Table 1. Thyroid Function indices in the ... - Clinical Chemistry

0 downloads 0 Views 511KB Size Report
18.5 and 17.5 mg/L; lots KOl and NOl yield positive results for minimum di- phenhydramine, concentrations of 170 and 52 mgfL respectively. Syva is also.

18.5 and 17.5 mg/L; lots KOl and NOl yield positive results for minimum diphenhydramine, concentrations of 170 and 52 mgfL respectively. Syva is also currently reformulating the assay such that its sensitivity to diphenhydramine is greatly diminished. Beginning with the 1984 lot R04, only concentrations exceeding 100 mg/L give positive responses. In all cases, only diphenhydramine in concentrations at or above the toxic range can elicit positive responses from the assay. Therefore, the probability of a false result for methadone due to diphenhydramine would have occurred only in situations of overdose in the earlier lots, and would be much less in later lots, in which cross reactivity to this drug is considerably less. References 1. Baselt RC. Disposition of Toxic Drugs and Chemicals in Man, Vol. I, Biomedical Publications, Canton, CT, 1978. 2. Baselt RC. Analytic Procedures for Therapeutic Drug Monitoring and Emergency Toxicology, Biomedical Publications, Davis, CA, 1980. Lorenzo

A. Ajel

Quality Assurance Dept. Syva Co. 900 Arastradero Rd. Palo Alto, CA 94303-0847

Thyroid-Function Indices In an Analbuminemic Subject Being Treated with Thyroxin for Hypothyroidism To the Editor:

In analbuminemia, interpretation of tests such as free thyroxin (free T4) or free thyroxin index (VII) is difficult, because of abnormal serum concentrations of several of the thyroid hormone-binding proteins, typically little (or no) serum albumin (13) but also unusually large amounts of thyroid-function

thyroxin-binding globulin (TBG) (1,2). Because TBG is an important contribu-

tor to T4 binding in serum, correspondingly high concentrations of T4 and triiodothyrornne (T3) are needed to yield free hormone concentrations typical of euthyroidism (2). However, the highly abnormal pattern of thyroid hormone-binding proteins seems otherwise to have no significant influences on normal thyroid function (1,2), or on general health (3). Here we describe the investigation of thyroid function in a 47-year-old woman who was discovered in 1975 to have a concentration of serum albumin of 2.7 gIL. Full investigations substantiated the diagnosis of analbuminemia. Her only clinical symptoms were a mild swelling of the legs, controlled by diuretics. The serum cholesterol had at all times been twice the normal value, as is a feature of analbuminemic patients, and in 1980 symptoms of intermittent claudication developed in both legs. In 1981, the patient presented with symptoms suggestive of hypothyroidism. After being examined and undergoing thyroid-function tests, the patient was diagnosed as being mildly hypothyroid and was given oral T4 medication; by 1984, a daily regime of 100 g had been established for two years. The patient now enjoys improved activity and general health and is judged to be clinically euthyroid. Results for the patient in 1981 (on first presentation with suspected hypothyroidism) and in 1984 (when T4 therapy was established) are shown in Table 1. Amersham International plc RIA kits were used for the thyroidfunction tests: Amerlex free and total T4, free and total T3, and thyrotropin (TSH), and the T3 uptake (MAA) assay. Table 1 also shows results for free T4 measured with an assay reformulated to minimize the effects of low serum albumin concentration, as described elsewhere (4-7). Free T4 concentrations in serum were also calculated from the concentrations of total T4, TBG, prealbumin, and albumin as described by Wilkins et al. (8). Concentrations of TBG, prealbumin, and albumin were measured by rocket immunoelectrophoresis (9).

