Evaluation of an EIA method for measuring serum levels of the estrogen metabolite 2-hydroxyestrone in adults

Two-hydroxyestrone (2OHE-1) and 16alpha-hydroxyestrone (16OHE-1) are two estrogen metabolites that may play important roles in the development or promotion of breast cancer. Our study assessed the reliability of a newly developed kit procedure for measuring 2OHE-1. Although under certain conditions the assay would not distinguish 2OHE-1 from estriol, or possibly 2-methoxyestrone, steroids such as 17beta-estradiol, estrone and 16OHE-1 should not interfere with the test. Our study evaluated the precision of this enzyme immunoassay (EIA) kit for measuring 2OHE-1 levels in serum obtained from healthy men and women. As a result of several replicate analyses of specimens obtained from 18 men and 20 women, we found that the within-run coefficients of variation (CVs) were approximately 20% and the among run CVs, 30%. Because the SD for the procedure is high, the limit of detection (LOD) was also high (130 ng/l). Nonetheless the assay could distinguish between 2OHE-1 levels in men (128 ng/l) and women (332 ng/l) because we performed a large number of analyses on each specimen. Improving the reproducibility of the assay would reduce the: 1. LOD; number of replicates needed to obtain reliable estimates of 2-OHE-1 levels; amount of time, effort, and cost for each analysis; and greatly improve the reliability of the method. Because the within-run variability is relatively smaller than the total variability (among run + within run), use of the assay for determining differences among groups could be justified only when measurements were made in a single run.

Stability of urine creatinine

The measurement of urinary creatinine is important in many situations, but perhaps most important when measuring the urinary content of substances other than creatinine. Because urine flow changes unpredictably during the day, but total creatinine output is generally constant, many investigators normalize their results to creatinine content (i.e. mg chromium/mg creatinine). Because the widespread use of creatinine levels make the reliability of their measurement important, we decided to test this by studying the effects of storage time and temperature on urine specimens obtained from 10 healthy adults. Our results showed that only prolonged storage time at high temperatures (30 days, 55 degrees C) could cause significant decreases in urine creatinine levels. When stored for 2 days at 55 degrees C the decrease in urine creatinine levels was < 3%. We conclude that in all but extreme cases urine creatinine is virtually unaffected by storage time and temperature.

Integrating smoking cessation into routine public prenatal care: the Smoking Cessation in Pregnancy project

OBJECTIVES

In 1986, the state health departments of Colorado, Maryland, and Missouri conducted a federally-funded demonstration project to increase smoking cessation among pregnant women receiving prenatal care and services from the Women, Infants, and Children (WIC) program in public clinics.

METHODS

Low-intensity interventions were designed to be integrated into routine prenatal care. Clinics were randomly assigned to intervention or control status; pregnant smokers filled out questionnaires and gave urine specimens at enrollment, in the eighth month of pregnancy, and postpartum. Urine cotinine concentrations were determined at CDC by enzyme-linked immunosorbent assay and were used to verify self-reported smoking status.

RESULTS

At the eighth month of pregnancy, self-reported quitting was higher for intervention clinics than control clinics in all three states. However, the cotinine-verified quit rates were not significantly different.

CONCLUSIONS

Biochemical verification of self-reported quitting is essential to the evaluation of smoking cessation interventions. Achieving changes in smoking behavior in pregnant women with low-intensity interventions is difficult.

Passive smoking alters lipid profiles in adolescents

Although cigarette smoking is associated with elevation of plasma lipid levels and changes in lipoprotein distribution, it is not known whether passive smoking is associated with an alteration in lipid profiles. The relation between plasma cotinine, a marker of exposure to tobacco smoke, and lipid profiles was studied in healthy adolescents from a suburban New York high school district who were undergoing preparticipation sports physicals. Forty-four percent of the adolescents reported that one or both parents currently smoked. Eleven percent of the adolescents had plasma cotinine concentrations greater than or equal to 2.5 ng/mL, the level considered indicative of exposure. Adolescents with two smoking parents had significantly higher plasma cotinine concentrations after adjustment for other factors than adolescents whose parents did not smoke. Plasma cotinine concentration greater than or equal to 2.5 ng/mL was associated with an 8.9% greater ratio of total cholesterol to high-density lipoprotein cholesterol (P less than .003) and a 6.8% lower high-density lipoprotein cholesterol (P less than .03). These results suggest that passive smoking, like active smoking, leads to alterations in lipid profiles predictive of an increased risk of atherosclerosis.

