The Peripheral Conversion of Thyroxine (T4) Into Triiodothyronine

The Peripheral Conversion of Thyroxine (T4) into i • Triiodothyronine (T3) and Reverse triiodothyronine (rT3)

3 The Peripheral Conversion

of Thyroxine (T4) into Triiodothyronine (T3) sts- a», and Reverse triiodothyronine (rT3)

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor in de Geneeskunde aan de Universiteit van Amsterdam,

?••!• op gezag van de Rector Magnificus dr. J.Bruyn, hoogleraar in de Faculteit der Letteren, in het openbaar te verdedigen in de aula der Universiteit (tijdelijk in de Lutherse Kerk, ingang Singel 411, hoek Spui), op donderdag 15 november 1979 om 16.00 uur precies

door

WILLEM MAARTEN WIERSINGA

geboren te Leiden

AMSTERDAM 1979 I; Promoter : Dr. J.L. Touber g Coreferent : Prof. dr. A. Querido 'l Copromotor : Prof. dr. M. Koster

Dit proefschrift is bewerkt op de afdeling endocrinologie (Dr. J.L. Touber) van de kliniek voor inwendige ziekten (Prof.Dr. M. Koster en Prof.Dr. J. Vreeken) van het Wilhelmina Gasthuis, Academisch Ziekenhuis bij de Universiteit van Amsterdam. Het verschijnen van dit proefschrift werd mede mogeiijk gemaakt door steun van de Nederlandse Hartstichting en van ICI Holland B.V. I i*'

Ii: VOORWOORD I Dit proefschrift is niet opgedragen aan iemand in het bijzonder, daar het tv, verrichte onderzoek in eerste instantie mijn eigen nieuwsgierigheid heeft 1} bevredigd en ik de indruk heb dat het meer mijn eigen genoegen heeft r ' gediend dan dat van anderen. Een woord van dank aan de velen die in de : afgelopen jaren bijgedragen hebben tot deze studies, is wel op zijn plaats, '. daar het onderzoek zonder hun steun niet mogelijk zou zijn geweest, en ; vooral ook omdat ik met plezier terugdenk aan de onderlinge , samenwerking. - \ In de eerste plaats wil ik graag Dries Schellekens bedanken voor zijn ¡ betrokkenheid bij de radioimmunologische bepalingen in al deze jaren, [ evenals Theo van Groenestein die mij ook de vereiste practische handvaardigheid bijbracht. Toen ik kwam, had Bertie Hensgens al de basis gelegd voor de T3-bepaling; Reinier Dalhuisen heeft later de T4-RIA ontwikkeld. Van de vele analisten op ons laboratorium die ooit 'in de schildklier' hebben gezeten, wil ik met name bedanken Nel Verhaest-de Jong (die het grootste deel van de T3-bepalingen heeft uitgevoerd, en die ook als eerste de prognostische waarde van 'haar' bepaling inzag- cf. figuur V. 1) en Margreet Broenink (voor het in record tijd bepalen van rT3 in een zeer grote hoevelheid monsters). Bloedmonsters krijg je niet zonder te prikken, en Doris de Haan prikte dat het een aard had, ook al deed het haar meer pijn dan degenen die geprikt werden. De jacht op normale personen, van wie m.n. het jongere vrouwelijke deel tegenwoordig toch al dun gezaaid is, is in een universiteitsziekenhuis steeds in volle gang; des temeer waardeerde ik hun medewerking als ik hen opgespoord had. Grote steun in het uitbreiden van de groep normalen kreeg ik van de collegae A.J.M. Fabius ('St. Willibrord', Heiloo) en P.A. Ykelenstam ('De Buitenhof, Amsterdam). De patiënten wil ik extra bedanken voor hun medewerking in situaties waarin zij het toch al moeilijk hadden. De vele testen werden uitgevoerd in de beste harmonie met Hanneke Assies. De bijzonder hartelijke medewerking aan mijn onderzoek die ik ondervond op de afdelingen inwendige geneeskunde (Prof. M. Koster, Prof. J. Vreeken), infectieziekten (R.M. van der Heide) en cardiologie (Prof. D. Durrer) was voor mij een extra stimulans. Een uitermate prettige bijkomstigheid was de medisch-politieke scholing door Henk Lie op de cardiologie. Het dierexperimenteel onderzoek is verricht door Ron van Noorden in een sfeer van arcadische eendracht, die alleen verstoord werd als de ratten hem L achtervolgden. C. Schuyt isoleerde de rattelevercellen in het Histologisch F Laboratorium volgens een door Prof. J. James ontwikkelde techniek. Van % de kritische open aanmerkingen van Prof. P. Borst, L. Offerhaus en Prof. ï P.A. van Zwieten bij deze dierstudies, hebben wij dankbaar gebruik l gemaakt. Veel heb ik geleerd van Herman Sallé en Ir. J. Oosting(Instituut ; voor Medische Fysica) die het grootste deel van de gegevens statistisch f hebben bewerkt. Hun logica bleek uiteindelijk steeds sterker dan de mijne. Van hen begreep ik dat medische tijdschriften voor statistici vaak een bron zijn van vermaak. Ellen Hunter wil ik hier graag hartelijk bedanken voor haar hulp bij het typen, Chris Bor en anderen van de centrale foto- en filmdienst voor het mooie tekenwerk en uitgeverij Rodopi voor de grote voortvarendheid waarmee dit proefschrift werd gedrukt. Over mijn promotor, Jan Touber, is natuurlijk veel te zeggen; het meest heb ik wel genoten van zijn scherpe analytische geest. Prof. A. Querido en Prof. M. Koster ben ik zeer dankbaar dat zij bereid waren het manuscript door te lezen; hun kritische opmerkingen zijn de uiteindelijke inhoud van het proefschrift op diverse punten ten goede gekomen. Tot slot wil ik graag diegenen bedanken van wie ik mijn opleiding tot internist heb ontvangen: Th.J. van Baar (Marine Hospitaal, Overveen), Th.J. Vervaat en H.B. Schreuder (St. Lucas Ziekenhuis, Amsterdam) en Prof. M. Koster en Prof. J. Vreeken (Wilhelmina Gasthuis, Amsterdam).

Amsterdam, september 1979

I ^ CONTENTS H-:' Chapter I. Introduction 1 -^ • 1.1. Definition and history of the peripheral conversion of T4 1 |x 1.2. Structure and biological activity of iodothyronines 3 1.3. Production of T3 and rT3 5 1.4. Nature of the deiodination of T4 6 1.5. Brief outline of the present study 8 1.6. References 8 Chapter II. Assays 12 II. 1 Introduction 12 - References 16 11.2. Thyrotropin (TSH) 17 2.1. Ingredients 17 2.2. Assay procedure 19 2.3. Results 22 - 2.4. Discussion 26 2.5. References 29 11.3. Iodothyronines 31 3a. Reverse triiodothyronine (rT3) 34 1' 3a. 1. Ingredients 34 1" 3a.2. Assay procedure 36 3a.3. Results 40 3a.4. Discussion 45 3a.5. References 49 3b. Triiodothyronine (T3) 52 3b. 1. Ingredients 52 3b .2. Assay procedure 52 3b.3. Results 53 3b.4. Discussion 54 3b.5. References 58 3c. Thyroxine (T4) 60 3c. 1. Ingredients 60 3c.2. Assay procedure 60 3c.3. Results 61 3c.4. Discussion 62 .2 3c.5. References 63 H.4. Triiodothyronine uptake ratio (T3U) 67 •y - References 69 ':'{

'% V

1 U.S. Sample handling and storage 70 Sa. Plasma versus serum samples 70 5b. Time of sampling 70 Sc. Technique of blood collection 70 t-. Sd. Storage of plasma samples 70 5e. References 72 Chapter III Studies in euthyroid volunteers 73 111.1. Introduction 73 111.2. Subjects and methods 73 111.3. Results 75 3a. Basal values 75 3b. TRH-test 78 3c. Cord blood 84 111.4. Discussion 87 4a. Thyroid function tests: basal values 87 - distribution 87 -effect of sex 87 -effect of age 87 - effect of body weight 88 - effect of circadian and circannual rhythms 88 4b. The TRH test 90 - side effects 90 - dosages and/or routes of administration 90 - control tests 90 - the TSH response to TRH 91 - the T3 response to TRH 93 4c. Cord blood 93 111.5. References 93 Chapter IV. Studies in patients with thyroid diseases 99 IV. 1. Introduction 99 IV.2. Patients and methods 99 IV.3. Results 99 IV.4. Discussion 109 4a. The spectrum of thyroid function disorders 109 4b. Hypothyroidism 112 -the TSH response to TRH 113 - the T3 response to TRH 114 4c. Hyperthyroidism 114 - T3-toxicosis 114 - the TSH response to TRH 115 - the T3 response to TRH 116 ~ T4-toxicosis 117 I IV.5. References 118 Chapter V. Studies in patients with acute non-thyroidal diseases 124 V.I. Introduction 124 V.2. Patients and methods 124 V.3. Results 128 V.4. Discussion 131 — incidence of subnormal plasma T3 131 — modulation of the peripheral conversion of T4 131 -relation between the severity of the disease and plasma T3133 — setpoint of the pituitary-thyroid axis 133 — role of thyroxine 134 V.5. References 136 Chapter VI. Studies in patients with acute myocardial infarction 139 VI. 1. Introduction 139 VI.2. Patients and methods 139 VI.3. Results 142 VIA Discussion 150 — reciprocal changes in plasma T3 and rT3 ISO — quantitative relationship between infarct size and the changes in the conversion of T4 ISO — fate of thyroxine 151 — sequence of events in thyroid hormone metabolism in acute myocardial infarction 153 VI.5 References 153 Chapter VII. Studies on the effect of propranolol in human subjects 156 VII. 1. Introduction 156 VII.2. Patients and methods 157 VII.3. Results 160 VII.4. Discussion 163 -propranolol inhibits the 5 '-deiodination of T4 164 - increase of T4 and TSH during propranolol 164 - nature of the inhibitory effect of propranolol and its biological significance 165 VH.5. References 166 Chapter VIII. Studies on the effect of propranolol in isolated rat liver parenchymal cells 170 VlII.l. Introduction 170 VIII.2. Materials and methods 171 VIII.3. Results 172 VIII.4. Discussion 174 VIII.5. References 176 1 Chapter IX. General discussion 178 IX. 1. Physiological modulation of the peripheral conversion of T4 178 IX.2. Pathophysiological modulation of the peripheral conversion of T4 179 IX.3. Pharmacological modulation of the peripheral conversion ofT4 180 IX.4. The peripheral conversion of T4 and the pituitary- thyroid axis: regulatory mechanisms 182 IX.5. The biological significance of the peripheral conversion ofT4 184 IX.6. Epilogue 186 IX.7. References 187 Summary 194 Samenvatting 197 Samenvatting voor de leek 201 Abbreviations 204 I "£'•

CHAPTER I. i INTRODUCTION

I.I. Definition and history of the peripheral conversion of T4.

The peripheral conversion of the thyroid hormone tetraiodothyronine or thyroxine (T4) into triiodothyronine (T3) and reverse triiodothyronine (rT3) is defined as the extrathyroidal deiodination of thyroxine, by which r. — depending on the site of removal of an iodine atom — either T3 or rT3 arises. T4 was demonstrated to be present in human plasma by Taurog and Chaikoff in 1948, and T3 by Gross and Pitt-Rivers in 1953. Although it v/as already shown by Albright et al in 1954 in vitro and by Pitt-Rivers et al in 1955 in vivo in man that T4 is deiodinated to T3, conversion of T4 into T3 could not be demonstrated in human subjects in a study of Lassiter and Stanbury in 1958. Interest in the deiodination of T4 thus diminished, also due to the technical problems of the measurement of T3. With the advent of less cumbersome techniques for the measurement of T3 in the late 1960's,

Figure I.I. Sequential deiodination of T4, according to Chopra et al (1978). Outer ring deiodination is indicated by open circles, inner ring deiodination by solid circles. Dashed lines indicate proposed deiodination pathways which have not yet been demonstrated conclusively. and especially since the introduction of the T3 radioimmunoassay in the early 1970's, the fate of T3 in health and disease was studied extensively. Braverman et al (1970) demonstrated conclusively that T3 was generated from T4 outside the thyroid in athyreotic human subjects, in whom substitution therapy with T4 was the only possible source of T3. In the same year conversion of T4 to T, was also found in normal human subjects (Sterling et al 1970). Many other reports have appeared, all indicating a substantial extrathyroidal production of T3 from T4 (Pittman et al 1971; Nicoloff et al 1972; Braverman et al 1973; Surks et al 1973; Inada et al 1975; Nomura et al 1975). Reverse T3 (rT3) was first shown to be present in human plasma by Chopra (1974), and the extrathyroidal production of rT3

NH2

CH2-CH COOH PHENOLIC j/3- RING ETHER j tt-RING ALANYL HYDROXYL |OUTER RING BRIDGE INNER RING SIDE CHAIN GROUP I PHENOLIC RING iTYROSYL RING] I

thyronine position

3 5 3' 5'

T4 I I I T3 I I H rT3, H I I 3,3'-T2 H I H I H H 3,5- H H I I 3-Tt H H I H 3-T» I H H H 1 To H H H H l\ Figure 1.2. The constitution of the thyronine molecule and its iodinated derivatives (I is iodine, H is hydrogen). in man has been demonstrated convincingly (Chopra 1976; Gavin et al 1977; Smallridge et al 1978; Suda et al 1978). The deiodination of T4 to T3 V.- and rT3 appears to be the first step in a cascade of sequential deiodinations of T4, which ultimately leads to the formation of To via several forms of triiodinated, diiodinated and monoiodinated thyronines as illustrated in figure I.I (Chopra et al 1978; Sakurada et al 1978; Rudolph et al 1978; Gavin et al 1978; Smallridge et al 1979; Geola et al 1979). The iodothyronines 3,5-T2, 3-3'-T2, 3',5'-T2 and 3-T, are already known to exist in human plasma (Chopra et al 1978; Burman et al 1978; Meinhold and Schiirnbrand 1978; Smallridge et al 1979; Faber et al 1979).

