FREQUENCY OF EUTHYROID SICK SYNDROME AS ASSESSED BY FREE THYROXINE INDEX AND A DIRECT FREE THYROXINE ASSAY A Limitation of FT4 Assays

1373

Hospital Practice

have become aware of a five-fold increase in the apparent incidence of ESS. We now report our experience of ESS as assessed by both conventional FTI and the amerlex direct FT4 assay in the Johannesburg Hospital.

FREQUENCY OF EUTHYROID

SICK SYNDROME AS ASSESSED BY FREE THYROXINE INDEX AND A DIRECT FREE THYROXINE ASSAY A Limitation of FT4 Assays

W. J. H. VERMAAK

W. J. KALK

W. J. ZAKOLSKI

Department of Chemical Pathology, School of Pathology, The South African Institute for Medical Research and University of the Witwatersrand; and Endocrinology Unit, Department of Medicine, University of the Witwatersrand, Johannesburg, South Africa

frequency of the euthyroid sick syndrome (ESS) was assessed in 3444 hospital patients. For the first 6 months the free thyroxine index (FTI) was used: 3·6% of 1674 patients had ESS compared with 17·3% of 1770 in the second 6 months diagnosed by a direct free thyroxine (FT4) assay (’Amerlex’). The relative frequencies of each type of ESS (low triiodothyronine [T3] only, low T3 plus low FT4 or FTI, and high FT4 or FTI with normal or low T3) and the frequencies of hypothyroidism (high TSH) and hyperthyroidism (high FTI/FT4 and T3) were similar in both periods. In the FT4 period requests for T3 assays increased by 300%. Measurements of both FT4 and in 46 patients with ESS confirmed that ESS is diagnosed more frequently with the FT4 assay: the distribution of normal and abnormal T3 and reverse T3 levels were similar in FT4 ESS and FTI ESS. FT4 is labile in illness and thus direct FT4 measurements also fluctuate, not necessarily reflecting thyroid status. Summary

The

INTRODUCTION

THE typical clinical manifestations of both hyperthyroidism and hypothyroidism can be obscured by intercurrent illness, and the presentation of both disorders may be atypical in the elderly. In these circumstances the accurate diagnosis of thyroid status may be particularly important, and greater reliance is placed on screening tests of thyroid dysfunction. Lack of familiarity with the effects of non-thyroid illness on standard tests of thyroid function.adds to the difficulties in making a firm assessment of thyroid functional status in ill patients, since high, normal, or low thyroxine (T 4) and free thyroxine index (FTI) values and suppressed triiodothyronine (T3) levels may be found.’-’ The euthyroid sick syndrome (ESS) is the most common biochemical abnormality of endocrine function among medical inpatients; the incidence may be as high as 70%.8 ESS may be defined as a variable array of abnormal thyroid function tests found in clinically euthyroid patients with acute or chronic systemic illnesses. Total T3 concentrations are often below normal and T4 and FTI values may be low, normal, or high, while serum thyroid-stimulating hormone (TSH) levels usually remain within normal limits. Reverse T3 (rT3) levels are often slightly or greatly above normal. 6,7,10 Assays for the direct measurement of free thyroxine (FT4) in serum have become widely available; reports have been on the effects of acute and chronic illnesses on several direct FT4 assays, but effects differ with the assay.l0-13 Little information is available on the influence of illness and drugs on the ’Amerlex’ FT4 assay (Radionuclear Centre, Amersham, UK). 14,15 Since introducing this assay as the screening test for thyroid function in our hospital, we