Liquid-chromatographic measurement of biopterin and neopterin in serum and urine

We report an improved "high-performance" liquid-chromatographic (HPLC) method for measuring biopterin and neopterin in serum and urine. Specimens are acidified, treated with iodine in 0.2 mol/L trichloroacetic acid, party purified on Bio-Rad MP-50 cation-exchange columns, and analyzed by reversed-phase HPLC with fluorometric detection. The minimal concentration of biopterin detectable is 0.3 micrograms/L in a 50-microL injection. The total CV is less than or equal to 10%. Improvements over other reported methods include the use of a single, simplified sample-preparation step with a Baker-10 SPE System, and a guard column to increase analytical column stability and analyte recovery. The assay is semiautomated to reduce technician time and improve precision. Mean observed values for biopterin and neopterin in sera of normal human adults were 1.64 and 5.52 micrograms/L, respectively. The mean ratio of neopterin to biopterin in acidified adult urine samples was lower than that found in matched nonacidified samples (n = 10). Serum specimens from diagnosed phenylketonuric (PKU) and hyperphenylalaninemic patients were also analyzed for biopterin and neopterin; the findings agreed with reported values for similar patients. One patient, previously identified as an atypical PKU patient, showed serum values of neopterin and biopterin suggestive of a defect in biopterin synthesis.

Liquid-chromatographic measurement of biopterin and neopterin in serum and urine

We report an improved "high-performance" liquid-chromatographic (HPLC) method for measuring biopterin and neopterin in serum and urine. Specimens are acidified, treated with iodine in 0.2 mol/L trichloroacetic acid, party purified on Bio-Rad MP-50 cation-exchange columns, and analyzed by reversed-phase HPLC with fluorometric detection. The minimal concentration of biopterin detectable is 0.3 micrograms/L in a 50-microL injection. The total CV is less than or equal to 10%. Improvements over other reported methods include the use of a single, simplified sample-preparation step with a Baker-10 SPE System, and a guard column to increase analytical column stability and analyte recovery. The assay is semiautomated to reduce technician time and improve precision. Mean observed values for biopterin and neopterin in sera of normal human adults were 1.64 and 5.52 micrograms/L, respectively. The mean ratio of neopterin to biopterin in acidified adult urine samples was lower than that found in matched nonacidified samples (n = 10). Serum specimens from diagnosed phenylketonuric (PKU) and hyperphenylalaninemic patients were also analyzed for biopterin and neopterin; the findings agreed with reported values for similar patients. One patient, previously identified as an atypical PKU patient, showed serum values of neopterin and biopterin suggestive of a defect in biopterin synthesis.

Urinary porphyrins in children exposed transplacentally to polyhalogenated aromatics in Taiwan

In 1979, there was a large (greater than 2,000 cases) outbreak of poisoning due to contaminated rice oil in central Taiwan. The causal agent was a mixture of thermally degraded polychlorinated biphenyls (PCBs), polychlorinated quaterphenyls, and polychlorinated dibenzofurans, which had become mixed with the oil during processing. Patients remained symptomatic for several years afterward, and the chemicals persisted in their tissue. Women who became pregnant had children with high perinatal mortality and a dysmorphic syndrome. We examined urines from 75 children born to exposed mothers after the oil was confiscated, 74 controls, and 12 sibs of the exposed children. Four of the transplacentally exposed children, 2 controls, and 1 sib had a type B hepatic porphyria (i.e., uroporphyrin greater than coproporphyrin); total porphyrin excretion was elevated in the exposed children as a group (95 vs. 81 micrograms/L); and 8 of the 75 exposed children and 2 controls had total urinary porphyrin concentrations of greater than 200 micrograms/L.