1.2. Structure and biological activity of iodothyronines.

The structure of the various iodothyronines (or according to Cody (1978) more precisely the "constitution", which refers to the connectivity of the molecule) is presented in figure 1.2. Triiodothyronine (or 3,S,3'-T3) has two iodine atoms in the inner ring and one in the outer ring; just the reverse is true for 3,3',5'-triiodothyronine of rT3, hence the adjective 'reverse'. The various iodothyronines differ greatly in biological activity. The metabolic activity by weight of T3 is three to four times greater than that of T4 as tested in the bioassay, and is virtually absent in rT3 (Money et al 1960; Pittman and Barker 1962; Barker et al 1965). It is natural to explain this difference from structure-activity relationships. The molecular constitution and conformation (according to Cody (1978) referring to the total geometric distribution or disposition of the atoms in three dimensions) of T4, T3 and rT3 are depicted in figure 1.3. The structural requirements for thyroid hormone activity are the presence of nonpolar groups (halogen or methyl) on the 3 and S positions, and the minimal energy conformation of mutually perpendicular planes of the phenyl rings and the 120° angle between the rings; the activity of this primary structure is enhanced by a halogen atom in the 3' position, but a second substituent in the 5' position reduces activity (Jorgensen 1978). Thus, removing one inner-ring iodine atom (as is the case in rT3) results in greatly diminished activity. Indeed, Koerner et al (1975) found that the binding to nuclear receptor sites is moderately strong for T4 and very weak for rT3 in comparison to T3. This does not mean that rT3 is without biological effect. In fact, rT3 has been shown to be more potent than T3 in some in vitro studies: it stimulates the erythropoietin-facilitated erythropoiesis (Golde et al 1977), the activity of L-T3 aminotransferase (Fishman et al 1977), the amino acid uptake by thymocytes (Goldfme et al 1976) and the growth hormone secretion by GH, cells (Papavasiliou et al 1977). An anti-thyroxine and antimetabolic effect of rT3 has also been demonstrated (Pittman and Barker 1959; Pittman et al I960). However, in all these studies pharmacological doses of rT3 were used, and the physiological significance of these actions is therefore questionable. & £

I ' HO CH, -CH *• I COOH

NH,

HO °\ VCH2-CH COOH I

Figure 1.3. The molecular constitution (left panels) and conformation (middle and right panels) of T4, T3 and rTj. (The support given by Dr. R. Dalhuisen in the construction of these models is grate- fully acknowledged). U. Production of T3 and rT3.

The production of T3 and rT3 occurs in the thyroid gland and extrathyroidally. Precise estimation of the relative contributions of the thyroid gland and of the peripheral tissues to the daily production of T3 and rT3 is difficult because published data on the plasma concentrations, the production rates (PR) and the metabolic clearance rates (MCR) differ. Nevertheless, some conclusions can be drawn from these kinetic studies tl; (figure 1.4). As calculated from the plasma concentration and the MCR the daily production of T3 is 34-72 nmol (representing the lowest and highest values), that of rT3 is 23-100 nmol. If it is assumed that the thyroidal secretion of T3 and rT3 is related to the proportion in which they are present in thyroglobulin (Inoue et al 1967; Abrams and Larsen 1973), the daily thyro-ial secretion of T3 (and rT3) can then be assessed from the product of the T3/T4 (and rT3/T4) ratio in thyroglobulin and the daily secretion of T4. As far as presently known, the thyroid gland is the only source of T4. The daily production (- secretion) of T4 is appr. 116 nmol,

T3/T4= 0.051-0.096 rT3/T4= 0.013-0.047

/ \

5.9-11.1 nmol T3/day 1.5-5.5 nmol rT3/day

(8.2-32.7% PR - T3) (1.5-23.7* PR-rT3)

/ \ r \ PR-T3 PR-Ta- PR-rT3 22-66 nmol T3/day 17-98 nmol rT3/day 34-72 -vll6 23-100 (65-92* PR-T ) nmoi/day 3 nmol/day (74-98% PR-rT3) nmol/day

SERUM T3 1.69-2.77 nmol/1 SERUM rT3 0.28-0.92 nmol/1 HCR 20-26 l/70kg/day MCR 82-108 l/70kg/day

Figure 1.4. Quantitation of the relative contributions of the thyroid gland and of the peripheral tissues to the daily production of T3 and rT3 (according to Chopra et al (1978) withminor modifications; PR=production rate, MCR=metabolic clearance rate). 8 33% 46% 38 nmol/day 56 nmol/Uay

CH2-CH

(COOH) r I CONJUGATION DEAMINATIO(TION (glucuronide, sulphate) DECARBOXYLATIODXY N I | NON-DEIODINATION RCTHWAYS: 19%. 22 nmol/day

7.5. The molecular constitution of thyroxine with an estimate of its various routes of metabolism (daily T4 production -"116 nmol).

and the thyroidal secretion of T3 is calculated as 5.9-11.1 nmol/day, and that of rT3 as 1.S-S.5 nmol/day. Consequently, the peripheral conversion of T4 generates 22-66 nmol T3 and 17-98 nmol rT3 per day. It can therefore be concluded that the bulk of the daily production of T3 and rT3 is generated extrathyroidally, and that thyroidal secretion of T3 and especially of rT3 contributes relatively minor amounts to their daily production. It thus appears that deiodination is a major mechanism in the degradation of T4; about 80-90% of T4 is converted to T3 and rT3 (Chopra 1976; Gavin et al 1977). Other routes in the metabolism of T4 (figure 1.5) are excretion in urine and bile as T4 and T4 conjugates accounting for appr. 15% of the daily PR of T4 (Oddie et al 1964), and deamination and decarboxylation accounting for appr. 4% (Chopra et al 1978).

1.4. Nature of the deiodination of T4.

The conversion of T4 into T3 (and rT3) in vitro has now been demonstrated in several human tissues — in fibroblasts (Refetoff et al 1972), in liver and kidney cells (Sterling et al 1973), in lymphocytes (Holm et al 1975) and in polymorphonuclear leucocytes (Woeber 1978) — and in a wide variety of animal tissues (Albright and Larsen 1954; Rabinowitz and Hercker 1971; Hesch et al 1975; Visser et al 1975; Chopra 1977; Chiraseveenuprapund et al 1978). In fact, the in vitro conversion of exogenously added T4 by tissues has been used as a model in the study of the nature of the deiodination of T4 and the physiological, pathological

iD f$ and pharmacological factors involved. Considerable differences in the |T capability to deiodinate T4 between various tissues have been found, the W liver and kidneys demonstrating the greatest deiodination capacity fv (Chopra 1977). This does not mean that other tissues contribute little to the j?" extrathyroidal production of T3 and rT3, since e.g. muscles by their great I mass might be an important source and may quantitatively equal the y kidneys and liver in this respect. It may be postulated that all thyroid V: hormone responsive tissues have the capacity for the deiodination of T4. : , The deiodination of T4 seems to be enzymic in nature since the in vitro process is dependent on temperature, pH and substrate concentration (Visser etal 1975; Chopra 1977; Hiifneretal 1977; Kaplan and Utiger 1978; Chiraseveenuprapund et al 1978; Hoffken et al 1978; Woeber 1978; Chopra et al 1978; Hiifner and Grussendorf 1978; Visser et al 1978). i However, so far all attempts to isolate the deiodinating enzymes in pure form have failed. The enzymic activity for the outer ring deiodination has been located in the plasma membranes of kidney cells (Chiraseveen- uprapund et al 1978; Leonard and Rosenberg 1978) and in the ; endoplasmatic reticulum of liver cells (Visser et al 1976; Maciel et al 1979; Fekkes et al 1979; Kdhrle and Hesch 1979). Evidence is accumulating that • the complete cascade of deiodination reactions of T4 is governed by an outer ring or 5'-deiodinase and by an inner ring or 5-deiodinase, and that these two enzymes are not the same. The existence of two different enzymes is further suggested by clinical data: after birth plasma T3 increases within hours to values comparable to those found in adults whereas rT3 remains elevated for several days (Chopra et al 1975), and prolonged reduction of dietary intake is associated with a transient increase in rT3 whereas T3 remains low (Visser et al 1978).

The presence of sulfhydryl groups is essential for the deiodination of T4 ; to T3 and of rT3 to 3,3'-T2 (Visser et al 1976; Chopra 1978; ] Chiraseveenuprapund et al 1978; Visser et al 1978; Chopra et al 1978; \ Woeber 1978). The deiodinating activity is stimulated by reduced \ glutathione and several thiol agents (dithiothreitol, mercaptoethanol), and : it is inhibited by oxidized glutathione, sulfhydryl-oxidizing agents (diamide) and sulfhydryl-binding agents (N-ethylmaleimide, mercuric chloride). The sulfhydryl groups acts as cofactor, and recently Visser et al r (1979) have proposed the following 'ping-pong mechanism' for the ;| deiodination of T4: • 1 - i \•j T4 + E-SH 5=t T4. E-SH —r*E-S\ I A» » E-SH + HI fI i. T3 2SH S-S $ :'i I where E-SI is the intermediate sulphenyl iodide of the enzyme, which is 'j I reduced by the cofactor. '{ 8 Other regulatory mechanisms in the deiodination proces have been proposed, such as changes in temperature and pH (Hflffken et al 1978), substrate-induction of the deiodinating enzymes (Grussendorf and Hiiffner 1977) and the concentration of iodothyronines themselves. It is highly questionable, however, whether all these factors are operative in vivo. Although it has been shown in vitro that the deiodination reactions are dependent on temperature and differ in their optimal pH (varying between 6.5 and 9.5), such wide changes are highly unlikely to occur in vivo. Also, the physiological significance of the depressive action of rT3on the T4-T3 conversion and of T4 on the rT3*3,3'-T2 conversion as demonstrated in vitro (Chopra 1977; Kaplan and Utiger 1978; Hoffken et al 1978) remains questionable.

1.5. Brief outline of the present study.

The aim of this study is to delineate several physiological, pathological and pharmacological factors involved in the peripheral conversion of T4. The assays used are described in chapter II. The determination of normal values of these tests under basal circumstances and after stimulation with thyrotropin-releasing-hormone (TRH) is presented in chapter III, and some physiological factors which may modulate the conversion of T4 are discussed. The results of the thyroid function tests in patients with thyroid disease are presented in chapter IV. The test results are interpreted both in terms of their value in the diagnosis of disordered thyroid function and of altered conversion of T4.

The peripheral conversion of T4 was further evaluated in patients with acute non-thyroidal diseases. The incidence and cause of subnormal plasma T3 concentrations in such patients was investigated, together with its effect on the pituitary-thyroid axis. A relationship (both qualitatively and quantitatively) was sought between the changes in the conversion of T4 and the severity of the disease (chapters V and VI). The effect of propranolol on the deiodination of T4 in human subjects and in animals is described in the chapters VII and VIII.

