Ten commercial kits compared for assay of thyrotropin in the normal and thyrotoxic range

I developed a standard procedure for assessing sensitivity, intra-assay precision, parallelism of sample dilutions, and assay drift, using a single trial assay kit. I used this to compare the performance of 10 commercial kits with an in-house radioimmunoassay for human thyrotropin. No one kit stood out as clearly superior overall. Differences in calibration between kits were evident in a comparison of thyrotropin concentrations measured in clinical samples, and by direct comparison of standards from different kits in a single assay. However, all kits gave clinically consistent results with respect to their published normal reference interval. Moreover, the performance of human thyrotropin assay kits has improved during the 14 months of this study.     

Measurement of alpha-fetoprotein in maternal serum: three commercial radioimmunoassay kits and two non-commercial radioimmunoassays compared

We evaluated three commercial radioimmunoassay kits (Amersham, Dainabot, Clinical Assays) and two non-commercial methods for determining alpha-fetoprotein in maternal serum during pregnancy. All five procedures were found to be acceptable with respect to practicability, sensitivity, linearity, and precision. Similar results were obtained with Dainabot, Clinical Assays, and the two non-commercial methods, but the Amersham method revealed a proportional error, results being about 20% lower than those by the other methods. Use of the international unit system is suggested for reporting results for AFP, to facilitate comparison between methods and laboratories.     

Comparison of countercurrent immunoelectrophoretic assay with commercial radioimmunoassay kits for measuring prostatic acid phosphatase

We evaluated and compared five commercial radioimmunoassay kits with a standard counter-immunoelectrophoretic assay for the measurement of prostatic acid phosphatase in serum. Four of the five radioimmunoassays performed as described by the supplier with respect to sensitivity, stability, precision, linearity, analytical recovery, and expected values for the normal male population. None of the radioimmunoassays was more clinically sensitive then the counter-immunoelectrophoretic assay for detecting increased prostatic acid phosphatase in serum. Results obtained by counter-immunoelectrophoretic assay agreed with results obtained by radioimmunoassay in 96% of the tests. The proportion of positive results in patients with confirmed prostatic adenocarcinoma increased with disease progression. The fewer positive tests in localized adenocarcinoma (Stages A and B) suggests that neither the counter-immunoelectrophoretic assay nor the radioimmunoassay procedures are useful for screening unselected populations for adenocarcinoma of the prostate. The high percentage of normal values found in those patients clinically free of disease after treatment is encouraging and supports the use of the prostatic acid phosphatase immunoassays in prospectively monitoring the treatment of prostatic cancer patients.     

New immunoenzymatic assay for human thyrotropin compared with two radioimmunoassays

We have evaluated a new immunoenzymatic assay for human thyrotropin involving three monoclonal antibodies (Abbott HTSH EIA) and compared the results with those of two conventional nonequilibrium double-antibody radioimmunoassay (RIA) methods: Clinical Assays' RIA and a research RIA (J Clin Endocrinol Metab 1975; 41:676). Mean values for thyrotropin in 100 euthyroid serum samples were similar in the Abbott and Clinical Assays methods, but both sets were significantly higher than those by the research RIA. By all methods, values for hypothyroid patients were clearly higher than values for euthyroid subjects. Results for hyperthyroid and euthyroid subjects were resolved slightly better with the research RIA than with the Abbott kit. The new Abbott assay was far more sensitive than either the Clinical Assays RIA or our research RIA. The correlation of results of the Abbott assay with those of the Clinical Assays and the research RIA exceeded 90% for samples from hypothyroid patients. The Abbott assay replaces radioisotope counting with spectrophotometric detection.     

Results with commercial radioassay kits compared with microbiological assay of folate in serum and whole-blood.