Quantitative differences among various proteins as blocking agents for ELISA microtiter plates

We tested instantized dry milk, casein, gelatins from pig and fish skin, serum albumin and several other proteins for their abilities to block non-specific binding (NSB) of a peroxidase-conjugated immunoglobulin to polystyrene microtiter plate wells. Each blocking protein was tested across a million-fold concentration range, both in simultaneous incubation with the peroxidase conjugate and as a pretreatment agent where excess protein was washed away before incubation with the conjugate. Overall, instantized milk and casein were the most effective proteins tested: they inhibited NSB by over 90% in both the simultaneous and pretreatment modes at far lower concentrations than most of eight other proteins. Enzymatically hydrolyzed porcine skin gelatin was the least effective protein tested: it did not reduce NSB by more than 90% even at its highest concentrations; its blocking ability fell rapidly upon dilution; and it was almost useless as a pretreatment agent. Fish skin gelatin showed much better blocking activity than hydrolyzed porcine gelatin, and it still had the practical advantage of remaining fluid even under refrigeration. Our results suggest that some proteins (such as casein) block NSB to plastic primarily through protein-plastic interactions, while others (such as porcine skin gelatin) block primarily through protein-protein interactions. Although the optimal blocking agent for any particular ELISA system must be determined by empirical testing, these results should be helpful in selecting the best possible candidate proteins for further evaluation.

Development and evaluation of a microtiter plate enzyme immunoassay for antibodies to 3,3'-dichlorobenzidine

The authors obtained and evaluated antisera from rabbits injected with a derivative of a potent bladder carcinogen, dichlorobenzidine (DCB), conjugated to bovine serum albumin (BSA). A 14C-radioimmunoassay (RIA) was able to detect the presence of DCB antibodies, but its relative insensitivity led to the development of a more sensitive enzyme immunoassay (EIA). The EIA test was a "sandwich" method in which a second antibody, labeled with an enzyme (horseradish peroxidase), was used to measure antibody binding to transferrin (Tf)-conjugated DCB immobilized on a microtiter plate. Antibody titers measured by RIA were approximately 1:40; when measured by EIA, they were approximately 1:40,000. Antibody specificity was assessed by comparing the antibody binding activities of DCB, BSA, Tf, BSA-conjugated to DCB, and a number of N-substituted aromatic compounds that included benzidine (Bz). Among the compounds tested, the rabbit antiserum reacted only with DCB and the carrier protein, BSA. Moreover, antibody binding activity to Tf-conjugated DCB was significantly inhibited by unconjugated DCB concentrations between 30 and 500 ng/mL. The precision of antibody binding activities as a function of DCB concentration (expressed by the CV) ranged from 9% for low (30 ng/mL) DCB levels to 12% for higher (500 ng/mL) levels. This evaluation suggests that the antiserum obtained would be appropriate for detecting DCB levels at the ng/mL level.

Phenylalanine analyses of blood-spot control materials: preparation of samples and evaluation of interlaboratory performance

Aliquots (0.1 mL) of whole-blood pools prepared to contain various concentrations of phenylalanine were applied to filter-paper collection cards, dried, and stored in sealed bags. We measured the phenylalanine content of the dried blood spots by bioassay, fluorometry, and "high-performance" liquid chromatography, and found that the concentrations remained constant for two years when samples were kept at -20 degrees C or lower. Intra- and interlaboratory studies showed that results for phenylalanine were greater for laboratories using bioassay procedures than for those using fluorometric procedures. Further, CVs (both among- and within-laboratory) obtained with fluorometric procedures were nearly half as great as the CVs obtained by laboratories using bioassay techniques.

Phenylalanine analyses of blood-spot control materials: preparation of samples and evaluation of interlaboratory performance

Aliquots (0.1 mL) of whole-blood pools prepared to contain various concentrations of phenylalanine were applied to filter-paper collection cards, dried, and stored in sealed bags. We measured the phenylalanine content of the dried blood spots by bioassay, fluorometry, and "high-performance" liquid chromatography, and found that the concentrations remained constant for two years when samples were kept at -20 degrees C or lower. Intra- and interlaboratory studies showed that results for phenylalanine were greater for laboratories using bioassay procedures than for those using fluorometric procedures. Further, CVs (both among- and within-laboratory) obtained with fluorometric procedures were nearly half as great as the CVs obtained by laboratories using bioassay techniques.