1.6. References.

Abrahams GM, Larsen PR. Triiodothyronine and thyroxine in the serum and thyroid glands of iodine deficient rats. J Clin Invest 1973; 52:2S22-31. Albright EC, Larson FC, Tust RH. In vitro conversion of thyroxine to triiodothyronine by kidney slices. Proc Soc Exp Biol Med 1954} 86: 137-40. Barker SB, Shimada M, Makiuchi M. Metabolic and cardiac responses to thyroxine analogs. Endocrinology 1965; 76: 115-21. BravermanLE, IngbarSH, Sterling K. Conversion of thyroxine (T4) to triiodothyronine (Tj) in athyreotic human subjects. J Clin Invest 1970; 49: 855-64. Braverman LE, Vagenakis A, Downs P, Foster AF, Sterling F, IngbarSH. Effects of replacement doses of sodium-L-thyroxine on the peripheral metabolism of thyroxine and triiodothyronine in man. J Clin Invest 1973; 52: 1010-7. * v: Burman KD, Wright FD, Smallridge RC, Green BJ, Georges LP, Wartofsky L. A radio- ¡mmunoassay for 3',5'-diiodothyronine. J Clin Endocrinol Metab 1978; 47: 1059-64. Chiraseveenuprapund P, Buergi U, Goswami A, Rosenberg JN. Conversion of L-thyroxine to triiodothyronine in rat kidney homogenate. Endocrinology 1978; 102: 612-22. Chopra IJ, A radioimmunoassay for measurement of 3,3',5'-triiodothyronine (reverse T3). J Clin Invest 1974; 54: 583-92. Chopra IJ, Sack J, Fisher DA. Circulating 3,3',5'-triiodothyronine (reverse T3) in the human newborn. J Clin Invest 1975; 55: 1137-41. Chopra IJ. An assessment of daily production and significance of thyroidal secretion of 3,3',5'-triiodothyronine (reverse T,) in man. J Clin Invest 1976; 58: 32-40. Chopra IJ, A study of extrathyroidal conversion of thyroxine (T4) to 3,3',5'-triiodothyronine (T3) in vitro. Endocrinology »977; 101: 453-63. Chopra IJ, Wu S-Y, Nakamura Y, Solomon DH. Monodeiodination of 3,5,3'-triiodothyro- nineand3,3',5'-triiodothyronineto3,3'-diiodothyroninein vitro. Endocrinology 1978; 102: 1099-1106. Chopra LI, Geola F, Solomon DH, Maciel RMB. 3',5-Diiodothyronine in health and disease: studies by a radioimmunoassay. J Clin Endocrinol Metab 1978; 47: 1198-1207. Chopra IJ, Solomon DH, Chopra U, Wu SY. Fisher DA, Nakamura Y. Pathways of metabolism of thyroid hormones. Recent Prog Horm Res 1978; 34: 521-67. Chopra IJ. Sulfhydryl groups and the monodeiodination of thyroxine to triiodothyronine. Science 1978; 199: 904-6. Cody V. Thyroid hormones: crystal structure, molecular conformation, binding, and structure-function relationships. Recent Prog Horm Res 1978; 34: 437-75. Faber J, Kirkegaard C, Lumholtz IB, Siersbaek-Nielsen K, Friis T. Measurements of serum 3\5'-diiodothyronine and 3,3'-diiodothyronine concentrations in normal subjects and in patients with thyroidal and nonthyroid disease: studies of 3',5'-diiodothyronine metabolism. J Clin Endocrinol Metab 1979; 48: 611-7. Fekkes D, Overmeeren E van, Visser TJ. Endoplasmic reticulum location of rat liver iodothyronine deiodinases; chromatography of solubilized enzymes (abstract). Ann d' Endocrinol 1979; 40: 22A. Fishman N, Huang YP, Tergis DC, Rivlin RS. Relation of triiodothyronine and reverse triiodothyronine administration in rats to hepatic L-triiodothyronine aminotransferase activity. Endocrinology 1977; 100: 1055-9. Gavin L, Castle J, McMahon F, Martin P, Hammond M, Cavalieri RR. Extrathyroidal conversion of thyroxine to 3,3',5'-triiodothyronine (reverse-T3) and to 3,5,3'-triiodothyronine (Tj) in humans. J Clin Endocrinol Metab 1977; 44: 733-42. Gavin LA, Hammond ME, Caste JN. Cavalieri RR. 3,3-Diiodothyronine production, a major pathway of peripheral iodothyronine metabolism in man. J Clin Invest 1978; 61: 1276-85. Geola F, Chopra IJ, Solomon DH, Maciel RMB. Metabolic clearance and production rates of 3',5'-diiodothyronine and 3,3'-diiodothyronine in man. J Clin Endocrinol Metab 1979; 48: 297-301. Golde DW, Bersch N, Chopra I J, Cline MJ. Thyroid hormones stimulate erythropoiesis in vitro. Brit J Haematol 1977; 37: 173-7. I') Goldfine ID, Smith GJ, Simons CG, Ingbar SH, Jorgensen EC. Activities of thyroid hormones and related compounds in an in vitro thymocyte assay. J Biol Chem 1976; 251: 4233-8. Gross J, Pitt-Rivers R. The identification of 3:5:3'-triiodothyronine in human plasma. Lancet 1952; 1: 439-41. I Grussendorf M, Hflfner M. Induction of the thyroxine (T4) to triiodothyronine (T3) con- 10 verting enzyme in rat liver by thyroid hormones and analogs. Clin Chim Acta 1977; 80: 61-6. Hesch RD, Brunner G, Sitting HD. Conversion of thyroxine (T4) and triiodothyronine (T3) and the subcellular location of the converting enzyme. Clin Chim Acta 197S; 59:209-11. Heffken B, Ktidding R, von zur Milhlen A, Hehrmann T, JUppner H, Hesch R-D. Regula- tion of thyroid hormone metabolism in rat liver fractions. Biochim biophys Acta 1978; 539: 114-24. Holm A-Ch, Lemarchand-Beraud Th, Scazziga B-R, Cuttelod S. Human lymphocyte binding and deiodination of thyroid hormones in relation to thyroid function. Acta Endocrinol 1975; 80: 642-56. Hilfner M, Grussendorf M, Ntokalou M. Properties of the thyroxine (T4) monodeiodinating system in rat liver homogenate. Clin Chim Acta 1977; 78: 251-9. Hilfner M, Grussendorf M. Investigations on the deiodination of thyroxine (T4) to 3,3'- diiodothyronine (3,3'-T2) in rat liver homogenate. Clin Chim Acta 1978; 85: 243-51. Inoue K, Grimm Y, Greer MA. Quantitative studies on the iodinated compounds secreted by the thyroid gland as determined by in situ perfusion. Endocrinology 1967; 81: 946-64. Inada M, Kasagi K, KurataS et al. Estimation of thyroxine and triiodothyronine distribution and of the conversion rate of thyroxine to triiodothyronine in man. J Clin Invest 1975; 55- 1337-48. Jorgensen EC. Thyroid-hormone structure-function relationship. In: Werner SC, Ingbar SH, eds. The thyroid: a fundamental and clinical text. 4th ed. Hagerstown: Harper and Row, 1978: 125-37. Kaplan MM, Utiger RD. Iodothyronine metabolism in rat liver homogenates. J Clin Invest 1978; 61: 459-71. Koerner D, Schwartz HL, Surks MI, Oppenheimer JL. Binding of selected iodothyronine analogues to receptor sites of isolated rat hepatic nuclei: high correlation between structural requirements for nuclear binding and biological activity. J Biol Chem 1975; 250: 6417-23. Kohrle J, Hesch RD. Thyroxine deiodination by subfractions of the endoplasmatic reticulum of rat liver (abstract). Ann d'Endocrinol 1979; 40: 22A. Lassiter WE, Stanbury JB. The in vivo conversion of thyroxine to 3:5:3'-triiodothyronine. J Clin Endocrinol Metab 1958; 18: 903-6. Leonard JL, Rosenberg IN. Subcellular distribution of thyroxine 5'-deiodinase in the rat kidney: a plasma membrane location. Endocrinology 1978; 103: 274-80. Maciel RMB, Ozawa Y, Chopra IJ. Subcellular localization of thyroxine and reverse triiodothyronine outer ring monodeiodinating activities. Endocrinology 1979; 104:365- 71. Meinhold H, Schilrnbrand P. A radioimmunoassay for 3,5-diiodothyronine. Clin Endo- crinol 1978; 8: 493-7. Money WL, Kumaoka S, Rawson RW, Kroc RL. Comparative effects of thyroxine analogues in experimental animals. Ann NY Acad Sci 1960; 86: 512-44. Nicoloff JT, Low JC, Dussault JH, Fisher DA. Simultaneous measurement of thyroxine and triiodothyronine peripheral turnover kinetics in man. J Clin Invest 1972; 51: 473-83. Nomura S, Pittman CS, Chambers Jr JB, Buck MW, Shimizu T. Reduced peripheral conversion of thyroxine to triiodothyronine in patients with hepatic cirrhosis. J Clin Invest 1975; 56: 643-52. Oddie TH, Fisher DA, Rogers C. Whole body counting of l3lI-labeled thyroxine. J Clin Endocrinol Metab 1964; 24: 628-37. Papavasiliou SS, Martial JA, Lathem KR, Baxter JD. Thyroid hormonelike actions of 3,3',5'-L- triiodothyronine and 3,3'-diiodothyronine. J Clin Invest 1977; 60: 1230-9. Pitt-Rivers R, Stanbury JB, Rapp B. Conversion of thyroxine to 3-5-3'-triiodothyronine in vivo. J Clin Endocrinol Metab 19S5; 15: 616-20. Pittman CS, Barker SB. Antithyroxine effects of some thyroxine analogues. Am J Physiol 1959; 197: 1271-4. f 1 if 11 Pittman CS, Barker SB. The calorigenic effects of some thyroxine analogs on intact animals and on excised tissues. J Clin Invest 1962; 41: 696-701. If Pittman CS, Chambers Jr JB, Read VH. The extrathyroidal conversion rate of thyroxine into t triiodothyronine in normal man. J Clin Invest 1971; SO: 1187-96. Pittman JA, Tingley JO, Nickerson JF, Hill Jr SR. Antimetabolic activity of 3,3',5'-triiodo- DL-thyronine in man. Metabolism 1960; 9: 293-5. Rabinowitz JL, Hercker ES. Thyroxine: conversion to truodothyronine by isolated perfused rat heart. Science 1971; 173: 1242-3. Refetoff S, Matalon R, Biaggi M. Metabolism of L-thyroxine (T4) and L-triiodothyronine (T3) by human fibroblasts in tissue culture: evidence for cellular binding proteins and conversion of T4 to T3. Endocrinology 1972; 91: 934-47. Rudolph M, Sakurada T, Fang S-L, Vaganakis AG, Braverman LE, Ingbar SH. Appearance of labeled metabolites in the serum of man after the administration of labeled thyroxine, triiodothyronine (T3), and reverse triiodothyronine (rT3). J Clin Endocrinol Metab 1978; 46: 923-8. Sakurada T, Rudolph M, Fang S-L, Vaganakis AG, Braverman LE, Ingbar SH. Evidence that triiodothyronine and reverse triiodothyronine are sequentially deiodinated in man. J Clin Endocrinol Metab 1978; 46: 916-22. Smallridge RC, Wartofsky L, Desjardins RE, Burman KD. Metabolic clearance and production rates of 3,3',5'-triiodothyronine in hyperthyroid, euthyroid, and hypothyroid subjects. J Clin Endocrinol Metab 1978; 47: 345-9. Smallridge RC, Wartofsky L, Green BJ, Miller FC, Burman KD. 3'-L-Monoiodothyronine: development of a radioimmunoassay and demonstration of in vivo conversion from 3',5'-diiodothyronine. J Clin Endocrinol Metab 1979; 48: 32-6. Sterling K, Brenner MA, Newman ES. Conversion of thyroxine to triiodothyronine in normal human subjects. Science 1970; 169: 1099-1100. Sterling K, Brenner MA, Saldanha VF. Conversion of thyroxine to triiodothyronine by cultured human cells. Science 1973; 179: 1000-1. Suda AK, Pittman CS, Shitnizu T, Chambers Jr JB. The production and metabolism of 3,5,3'-triiodothyronine and 3,3',5'-triiodothyronine in normal and fasting subjects. J Clin Endocrinol Metab 1978; 47: 1311-9. Surks MJ, Schadlow AR, Stock JM, Oppenheimer JH. Determination of iodothyronine absorption and conversion of L-thyroxine (T4) to L-triiodothyronine (T3) using turnover rate techniques. J Clin Invest 1973; 32: 805-11. Taurog A, Chaikoff IL. On the nature of the circulating thyroid hormone.J Biol Chem 1948; 176: 639-56. Visser TJ, Does-Tobe I vd, Docter R, Hennemann G. Conversion of thyroxine into triiodothyronine by rat liver homogenate. Biochem J 1975; 150: 489-93. Visser TJ, Does-Tobe I vd, Docter R, Hennemann G. Subcellular localization of a rat liver enzyme converting thyroxine into tri-iodothyronine and possible involvement of essential thiol groups. Biochem J 1976; 157: 479-82. Visser TJ, Lamberts SWJ, Wilson JHP, Docter R.Hennemann G. Serum thyroid hormone concentrations during prolonged reduction of dietary intake. Metabolism 1978; 27:405-9. I Visser TJ, Fekkes D, Docter R, Hennemann G. Sequential deiodination of thyroxine in rat liver homogenate. Biochem J 1978; 174: 221-9. Visser TJ, Fekkes D, Overmeeren E van, Docter R, Hennemann G. Mechanism of action of rat liver enzymes catalysing the reductive deiodination of iodothyronines. J Endocrinol 1979; 80: 24P-5P. Woeber KA, L-Triiodothyronine and L-reverse-triiodothyronine generation in the human polymorphonuclear leucocyte. J Clin Invest 1978; 62: 577-84. I I

CHAPTER II £••'

ASSAYS I,

//./. Introduction.

The laboratory tests used in this study are competitive protein-binding assays, mainly radioimmunoassays. The competitive inhibition principle of radioimmunoassay is presented by the following competing reactions (Yalow and Berson 1971):

labeled antigen specific antibody labeled antigen- antibody complex Ag* Ab An* Ah (F)

Ag unlabeled antigen in known standard solutions or I! unknown samples AgAb unlabeled antigen- antibody complex

Unlabeled hormone competes with labeled hormone for specific binding sites on the antibody, thereby inhibiting the binding of labeled hormone. Consequently, as the concentration of the unlabeled hormone increases, the ratio of antibody-bound (B) to free (F) labeled hormone will diminish. The concentration of hormone in an unknown sample is obtained by comparing the inhibition observed with that produced by standard solutions containing known amounts of hormone ("standard hormone"). This is conveniently done by plotting the B/F ratio (or the percentage bound B) versus the hormone concentration in the standard solutions. From the obtained "standard curve" one can find the hormone concentration that corresponds to the B/F ratio observed in the unknown sample.

ss. 13 I; I I:

hormone concentration

Figure ILL Standard curve and equations, giving definition of precision and sensitivity, i.e. the detection limit (modified according to Ekins 1974).

ABO = error in BQ ABX = error in Bx

AXO = error in B (when x=o) AX = error in X

AB0 = ,the -,the

slope at Bx slope at Bo precision of X detection limit Since the purpose of the radioimmunoassay method is to measure unknown hormone concentrations accurately, the method must be both precise and not influenced by non-specific factors (Ekins 1974). The concept of precision is illustrated in figure II. 1. The precision of the measurement of a concentration of hormone X is dependent both on the slope of the standard curve and on the error in the measurement of the response metameter (i.e., the percentage bound B) at the corresponding point. The magnitude of the error in measuring B can be calculated as the difference of x+2. SD and x - 2. SD for a number of measurements of B (AB).When X is zero, the corresponding value of AX(i.e.,A B0)defines the detection limit or "sensitivity" of the system. Since both the slope and AB change at different points on the standard curve, AX will vary as a function 1 of X. The intra-assay coefficient of variation($Bx!00%) can serve as an 3 indicator for the precision of the measurements on each point of the standard curve.The error AB is the net result of two independent errors, i.e. the "experimental" errors arising from pipetting and all other manipulations during the assay, and the counting errors inherent in 14 radioactivity measurements. The choice of the many variables in each assay (e.g. concentration of reagents, times of incubation and counting) may be facilitated by mathematical approximation of the optimal conditions for reacting equilibrium between antigen and antibody binding sites. We have determined the optimal conditions by trial and error.

Since radioimmunoassays compare an unknown concentration of the hormone in the system with known concentrations of the same hormone in a set of standards, the incubation mixtures should theoretically be identical in every respect with the exception of the concentration of the hormone to be measured. Addition of plasma without the hormone under test to the standard tubes is a first step in meeting this requirement, but complete identity between the unknown and standard incubation mixtures is impossible to achieve in practice, and hence non-specific effects may be encountered. Due to the immunological nature of the reactions, cross- reactions may occur with closely related antigens and with hormonal fragments. Due to the chemical nature of the reactions the result may be influenced by the composition of the medium, by the presence of other substances reacting with antigen or antibody, and by the separation of bound and free hormone. The non-specific effects in radioimmunoassay of hormones may thus be due to (Berson and Yalow 1973) hormonal cross- reactivity and non-hormonal effects (pH, ionic environment, anti- coagulant, temperature, variable damage of labeled hormone). The labeled hormone molecule may disrupt by decay of a radioiodine atom. Labeled fragments or free radioiodine are produced, and the radioactivity is no longer associated only with unaltered molecules. This phenomenon of decay catastrophe (Berson and Yalow 1973) shortens the shelf life of the labeled hormone and diminishes immunoreactivity of the tracer. Peptide hormones are subject to proteolytic damage by enzymes in blood. Damage to the labeled hormone is therefore in general greater in plasma (particularly when hemolysed) than in buffer solution. This effect can be minimized by using a low plasma concentration in the incubation mixture. Since plasma vary in their proteolytic effects on hormones, control incubations must be performed. Control mixtures or "blanks" contain no antiserum but are otherwise identical to the incubation mixtures of the plasma sample to be assayed, and are treated in the same manner. Controls of the standard curve are also performed. If for example 93% of the labeled hormone in the control tube separates with the free fraction at the time of starting the assay, the remaining 7% has thus aspecifically been bound, due to damage to the label and to incomplete separation of bound and free hormone fraction. If at the end of the incubation 87% separates with the free fraction, another 6% of the labeled hormone is aspecifically bound due to incubation damage. If approximately the same results are obtained in the controls for standards 15 and for unknown samples, no correction may be needed for incubation damage and/or lack of complete separation. Assays in which the damage to labeled hormone exceeds 10% probably should be discarded. Uncorrected damage of less than 10% to labeled hormone however can result in great errors in the measurements of unknown samples (Berson and Yalow 1973).

Furthermore, tests for qualitative and quantitative specificity must be carried out. First, similarity, or rather the absence of clear nonlinearity I must be demonstrated between the standard curve and a dilution curve of an unknown sample, i.e. the hormone concentration in the unknown sample must be a linear function of the dilution factor (parallelism). Although parallelism is a condition which has to be met, it is in itself not sufficient to establish immunochemical identity between the standard hormone and the hormone in the unknown sample. Second, exogenous hormone added to the incubation mixture, should be recovered quantitatively in the assay (recovery).

The results of hormone determinations by radioimmunoassay methods must be subject to quality control (Jeffcoate 1978). In our laboratory each sample is assayed in duplicate or triplicate, the result being rejected when their coefficient of variation is greater than the mean coefficient of variation + 2 SD of plasma with a corresponding hormone concentration, as determined in previous experiments. A drift within an assay is detected by putting three samples of the same plasma at different points in one assay run; the position of the tubes in the assay should not influence the outcome. Reproducibility is tested by including three known plasma samples (with a low, a normal and a high hormone concentration respectively) in each assay; the assay is considered for rejection if by inter- assay variation the concentration of hormone measured is outside the range of the mean hormone concentration ± 2 SD of that plasma, as determined in the previous experiments. The discrepancies between laboratories one to another may seriously hinder comparison of the results obtained in different laboratories. Inter-laboratory variation can be evaluated by the measurement of hormone contents in the same plasma by different laboratories. Such a program has been initiated in the Netherlands by the national working group on binding analysis (LWBA). The results so obtained for the TSH, T4 and T3 assays however are beyond the scope of this study.

Last but not least, the most important factor in the validation of a radioimmunoassay is clinical validation of the assay. That is, the results of the hormone determinations must make sense. The hormone level must rise and fall after the addition of known stimulants and suppressants; low levels must be found in samples from patients known to be deficient in the 16 hormone, and high levels in those with excessive secretion. The clinical validation of the assays to be described can be judged by the obtained results as presented in chapters III and IV.