We compared results with three commercial folate radioassay kits [Bio-Rad, New England Nuclear (NEN), and RIA Products] with those by microbiological assay for more than 200 samples of human serum and whole blood. All but one kit (NEN) compared favorably with the microbiological assay for serum samples, although there were notable diagnostic discrepancies. Two kits (NEN and Bio-Rad) were tested on whole-blood samples; both yielded values significantly higher than those by microbiological assay. The frequency distributions of erythrocyte folate data differed strikingly between the two kits; the NEN method yielded a much narrower range of normal values than did either the Bio-Rad or the microbiological assay. Radioassay kits appear to be suitable diagnostic agents for serum folate, if the behavior of a particular kit is investigated thoroughly before its routine use. However, the diagnostic value of radioassays of erythrocyte folate needs to be validated.     

Determination of osteocalcin in human serum: results with two kits compared with those by a well-characterized assay

Osteocalcin, the vitamin K-dependent protein in bone, can also be detected in serum and is receiving increased attention as a marker for bone turnover in evaluating patients with metabolic bone disease. We compared results for patients as determined with two commercial radioimmunoassay kits (Immuno Nuclear Corp. and Seragen Inc.) and with our in-house radioimmunoassay (Methods Enzymol 107: 517, 1984). Results by our assay correlated well (r greater than 0.9) with those by both kits, but the values by the Immuno Nuclear and Seragen methods were respectively 40% and 10% lower than by our radioimmunoassay. Within-run variation (CV) for the two kit methods was respectively 7.3% and 9.8%, run-to-run CV was 9.7% and 8.6%. The standard curve was linear from 1 to 25 micrograms/L for the Immuno Nuclear kit, from 4 to 100 micrograms/L for the Seragen equilibrium method, and from 1 to 25 micrograms/L for the Seragen nonequilibrium method. A second freeze-thaw cycle reduced the serum values by 20% to 40% for both kit methods. A third freeze-thaw cycle further reduced values and eliminated any correlation among methods.     

Parathyrin assay. An analytical evaluation of two commercial Immuno Radiometric Assay kits

We have performed a comparative evaluation of two Immuno Radiometric Assay (IRMA) kits for parathyrin against an existing Radioimmuno Assay (RIA) technique for the measurement of intact parathyrin. The analytical evaluation which was performed in line with the ECCLS recommended kit evaluation protocol showed a marginal improvement in precision with the new assays. There was a substantial improvement in the theoretical limit of detection utilising the IRMA kits although it may prove difficult to realize this improvement in practice with individual samples because of differing protein matrices. The evaluation also demonstrated the degree of parallelism and range of linearity of both kits as well as inaccuracies when compared with the International Reference Preparation for parathyrin (IRP 79/500) as the accepted standard. The demonstration of lack of agreement between measured and kit assigned results for standards when cross-over studies between kits were performed may highlight a possible contributing factor to inaccuracy. Alternatively there may be a difference of antisera avidity within the kits for intact parathyrin. Whilst minor differences in sample stability were demonstrated between the kits, the sample stability was much improved compared to that for intact parathyrin measurement by RIA. The correlation studies showed a degree of correlation consistent with other comparisons similarly performed.     

Diagnostic value of thyrotropin concentrations in serum as measured by a sensitive immunoradiometric assay

A new, highly sensitive immunoradiometric thyrotropin (TSH) assay involving solid-phase-coupled monoclonal antibodies (Boots-Celltech Sucrosep IRMA-TSH) has been evaluated in a wide variety of patients with thyroidal and nonthyroidal illnesses and the results compared with those obtained by conventional diagnostic TSH RIAs. The sensitivity of the present assay ranged from 0.036 to 0.1 milli-int. unit/L (mean 0.056). TSH, measurable in serum of each of 128 euthyroid patients, ranged from 0.1 to 6.3 milli-int. units/L (mean 1.7, SD 1.1). Similar concentrations were found in 15 healthy pregnant women. TSH was undetectable in 27 hyperthyroid patients, of whom six were tested with thyroliberin stimulation and failed to respond. The mean TSH concentration measured in 62 seriously ill hospital patients of 2.7 (SD 2.5) milli-int. units/L was significantly higher (p less than 0.05) than in the euthyroid patients. Basal values and peak TSH responses to thyroliberin testing correlated well (r = 0.63, n = 48), irrespective of clinical diagnosis. We conclude that the present assay readily discriminates between euthyroid and hyperthyroid patients and should replace conventional TSH RIAs in diagnostic laboratories.     