Her serum concentrations of TBG (two to threefold normal) and albumin (very low) were typical of an analbuminemic subject (1,2); prealbumin concentrations were within the normal range. The serum concentration of TBG was rather higher in the 1984 sample, but this may be a secondary effect of changes in health over the three-year period. Concentrations of total T4 were increased in both samples, and were considerably above the normal range in therapy. Nevertheless, because TBG concentrations were so high, the T4/TBG ratio before and after T4 therapy was below normal. The Amerlex free T4 values were below normal before therapy and normal afterwards, though the results before T4 therapy was begun indicated a more extreme hypothyroid state. This result was not in agreement with calculation, which suggested a borderline condition. The free T4 values from the reformulated assay were more in harmony with the calculated values, both before and after the beginning of therapy. Serum TSH concentrations were just above the normal range before therapy, indicating a borderline hypothyroid condition, but were well within the normal range when therapy was established. The FF1 was incapable of correcting for the high TBG content of this serum: before therapy, the FF1 value was high-normal, and afterwards, was well above normal, falsely indicating hyperthyroidism. Interestingly, the total T3 concentrations at both times were within the normal range; because of the high TBG values, however, the T3ITBG ratio was extremely low, forecasting a correspondingly low concentration of free T3. This low free T3 was found by use of the Amerlex kit, though the values during therapy were higher than those before. The results from the tests (with the exception of F1’I) are in basic agreement with the patient’s clinical condition on first presentation for hypothyroidism and also after euthyroidism had been established by T4 therapy. The effect of grossly decreased serum albumin concentrations on the original

Table 1. Thyroid Function indices in the Analbuminemic Subject TBG






T3 uptake,








TSH, mliii-
























After therapy with T4 47







11.18 l3.2’

14.2C Reference intervals 6-16


PA, prealbumin.




From the original Amerlex kit.






From the reformulated assay, now marketedas the Amerlex-M”

FT4 AlA kit. C Calculated values.

CLINICAL CHEMISTRY, Vol. 31, No. 2, 1985


Amerlex free T4 values can be roughly quantified as giving about 60% of the “true” value when albumin is absent. Much of this distortion is removed in the reformulated assay, giving results closer to theoretical expectation and to the patient’s clinical status. These results conflict to some extent with those of Stockigt et al. (2) for another analbuminemic subject, for whom no detailed clinical background was given. The degree of distortion of the Amerlex free T4 results in our patient was considerably less than reported by those workers (2), even though other determining free T4 (e.g., total T4 and TBG) were present in similar concentrations. There is no doubt that, in these extremely rare subjects, Axnerlex free T4 values are affected by the decrease in serum albumin, but there is equally no doubt that methods such as FF1 are even less valid; moreover, estimates of free T4 by equilibrium dialysis can be affected by serum fatty acids (1). By suitable kit reformulation, much of the distortion caused by low serum albumin has been removed, allowing a more nearly accurate overall assessment of the clinical state of this unique subject, and suggesting appropriate therapeutic action. Thus, other subjects with decreases in serum albumin, e.g., those with nonthyroidal illness, should also be more validly assessed by the improved method. We thank Dr. P.G.H. Byfield and Mr. M.R.A. Lalloz, Clinical Research Centre, Harrow, U.K., for kindly measuring the concentrations of thyroxin-binding protein in serum. References 1. Hollander CS, Bernstein G, Oppenheimer JH. Abnormalities of thyroxin binding in analburninemia. J Clin Endocrinol Metab 28, 1064-1066 (1968). 2. Stockigt JR, Stevens V, White EL, Bar-

low JW. “Unbound analog” radioirnmunoassays forfree thyroxin measure the albuminbound hormone fraction. Clin Chem 29, 1408-1410 (1983). 3. Tarnoky AL. Genetic and drug-induced variation in serum albumin. Adv Clin Chem 21, 101-146 (1980). 4. Stockigt JR, White EL, Barlow JW. What do radioimmunoassays for free thyroxin using “unbound analogues” actually measure? Lancet ii, 712 (1982). Letter. 5. Amino N, Nishi K, Nakatani K, et al. Effect of albumin concentrations on the assay of serum free thyroxin by equilibrium radioimmunoassay with labeled thyroxin analog (Atnerlex#{174} Free T4). Gun Chem 29, 321-325 (1983). 6. Bayer MF. Free thyroxin results are affected by albumin concentration in nonthyroidal illness. Clin Chim Acta 130, 391396 (1983). 7. Ooi DA, Sorisky A. Effect of albumin on results of analog-type assays for thyroxin. Clin Chem 30, 1109-1110 (1984). 342