For the sake of clarity I have adopted the nomenclature for tests of thyroid hormones in serum as recommended by a committee of the American Thyroid Association (Solomon, Benotti and DeGroot 1972; 1976). I hope that this nomenclature may be uniformly adopted by all Dutch thyroidologists and clinical chemists. The nomenclature of peptide hormones is according to the recommendations of the IUPAC-IUB Commission on Biochemical Nomenclature (1975). Furthermore, despite the pamphlet "De blinde mol" (Lindeboom 1971) I used the units of mol and liter for expression of the measured hormone concentrations where possible.

References

Berson S A, Yalow RS. Radioimmunoassay. In: Berson S A, ed. Methods in investigative and diagnostic endocrinology. Volume 2a, Berson S A, Yalow R S, eds. Peptide hormones. Amsterdam-London: North-Holland Publishing Company 1973: 84-120. Dijkbaer R. An international recommendation for quantities and units in clinical chemistry. Ann Int Med 1968; 69: 621-3. Ekins RP. Basic principles and theory of radioimmunoassay and saturation analysis. Brit Med Bull 1974; 30:3-11. Ekins R P. Basic concepts in quality control. In: Radioimmunoassay and related procedures in medicine 1977, volume II. Vienna: International Atomic Energy Agency, 1978: 6-20. International Union of Pure and Applied Chemistry — International Union of Biochemistry Commission on Biochemical Nomenclature. The nomenclature of peptide hormones. Biochim Biophys Acta 1975; 404: 152-5. JeffcoateS L. The use of quality control within a laboratory. In: Radioimmunoassay and related procedures in medicine 1977, volume II. Vienna: International Atomic Energy Agency, 1978: 57-79. Lindeboom G A. De blinde mol. Kritische beschouwingen over het door klinisch-chemici aanbevolen eenhedenstelsel vanuit de kliniek der inwendige geneeskunde. De Erven F. Bohn N.V.: Haarlem, 1971. Solomon D H, Benotti J, DeGroot L J. A nomenclature for tests of thyroid hormones in serum: report of a committee of the American Thyroid Association. J Clin Endocrinol Metab 1972; 34: 884-90. The Committee on Nomenclature of the American Thyroid Association. Revised nomenclature for tests of thyroid hormones in serum. J Clin Endocrinol Metab 1976; 42: 595-8. Yalow RS, Berson S A. Introduction and general considerations. In: Odell WD, Daugha- dayWH, eds. Principles of competitive protein-binding assays. Philadelphia and Toronto, J.B. Lippincott Company, 1971: 1-24. 17 II.2 Thyrotropin (TSH) ••} i The radioimmunoassay of thyrotropin in human blood samples that we '. have applied, is essentially the method described by Odell and co-workers . *f (Odell et al 1965). Materials of human origin (human TSH for radioiodina- , | tion, anti-human TSH serum, purified human pituitary TSH standard) are | required for maximum assay sensitivity, since the structure of TSH is species- f specific. TSH antisera react to some extent with other glycoproteins such as , | follitropin (FSH) and lutropin (LH), since these hormons share a simflar alpha - subunit. Absorption of TSH antisera with a source of alpha subunits, e.g. [ choriogonadotropin (CG) decreases crossreactivity to FSH and LH.

"• 2.1. Ingredients. i a. Buffer-solution. Phosphate buffered saline (PBS) (0.01 M phosphate \ buffer pH = 7.6, containing 0.15 M NaCl) was used as buffer solution. ; { Merthiolate 10"4 M and bovine serum albumin 0.25% (BSA) was added j to the PBS unless stated otherwise. b. TSHStandard. As standard served the Medical Research Council's human pituitary thyrotropin 68/38, which was kindly provided by Dr. D. Bangham (National Institute for Biological Standards and Control, Lon- don). It was dissolved and further diluted in PBS, and stored at -20°C. The approximate estimate of potency by bioassay of this international reference standard is 3 U/mg (Bangham and Cotes 1971). The loss of immunopotency during storage at -20°C is low, and has recently been estimated to be about 0.56% per year (Cotes et al 1978). c. TSH-Antiserum. Rabbit antiserum to human TSH, preabsorbed with human choriogonadotropin (hCG), was purchased from Calbiochem, San Diego, California (lot no. 389523). The lyophilized antiserum was dissolved in PBS and stored at -20°C in a dilution 1=1000. d. TSH-Tracer. Human lyophilized TSH was purchased from Calbiochem (lot no. 389506). lit was solubilized and diluted in PBS (without BSA), and stored at -20°C in ampoules containing 2 jug TSH in 10 jt/1. Radio- ' labeled thyrotropin was prepared by the chloramine-T method (Green- ; wood and Hunter 1963). In this procedure phosphate buffer is em- | ployed to control the pH of the iodination reaction, and the volumes \ are kept low in order to achieve a high percentage incorporation of \ \ 12SI into protein, the yield being strongly dependent on the concen- ,.> ; tration of the protein. Chloramine-T is used to oxidize the Na 12SI; J ; the oxidization is terminated by addition of sodium metabisulphate. ":l - Carrier protein is sometimes added in larger volumes to stop the iodi- ?| nation of the peptide hormone and to minimize iodination damage '! 1 (Hunter 1974). } j In practice our procedure is as follows: 10 /A 0.5 M PBS (pH = 7.6; no i 18

60 60 70 23 24 25 FRACTION NO. FRACTION NO.

Figure 11.2. Elution pattern of TSH iodination reaction mixture on Sephadex G-75 fine (column size 16x1 cm; flow rate 4 drops/minute; fraction volume 0.45 ml). Radioactivity in each fraction is expressed as counts per minute (c.p.m.).

Figure II.3. TSH-tracer test of fractions 16-25 (see figure H.2.). Radioactivity of each fraction is presented in the lower curve, and immunoreactivity in the upper curve (10.000 c.p.m./ 50 Ml of each fraction is incubated with 750 MI PBS and 50 y\ TSH-antiserum FD 1= 16.000 for 22 hours at room temperature; separation of the bound hormone fraction by the DASP technique).

BSA or merthiolate is added to buffer solutions used for iodination), 30 #g chloramine-T in 10 jul 0.5 M PBS, 10 jul 0.5 M PBS and 0.5 mCi Na * 2SI in a volume of about 5 #1 (sodium iodide (125I) for protein iodination IMS.30, specific activity > 14 mCi/jug, in NaOH solution, pH 8-11, 100 mCi/ml, purchased from The Radiochemical Centre, Amersham) are very rapidly added to an ampoule containing 2 ng TSH in 10 jul PBS; the reaction mixture is swirled continuously. Twenty seconds later 60 jug Na2S2Os in 20 jul 0.5 M PBS is added, followed by 200 jul 0.2% BSA in 0.5 M PBS fourty seconds after initiation of the iodination reaction. The reaction mixture is chromatographed on a Sephadex G-75 fine column (size 16x1 cm), which has been previously equilibrated with 1 ml 2% BSA in PBS to minimize adsorption of the 12SI-labeled protein. As elution buffer 0.01 M PBS (pH=7.6) is used, and fractions of 0.45 ml are collected at a flow rate of 4 drops/minute. A typical elution pattern is presented in figure II.2. The second peak of radioactivity (fractions 35AT) is not precipitable with 12.5% trichloroacetic acid, and these fractions do not bind specifically to the antibody. 19 In the first peak radioactivity is precipitated by 12.5% trichloroacetic acid, and the fractions bind specifically to anti-hTSH antiserum. These fractions thus contain the radiolabeled thyrotropin; the second peak of radioactivity constitutes that portion of 12SI not incorporated into protein. The percentage incorporation of the label into the protein, as judged from the surfaces of the two peaks, is 30% in this experiment, and varied from 30 to 57% in four iodinations, with a mean of 43%. The percent tracer specifically bound to antiserum in the fractions 16-25 of the iodination procedure illustrated in figure II.2, is depicted in figure II .3. Maximal binding is achieved in fractions at the trailing edge of the TSH peak, and this is a consistent finding in all our iodination experiments. We selected the fraction with greatest binding for use in our assay, and never employed the peak or earlier fractions. The tracer, diluted in PBS to 10.000 c.pjn./50 fd, was stored at -20°C. The shelf life appeared to be about 3 weeks. At this time, the second best fraction obtained at the original gel filtration, was rechromatographed under the same conditions. The repurified tracer could be conveniently used for another 3 weeks. Thus, in our hands it sufficed to produce radiolabeled thyrotropin only once in six weeks. The specific activity of the tracer, as tested in self-displacement experi- ' ments was on average 130 fiCi/ng, varying from 99 to 161 in six experiments.

2.2. Assay procedure. . a. Selection of "milieu" in standard curve. To ensure that the milieu in the tubes containing the standards is as similar to that of the test samples, it is necessary to add hormone-free serum to the standard tubes. Thyrotropin-free serum can be obtained from hypopituitary patients, - from thyrotoxic patients or from subjects pretreated with pharmacolog- I ical doses of T3 (T3-suppression sera). : Alternatively, one can "strip" the hormone from serum, or employ blood from species whose hormones do not cross-react in the assay (Golstein and Vanhaelst 1973; Jacobs and Lawton 1974). Sluiter and co-workers (1972) added 5% bovine y-globulin in buffer in the standard tubes, which minimized non-specific protein-protein interactions. We '' could confirm then* findings, in that standard curves made in 5% BS7G equalled the curves in T3-suppression sera, while employing 2% BS7G or 20% BSA all resulted in a higher percentage of binding. However, we preferred to use heterologous serum in our homologous test system. Pig ; i serum could be obtained free of charge from the local slaughter-house. 1 I Figure II.4 demonstrates identical standard curves employing either Jj I 400 pi T3 -suppression serum or 400 jul pig serum. The advantage of this '}• I procedure is its low cost. The disadvantage is that we encountered some j; i-f 20

KB J5

6 8 10 AlU TSH/lube reciprocal final dilution antiserum

Figure 11.4, TSH-standard curve, made in human T3-suppression serum (•) or pig serum (x). Reac- tion conditions: SO jul standard, 400 M! serum, 50 nl tracer, 50 fil AS (FD 1-16.000) and buffer up to 800 Ml are incubated for 4 days at 4°C; separation by DASP.

Figure 11.5. TSH-antiserum dilution curve (incubation of antiserum and 50 jul TSH* in buffer for 4 days at 4°C; separation by DASP).

differences from time to time with serum from different pigs. There- fore each new batch of pig serum has to be checked carefully. Finally, non-specific binding of tracer varies from one plasma sample to another, and therefore one may occasionally find a percentage bound greater than the "initial binding", i.e. the percentage of tracer bound to the antibody in the first zero-standard point of the standard curve. b. Selection of tracer concentration. When the reaction mixtures were incubated for 22 hours at room temperature, standard curves employing SO jul TSH* (10.000 c.p.m.) showed higher initial binding and a steeper slope as compared with curves using 25 or 100 JJI TSH*. HOW ever, nearly identical standard curves were obtained with 25, SO or 100 ptl TSH* when the incubation was performed for 4 days at 4°C. Since we decided to incubate for 4 days at 4°C (see below), any one of these tracer volumes could be used. For reasons of pipetting convenience, counting time efficiency and tracer economy the volume of SO JLCI TSH* was chosen. It is calculated from the self-displacement experiments that SO jul radiolabeled thyrotropin contained on average 0.16 MU TSH. c. Selection of antiserum dilution. Tracer binding to the antiserum at various dilutions is depicted in figure II .5. For economical reasons binding of the tracer with undiluted antiserum was not tested. Three 21

7 .i Si

—incubation 24hours, room tetnp 4days .

6 8 10 JUU TSH/lube

FjgareJT.6. TSH-standard cuives under different incubation conditions. Reaction mixture: 50 »xl tracer, SO pi AS (FD 1=16.000), 400 jul pig serum, SO p\ standard, buffer up to 800 Ml; separation by DASP.

different dilutions of antiserum were tested, i.e. final dilutions (FD) of 1=8000, 1=12000 and 1=16000. Although the initial binding was highest in FD 1=8000 and lowest in FD 1=16000, the slope of the standard curves employing a FD 1=16000 was the steepest, and consequently sensitivity was greatest at this final dilution. This was true under the two incubation conditions tested, i.e. 4 days at 4°C, or 22 hours at room temperature. Sensitivity decreased at final dilutions of antiserum 1=20000 and greater. d. Selection of incubation conditions. We compared a long incubation (4 days at 4°C) with a short incubation (22 hours at room temperature). Sensitivity was consistently greater under the longer incubation conditions (figure II.6), and this was true for the three different concentrations of both tracer and antiserum tested. Therefore 4 days at 4°C was chosen. Preincubation in the absence of tracer (which in some radioimmunoassays is known to increase sensitivity) during one or more days did not increase sensitivity substantially. e. Separation of bound and free hormone. The double-antibody solid- phase method was chosen as separation technique. Immunological precipitation of bound fraction is achieved in this method by sheep anti-rabbit 7-globulin antibodies coupled to cellulose (anti-rabbit DASP®, Organon, Holland). Tween 20 (0.15% v/v) was added to reduce aggregation of micro-particles and non-specific adsorption of tracer (Ratcliffe 1974). f. Assay procedure. The assay procedure as finally determined, starts with t- 22 pipetting consecutively the following reagents in 4 ml polystyrene tubes: x Ml PBS up to a final volume of 800 JKI 10-50 Ml standard (hTSH, MRC 68/38) in standard curve tubes only (0.5-10 MU TSH per tube). i 400 MI unknown plasma sample, or pig serum in standard curve tubes 50 Ml tracer, containing ± 10.000 cp.m. ~ 0.16 iiV TSH. 5c 50 Ml antiserum 1=1000(rabbit anti-hTSH,Calbiochem), final dilution 1=16000 After mixing, the tubes are incubated at 4°C for 4 days. Then, to each ?••: test tube 1 ml of DASP® suspension (containing one 5 ml vial DASP, Tween 20 0.15% v/v and PBS up to 50 ml) is added followed by rotating incubation at 4°C for 24 hours. The antibody-bound hormone fraction is collected by centrifugation, the supernatant being sucked off, and the radioactivity in the pellet is counted after a single wash with physiological saline in a gamma-scintillation camera. All samples are also incubated in the absence of antiserum (control tubes or "blanks") in order to correct for non-specific binding of the tracer. Standards are run in triplicate, unknown plasma samples in duplicate.