Implications of using different tissue ferritins as antigens for ferritin in serum: four radioimmunoassay kits compared

We compared three serum ferritin radioimmunoassay kits and one noncommercial RIA for ferritin quantitation, with regression analysis of results for sera from 35 ostensibly healthy subjects. There was a good correlation (p less than 0.001) between these various RIAs, but the slope of the regression line varied widely, most probably because of lack of standardization of the serum ferritin assay. Determination of the ferritin content of purified samples of tissue ferritin revealed that the kits differ in specificity, differences for purified human spleen ferritin and human liver ferritin being larger than those for normal sera. Removing the iron from purified liver ferritin increased antiserum binding in two of the kits by twofold, but had no effect in two other kits. We conclude that commercial RIA methods for serum ferritin differ in specificity, and that this difference is related to the source of ferritin used in the production of the antibodies.     

Specificities compared for a radioreceptor assay and a radioimmunoassay of atrial natriuretic peptide

The main active form of atrial natriuretic peptide (ANP) in the circulation is alpha ANP(1-28) [gamma-ANP(99-126)], although some active and inactive metabolites are also though to be present. Some immunoassays for alpha ANP reportedly detect inactive metabolites, as determined by their comparison with results by receptor assay, which would lead to misleading results for alpha ANP(1-28) measurements. Here we compare a radioreceptor assay (RRA) and a radioimmunoassay (RIA) for alpha ANP and show that their results correlate very well with regard to specificity and that the RIA appears to detect only the active circulating form of alpha ANP(1-28).     

The effect of pre-incubation and direct procedure for the alanine transaminase assay in two kits compared.

The kinetic measurement of serum alanine transaminase was carried out on two enzyme kits (Roche and Calbiochem), done as a direct method and a pre-incubation method. Variable results from commercial control sera and patient sera by these four procedures were obtained. Pre-incubation enzyme values were generally lower than the direct values (mean difference 10%), but there were marked differences between the results from the two kits. This finding has important significance especially for laboratories participating in external QC Programmes. Frequent precision checks of the method and a local reference range are therefore necessary.     

Substantial discrepancies in 17beta-oestradiol concentrations obtained with three different commercial direct radioimmunoassay kits in rat sera

The extensive use of oestrogen for contraception and amelioration of post-menopausal symptoms has made it the subject of substantial recent research efforts, and ovariectomized (ovx) rats treated with exogenous ovarial hormones are important when investigating the effects and mechanisms of oestrogen actions. The crucial need to control and monitor plasma levels of 17beta-oestradiol calls for accurate, precise and robust assay methods. The performance of direct radioimmunoassays (RIAs) in measurement of 17beta-oestradiol has been reported previously for human samples, but to our knowledge not for rat samples. In the current study, 552 serum samples from ovx, native and hormone-treated rats were used to compare the performance of three commercially manufactured direct RIAs from the companies DPC (Siemens Healthcare Diagnostics Inc., formerly Diagnostic Products Corporation), DSL (Diagnostic Systems Labs) and MPB (MP Biomedicals, formerly ICN Biomedicals). Substantial differences in results between the three assay methods were found when measuring serum 17beta-oestradiol concentrations. The following formulas describing the relation between the different methods were obtained using weighted Deming's orthogonal regression (based on pg/mL): DSL = 0.43*DPC+12.3, MPB = 2.1*DPC+84.7 and DSL = 4.8*MPB+22.2. Furthermore, a preceding diethyl ether extraction step of the serum appears to impair the performance of the RIAs in the present samples (based on pg/mL): DPC(ex) = 0.39*DPC(unex)+0.76, DSL(ex) = 0.32*DSL(unex)-1.7 and MPB(ex) = 0.22*MPB(unex)+1.4.     

