of glucosylated haemoglobin by isoelectric focusing and cation-exchange chromatography on minicolumns

The labile intermediate, pre A1c, formed in the glycosylation of haemoglobin A is a potential contaminant in the measurement of glycosylated haemoglobin when this is determined as the amount of HbA1c present in the sample. By isoelectric focusing on polyacrylamide gel plates this contamination could be avoided either by excision of the HbA1c leaving the neighbouring pre A1c behind on the slab, or by converting the pre A1c to HbA in glucose-free medium before electrophoresis. In cation-exchange chromatography on minicolumns (from Bio-Rad) the pre A1c was removed in the haemolysis process by borate-induced transformation to the non-interfering HbA. The chromatographic method nevertheless gave about 10% higher values than isoelectric focusing. The linearity between paired results of the electrophoretic and chromatographic methods was not perfect (p less than 0.05). Both methods measured decreasing concentrations of HbA1c equally well and with the same precision at both high and low levels (CV less than 5%). All HbF was simultaneously determined in the chromatographic method, while HbF did not interfere in the electrophoretic method. The HbA1c in whole blood samples was stable at 4 degrees C for up to 1 week. Carbon monoxide treatment made the HbA1c in haemolysates stable for at least 3 months at -70 degrees C making possible long-term control by both methods.

Evaluation of a mini-column chromatographic procedure for the measurement of hemoglobin A1c

An ion-exchange chromatographic procedure is described which facilitates the determination of beta-chain aminoterminal modified glycated hemoglobin. The procedure includes an erythrocyte lysis reagent which eliminates the labile aldimine component (pre-A1c) and a two-stage elution step which separates HbA, 1a + b from HbA1c. This procedure also includes calibrator material which aids in correcting for temperature fluctuations during the analysis. Within-run CV's for samples with HbA1c levels between 4.0% and 13.7% were 1.4 to 3.2%. The between-run CV for an HbA1c control was 5.5%. A comparison of the present test to an ion-exchange HPLC method yielded the equation: HPLC = 0.96 (present method) -0.2% (n = 101 and r = 0.984). Two separate reference range studies yielded comparable results (n = 220/65, mean = 4.77/4.78%, S.D. = 0.68/0.55). Studies with pooled erythrocytes and various lipemic plasmas did not reveal any assay interferences. Various abnormal hemoglobins were studied for their effect on the assay.

Description of analytical problems arising from elevated serum solids

There has always been a problem with the collection of data and interpretation of the results obtained from any biological fluid in which the solids content was increased to a great extent. Of these solids, the triglycerides of the lipids may cause a plasma (serum) to vary in appearance from opalescent to milky. This condition of the specimen and the concomitant turbidity upon its addition to reagents creates the well-documented optical aberrations of spectrophotometric measurements. In addition, the lipids, in conjunction with the proteins, can act as diluents when they are elevated, thereby decreasing what might be termed the residual true plasma volume. Thus the water content of an aliquot sampled for a particular analytical procedure is diminished, and by that means a situation is created in which a short sample is drawn. This dilution effect by the solids results in a lowering of the assay values obtained for the measured constituents of such a serum sample. An associated phenomenon of high concentrations of solids, especially proteins, is the increase in viscosity of a specimen, a condition that also causes an error of short sampling when certain peristaltic pumping devices are used. This review considers several aspects of problems encountered when dealing with a number of circumstances that are critical to the measurement of analytes in severely hyperlipemic and/or hyperproteinemic specimens. These include the problems of short sampling; the potential amelioration of the problem by corrective mathematics, extraction of the lipids, or ultracentrifugation of the true plasma from the lipids; the important need to include most analytes into our considerations; the difference in reference base values for the calculation of concentrations of lipids of serum versus other analytes; the concept of the use of ratios when the reference base values differ, numerator analyte from denominator analyte; and the problems of using serum blanks when necessary corrective action for the solids volume is neglected. Thus, in the final analysis, problems with underestimated volumes of samples used for many spectrophotometric determinations are considered here along with the other difficulties encountered when the need to measure analytes in serums with extremely high solids content presents.

Recent advances in glycosylated hemoglobin measurements

Glycosylated hemoglobins have gained wide acceptance as an accurate index of long-term blood glucose control in diabetes mellitus. A variety of glycosylated hemoglobin assays is available. There is a high degree of correlation between results determined by these assays. The ideal laboratory method for measuring glycosylated hemoglobin in the diabetic should be accurate, precise, easily standardized, inexpensive, and rapidly performed. Unfortunately, none of the currently used methods meet all of the criteria necessary to be considered the ideal laboratory method. The most widely used methods for quantitating glycosylated hemoglobins--including ion exchange chromatography, electrophoresis, isoelectric focusing, thiobarbituric acid colorimetry, and affinity chromatography--are reviewed with respect to the important advantages and disadvantages of each method for the clinical laboratory. Techniques for quantitating glycosylated proteins other than hemoglobins, such as albumin, are also discussed.