2.3.Results. a. Sensitivity and precision. The precision of the standard curve was determined by the assay of each standard point in ten-fold in one experiment. The results are given in table U.I and figure II.7. Sensitivity is 0.10 uU TSH/tube, or 0.25

%B 25

0 12 4 6 8 10 UU TSH/luta

Figure 11.7. Sensitivity and precision of the TSH-standard curve (the mean ± 2.SD of the standard points, each assayed in tenfold, is depicted). 23 Table ILL Precision of the TSH standard curve (each point assayed in tenfold). standard point percentage bound coefficient of variation 0*U TSH/tube) mean SD (^°-x 100%) ft f 0 23.84 0.39 1.64 O.S 20.16 0.34 1.69 1 17.35 0.39 2.25 2 13.32 0.39 2.93 4 9.98 0.31 3.11 6 8.01 0.27 3.37 8 6.75 0.34 5.04 10 6.16 0.20 3.24

mU/1 plasma. The detection limit of the assay therefore is rounded off at 0.3 mU/1 plasma. There is no complete discrimination between the last two points of the standard curve. Plasma samples with a TSH concentration of > 20 mU/1 (i.e. > 8 juU/tube) must consequently been diluted and assayed again. The precision of measuring TSH concentration in unknown plasma samples is presented in table II .2. Intra-assay variation is lowest at hormone levels between 2 and 8 mU/1, i.e. measurement seems to be most precise at hormone levels of 0.8 to 3.2 juU/ tube, b. Specificity. Non-specific adsorption in the assay system (binding of tracer in the absence of antibody) was low, and seldom exceeded 3%. Similarity between standard curves and dilution curves of plasma samples could be

Table 11.2. Intra-assay variation of unknown plasma samples in the TSH radioimmunoassay (each sample assayed in tenfold). unknown plasma TSH mU/1 coefficient of variation mean SD (^-x 100%)

A 1.8 0.11 6.11 B 2.0 0.08 4.00 ..,'•3 C 2.9 0.11 3.79 D 3.0 0.18 6.00 E 8.0 0.36 4.50 F 10.5 1.21 11.52 G 21.8 3.02 13.85 24

%B jui unknown ptawna %6 «l unknown plasma SO tOO 200 50 100 30 30

4 6 8 2 4 juU TSH/lube uU TSH/lube

Figure II.8. Paiallelism between standaid cuive and dilution cuive of unknown plasma sample in the TSH ladioinununoassay.

Figure II. 9. Parallelism between standaid cuive and dilutions of unknown plasma samples (taken before and after TRH stimulation in a patient) in the TSH radioimmunoassay.

0.S 8 10 20 40 80100 hFSH.hLH ng/tube 40 100 200 400 1000 2000 4000

%B 30

hTSH

0.5 4 6 8 10 20 100 hTSH jiU/tUbe

Figure n.10. : \ Cross-reactivity of hCG (Pregnyl®, (Organon, Holland), hLH and hFSH (butt. CPDS/2, obtained by courtesy of Dr, Lequin and Dr. Hennen) in the TSH radioimmunoassay (50% displacement of the TSH tracer is effectuated by 0.66 ngTSH/tube (2 (iV), or by 5ngFSH/tube). 25 Table 11.3. Expected and found concentrations of TSH (mU/1) in recovery experiments with three different plasma samples. added TSH (mU/1) 0 1 2 4 6 8

plasma K: expected 1.4 2.4 4.4 6.4 8.4 found 0.4 1.3 2.55 4.3 6.15 10.1 % recovery 93 106 98 96 120 plasma L: expected 2.05 3.05 5.05 7.05 9.05 found 1.05 2.05 3.4 4.8 6.9 9.9 % recovery 100 111 95 98 109 plasma M: expected 3.55 4.55 6.55 8.55 10.55 found 2.55 3.45 4.1 7.0 10.0 9.0 % recovery 97 90 107 117 85

demonstrated (see figures II.8 and II.9), i.e. there was no obvious deviation from parallelism between these lines. Hormonal cross-reactiv- to ity was tested for the related human glycoproteins choriogonadotropin, follitropin and lutropin (figure 11.10). The preparation of follitropin i used in this experiment reacts to a considerable extent in the TSH assay, and in the high concentration range of FSH parallelism is observed with the TSH standard curve. Plasma TSH concentration was measured in 10 postmenopausal and in 10 premenopausal women; there was no difference between the two groups (1.0 ±1.0 vs. 1.0 ± 0.5 mU/1). The recovery of exogenously added TSH to plasma samples with a known TSH concentration is given in table II.3. The mean percentage recovery in these fifteen experiments is 101.5%. c. Quality control. The coefficient of inter-assay variation is about 10% (see table II .4).

Table 114. Inter-assay variation of plasma pools in the TSH radioimmunoassay (n - number of assays). plasma pool TSH mU/1 coefficient of variation mean SD (SJ2 x 100%)

P 18 1.6 0.17 10.43 Q 14 1.9 0.16 8.42 R 19 2.3 0.25 10.87 S 16 7.4 0.85 11.49 T 6 27.0 2.97 11.00 26 - t 2.4. Discussion. ^ ••iz The reaction conditions chosen in our TSH radioimmunoassay were die- H tated by the goal of reaching maximal sensitivy. The slightly better sensitivity . }i| achieved byPekaryetal. 1975(0.2mU/lvs.025mU/linourassay)seemsmainly [ « fe« due to the relatively high affinity constant of their hTSH antibody, which per- ' P? mits a final dilution of the antiserum of 1=1.000.000 in their assay. Preparation I K and selection of optimally radiolabeled TSH is of critical importance for the I ; £='•. sensitivity and validation of the TSH assay. Although it has been described j ••-' |; that human 12 s I-labeled TSH prepared by lactoperoxidase shows less damage b on storage, an increased shelf-life, and improved binding at different anti- r serum dilutions as compared to chloramine-T labeling (Burrin 1976), all ;• authors reporting high sensitivity in the TSH assay use chloramine-T labeling; fc we have used this method exclusively. Thyrotropin tracers of high specific , \. activity are less immunoreactive and consequently decrease sensitivity as ;. opposed to tracers of low specific activity (Erhardt et al 1973; Moser and j | Hollingsworth 1975). Tracers of low specific activity however decrease count- i ing efficiency. Therefore as a compromise iodination conditions were chosen ; | resulting in radiolabeled TSH of intermediate specific activity (specific I activities are in our assay ±130 ^Ci/jug, and 220 -123 - 43 in the experiments of Erhardt et al). The shelf-life of our tracer is in the same order of magnitude as those reported by others (Hershman and Pittman 1971: 2 months if repeatedly purified by gel filtration; Erhardt et al 1973: 3 weeks without purification; Burrin 1976: 3 weeks). We find maximal binding to the antibody in tracer fractions from the trailing edge of the TSH* peak. This is in agreement with Gordin and Saarinen (1972) who report that selection of a specific fraction after gel filtration of the iodination mixture can be critical for the sensitivity and validation of the TSH assay. Thus, parallelism between the standard curve and a dilution curve of a serum sample could be demonstrated using fractions from the trailing edge of the TSH* peak; however parallelism disappeared when the peak fraction was used (Jacobs and Lawton 1974). This phenomenon may be explained by the presence of damaged components of the labeled peptide hormone, which due to their frequent binding to proteins, particular albumin and a-globulins, elute earlier than the purified labeled hormone (Berson and Yalow 1973). Optimal incubation conditions have been investigated in detail by Erhardt et al (1973). Equilibrium was not reached in his experiments at any of the temperatures under test (4 - 12 - 20 - 28 - 32 - 37°C), even not after 140 hours incubation. Initial binding was highest at 20°C incubation, and reaction rate at this temperature was twice as great as at 4°C. Non specific binding was 2.5%, being constant for all temperatures tested. In contrast, maximal binding was the same at each temperature (4,20 and 37°C) and in all groups equilibrium was reached after 4 days in the experiments of Gordon and Table 115 Main features of TSH radioimmunoassays, as reported in literature.

REFERENCE TRACER ANTISERUM STANDARD INCUBATION CONDITIONS SENSITIVITY NORMAL VALUES mU/l spec. act. pg added F.D. pro separation MCi/jug in assay incubation incubation BandF mU/l n mean ± SD range

Hersman 1971 1=175.000 Int. Human 5 days DA 1.25 173 3.9 ± 2.0 TSH Standard 5"C Patel 1971 150-400 25 1=300.000 MRC A 24 hours 4 days DA 0.4 69 1.0 <0.4-3.1 4°C 4°C Hall 1971 200-300 100 1=600.000 MRCA 48 hours 5 days DA 0.5 29 1.6 ± 0.8 <0.5-4.2 4°C 4°C Sluiter 1972 ± 150 1=4.000 MRC 68/38 48 hours 5 days DASP 0.5 37 2.0 ±1.53 4°C 4°C Gordin 1972 ± 100 1=300.000 MRCA DA 1.0 71 3.6 ± 1.4 1.6-8.8 Erhardt 1973 36 1250 1=180.000 MRC 68/38 44 hours 2.5 days DA 0.7 0-3.8 4 18-25°C 18-25°C 140 hours 2.5 days DA 0.4 Moser 1975 35 MRC 68/38 72 hours 4 days DA 29 5.7 2.8-11 Wood 1975 140-160 30 1=40.000 MRC 68/38 48 hours 5 days DA 0.4 80 1.9 <0.5-5.3 I 4°C 4°C i 6 hours 1 day DA 0.4 '• 37°C 37°C Pekary 1975 41-49 1=1.000.000 MRC 68/38 36 hours l.Sday DA 0.2 54 1.5 ± 1.0 4°C 4°C Nisula 1978 150-250 1=60.000 MRC 68/38 5 days DA 0.56 9 <0.56-4.0 4°C Wiersinga 1979 ±130 ±53 1=16.000 MRC 68/38 4 days DASP 0.25 63 1.2 ±1.1 <0.3-4.5 4°C j "3—

28 Saarinen (1972). In our hands, incubating at 4°C for 4 days resulted in higher initial binding and greater sensitivity as compared to incubation at 20°C for 22 hours, and was therefore preferred. Preincubation by adding tracer hormone at a later time may increase sensitivity, and this procedure can be of value for the TSH assay as reported by Erhardt et al (1973): sensitivity at preincubation times of 0, 44 and 140 hours was respectively 2.0, 0.7 and 0.4 juU/ml. Most authors claiming great sensitivity of their TSH assay, employ preincubation without tracer (see table II.5). Preincubation in our system however did not increase sensitivity consistently to an extent which warranted routine use of this time consuming procedure, and was therefore discarded. Separation of bound and free hormone was performed by the double-antibody solid-phase technique. We found this method as reliable and easy to handle as Sluiter et al (1972). The time required for precipitation of the bound hormone fraction by the second antibody may be much shorter than 24 hours (Erhardt et al 1973), but this point was not further investigated. Thyrotropin, follitropin, lutropin and choriogonadotropin are structurally related glycoproteins, each composed of an a- and 0-subunit. The chemical structure of the a-units are very similar in these four peptides, the /3-units giving the molecule its specific biological properties. Consequently, hCG, hLH and hFSH may cross-react in the hTSH-RIA to a considerable extent. Cross-reaction to these peptides may be prevented by adding hCG to the test tubes or preadsorption of the antiserum with hCG (Odell et al. 1965). Never- theless, we observed strong interference of FSH in the TSH assay, apparently due to cross-reactivity. The parallel path of the FSH curve and the TSH standard curve, however, indicates contamination of the FSH preparation with TSH rather than real cross-reaction to FSH. This possibility is even more likely in the light of reports about the hTSH contamination of "immunochemical grade" hFSH preparations, which in one case was 22% by weight (Parlow 1975). The most convincing evidence that FSH does not cross-react in our TSH assay to any extent of clinical importance, is the absence of differences in TSH levels between postmenopausal women (with high FSH plasma concentrations) and premenopausal women. The described TSH assay is peculiar in its use of a relatively high concentration of plasma (400 jil in a final volume of 800//!), by which non-specific effects may arise more frequently. In practice this has not been encountered. The sensitivity of the assay decreases to 0.5 mU/1 plasma if 200 jul plasma is used.

The main features of some TSH radioimmunoassays described in the literature are presented in table II.5. We may conclude that our TSH assay is very sensitive, precise and meeting the requirements for qualitative and quantitative specificity and good quality. Although most euthyroid individ- 29 uals have detectable TSH plasma levels and most hyperthyroid patients have not, there is some overlap in the radioimmunoassay of TSH between the euthyroid and hyperthyroid range for plasma TSH levels (Hall 1972). Since sensitivity seems to be near its maximum in the reported radioimmunoassays of TSH, other techniques in which sensitivity can be dramatically increased are needed for complete separation of euthyroid and hyperhyroid subjects in the TSH assay, A newly designed cytochemical bioassay for TSH based on quantitative microhistochemistry (measuring the TSH-induced increase in lysosomal membrane permeability in thyroid follicle epithelial cells) is promising in this respect, since this assay is 104 times more sensitive than current radioimmunoassays (Bitensky et al (1973). Nevertheless, with use of this assay the overlap in TSH values of euthyroid and hyperthyroid individuals is still present (Petersen et al 1975).

2.5. References.

Bangham D R, Cotes P M. Reference standards for radioimmunoassays. In: Kirkham K E, Hunter W M, eds. Radioimmunoassay methods. Edinburgh and London: Churchill Livingstone, 1971: 345-68. Berson S A, Yalow R S. Radioimmunoassay. In: Berson S A, ed. Methods in investigative and diagnostic endocrinology. Volume 2a, Berson S A, Yalow R S, eds. Peptide hormones. Amsterdam - London: North-Holland Publishing Company 1973: 84- 120. Bitensky L, Alaghband-Zadek J, Chayen J. Studies on thyroid-stimulating hormone and the long-acting thyroid stimulating hormone. Clin Endocrinol 1973; 3: 363-74. Burrin J M. Comparison between human' *s I-labelled TSH labelled with lactoperoxidase or chloramine-T; advantages in the use of enzymically labelled antigen in a radioimmunoassay for human TSH. Clin Chim Acta 1976; 70:153-9. Cotes P M, Gaines Das R E, Kirkwood T B L, Bennie J G, Hunter W M. The stability of standards for radioimmunoassay of human TSH: research standard A and the international reference preparation initially 68/38. Acta Endocrinol 1978; 88: 291-7. Ergardt F, Marschner I, Pichardt R C, Scriba P C. Improvement and quality control of the radioimmunological TSH determination. J ClinChem Clin Biochem 1973; 11: 381-7. Golstein J, Vanhaelst Z. Influence of thyrotnrain-free serum on the radioimmunoassay of human thyrotropin. Qin Chim Acta 1973;49:141-6. Gordin A, Saarinen P. Methodological study of the radioimmunoassay of human thyrotropin. Acta Endocrinol 1972; 71: 24-36. Greenwood F C, Hunter W M, Glover J S. The preparation of 131 I-labelled human growth hormone of high specific radioactivity. Biochem J 1963; 89:114-23. Hall R, Amos J, Ormston B J. Radioimmunoassay of human serum thyrotropin. Brit MedJ 1971; 1:582-5. Hall R. The immunoassay of thyroid-stimulating hormone and its clinical applications. Clin Endocrinol 1972; 1:115-25. Hershman J M, Pittman Jr J A. Utility of the radioimmunoassay of serum thryrotrophin hi man. Ann Int Med 1971; 74:481-90. Hunter W M. Preparation and assessment of radioactive tracers. Brit Med Bull 1974; 30: 18-23. •~n—