published

PATIENTS AND METHODS

During a 12-month period, 3444 consecutive thyroid hormone profiles were carried out- 1674 by FTI in the 6 months before the laboratory changed to the FT assay and 1770 in the subsequent 6 months with FT4 as the screening test. 816 abnormal results were found and the case records of each of these patients were scrutinised. The patients with abnormal results formed three groups: those, obviously hypothyroid (high serum TSH); those clinically and biochemically thyrotoxic (raised serum T3); and those judged clinically euthyroid, the ESS group. This group was further subdivided: T3 below normal but FTI/FT4 normal ; FTI/FT4 and total T3 below normal; and FTI/FT4 high but T3 normal or below normal. Serum concentrations of TSH were normal in all the ESS patients. Serum samples from 46 clinically euthyroid patients who had abnormal FT4 but normal TSH and normal or low T3 values (<2 - 5nmol/1) were assayed for T4, T3 resin uptake (T3RU), FTI, and rT3 in addition. Commercial radioimmunoassay kits were used for all thyroid hormone measurements, and the manufacturers’ protocols were used in all the assays. Total T4 was measured by solid-phase ’Gamma Coat’ tubes (Clinical Assays, Cambridge, Mass) and kits were used to measure T3RU, total T3 (Clinical Assays), rT3 (Serono-Biodata Diagnostics, 1267 Coisins, Switzerland), and TSH (Becton Dickinson, Orangeburg, New York). FTI was calculated from T4 x T3RU/ 100. All serum samples in the study were part of the normal routine analysis in the hospital radioimmunoassay laboratory. Results were accepted only if all the internal qualitycontrol samples fell within acceptable limits in each assay run. Reference ranges were: total T4 60-150 nmol/1; T3RU 30-40%; total T3 1 .5—3 -5nmo1/1 ; TSH 0-8 mUll; FT4 10-25 pmol/1; and rT3 0139-0’539 nmol/1. Chi-squared analysis was used to compare the frequencies of the various thyroid hormone disturbances during the two study periods. ,

RESULTS

298

(17 . 8°70) patients

assessed

by

FTI had abnormal

results, compared with 518 (29.3%) abnormal by the direct FT4 assay (A2=61.9 ; p<0.001). The numbers of patients

diagnosed as being definitely hypothyroid or hyperthyroid by each method were similar (table I). In contrast there was a five-fold increase in the frequency of the ESS in the second Thus, by the FTI, 21% of period (A2= 166 ° 8; p<0.001). with results had abnormal ESS, whereas by the direct patients FT4 assay 58’6% were diagnosed as having ESS (X2 = 104.8; p<0. 001). Within the ESS groups the distribution of patients with low T3 or high or low FTI or FT4 values was similar

(table II). Of 46 consecutive ESS cases diagnosed on the basis of an abnormal FT, normal TSH, and normal or low T3 levels, FT4 was high in 34 and low in 12. The FTI was also measured in these samples; FTI was elevated in only 16 (47%) and low in 8 (67%). Although this finding was not representative of the larger group because of the lower TABLE

I-FREQUENCIES

OF

HYPOTHYROIDISM, HYPERTHYROIDISM,

AND ESS DURING FTI AND

FT4 PERIODS

1374 TABLE

II-FREQUENCIES OF SUBTYPES OF ESS IN

FTI AND

FT4

PERIODS

TABLE

rr3

III-FREQUENCIES

AND

IN THE

frequently

in the second 6

months, after FT,

was

introduced.

OF NORMAL AND ABNORMAL LEVELS OF

46 PATIENTS SELECTED ON THE BASIS OF AN

T3 FT4 CONCENTRATION BUT NORMAL OR LOW T3 LEVELS

ABNORMAL

*T3, was assayed in 27 patients with high FT4

more

levels.

of patients with lower than normal thyroxine levels, ESS was nevertheless diagnosed almost twice as frequently by FT4 than by FTI. The proportions of subjects with normal or low T3 levels or with normal or high rT33 levels in the FT and FTI ESS groups were similar (table III).

frequency

DISCUSSION

It appears from our findings that the amerlex FT4 assay was substantially more sensitive than the FTI in detecting ESS. The observation was confirmed by the detailed study of the 46 consecutive subjects with the diagnosis of ESS based on abnormal FT4 concentrations. Both high-T4 and low-T4 ESS were diagnosed more frequently with the FT4 assay than with the FTI. In addition in the patients in this group with either high FT4 or high FTI values, the proportions with low or normal T3 levels and with normal or high rT3 levels were similar. Measurements of rT3 were helpful in confirming the diagnosis of ESS only in patients with low FT4 and FTI values. Since high circulating rT3 concentrations may be found in both ESS and in hyperthyroidisml,5,6 measurements of this metabolite are not helpful in differentiating the two conditions. Much has been written about the difficulties of interpretation of low T4 and T3 levels and indices in patients with various acute and chronic illnesses; with few exceptions serum TSH levels are normal in such patients. The diagnosis of suspected primary hypothyroidism can easily be differentiated from low-T4 ESS by the finding of high serum TSH concentrations, even in the presence of intercurrent illness 13 including myxoedema coma. In our experience highT ESS is the least common form of this syndrome, but raises the greatest difficulties in the differentiation of the ESS from thyrotoxicosis in ill patients. High T levels may occur in euthyroid sick patients as a result of their illness, contrast radiology and iodine-containing drugs, therapy, or because of high levels of binding proteius.I,7,l6-18 High FT4levels may be found in both low-T4 and high-T4 ESS6,7,9,10-the frequency depends on the assay used to measure FT 4.10,12,13.15 A proportion of our patients with presumed high-FTI or high-FT ESS may have been thyrotoxicl and the normal or low T3 levels the result of intercurrent severe illnesses.l9 Thus some patients with mild biochemical thyrotoxicosis, who were judged clinically euthyroid, may have been missed in both periods. It seems unlikely, however, that the diagnosis of thyrotoxicosis was missed five times