Evaluation of four reagents for delipidation of serum

Four reagents, Aerosil 380, Freon 113, Dextran sulfate 500-S, and a mixed organic solvent were tested for their abilities to produce optically clear, pooled human serum. Aerosil-380, a silicon dioxide, removed 95% of serum cholesterol and triglycerides, and 80% of the free fatty acids. A mixed organic solvent (n-butanol:diisopropyl ether) was equally effective, but also removed nearly all endogenous alkaline phosphatase and lactate dehydrogenase. Freon-113 and Dextran sulfate 500-S removed about half of the serum cholesterol and triglycerides. The serum content of several non-lipid components was unaffected by Aerosil-380, Freon-113, and Dextran sulfate treatments; however, the mixed organic solvent removed 69% of the endogenous calcium. Light scattering data revealed that treatment with all reagents except the mixed organic solvent resulted in optically-clear serum products.

Liquid-chromatographic measurement of phenylalanine and tyrosine in serum

With phenylalanine ammonia-lyase (EC 4.3.1.5) we converted phenylalanine (Phe) and tyrosine (Tyr) to transcinnamic acid and p-coumaric acid, respectively. These were separated by "high-performance" liquid chromatography and detected at 280 nm. We measured the Phe and Tyr content of human serum by adding 100 mU of the enzyme to a 20-microL serum aliquot, mixing for 2 h at 24 degrees C, then stopping the reaction with 1 mL of cold methanol. Precipitated proteins were removed by centrifugation and the separated clear supernates were stored at -20 degrees C. For chromatographic separation, detection, and quantification, we used a system equipped with a C-18 reversed-phase column, a variable-wavelength spectrophotometer, a printer-plotter, and a microcomputer. The mobile phase was a mixture of dilute aqueous (50 g/L) acetic acid and CH3CN (80/20, by vol). CVs for specimens containing 100 mg of Phe or Tyr per liter varied from 5 to 10%. Analytical recoveries were near 100%.

Liquid-chromatographic measurement of phenylalanine and tyrosine in serum

With phenylalanine ammonia-lyase (EC 4.3.1.5) we converted phenylalanine (Phe) and tyrosine (Tyr) to transcinnamic acid and p-coumaric acid, respectively. These were separated by "high-performance" liquid chromatography and detected at 280 nm. We measured the Phe and Tyr content of human serum by adding 100 mU of the enzyme to a 20-microL serum aliquot, mixing for 2 h at 24 degrees C, then stopping the reaction with 1 mL of cold methanol. Precipitated proteins were removed by centrifugation and the separated clear supernates were stored at -20 degrees C. For chromatographic separation, detection, and quantification, we used a system equipped with a C-18 reversed-phase column, a variable-wavelength spectrophotometer, a printer-plotter, and a microcomputer. The mobile phase was a mixture of dilute aqueous (50 g/L) acetic acid and CH3CN (80/20, by vol). CVs for specimens containing 100 mg of Phe or Tyr per liter varied from 5 to 10%. Analytical recoveries were near 100%.

Interlaboratory comparison of serum thyroxine analyses

Approximately 300 clinical chemistry laboratories participated in a survey of measurements of the thyroxine (T4) concentration in 13 lyophilized serum specimens prepared by spiking a base serum pool with several different levels of T4. Of 35 commercially available kit methods, 4 kits were used by 30 or more laboratories, and 9 by 10 or more laboratories. Values obtained with the Abbott or Nuclear Medical laboratories, kits, which were used by one-third of all laboratories that named a specific kit, averaged 1 to 2 micrograms/dL greater than those obtained with the other methods. Within-run coefficients of variation were 10 to 20% for specimens containing less than 3 micrograms/dL of T4 and 3 to 6% for specimens with more elevated T4 levels.