--30 •"-,._:-.:,...= Jacobs H S, Lawton N F. Pituitary and placental glycopeptide hormones. Brit Med Bull 1974; 30: 55-61. Moset R J, Hollingsworth D R. Examination of some factors affecting sensitivity and reproducibility in radioimmunoassay of thyrotropin. Clin Chem 1975; 21:347-52. Nisula B C, Louvet J P. Radioimmunoassay of thyrotropin concentrated from serum. I Clin Endocrinol Metab 1978; 46: 729-33. Odell W D, Wilber J F, Paul WE. Radioimmunoassay of thyrotropin in human serum. J Clin Endocrinol Metab 1965;25: 1179-88. Parlow A F. The immunoreactive hTSH contamination of "immunochemical grade" , hFSH preparation. J Clin Endocrinoi Metab 1975; 41: 189-90. - Patel Y C, Burger H G, Hudson B. Radioimmunoassay of thyrotropin: sensitivity and specificity. J Clin Endocrinol Metab 1971; 33: 768-74. Pekary A E, Hershman J M, Parlow A F. A sensitive and precise radioimmunoassay for human thyroid-stimulating hormone. J Clin Endocrinol Metab 1975; 41: 676-84. Petersen V, Smith B R, Hall R. A study of thyroid stimulating activity in human serum with the highly sensitive cytochemical bioassay. J Clin Endocrinol Metab 1975; 41: 199-202. Ratcliffe J G. Separation techniques in saturation analysis. Brit Med Bull 1974; 30: 32-7. Sluiter W J, Kersen F van, Zanten A K van, Beekhuis H, Doorenbos H. A radioimmunoassay of human TSH, employing a solid phase second antibody, and a purified globulin preparation for standardization of nonspecific protein interactions. Clin Chim Acta 1972; 42: 255-62. Wood P J, Smith J, Marks V. Notes on the radioimmunoassay of human thyrotropin. Ann clin Biochem 1975; 12:196-9. v. 31 i £.? //.J. Iodothywnines. ? fl | §" The chemical structure of the various iodothyronines differs only in the f| I number and/or position of the iodine atoms. Due to the similarity in chemical | '•• structure it is not surprising that so-called "specific" antisera raised against a • .. ' | f particular iodothyronine may cross-react considerably with other iodothyro- •• I; I nines. With the techniques cunently available an absolute specific antiserum £ directed against just one iodothyronine cannot be obtained; only the condi- ; [-'. tions chosen for the performance of a radioimmunoassay make the assay -j p rather specific for the iodothyronine under test. This is illustrated by the | following set of experiments. ! 12S ; Radiolabeled iodothyronines, either commercially bought ( I-T4 and jt 12 S 12s I-T3, Abbott Laboratories) or self prepared ( I-rT3 by iodination of f. 3,3'-T2; for details see below), were purified by Sephadex G-25 fine chroma- i. tography (column size 40 x 2.5 cm, volume 135 ml; elution buffer 0.01 M i NaOH in distilled water pH=12; fractions of 2.85 ml each were collected at a ; flow rate of 4 drops/min). The radioactive elution patterns are depicted in figure n.l 1. Two comments may be made. Firstly, pure tracers of T4, and T3 and rT3 are easily obtained by chromatography of the iodination mixture. Secondly, iodothyronines with a higher molecular weight have a larger elution volume than those with a lower molecular weight, and different iodothyronines with the same molecular weight (T3 and rT3) elute at different volumes. Therefore Sephadex chromatography separates iodothyronines not by gel filtration (a process in which larger molecules elute before the smaller ones) but by differential adsorption of iodothyronines onto the gel. From these and other experiments using smaller columns, we calculated the relative retardation of iodothyronines by Sephadex. Mean ratios Vx/Vt (Vx = elution 25 volume of substance x; Vt = total volume of the column) are: * I-albumin 12S 12s 12S 12S I2S 0.39, I 0.66, I-3^'-T2 1.33, I-T3 1.70, I-rT3 2.97, I-T4 3.88. The adsorption of iodothyronines onto the Sephadex gel is illustrated by the fact that the ratio Vx/Vt is > 1. It has been reported that column adsorption may be influenced by the phenolic hydroxyl group and by the number and location of iodine atoms in the iodothyronine molecule: in general more retardation by the gel is observed at a greater number of iodine atoms, and less iodine in the outer ring is associated with a smaller elution volume (Lissitzky et al 1962; Mougey and Mason 1963; Blasi and De Masi ; 1967). • The purified tracers were incubated separately with various dilutions of the best (i.e. most specific) antisera against T4, T3 and rT3. Tracer hormones I were added at volumes of 50 jil (10.000 c.pjn.), containing either 32 pg T4, I 85 pg T3 or 4 pg rT3. Antisera were raised in rabbits against bovine thyro- % : globulin (T4 -antiserum 3045), against a conjugate of T3 -bovine thyroglobulin : (T3-antiserum 137) and against a conjugate of rT3 -BSA (rT3 -antiserum 3686) I 32

1.0

- V, If 0.5 ri. d

. .-7

I- T2* 2.0 o X

1.0 • IJ

0.5 -

n J 0 20 40 60 80 100 120 140 160 180 200 FRACTION NO Figure n.lL ll5 Purification of Mabeled T4 (top), Ts (middle) and rT, (bottom) by Sephadex chromatography (for details see text).

•rrry 33

T4- antiserum 3045

f.r

101 103 TO"

T3-antiserum 137

104

rT3-antiserum 3686 100 r

10Z 103 reciprocal final dilution antiserum Figure 11.12. Binding of radiolabeled T4, T, and rT3 to each of the so-called "specific" antiseia used in the radioimmunoassay of T4, T, and rT3 respectively. 34 (for further details see below). It is obvious (figure 11.12) that each "specific" antiserum also binds the other two iodothyronines, albeit at a higher concentration of the antiserum. All three iodothyronines are present in thyroglobulin, and consequently immunisation by thyroglobulin raises polyclonal antisera. However, the rT3-antiserum also cross-reacts with T4 and T3. There- fore, these antisera are not really specific, and in principle each of these three antisera might be used in the assays of both T4, T3 and rT3, if the appropriate dilution is chosen. In the assay of plasma iodothyronines this is not always practical, because the molar concentration of rT3 in human plasma in 8 times less than of T3, and 460 times less than T4. Therefore, the radioimmunoassay of rT3 is the most critical one, and will be discussed in detail. The same difficulties and procedures apply to the T3 and T4-RIA's, and these assays will be described only briefly.

11.3a. Reverse triiodothyronine (rT3).

3a. 1. Ingredients. a. Buffer-solution. The composition of the buffer is: sodium barbital 0.075 M, merthiolate 10"4 M, bovine serum albumin 0.25% in distilled water, pH=8.6. Sodium barbital is added to block the binding of iodothyronines to prealbumin and albumin, merthiolate to prevent bacterial growth, bovine serum albumin to prevent adsorption of proteins, and 1 M HC1 as much as needed for reaching a pH=8.6 since the antigen-antibody reaction for iodothyronines in general is optimal at an alkaline pH below 9 (Hufner and Hesch 1973). Alkalinization also inhibits binding of iodothyronines to plasma proteins. Unless stated otherwise this buffer is used throughout the assays of iodothyronines.

b. rT3-standard. 3,3',5'-triiodo-L-thyronine (rT3, lot no. 545/24B) was obtained by courtesy of Henning Berlin GmbH. Contamination of this agent with T3 and T4 is < 0.1%. Standard preparations were made by dissolving rT3 in 10 ml 0.01 N NaOH/propylene-glycol (1:2) solution, and further dilution in buffer. Storage was at -20°C.

c. rT3-antiserum. Six white rabbits ("Vlaamse reuzen") were injected intramuscularly at four sites on the thighs and the back with 0 J5 mg rT3 -BSA conjugate (courtesy of Dr. H. Steinmaus, Henning Berlin GmbH) in a suspension of physiological saline in complete Freund's adjuvant (l:3,v/v). A mixture of saline, incomplete and complete Freund's adjuvant (l:1.5:1.5,v/v) was used at the second and following immunizations, which took place at three week intervals. Blood was withdrawn two weeks after the third and fourth immunization. All rabbits developed antibodies to rT3 of good avidity. However, cross-reaction to.T4 was in no case lower than 0.17%, which we considered inacceptably high. Absorption of antisera with T4 decreased crossreactivity to 0.08%, but was associated with 35 7ViWe//.6. Tlieorctical maximal specific activity for radioiodothyronines.

number of'2 * 1 atoms/molecule

1 2 I ft- 3000 6000 3500 7000 3500 4400

considerable loss in avidity and sensitivity. We are most grateful to Dr. T.J. Visser (Rotterdam) who donated rT3-antiserum with a lower cross- reactivity to T4; this antiserum (no. 3686) is also raised in rabbits against the rT3-BSA conjugate of Henning (Visser, Docter and Hennemann 1977). d. rTS-tracer. Tracers of high specific activity are required for the development of sensitive assays for iodothyronines. The theoretical maximal specific activity for radioiodothyronines (according to Kochupillai and Yalow 1978) is given in table II.6. Radioiodothyronines can be produced by exchange of iodine atoms, which results in lower specific activity tracers, and by addition of iodine atoms, which results in a very high specific activity (Nakamura, Chopra and Solomon 1977; Kochupillai and Yalow 1978). Addition of an iodine atom to the outer-ring occurs much more frequently than exchange labeling of outer-ring iodine atoms. There- fore, outer-ring radiolabeled iodothyronines of high specific activity are easily prepared when the substrate for iodination has iodina atom less in the outer ring than the compound desired (Nakamura, Chopra and Solo- mon 1977; Kochupillai and Yalow 1978). Thyronines labeled in the inner ring of the molecule have never been encountered in these procedures.

Consequently, radiolabeled rT3 is prepared in our laboratory as follows. Volumes of 20 jul 0.5 M phosphate buffer pH=7.5,20 ;ul 0.5 M phosphate buffer containing 10 /ig chloramine-T (freshly made) and 10 jul 12SI (1 mCi) are very rapidly added to 0.5 jug 3,3 '-T2 (obtained from Henning Ber- lin GmbH, dissolved in 0.05 M phosphate buffer pH=7.5); the iodination reaction is stopped after 30 seconds by addition of 100 jug ex tempore prepared Na2S2Os in 20 jtd 0.5 M phosphate buffer and the reaction mixture is applied on a small Sephadex LH-20 column of 1.5 ml. Unreacted 12 s I is removed by washing the column in succession with 1 ml of 0.05 M phosphate buffer, 4 ml of distilled water and 1.2 ml of 99.1 (methanol): 36 (2N NH4OH). The radiolabeled iodothyronines are than eluted with 4 ml of 99:1 (methanol): (2N NH4OH), and the eluate is dried under a stream of nitrogen. The residue is redissolved in 1 ml of propylene glycol - 0.01 M NaOK (2:1), and chromatographed on a Sephadex G-25 fine column, size 15x1 cm. The elution buffer is 0.05 M phosphate buffer pH=7.5, and the fraction volume 3ml. The elution profile consists of three peaks, provisionally identified as iodine 3,3'-T2 and rT3. The best fractions of the rT3-peak are pooled, diluted in buffer to 10.000 c.pjn./25 jul, and stored at -20°C. 12S In three successive iodinations specific activity of I-rT3 was found to be 3070,3024 and 3385 juCi/jug. These values approximate the theoretical maximum. In our hands the shelf life of the rT3 is about 10-12 weeks.

310.2. Assay procedure. a. Selection of milieu. Thyroid hormones are transported in plasma by ; j * thyroid hormone binding proteins (TBP), and over 99% of thyroid hormones circulating in the human plasma is bound to these proteins. In determining thyroid hormone content of human plasma samples, the binding capacity of TBP must be neutralized in order to prevent competition between TBP and antiserum for thyroid hormones. This can be done by extraction of the thyroid hormones (by ethanol as applied to the rT3 assay by Chopra (1974) and Meinhold (1975), or most elegantly by extraction of iodothyronines onto small Sephadex columns in an alkaline medium as described for rT3 by Faber et al in 1978). If one wants to avoid the more laborious extraction procedures, it is necessary to block the binding of iodothyronines to TBP in all tubes. This can be done by heating (Sterling and Milch 1974), by alkalinization up to a pH > 11.0 (Hiifner and Hesch 1973), or by addition of agents which displace iodothyronines from their binding sites on TBP (so-called blocking agents). We preferred to avoid extraction procedures, and proceeded as follows. 1. Preparation of iodothyronine-free plasma. Iodothyronine-free plasma is prepared by repeated washings of a human plasma pool with the resin AG 1-X2 (Analytical Grade Anion Exchange Resin AG 1-X2,200400 mesh chloride form, Bio-Rad catalog no. 1401251). The procedure in detail is : 200 gram of AG 1-X2 is washed consecutively ten times with 600 ml of distilled water, three times with 400 ml 0.225 M sodiumbarbital pH=8.6, and ten times with 0.075 M sodiumbarbital pH=8.6; 500 ml of pooled human plasma is mixed with 50 gr. AG 1-X2 under rotating conditions for one hour; after centrifu- • [ I gation the same procedure is applied to the supernatant for another three \ times. The effectiveness of the procedure is checked by performing ex- l pertinents in which known amounts of radiolabeled iodothyronines are 37

%B 60 m

%B 70

Q5 1.0 290 500 fct jjg MIS/tuba

Figure 11.13. rTj -standard curve, made in iodothyronine-free plasma prepared by the resin technique (•) and in plasma of a severely hypothyroid patient (o). Reaction conditions: 50 jil z1 standard, 50 it\ ANS (300 ftg), 50 jul plasma, 50 nl antiserum (FD 1=100.000), 25 nl rTs *, 775 ii\ buffer; incubation for 5 days at 4°C; separation by PEG.

Figure 11.14. Initial binding of rT, as function of 8-anilino-l-naphthalene sulphonic acid (ANS) concentration in the rT,-assay (in the presence of iodothyronine-free plasma).

added to the plasma pool. We found that this procedure removed 99.2% of the rT3-tracer, 99.5% of the T3-tracer and 99.3% of the T4-tracer, that is the prepared plasma contains less than 1% of its original iodothyronine content. In our hands the resin technique was easier to handle and slightly more effective than stripping the plasma pool with charcoal. Standard curves made in iodothyronine-free plasma prepared by the resin technique, or in plasma from a severely hypothyroid patient supposed to have undetectable plasma levels of rT3 (figure n.13) were superimposable, and this is accepted as further proof of the validity of the procedure. 2. Selection ofTBP-binding blocking agent. Several agents which block the binding of iodothyronines to TBP have been used in the assays of iodothyronines, especially salicylate (Larsen 1972), merthiolate (Hufner and Hesch 1973) and 8-anilino-l-naphthalene sulphonic acid (ANS) (Chopra 1972; Mitsuma et al 1972). Salicylate is mainly used in the radioimmunoassay of T3, since very large quantities are needed for full dissociation of T4 from its binding to TBP. It is reported that merthiolate is equally effective (Hufner and Grussendorf 1976) or less 38

effective (Ratcliffe et al 1976) than ANS in blocking rT3-binding to TBP. Since we were already accustomed to the use of ANS in the T4-assay, we therefore decided to use ANS also in the rT3 -assay. Figure 11.14 depicts initial binding of rT3* as a function of ANS concen- rt- tration in the incubation mixture; the experiments were performed in the f. presence of iodothyronine-free plasma. Clearly, without the addition of ANS the labeled hormone fraction specifically bound to the antibody is low, since the tracer is not prevented to bind to the TBP in the iodothyronine-free plasma. At concentrations of 31.2 and 62.5 /ag ANS/tube the percent hormone bound to antibody is highest, whereas a decrease is seen with increasing ANS concentrations. Thus, ANS in high concentration decreases the binding of tracer to antibody in the rT3-RIA. Since in plasmas with high TBG levels (e.g. pregnant women, subjects on estrogen medication) the lower concentrations of ANS did not completely block the rT3-TBP binding, we routinely used 300 jug ANS/tube. However, at very high TBP-concentrations no complete blocking of rT3-TBP binding may occur. b. Selection of tracer concentration. Due to the high specific activity of the radioiodinated rT3, quantities as low as 0.00312 pmol or 2 pg rT3*/tube can be used in a volume of 25 jul (10.000 cpm.) The amount of tracer hormone added is two times less than the hormone concentration of the first standard point (0.0062S pmol or 4.06 pg rT3). Addition of 50 or 100 nl rT3 * resulted in slightly lower sensitivity.