We can give theoretical explanations for the higher incidence of ESS as diagnosed by FT measurements. FTI is in dynamic equilibrium with bound hormone; approximately 0-016% of the total exists in the free form. Any disturbance of this equilibrium may therefore cause a profound effect on the free hormone fraction. Interference with T binding to carrier proteins, with consequent alterations of the free T4 fraction, is well recognised in illness: changes in the free fraction may be induced by the illness itself,lo.2O possibly by metabolic disturbances,7,21 or by therapy.15,17,22 Thus in sick patients who are treated with a variety of drugs and may have associated quantitative and qualitative abnormalities in the thyroxine-binding proteins,7,18,19 FTI values are frequently outside the normal range. For example, heparin-induced high levels of circulating free fatty acids decrease or increase FT4 concentrations, depending on the type of FT4 assay used.ITS Precisely what is being measured by the various FT4 assays remains to be clarified.I2,z3The clinical objective m measuring circulating thyroid hormone levels is to ascertain the functional status of the thyroid gland itself. Therefore, the ideal test should reflect only the thyroid’s functional status and should not be influenced by changes in the peripheral economy of the circulating thyroid hormones, which is especially liable to changes induced by acute and chronic illness and drugs. The frequency of abnormal FT4 levels in euthyroid patients with non-thyroid illness, measured by a variety of methods,I1,12 casts doubt on the suitability of any FT measurements as screening tests for thyroid functional status in sick patients. Although the amerlex FT assay appears to give a more sensitive measure of thyroid function than the FTI in healthy individuals,24 the high incidence of ESS diagnosed by this method at our hospital suggests that it too is an unreliable test in acute and chronic illness. Furthermore, after the introduction of the FT assay we observed a 300% increase in the number of requests for T3 measurements, presumably in an attempt to clarify the large number of abnormal FT4 results. Requests for additional tests reduced the cost-effectiveness of the single FT4 assay. In our experience, the conventional FTI appears to be a more robust test than the FT in sickness and diagnoses fewer patients with ESS. Moreover, the inclusion of the T3RU test may be useful as an independent marker of low- T4 ESS, because values are often high in this syndrome. 6,7 At present, there seems to be no clear advantage to the routine use of FT4 assays in sick people. Indeed, in patients with non-thyroid illnesses, high- T or low- T ESS may compound the diagnostic difficulties. We conclude that the usefulness of T screening tests for thyroid function in sick patients is debatable, whether FTI or FT measurements are used; more specific tests will often be necessary to establish a definite diagnosis of thyroid gland dysfunction in patients who are likely to have ESS. The differentiation of low- T ESS from hypothyroidism may be of more than theoretical importance: the changes in circulating thyroid hormones in non-thyroid illness may be adaptive and treatment with thyroxine might be harmful.=’ In primary hypothyroidism serum TSH concentrations remain high during intercurrent illnesses. 13The TSH response to intravenous thyrotropin-releasing hormone (TRH) is probably the best test for excluding the diagnosis of hyperthyroidism. In most acutely ill patients there is a normal TSH response to TRH, although peak elevations may be reduced and delayed;20,26 a response excludes the diagnosis of

1375

hyperthyroidism. The TSH response may be absent in euthyroid patients with some chronic illnesses,2,27,28 but have low- T ESS. Therefore, we suggest the use of the TRH test, with several blood samples, to confirm or exclude the diagnosis of suspected hyperthyroidism in sick patients, including those with possible high-T4 ESS, although data on the TRH test in this condition are limited.29 these

Correspondence should be addressed to W. J. H. V., Department of Chemical Pathology, SAIMR, PO Box 1038, Johannesburg 2000, South Africa. REFERENCES

M, Burer T, Busset R, Burger A Diagnosis of hyperthyroidism when thyroxine alone is raised. Lancet 1977; ii: 53-56. Burrows AW, Cooper E, Shakespear RA, et al. Low serum L-T, levels in the elderly sick: protein binding, thyroid and pituitary responsiveness, and reverse T,

1. Birkhauser serum

2

concentrations.

3.

4. 5

6.