The effect of temperature and wavelength on the measurement of creatinine with the Jaffe procedure

Studies were conducted to determine the effect of temperature and wavelength on the absorbance of alkaline solutions of picric acid in the presence and absence of creatinine. Absorbance values of an alkaline solution of picric acid were found to be influenced by temperature. At wavelength settings between 475 and 520 nm, absorbance values increased as the temperature increased. The magnitude of the thermochromic response (temperature-induced increase in absorbance) was found to be a function of wavelength: At 490 nm, the response was about three times greater than it was at 500 nm and about fifteen times greater than it was at 520 nm. Other experiments demonstrated that the response was: quantitatively related to picric acid concentration, reversible, rapid, and independent of creatinine concentration.

Linearization of data for saturation-type competitive protein binding assay and radioimmunoassay

Most of the commonly-performed competitive protein-binding radioassay methods utilized in the clinical laboratory are based on the principle of saturation analysis. Although many different methods for linearization of saturation-type assays have been proposed, the algebraic equivalency of all of these methods has not been adequately documented. In this manuscript we have shown the physical and mathematical basis for various methods for linearlization of saturation type assays and the algebraic equivalency of these linearization methods. We have also shown that key parameters such as slope and intercept may be dependent on different components of the assays system with differnt linearization methods. An understanding of these key parameters can help the analyst to evaluate changes in these key parameters and to integrate these parameters in a complete quality control system.

Factors affecting T3 and T4 proficiency testing

Problem areas within a proficiency testing (PT) program are performance evaluation and sample stability. The different units used in the various T3 uptake methodologies make performance evaluation complex. To facilitate this evaluation, a normalization method for T3 uptake performance evaluation has been developed. Sample stability studies for T3 uptake indicate that, at room temperature, sample values increase after storage for about seven days. Room temperature sample stability studies for T4 using a competitive protein binding (CPB) method indicate that the apparent T4 content of pooled serum increases after about one week. Fatty acids are shown to be an interfering substance in the T4 CPB method as well as the T4 radioimmunoassay (RIA) method. This interference increases with a decrease in carbon chain length from C18 to C12 and with an increase in unsaturation of fatty acids. The B/B0 ration for arachidonic acid at a concentration of 0.48 micronMoles per tube is 17.4 in a CPB method and 87.1 in a radioimmunoassay method indicating that the greater effect is in the CPB method. The increase in T3 uptake values are probably also due to the interfering effect of fatty acids.

Optimization of T assay: a model study

Factors involved in the optimization of competitive binding radioassays have been analyzed using a thyroxine (T4) radioassay as a model system. The effects of kinetic errors are minimized by the use of the lowest possible incubation temperature for the separation step and by the use of labeled T4 of high purity (which is not consistently available from commerical sources). The use of labeled T4 containing significant quantities of labeled triiodothyronine (T3) can lead to a marked increase in bias and a decrease in precision with relatively small errors in the separation step due to the relatively short half-life of the T3-binder complex. The determination of the dissociation rate constant (not to be confused with the affinity constant) allows one to make useful estimates of the tolerances that can be allowed in the separation step in order to achieve a desired level of accuracy and precison.

An evaluation of three commercially prepared anion-exchange resin columns for separation of tetraiodothryonine in serum

Three different commercially prepared anion-exchange resin columns for thyroxine (T4) separation from serum for the "T4 by column" assay were evaluated using the protocol of the supplier. The claims by the suppliers for the distribution and recovery of T4 in the first and second thyroxine-containing eluates were experimentally evaluated by the addition of a tracer quantity of purified [125I]T4. The average experimentally determined elution ratios of recoverable T4 (sum of T4 in first and second eluates) with these commercial columns were: Bio-Rad, 92.2:7.8 (90:10 claimed); Oxford, 87.8:12.2 (94:6 claimed); and Curtis Nuclear, 88.7:11.3 (80:20 claimed). The percent recoveries of T4 in the first thyroxine-containing eluate were: Oxford, 82.3 + 9.70 (x +/- 1 S.D.); Bio-Rad, 91.75 +/- 2.09; and Curtis Nuclear, 74.20 +/- 6.14. Mean serum T4 values obtained by the column method with all commercial columns tested were lower than competitive protein binding radio-assay (CPBR) values if the former values were not corrected for recovery. When individual recovery correction factors were applied to column results, improved correlation and better correspondence of "T4 by column" mean values with the CPBR values were noted. It is concluded that the largest part of the total variability of the "T4 by column" assay is contributed by the chromatographic step when the colorimetric step is performed with an automated technique.