%B 50

2.0 rTjnrol/l

Figure 11.15.

rT3-standard curves under different incubation times (1,3,5 and 7 days respectively at 4°C). Reaction conditions otherwise as stated under assay procedure. 39

c. Selection of antiserum dilution. Binding of radioidodinated rT3 to rT3- antiserum no. 3686 in various dilutions is depicted in figure 11.12. More i than 90% of the tracer is specifically bound by excess antiserum. The I optimal antiserum dilution was determined in several experiments. The slope of the standard curve is very similar when employing final antiserum dilutions 1=100.000 and 1=160.000, although sensitivity is slightly better in the lower dilution. Further dilution of the antiserum results in a substantial decrease of sensitivity. d. Selection of incubation conditions. Several incubation times were tested (figure 11.15). Sensitivity is highest at the longest incubation time (7 days), it is equal at incubation times of 5 and 3 days, and is the lowest after one day of incubation. The slope of the standard curve up to 0.S nmol/1 (upper limit of normal range) is steepest at an incubation time of 5 days. Incu- bation for 5 days at 4°C is therefore preferred. e. Separation of bound and free hormone fraction. We compared immunological precipitation of bound fraction by the double-antibody solid phase technique (DASP) (Den Hollander and Schuurs 1971) vs. fractional precipitation of bound fraction using polyethylene glycol (Polyaethylenglycolum 6000, Brocacef PO 431) (PEG) (Desbuquois and Aurbach 1971). We found separation by PEG better and easier to handle due to the firm pellet obtained than separation by DASP. The PEG technique depends on precipitation of immunoglobulin at a critical concentration of precipitant, which is largely determined by electrostatic forces, influenced by many factors such as temperature, pH, protein and salt concentration. Optimal separation is obtained by the addition of a carrier-7-globulin (e.g. bovine 7-globulin). Instead of bovine 7-globulin we used pig serum which could be obtained free of charge. Optimal conditions were determined experiment- ally. In comparison with separation by 12.5% PEG and 0.2% bovine 7-globulin optimal separation in our hands was achieved with 30% PEG and adding 6 ml pig serum to every 100 ml PEG solution. The final concentration of PEG is thus 14% at an incubation volume of 1 ml. Optimal separation was reached between 10 and 60 minutes after addition of the PEG solution, but was bad at times shorter than 10 minutes. A separation time of IS minutes was therefore selected. Pig serum contains iodothyronines which might influence the result of the assay. However, addition of large amounts of iodothyronines to the pig serum did not interfere with the measurement of endogenous iodothyronine concentration in the unknown samples. Obviously, the short incubation with exogenous iodothyronines at a low temperature (4°C) has no effect on the antigen- antibody equilibrium reached at the end of the first incubation. The separation technique was found to be sensitive to changes in temperature, especially in the rT3 assay: a moderate increase in temperature was associated with a decrease in bound hormone. Therefore centrifugation was 40 carried out in a cooled, temperature-stabilized centrifuge at 4°C. f. Assay procedure. Reagents are pipetted into 4 ml polystyrene tubes: x /il buffer up to a final volume of 1 ml 50 Ml blocking agent (300 jug ANS diluted in buffer, freshly prepared) W SO ml plasma of unknown sample SO ill iodothyronine-free plasma (in standard curve tubes only) SO jul standard (in standard curve tubes only, resulting in concentrations of 0.125, 5* 0.25,0.5,0.75 and 1.0 nmol/l rT,) & 50 Ml rT,-antiserum (no. 3686; dilution 1=5000, final dilution 1=100.000) 2 5 IK 25 Ml' 1-rT3 (containing 10.000 c.p.m. and 0.003125 pmol = 2 pg) After mixing, the tubes are incubated for 5 days at 4°C. Separation of bound and free hormone is performed by addition of 1 ml of polyethylene glycol solution (30 gr. PEG in a final volume of 100 ml buffer to which is added 6 ml pig's serum). After a separation time of IS minutes at 4°C, the antibody-bound hormone is collected by centrifugating for IS minutes at 4°C and sucking off the supernatant. The radioactivity in the pellet is counted in a gamma-scintillation-counter. Each sample is assayed in triplicate. The percent bound hormone is corrected for nonspecific binding by simultaneously assaying samples in the absence of antiserum.

3a.3. Results. a. Sensitivity and precision. The precision of the standard curve was determined by the assay of each standard point in tenfold in one experiment. The results are given in table II.7 and figure 11.16. Sensitivity is 0.00125 pmol rT3/tube, or 0.025 nmol/1 plasma. The detection limit of the assay is rounded off at 0.03 nmol/1 plasma. There is discrimination between the

I I I

0 OSS 025 050 075 1.0 rT] nmol/l

Figure 11.16. Sensitivity and precision of the rT, standard curve (the mean ± 2SD of the standard points, each assayed in tenfold, is depicted). 41 Table 11.7.

Precision of the rT3 standard curve (each point assayed in tenfold). standard point percentage bound coefficient of variation (pmoIrT,/tube) mean SD (§B x 100%)

0 50.8 0.99 1.95 0.00625 42.0 0.63 1.50 0.0125 35.8 0.43 1.20 0.025 28.1 0.64 2.28 0.0375 23.3 0.45 1.91 0.05 20.2 0.73 3.61

last two points of the standard curve, and therefore only plasma samples with a rT3 concentration of > 1.0 nmol/1 have to be reassayed at a higher dilution. The precision of measuring rT3 levels in unknown plasma samples is presented in table II .8. b. Specificity. Non-specific adsorption in the assay system varies between 3 and 7%. There is no difference in non-specific adsorption between the controls of the standard curve and the controls of unknown plasma samples: 4/74 ± 0.36% (n=27) and 4.50 ± 0.43% (n=116) respectively (values as mean ± SD). Therefore, correction for non-specific adsorption in unknown plasma samples is performed by applying the percentage non- specific adsorption observed in the standard curve of the assay. No obvious deviation from parallelism between the standard curve and dilution curves of several unknown plasmas was noted (figure 11.17). Cross-reactivity to various related iodothyronines and iodotyrosines in the rT3 assay is presented in table II.9. Cross-reactivity in these experiments is determined by comparing the concentrations of rT3 and of the cross-reacting agent (x) 12S which both displace 50% of the I-rT3 tracer in the rT3 assay. The

[rT3jatl/2Bc percentage cross-reactivity is thus xl00%. [X] at 1/2 Bo Table 11.8.

Intra-assay variation of unknown plasma samples in the rT3 radioimmunoassay (each sample assayed in tenfold).

unknown plasma rT3 nmol/1 coefficient of variation mean SD (^P x 100%)

A 0.11 0.027 24.5 B 0.27 0.040 14.8 C 0.30 0.018 6.0 D 0.41 0.028 6.8 42

KB 40

0.50 0.75 tO 1.16 1,4 1.2 1.1 tT3 ranol/l dilution unknown plasma

Figure 11.17. Parallelism between standard curve and dilution curves of unknown plasma samples in therT3 radioimmunoassay.

Since on a molar base plasma levels of T4 are about 460 times higher than of rT3, cross-reactivity of T4 in the rT3 assay is most important. As can be seen in figure n.18, concentrations of T4 exceeding those of rT3 by more 12S than 3 orders of magnitude are needed to compete with I-rT3 to the same extent. Cross-reactivity to T4 in the rT3 assay as determined in these experiments, varied between 0.03 and 0.04%, and was the same using dif- Tabkn.9. Cross-reactivity of thyroid honnone analogs in the rT3 radioimmunoassay (AS 3686, FD 1=100.000) (cross-reactivity detennined at 50% displacement level of rT, tracer, expressed on a molar base; cross-reactivity to human thyroglobulin, obtained by courtesy of Dr. J.J.M. de Vijlder, is expressed by weight). thyroid hormone analog abbreviation % cross-reactivity

3,3',5' - triiodo - L - thyronine 100

3,5,3',5' - tetraiodo - L - thyronine T« 0.03 3,5,3\5' - tetraiodo -D- thyronine D-T4 0.03 3.5.3' - triiodo - L - thyronine T, < 0.003 3,5 - diiodo - L - thyronine 3,5-T, < 0.003 3,3' - diiodo - L - thyronine 3,3'-^ 0.65 3',5' - diiodo - L - thyronine 3',S'-Tt 1.33 3 - monoiodo - L - thyronine 3-T, < 0.003 3' - monoiodo - L - thyronine 3'-T, 0.06 L - thyronine To < 0.003 3,5,3',5' -tetraiodo -L-thyroaceticacid TETRAC 0.006 3,5.3' -triiodo -L-thyroaceticarid TRIAC < 0.003 3,5 - diiodo - L - tyrosine D1T < 0.003 3 - monoiodo - L - tyrosine MIT < 0.003 thyroglobulin Tg 0.0005 43

%B SO

0.125 0.25 050 1 25 10 20 50 100 200 500 1000 2000 iodolhyronine nmol/l

Cross-reactivity of T4,T3 and 3,3'-T2 in therT, radioimmunoassay.

ferent batches of T4 (T4 obtained from Henning Berlin GmbH no. 614200, and from Sigma - no. T-2376). Cross-reactivity to T4 in the rT3 assay was also evaluated by adding increasing amounts of T4 to four different plasma samples (table 11.10). Measured rT3 plasma concentrations are

Table 11.10.

Effect of exogenously added T4 on measured rT3 concentrations in four different plasma samples.

mean in- %cross crease in reactivity measured ofadded added T4 measured rT3 (nmol/l) rT3 (nmol/l) T4 inrT3 (nmol/l) plasma E plasma F plasma G plasma H assay

0 <0.03 0.27 0.31 0.39 2.5 <0.03 0.27 0.31 0.39 0 0 5 <0.03 0.27 0.31 0.39 0 0 10 <0.03 0.27 0.31 0.39 0 0 20 <0.03 0.30 0.31 0.39 0.0075 0.0375 50 <0.03 0.33 0.34 0.43 0.03 0.06 100 <0.03 0.37 0.37 0.49 0.07 0.07 200 0.125 0.44 0.42 0.56 0.14 0.07 500 0.25 0.62 0.58 0.80 0.32 0.06 44 Table 11.11.

Expected and found concentrations of rT3 (nmol/1) in recovery experiments with four different plasma samples.

added rT3 (nmol/l) -I: 0.125 0.25 0.50

IV plasma K: expected : - ^ 0.20 0.325 0.575 ; found 0.075 0.20 ; 0.37 0.60 % recovery 100 114 104

plasma L: expected 0.375 0.50 0.75 found 0.25 0.405 0.52 0.90 % recovery 108 104 120

plasma M: expected 0.375 0.50 0.75 found 0.25 0.41 0.57 0.89 % recovery 109 114 119

plasma N: expected 0.40 0.525 0.775 found 0.275 0.395 0.50 0.845 % recovery 99 95 109

consistently higher at additions of 50 nmol/l T4 and more, and at these levels exogenously added T4 cross-reacts for 0.06-0.07% in the rT3 assay. The quantitative specificity of the rT3 assay as judged from recovery experiments, is presented in table 11.11. The mean percentage recovery of exogenously added rT3 hi these twelve experiments is 108%. c. Quality control The coefficient of inter-assay variation is about 12% (table H.12).

Table 11.12. Inter-assay variation of plasma pools in the rT, radioimmunoassay (n= number of assays).

plasma pool rT3 nmol/l coefficient of variation 1 mean SD ($P x 100%) p 18 0.176 0.026 14.8 Q 19 0.276 0.032 11.6 R 18 0.336 0.040 11.9 S 18 0.419 0.041 9.8 45 3a.4. Discussion.

The rT3 radioimmunoassay as performed in our laboratory, is very sensitive (see table 11.14 for comparison with other assays reported in literature). The low detection limit of the assay has been made possible by the use of a high specific activity tracer and of an antiserum with high avidity. 2s Some comments should be made on the preparation of' I-rT3. Firstly, from the specific activity obtained in our radioiodination experiments one 11 can conclude that the rT tracer is produced mainly by monoradioiodination 3 \ (i.e. addition of one radioactive iodine atom) rather than by diradioiodination (i.e. exchange and addition of iodine atoms). This is in accordance with the r experience that radioiodination of iodothyronines is much more likely to occur by addition than by exchange of iodine atoms (Kochupillai and Yalow 1978; Nakamura et al. 1978). The rT3 tracer prepared from iodination of 3,3'-T2 is also more stable than radiolabeled rT3 prepared from iodination of 3-Tj, since monoradioiodinated radioiodothyronines are more stable than diradioiodinated radioiodothyronines (Kochupillai and Yalow 1978). The obtained high specific activity tracer increased the sensitivity of the assay. Secondly, by applying the reaction mixture of the radioiodination of 3,3'- T2 on a large Sephadex G-25 fine column (see figure 11.11) an additional 12S fourth peak was detected, eluting exactly in the volume of I-T3. In this respect it is of interest that Kochupillai and Yalow (1978) observed an unidentified shoulder on the rT3 peak, and that Ratcliffe et al (1976) also detected four peaks, the unidentified third peak eluting at approximately the

Table 11.13. Substrates and products of radioiodination of thyronines.

substrates radioiodothyronine products formed: by exchange by addition one li5I atom two iasI atoms

3,3'-!, 3,3'-Ta

3,5-T4 3-T, 3,3'-!,

3'-T, 3'-T, 3',5'-Ta 3'-T, 3',5'-T, v.Sh j-

Table II-14 Main features of iT, radioimmunoassays as reported in the literature

REFERENCE TRACER ANTISERUM STANDARD' rTj-TBP INCUBATION SENSI- % CROSS-REACTI- F.D. NEUTRAL- TIVITY VITY1 specact. pg added IZATION incubation separation nmol/1 T« 3,3'-^ MCi/Mg in assay B/F