Toxicology

patients usually

Clin Endocrinol 1977; 7: 289-300.

Mahrajan G, Etta KM, Singh A, Ahuja IS, Ahuja GK Thyroxine, triiodothyronine and thyrotrophin levels in meningococcal meningitis, typhoid fever and other febrile conditions., Clin Endocrinol 1978, 9: 401-06. Levy RP, Jensen JB, Laus VA, Agyle DP, Engel IM Serum thyroid hormones in abnormalities in psychotic disease Metabolism 1981, 30: 1060-64. Burger A, Nicod P, Suter P, Vallottan MB, Vagenakis A, Braverman L. Reduced active thyroid hormone levels in acute illness. Lancet 1976; i: 653-55. Chopra IJ, Solomon DH, Hepner GW, Morgenstein AA. Misleadingly low free thyroxine index and usefulness of reverse truodothyronine in non-thyroidal

illnesses. Arch Intern Med 1979; 90: 905-12. T, Liewendahl K Abnormal thyroid function tests in severe non-thyroidal illness: diagnostic and pathophysiologic aspects. Scand J Clin Lab Invest 1979; 39: 389-97. 8 Gooch BR, Isley WL, Utiger RD. Abnormalities in thyroid function tests in patients admitted to a medical service. Arch Intern Med 1982; 142: 1801-05. 9 Chopra IJ, Chopra U, Smith SR, Resa M, Solomon DH Reciprocal changes in serum concentrations of 3,3’,5’-triiodothyronine (reverse T,) and 3,3’,5 triiodothyronine (T,) in systemic illness. J Clin Endoerinol Metab 1975; 41: 1043-48. 10. Chopra IJ, Van Herle AJ, Chua Teco G, Nguyen AH. Serum free thyroxine in thyroidal and non-thyroidal illness: a comparison of measurements by radioimmunoassay, equilibrium dialysis and free thyroxine index. J Clin Endocrinol Metab 1980, 51: 135-43. 11. Witherspoon LR, Shuler SE, Garcia MM, Zollinger LA. An assessment of methods for the estimation of free thyroxine. J Nucl Med 1980; 21: 529-39 12. Kaptein EM, MacIntyre SS, Weiner JM, Spencer CA, Nicoloff JT. Free thyroxine estimates in non-thyroidal illness: comparison of eight methods. J Clin Endocrinol Metab 1981; 52: 1073-77 13. Melmed S, Geola FL, Reed AW, Rehavy AE, Park J, Hershman JM. A comparison of methods for assessing thyroid function in non-thyroidal illness. J Clin Endocrinol Metab 1982; 54: 300-06. 14. Stockigt JR, DeGarris M, Csicsmann J, Barlow JW, White EL, Hurley DM. Limitations of a new free thyroxine assay (Amerlex Free T4). Clin Endocrinol 1981; 15: 313-18. 15. McDougall R, Bayer MF, Nierenberg D, Lewis SJ Disparate effects ofheparin on free thyroxine as measured by two radioimmunoassays. J Nucl Med 1982; 23: 507-10 16. Fehcetta JV, Green WL, Nels WB. Inhibitors of hepatic binding of thyroxine: by cholecystographic agents. J Clin Invest 1980; 65; 1032-40. 17. Cooper DS, Daniels GH, Ladenson PW, Ridgeway C Hyperthyroxinemia in patients treated with high dose propranolol Am J Med 1982, 73: 867-71. 18. Kalk WJ, Kew MC, Danilewitz MD, Jacks F, Van der Walt LA, Levin J. Thyroxine binding globulin and thyroid function tests in patients with hepatocellular carcinoma Hepatology 1982; 2: 72-76. 19. Engler D, Donaldson EB, Stockigt JR, Taft P. Hyperthyroidism without triiodothyronine excess: an effect of severe non-thyroidal illness. J Clin Endocrinol Metab 1978; 46: 77-82. 20 Talwar KK, Sawhney RC, Rastogi GK. Serum levels of thyrotropin, thyroid hormones and their response to thyrotropin releasing hormone in infective febrile illnesses. J Clin Endocrinol Metab 1977; 44: 398-403. 21. Spaulding WS, Gregerman RI Free thyroxine in serum by equilibrium dialysis: effects of dilution, specific ions and inhibitors of binding. JClin Endocrinol Metab 1972; 34: 974-86. 22. Mardell R, Ganlan TR Artificial reduction in circulating free thyroxine concentrations by radioimmuno-assay. Lancet 1982; i. 973-74. 23. Tuttlebee JW. Further experience with free thyroxine assays with particular reference to pregnancy Ann Clin Biochem 1982; 19: 374-78. 24. Midgely JEM, Wilkins TA. The direct estimation of free hormones by a single equilibrium radioimmunoassay Amersham UK: Amersham International Ltd, 1981. 25. Werner SG. Thyroid function tests in non-thyroidal illness are diagnostically and prognostically important. Arizona Med 1981; 38: 356-59. 26. Demeester-Mirkine N, Kutnowski M, Futeral B, Brauman H, Corvilain J Thyroid status in elderly sick patients. J Clin Invest 1981; 4: 41-44. 27. Green JRB, Snitcher EJ, Mowat NAG, Ekins RP, Rees LH, Davidson AM Thyroid function and thyroid regulation in euthyroid men with chronic liver disease: evidence of multiple abnormalities. Clin Endocrinol 1977; 7: 453-61 28. Kalk WJ, Morley JE, Gold CH, Meyers A. Thyroid function tests in patients on regular 7. Helenius