Interference of fatty acids in the competitive protein-binding assay for serum thyroxine

An increase in apparent thyroxine values obtained by competitive protein-binding assay on storage of sera is well documented. We find that the major source of this positive bias is probably the unsaturated nonesterified fatty acids. Nonesterified fatty acids cause a positive basis in the competitive protein-binding assay for serum thyroxine because they inhibit the binding of radioactive thyroxine by the serum-binding reagent (probably thyroxine-binding globulin). This inhibition by fatty acids may be due to the formation of a fatty acid/thyroxine complex. The degree of inhibition caused by the fatty acids depends on the length of the carbon chain of the fatty acid and the degree of saturation. Short-chain fatty acids are more potent inhibitors of thyroxine binding than those with a longer chain, and unsaturated fatty acids are more potent inhibitors of thyroxine binding than are saturated fatty acids. The polyunsaturated fatty acid, arachidonic acid, was the most potent inhibitor of all the fatty acids tested. Triglycerides (triacylglycerols) insignificantly inhibit thyroxine binding in this assay.

Interference of fatty acids in the competitive protein-binding assay for serum thyroxine

An increase in apparent thyroxine values obtained by competitive protein-binding assay on storage of sera is well documented. We find that the major source of this positive bias is probably the unsaturated nonesterified fatty acids. Nonesterified fatty acids cause a positive basis in the competitive protein-binding assay for serum thyroxine because they inhibit the binding of radioactive thyroxine by the serum-binding reagent (probably thyroxine-binding globulin). This inhibition by fatty acids may be due to the formation of a fatty acid/thyroxine complex. The degree of inhibition caused by the fatty acids depends on the length of the carbon chain of the fatty acid and the degree of saturation. Short-chain fatty acids are more potent inhibitors of thyroxine binding than those with a longer chain, and unsaturated fatty acids are more potent inhibitors of thyroxine binding than are saturated fatty acids. The polyunsaturated fatty acid, arachidonic acid, was the most potent inhibitor of all the fatty acids tested. Triglycerides (triacylglycerols) insignificantly inhibit thyroxine binding in this assay.

Plasma Renin Activity: A Method for Rapid Screening by Radioimmunoassay and a Comparison of Two Commercially Available Kits

Twenty to thirty percent of patients with essential hypertension have subnormal plasma renin activity (PRA) and lower incidences of stroke and myocardial infarction. NaCl intake influences PRA, which should thus be compared with 24-h urinary sodium excretion. Because such specimens are difficult to collect accurately, we investigated the relation between PRA and sodium in casual (i.e., untimed) urine specimens and found none. However, PRA and urinary Na/creatinine ratio are inversely correlated for casual specimens. This method is useful for screening individuals whose salt intake is low. Mean PRA, measured by the Schwarz-Mann procedure, was 0.62 ± 0.12 (SD) ng/ml per hour for 59 normotensive volunteers, 35% lower by the Squibb procedure.

Interlaboratory and Intralaboratory Comparison of T3-Uptake Methodology

Triiodothyronine (T3)-uptake values in selected human serum pools were compared by use of Abbott's "Trisorb," Mallinckrodt Nuclear's "Res-O-Mat," Curtis Nuclear's "Tri-lonex," and a standard procedure. In addition, interlaboratory variation was measured among laboratories using the Abbott and Mallinckrodt procedures. Study of methodological variables indicated that incubation time and temperature affect T3 measurement. The Abbott procedure is most liable to variations caused by either time or temperature changes; the Mallinckrodt and Curtis methods are least liable to variations caused by changes in time and temperature, respectively. Between-run variation for sera with low T3-uptake values was greater than within-run variation as measured by the Abbott, Mallinckrodt, or standard procedure. For sera with high T3-uptake values, between-run variation and within-run variation are similar; with the Curtis method this was true for all serum pools. We and the reference laboratories found the Abbott method to be most sensitive in detecting changes in T3-uptake values of samples in the "hyperthyroid" range.