Serum thyroglobulin concentrations in the first weeks of life as measured with an immunoluminometric assay

In addition to the determination of thyrotropin (TSH) on the 5th day of life as a screening parameter for congenital hypothroidism, serum thyroglobulin concentrations were first measured with a commercially available immunoradiometric assay in those cases presenting with elevated thyrotropin levels. As this thyroglobulin assay required at least 400 microliter serum for a duplicate determination, it was decided to employ a sensitive immunoluminometric assay (detection limit 50 amol/tube) instead, the amount of serum needed being reduced to 100 microliter, the sensitivity to under 3 micrograms/l. We describe the serum thyroglobulin concentrations determined by the immunoluminometric assay in both cord and venous blood and in both full-term and pre-term babies divided into 4 main groups with respect to thyroid function. The criteria for the groups with thyroid dysfunction were determined as a result of a) isolated thyrotropin elevation measured in the dry blood spot test (thyrotropin above 4 mU/l), b) lowered thyroxine:thyroxine binding globulin ratio together with elevated thyrotropin both prior to and under substitution with L-thyroxine. In full-term babies thyroglobulin levels in serum fell steadily over the first months of life. The same effect was seen in pre-term babies. The effect of L-thyroxine on the suppression of serum thyroglobulin levels appeared to be dose-dependent, as newborns from experimental criterium group a) (defined above) showed no suppression of serum thyroglobulin levels when under partial substitution with L-thyroxine (median 84 micrograms/l), whereas those in group b), i.e. with full L-thyroxine substitution showed a noticeable suppression of serum thyroglobulin (median 27 micrograms/l).     

Effect of assay conditions on cross reactivity of digoxin-like immunoreactive substance(s) with radioimmunoassay kits

One or more digoxin-like immunoreactive substances (DLIS), most frequently present in serum of premature and full-term neonates. cross react to various extents with different digoxin immunoassay kit reagents. Mostly, this variation is attributed to the relative cross reactivity of DLIS with the antiserum in each kit. However, modification of standard assay procedures for digoxin can also greatly alter the relative cross reactivity of DLIS. Using sequential RIA kit methods for digoxin by 20 to 60% relative to the standard equilibrium RIA mode. Cross reactivity was decreased still more if the concentrations of antiserum (binding-site concentration) and tracer (125 l-labeled digoxin) were decreased, and conversely. Serum samples containing only digoxin, analyzed by the modified method, consistently yielded results well comparable with those obtained with the manufacturers' recommended procedures. We describe use of the different responses to digoxin and DLIS of standard and sequential radioimmunoassays and use of simultaneous equation to calculate the concentration of DLIS (in digoxin equivalents) in digoxin-containing samples.     

Serum 17α-hydroxyprogesterone in infants and children as measured by a direct radioimmunoassay kit

A direct 17 alpha-hydroxyprogesterone (17-OHP) radioimmunoassay kit was used for the assay of samples from 219 infants and children. The kit was used according to the manufacturer's protocol on unextracted serum or plasma and also on reconstituted material extracted from the serum with propanol-heptane. The extraction protocol recovers approximately 88% of 17-OHP. Patients were grouped as infants 3-90 days (96 subjects) or older children, adolescents and young adults 91 days-20 years (123 subjects). 17-OHP results by the direct and extraction protocols correlated but the slopes of the regression lines (0.43 and 0.63) differed in the two groups, indicating that only about 49% of the immunoreactive material measured by the kit in the infants was 17-OHP whereas the corresponding percentage was 72% in the older children. Despite this nonspecificity, the present antibody is much more specific for 17-OHP in the presence of neonatal plasma steroids than that used previously. Reference values were obtained for the two groups using the method with and without an extraction step. 17-OHP values on four untreated patients with CAH were clearly elevated at the time of diagnosis. It is recommended that when the kit is used with neonatal and infant samples, an extraction step should be incorporated to enhance specificity.