Characteristics of labile HbA1 in erythrocytes from normals and diabetics

Incubation of erythrocytes from diabetic patients and non-diabetic persons in 0.15 mol/l saline at 37 degrees C was followed by a decrease in ion-exchange microcolumn determined HbA1 only during the first 4 h. The fraction removed was denoted as labile HbA1 and the remaining fraction as stable HbA1. Incubation in saline overnight at 20 degrees C did not completely remove the labile fraction. Incubation in saline overnight at 4 degrees C yielded only a slight decrease in labile HbA1. Incubation in saline containing various glucose concentrations (up to 100 mmol/l) produced increasing amounts of HbA1 during the first 3-4 h after which an equilibrium was reached. In both normal and diabetic erythrocytes incubated for 3-4 h we found a linear relationship between changes in labile HbA1 and glucose concentration. The degree of increase in labile HbA1 was the same in normals and diabetics and not dependent on pre-existing normal or moderately increased stable HbA1. In the blood from non-diabetics (n = 30) labile HbA1 was 0.5 +/- 0.3% (mean +/- SD) and in diabetics (n = 80) 1.6 +/- 0.8%. The correlation between labile HbA1 and simultaneously determined blood glucose was r = 0.72.

Evaluation of an affinity chromatographic procedure for the determination of glycosylated hemoglobin (HbA1)

An affinity chromatographic method for the determination of glycosylated hemoglobin (HbA1) was evaluated. The procedure was shown to be precise, the within- and between-assay coefficients of variation being less than 5%. It was also shown to correlate well with electrophoresis (r = 0.968) and ion-exchange chromatography (r = 0.916). An inverse relationship was shown to exist between increasing temperature and HbA1 levels measured by affinity chromatography. A statistically significant difference was found for samples run at 20 degrees C and 25 degrees C respectively, suggesting that the method should be run in a temperature-controlled environment. The affinity procedure was also shown not to be affected by the type of anticoagulant, the concentration of hemoglobin in the hemolysate, and acetylation.

Serum lipid and lipoprotein levels and metabolic control in insulin-treated diabetics

Serum lipid and lipoprotein levels were evaluated in 50 insulin-treated diabetic out-patients (25 male and 25 female) and in 46 normal volunteers (22 male and 24 female). In these groups metabolic evaluation was carried out by assaying fasting plasma glucose, glucose in urine and glycosylated hemoglobin (G-HbA1). No differences were observed in the lipid and lipoprotein patterns between diabetic patients and normals. HDL values were significantly lower in male subjects, diabetic and normal, as compared to females, but there were no differences between the diabetic and the normal group. G-HbA1 was significantly correlated to fasting plasma glucose and glucose in urine, but also to WS-TG and VLDL-TG. Fasting plasma glucose too was correlated to WS-TG and VLDL-TG. Moreover, a negative correlation was found between HDL-Ch and WS-TG and VLDL-TG. These results show that sufficiently well-controlled insulin-treated diabetics do not have altered plasma lipid and lipoprotein levels. In particular, in these patients HDL-Ch values can be normal, because insulin levels are sufficient to activate lipoprotein lipase and to guarantee an adequate plasma VLDL clearance.

An update on electrophoretic and chromatographic methods in the diagnosis of hemoglobinopathies

This review primarily deals with methods for separations of hemoglobins. An introduction considers electrophoretic methods as well as those involving isoelectric focusing and chromatography. The main advantages or disadvantages of each procedure are discussed after each technical description. The chromatographic methods are mainly limited to those used in clinical biochemistry. The second section treats the main diagnostic problems typically met with in the field of the hemoglobinopathies and deals successively with the diagnosis of hemoglobinopathies in the adult and the newborn. Numerous variants have been described in the adult, and among them Hb-S and Hb-C variants are the most frequent. Unstable or high oxygen affinity variants of hemoglobin are also considered. Finally, a new strategy for diagnosis is proposed. A special section is devoted to the diagnosis of thalassemia syndromes. The prenatal diagnosis of hemoglobinopathies is also discussed in some detail with a view to preventing the birth of homozygous children. This update ends with a chapter on the interest of the assay of hemoglobins A1c in the pathology of diabetes mellitus.

A chromatographic method for measuring haemoglobin A1: comparison with two commercial kits

A modified chromatographic method for the separation of haemoglobin A1 is described. Its use for the estimation of the concentrations of the fractions is compared with those of two commercial kits.

Evaluation of a Commercial Kit for Measurement of Glycosylated Hemoglobin in Canine Blood

Total glycosylated hemoglobin (HbA(1)) was measured in canine blood samples by conventional macrocolumn ion exchange chromatography and with a commercial glycosylated hemoglobin kit. The two methods correlated well (r= 0.94, p < 0.001) and the reproducibility of the kit method was good. The commercial kit method is recommended for measurement of HbA(1) in canine blood.

A rapid microchromatographic method for determination of hemoglobin AI

Hemoglobin AI (HbAI) is a minor hemoglobin fraction, which is continuously formed by a glycosylation reaction throughout the life-span of the erythrocyte. The concentration of HbAI reflects the average blood glucose level during the preceding 1 - 2 months and HbAI-determinations have been found to be of value in long-term diabetes control. A simple and reliable microchromatographic method for HbAI-determinations, suitable for routine clinical analyses, is described. In a group of diabetic patients (n = 126) the mean (+ S.D.) S.D.) HbAI--value was 9.6 + 1.8%, in the reference group (n = 33) the corresponding value was 6.0 + 0.8%. A commercial kit (Bio-Rad) based on the same principle was compared with our method. It was of equal accuracy and gave comparable results.