8. Wilkins TA, Midgley JEM, Giles AF. Theoretical basis, computer simulation, optitnisation and technical validation of a new direct free ligand assay principle, with particular reference to the measurement of free thyroxin. In Radioimmunoassay and RelatedProcedures in Medicine, Int Atomic Energy Authority, Vienna, Austria, 1982, pp 221-240. 9. Byfield PGH, Lalloz MRA, Pearce CJ, et al. Free thyroid hormone concentration in subjects with various abnormalities of binding proteins: Experience with Amerlex Free-T4 and Free-T3 assays. Clin Endo-

populations and sampling procedures used. The position of any one method on the ICES scale based on reference range is therefore uncertain. Precise location of methods relative to each other and to the IFCC method using this approach would require that all reference ranges be established on one and the same subset of reference, indi-

crinol (Oxford) 19, 277-283



H S Platt Basingstoke Basingstoke,

District UK.


N. Barron A. F. Giles J. E. M. Midgley T. A. Wilkins Clin. Reagents Res. and Devel. Amersham International plc Amersham, UK.


Will Implementationof ICES Delay Method Standardization InClinical Enzymology? To the Editor:

In a Special Report, Bowers and McComb proposed “to promote comparability of currently incompatible numerical results for ASAT [aspartate aminotransferase] through the use of one ASAT scale of units,” the “International Clinical Enzyme Scale” (ICES) (1). This scale-unification concept “would permit all current methods, instruments, and temperature choices to be used for ASAT determinations in the daily working laboratory.” This philosophy is extended to other enzyme methods in the Discussion: “As shown here with a model for ASAT, the use of a hierarchy of stable reference, calibrator, and control materials carrying ICES values makes possible the direct calibration of any other working method at any other temperature at any loci to the one unifjring scale of ICES” (italics added). ICES. The scheme represented by the new acronym ICES professes to

reduce nearly all problems of enzyme methodology to being one of numerology and scales. The authors argue their point by use of an “ICES” figure showing “how results by the IFCC Reference Method for ASAT at the gallium reference temperature of 29.77 #{176}C are related to the many commonly used ASAT methods” (1). However, this figure disregards the considerable variation in upper and lower reference limits determined for each method, depending on the different reference

CLINICAL CHEMISTRY, Vol. 31, No. 2, 1985

viduals. Consequently,

an ICES figure comprising many different methodologies is idealized and should not be used for The authors (1) recalculated data from a 1976 CDC ASAT survey that included two controls, R31 and R30. The ASAT (IFCC) value of R31 was determined. For each laboratory (x) a calibration factor, F, was then calculated: F



R31 (IFCC) R31(x)

The second control, R30, was then calibrated by this factor. This “ASATI ICES” calibration reduced the apparent interlaboratory imprecision from the original 34% (CV) to 11% (CV). The philosophy of ICES. The ICES concept is an attempt to cure the symptoms rather than the disease, which is the unacceptable flora of widely different enzyme methods. ICES is based on the oversimplification that calibrators, controls, and patients’ sera react similarly under any reaction conditions, and that enzyme methodological problems are simply problems of numerical scales. Unconditional calibration is an illusion. Enzyme determinations by catalytic activity are method dependent.

Different methodologies not only yield different numerical values, they also vary considerably in analytical bias for different isoenzyme composition in the samples, activation of enzymes, and suppression of interfering enzymes. All these factors vary with individual patients’ sera. Consequently, while postcalibration values for specified reference and control materials may provide the illusion that a number of widely different methods function nearly identically, results with actual patients’ seru may still have unacceptable variations of measured activity. We provide some examples: ASAT (EC For a method without pyridoxal phosphate (PP) the catalytic activity depends on the concentration of (PP) in the sample (2). Calibration of an ASAT method without PP to (e.g.) the IFCC reference method by use of a calibrator with PP will produce too-low post-calibration values with patients’ sera, depending on the degree of PP deficiency in these. Calibrating by use of a calibrator without PP will tend to give too-high

Suggest Documents