haemodialysis. Nephron 1980, 25: 29 Lambert

173-78.

MJ, Burger AG, Galeazzi RL, Engler D.

Are selective

increases in serum

thyroxine (T4) due to iodinated inhibitors of T4 monodeiodination hypothyroidism? JClin Endocrinol Metab 1982; 55: 1058-65.

indicative of

IS THERE A PLACE FOR CIMETIDINE OR ETHANOL IN THE TREATMENT OF PARACETAMOL POISONING?

J. A. J. H. CRITCHLEY

E. H. DYSON A. W. SCOTT D. R. JARVIE L. F. PRESCOTT

University Department of Therapeutics and Clinical Pharmacology, and Department of Clinical Chemistry, Royal Infirmary, Edinburgh; and Royal Army Medical Corps, Ritchie Camp, Kirknewton, Midlothian In healthy volunteers, cimetidine had no effect on the metabolism of paracetamol (acetaminophen) whereas with ethanol there was a significant decrease in formation of the hepatotoxic metabolite.

Summary

INTRODUCTION

PARACETAMOL (acetaminophen) is converted by hepatic mixed function oxidase to a highly reactive metabolite. This is normally inactivated by conjugation with reduced glutathione and eventually excreted in the urine as the mercapturic acid and cysteine conjugates.1 After a toxic dose of paracetamol, hepatic glutathione is depleted and liver damage ensues. However, early administration of glutathione precursors such as N-acetylcysteine is remarkably effective in preventing liver damage.2 An alternative approach to therapy is inhibition of the metabolic activation of paracetamol. 3-8 Cimetidine inhibits the oxidative metabolism of many drugs9 and there is some evidence that in large doses it can protect rodents against acute paracetamol hepatotoxicity.3-7 On these grounds, cimetidine has been proposed as an antidote for paracetamol poisoning in man6-8 and there have been anecdotal reports implying efficacy.3 However, the assumption that cimetidine protects by inhibiting the metabolic activation of paracetamol does not seem to have been confirmed. Ethanol is another agent that inhibits drug metabolismlo and is frequently taken by self-poisoners. We have therefore investigated the effects of cimetidine and acute ethanol on the metabolic activation, and hence potential toxicity, of paracetamol in man. METHODS

On three occasions at least a week apart, ten fasting healthy male subjects consumed paracetamol (20 mg/kg) dissolved in 400 ml of ’Coca Cola’ over 2 min. Five were normal subjects who took ethanol in moderation only occasionally but five were heavy drinkers who consumed at least 8 pints (4-77 1) of beer or the equivalent amount of ethanol daily and showed evidence of alcohol-related problems on the Michigan Alcoholism Screening Test.1The mean age of the ten subjects was 37’ 1 years and mean weight 77’ 6 kg. On one occasion the paracetamol was taken alone, on another with oral cimetidine (800 mg 1 h before and further doses of 400 mg taken at 4, 8, and 12 h after the paracetamol), and on the other with ethanol (70° proof whisky). An initial dose of 0-6g/kg ethanol was taken 0 - 5-1 - 0 h before the paracetamol, followed by maintenance doses ofO-11 and 0-16 g/kg hourly for 8 h in the normal subjects and the

heavy drinkers, respectively. After paracetamol administration

venous blood was sampled at intervals for 8 h and urine collected for 24 h. Plasma and urinary concentrations of paracetamol and its metabolites were measured by high-pressure liquid chromatography, 12 with which cimetidine did not interfere. Plasma ethanol was estimated by gas-liquid