Chopra 1974 400-500 30-40 1=250 Meltzer ethanol 24 hrs. 4°C DA 0.15 0.07 8.1 extraction Meinhold 1975 ±3300 6 1=18.000 Henning ethanol 2 hrs. room temp. DA 0.03 0.025-0.1 1.0 extraction Nicod 1976 1=6.400 Cahnmann ANS 300Mg 2 hrs. -40°C QAE 0.15 0.07 0.5 16 hrs.-room temp. Sephadex 2 days-4° C A-25 Ratdiffe 1976 6 1=40.000 Henning ANS 250/ig 16 hrs. 4o C DA 0.05 0.14 0.42 Griffiths 1976 1=13.500 Henning ANS 175pg 2 hrs. room temp. DA 0.08 0.08 16-20 hrs. 4°C Hüfner 1976 1=25.000 Henning ANS 150Mg 20 hrs. 22° C DASP 0.09 ±0.06 <0.01 Visser 1977 ±2700 4 1=100.000 Henning ANS 500ng 2 days 10°C DA 0.03 0.035 0.57 Burrows 1977 1=315.000 Henning 0.12 0.02 Burman 1977 350-400 10 1=5.000 Cahnmann ANS 300/ig 18 hrs. room temp. (NH4)aS04 0.15 <0.01 0.08 Kaplan 1977 400-500 variable Meltzer ANS 500Mg 16 hrs. 4°C DA 0.06 0.05 0.4 Gavin 1977 ±500 1=10.000 Jorgensen ANS DA 0.08 0.07 3.1 Laurberg 1977 3200 2 1=1.450 Henning ANS 75/ig 18 hrs. room temp. DE cellulose 0.024 0.075 0.23 Kochupillai 1978 3500 1 1=3.000 Meltzer ANS 200Mg 16 hrs. charcoal 0.015 Faber 1978 1=800 Henning Sephadex 16 hrs. room temp. Sephadex 0.07 0.04-0.07 2.3 G-2S fine P-25 fine Mathur 1979 2 Henning ANS 250*ig 16 hrs. 4°C charcoal 0.015 "insignificant" Wiersinga 1979 ±3150 2 1=100.000 Henning ANS 300jig 5days4°C PEG 0.03 0.035 0.65

iJTSiS.' -J WkÄJUÄSi-iiAHi ' TV'

REFERENCE NORMAL VALUES (nmol/l) HYPERTHYROIDISM HYPOTHYROIDISM

mean ± SD range n mean ± SD range mean ± SD range Chopra 1974 0.62 ±0.16 0.42-0.9S 27 1.58 ± 0.75 0.83 - 3.54 22 0.29 ± 0.14 <0.15- 0.54 12

Meinhold 1975 0.28 ± 0.18 0.03-0.71 45 0.801 0.53 0.24 - 1.88 28 0.06 <0.03-0.19 15

Nicod 1976 0.69 ± 0.31 0.38-0.99 83 2.40 1.17-3.97 11 <0.15-0.61 12

Ratcliffe 1976 0.27 ± 0.06 0.15-0.42 44 1.26 ±0.76 0.41-4.66 58 0.10 ±0.04 0.05-0.25 31 0.47 ±0.25 0.21-1.18 153 <0.05 23 Griffiths 1976 0.68 ±0.16 0.43-0.90 67 1.18 ±0.49 0.71-2.50 18 0.19 ±0.07 <0.06-0.33 12 0.42 ±0.16 0.16-0.74 16s Hufner 1976 0.31 ± 0.14 <0.09 - 0.69 60 ±1.00 ±0.46 ±0.30-2.80 34 most<0.09 14 Visser 1977 0.22 Burrows 1977 1 Burman 1977 0.92 ±0.18 0.63-1.37 21 2.60 ± 0.26 0.91 - 3.96 19 0.29 <0.15-0.85 11 Kaplan 1977 0.35 ± 0.12 106 1.39 ± 0.75 22 0.22 ± 0.08 21 0.55 ± 0.20 53 Gavin 1977 0.37-0.65 6 Laurberg 1977 0.37 ± 0.099 0.18 - 0.58 32 1.35 i 0.37 0.81 - 1.98 17 <0.024 - 0.046 10 Kochupillai 1978 Pabcr 1978 0.74 ±0.14 0.45-1.09 58 2.48 ±0.77 1.60-4.74 30 0.37 ±0.06 0.25-0.42 12 0.86 0.82 - 0.88 Mathur 1979 0.29 ± 0.07 0.17-0.42 40 1.61 ±0.99 0.40-5.67 48 0.07 ±0.06 0.015-0.27 35 <0.015 5 Wiersinga 1979 0.24 ± 0.07 0.11 - 0.44 63 0.81 ± 0.54 0.26 - 4.05 55 0.06 ± 0.05 <0.03 - 0.17 33 0.42 ±0.16 0.22-0.69 II3

1 standard: Dr. Meltzer = DL-rT,, others L-rT, 2 on a molar base JTj-toxicosis

sfetiii-V'SsSf'-*«i;irfii*i4itLcSit'ii"--''^"i1"'.1* V-t Jn*»es.*"«i*J''5Vt*i' » ^C ^ .J. -.

:'^r'^~Tm*^7^^ & same volume as in our experiment. The radioiodothyronines which can be '£& fc produced by radioiodination of the various thyronines as substrate, are listed %/; in table 11.13. It must be noted that iodination in the inner ring does not || occur under these circumstances, probably caused by the steric structure of fj^ the thyronine molecule. It can be speculated that the unknown peak consists « || of labeled 3,5,3'-T3, which is formed by addition of radioiodine to 3,5-T2 fe present as contaminant in the 3,3'-T2 batch, since a) the elution volume is |- 12 S that of I-T3 b) the chance of radioiodination by addition is greater than by exchange, which favours 3,5-T2 over 3,5,3'-T3 as the contaminant present as substrate for T3 * (according to Henning Berlin GmbH the 3,3'-T2 preparation is pure for > 99% and contains < 0.5% T3) c) it is improbable that the 12S identity is I-3',5'-T2 (formed by addition from To or 3'-Ti, or by exchange from 3',5'-T2, present as contaminants in the 3,3'-T2 batch) as 3',5'- T2 elutes earlier than T3. The Henning L-rT3 preparation is used as standard in the rT3 assay. rT3 standards from various sources have been shown to behave differently in the same rT3 assay (Premachandra 1978). The slopes of the standard curves differed considerably, and employment of the Henning standard resulted in the steepest slope. The concentration of rT3 in one particular human serum as measured with the Henning standard was 0.31 nmol/1, and 0.65 nmol/1 when Jorgensen's L-rT3 standard was used. The DL-rT3 preparation of Dr. Meltzer has approximately half of the immunological potency as Henning's \\ rT3 (Burrows et al 1977). Premachandra argues that differences in techniques for the synthesis and purification of rT3, and not measurable T3 or T4 contamination of rT3 constitutes to the different results obtained with the various rT3 standards, since these standards even in high concentration did not bind to T3 - or T4 -antisera in the T3 - and T4 -RIA's. However this may be, it is clear that different rT3 standards can result in a different range for normal values in the rT3 assay. All authors performing the rT3 assay in unextracted serum or plasma, employ ANS to dissociate the binding of rT3 to TBP (see table 11.14). Omittance of ANS results in a substantial decrease of specific binding of the rT3 tracer (figure 11.14), and is in itself proof of the binding of rT3 to plasma proteins. The concentrations of ANS chosen will completely block the rT3 protein binding in most cases, but it cannot be excluded that incomplete blocking may occur at very high TBP concentrations. Separation of the bound and free hormone fractions in our rT3 assay is performed by polyethylene glycol, but judging from the literature other separation techniques work equally well (table 11.14). Cross-reactivity to T4 in our assay was 0.03-0.04% when determined by 50% displacement of the rT3 tracer, and 0.06-0.07% when determined by the effect of exogenously added T4 on measured rT3 concentration in a plasma sample. The parallelism between the 3,3'-T2 and T4 curves and the rT3 V. 49

standard curve (figure 11.18) may indicate contamination with rT3 of the 3,3'-T and T preparations rather than real cross-reactivity of the antiserum. 2 4 1 The percentage cross-reactivity to T4 is rather low when compared to the figures in the literature (table 11.14), although some authors report still lower I cross-reactivity. Burrows et al (1977) succeeded in obtaining a T4 -cross- reactivity of 0.02% by preabsorption of the antiserum with T4, albeit at the cost of a loss in sensitivity. In principle this procedure may also be applied to the antiserum 3686, since due to its very good avidity some loss in sensitivity may be acceptable if cross-reactivity substantially decreases. Other authors have sought to solve the problem of T4-crossreactivity in the rT3 assay by correction of the measured rT3 concentrations forT4-crossreactivity. Correc- tion for cross-reactivity however is a complicated matter which is not per- mitted unless the characteristics of the antibody are very meticulously tested (Mathur et al 1979), and we do not apply any correction for cross-reactivity. In practice, it is possible that the presence of T4 in a concentration of 100 nmol/1 in a plasma sample is measured as 0.03 nmol/1 in the rT3 assay. Changes in T4 concentration of SO nmol/1 and more may interfere with the - rT3 measurements, and must be taken into account when interpreting the rT3 levels. Crossreactivity of 3,3'-T2 is greater than of T4 in the rT3 assay, but this is of no practical importance since the 3,3'-T2 concentration in normal i; human serum is approximately 0.02 nmol/1 (Visser et al 1978). Even the highest reported normal values for 3,3'-T2 of 0.3 nmol/1 results in rT3 values ten times less than the detection limit of the assay. For similar reasons, the crossreactivity to 3',5'-T2 in the rT3 assay can be neglected.

The described assay for rT3 further meets the requirements of precision, reproducibility, and quantitative specificity (parallelism and recovery). The reported values for the normal range of plasma rT3 vary considerably (see table 11.14) and this may be due to different technical aspects of the assays (especially differences in sensitivity and specificity with regards to T4), to different rT3 standards and to different ways of correcting for T4 -cross- reactivity.

3a.5. References.

i Blasi F, De Masi R V. Adsorption properties of iodotyrosines and their derivatives on 8 Sephadex. J Chromatogr 1967; 28: 33-6. Burger A, Ingbar S H, Labelling of thyroid hormones and their derivatives. Endocrinology 1974:94:1189-92. Burman K D, Dimond R C, Wright F D, Earll J M, Bruton J, Wartofsky L. A radio-

immunoassay for 3,3',5'-L-triiodothyronine (reverse T3): assessment of thyroid gland content and serum measurements in conditions of normal and altered thyroidal economy and following administration of thyrotropin releasing hormone (TRH) and thyrotropin (TSH). J Clin Endocrinol Metab 1977; 44: 660-72. Burrows A W, Cooper E, Shakespear R A, Aickin C M, Fraser S, Hesch R D, Burke C W. Low serum L-T3 levels in the elderly sick: protein binding, thyroid and pituitary 1 h 50 . $>" f. i , lv |>, responsiveness, and reverse T3 concentrations. Clin Endocrinol 1977; 7: 289-300. *$•*. '('2 Chopra I J, A radioimmunoassay for measurement of thyroxine in unextracted scrum. ', -££ Sf J Clin Endocrinol Metab 1972; 34:938-47. ; *£ ?£ Chopra I J, A radioimmunoassay for measurement of 3,3'.5'-triiodothyronine (reverse jf;, $ Ts). JClin Invest 1974; 54: 583-92. i § § Desbuquois B, Aurbach G D. Use of polyethylene glycol to separate free and antibody- « p. \~ bound peptide hormones in radioimmunoassays. J Clin Endocrinol Metab 1971; |;';" ii 33:732-8. : $ £' Faber J, Friis T, Kirkegaard C, Siersbaek-Nielsen K. Radioimmunoassay of 3,3',5'- K I} triiodothyronine (reverse T3) on small reusable Sephadex columns. Acta Endocrinol J SJ 1978; 87: 313-9. Gavin L, Castle J, McMahon F, Martin P, Hammond M, Cavalicri R R. Extrathyroidal conversion of thyroxine to 3,3\5'-triiodothyronine (reverse T3) and to 3,5,3'- triiodothyronine (T3) in humans. J Clin Endocrinol Metab 1977; 44: 733-42. Griffiths R S, Black E G, Hoffenberg R. Measurement of serum 3,3',5'-(reverse) T3, ' with comments on its derivation. Clin Endocrinol 1976; 5: 679-85. | Hollander F C den, Schuurs AHWH. Discussion. In: Kirkham K E, Hunter W M, eds. ; ; Radioimmunoassay methods. Churchill Livingstone, Edinburgh: 1971: 419-22. • |- Hiifner M, Hesch R D. A comparison of different compounds for TBG-blocking used in .; j \ radioimmunoassay for triiodothyronine. Clin Chim Acta 1973; 44: 101-7. ( Hiifner M, Grussendorf M. Radioimmunoassay for 3.3',5'-triiodothyronine (reverse T3, ] r-T,) in unextracted human serum. Clin Chim Acta 1976; 69: 497-504. = Kaplan M M, Schimmel M, Utiger R D. Changes in serum 3,3',5'-triiodothyronine (rever- • se T3) concentrations with altered thyroid hormone secretion and metabolism. J Clin Endocrinol Metab 1977; 447-56. Kochupillai N, Yalow R S. Preparation, purification, and stability of high specific activity' "Mabeled thyronines. Endocrinology 1978; 102: 128-35. • Kodding R, Hesch R D, Miihlen A von zur. A radioimmunoassay for reverse triiodothyronine. Acta Endocrinol 1976; Suppl. 102:47-9. Larsen P R. Direct immunoassay, of triiodothyronine in human scrum. JClin Invest 1972; 51:1939-49. Laurberg P, Weeke J. Radioimmunological determination of reverse triiodothyronine in unextracted serum and serum dialysates. Scand J Clin Lab Invest 1977; 37: 735-9. .; Lissitzky S, Bismuth J, Rolland M. Separation des composes iodes du serum et de la ; thyroTde par filtration sur gel de dextrane (Sephadex). Clin Chem Acta 1962; 7: i 183-9. • Mathur H, Ekins R P, Brown B L, Malan P G, Kurtz A B. Correction for the presence ] of cross-reactants in saturation assays: application to thyroxine cross-reactivity in 3,3',5'-(reverse)-triiodothyronine radioimmunoassay. Clin Chim Acta 1979; 91: •• 317-27. i | Meinhold H, Wenzel KW, Schiirnbrand P. Radioimmunoassay of 3,3',5'-triiodo-L-thy- : !f. ronine (reverse T3) in human serum and its application in different thyroid states. . || JClin ChemClinBiochem 1975; 13;571-4. ; % Mitsuma T, Gershengorn M, Colucci J, Hollander CS. Radioimmunoassay of triiodo- j % thyronine in unextracted human serum. J Gin Endocrinol Metab 1971; 33: 364-7. \} Mougey E H, Mason J W. Separation of some iodoamino acids and iodide by gel filtra- i ;• tion. Anal Biochem 1963; 6:223-33. .. | Nakamura Y, Chopra I J, Solomon D H. Preparation of high-specific-activity radioactive | j iodothyronines and their analogues. J Nud Med 1977; 18:1112-5. \ ^ Nicod P, Burger A, Staeheli V, Valloton M B. A radioimmunoassay for 3,3',5'-triiodo- ] ., L-thyronine in unextracted serum: method and clinical results. J Clin Endocrinol i Metab 1976; 42:823-9. i -.•M-J--

t3 51 Premachandra B N, Radioimmunoassay of reverse triiodothyronine. J Clin Endocrinol Metab 1978; 47: 746-750. Ratcliffe W A, Marshall J, Ratcliffe J G. The radioimmunoassay of 3,3',5'-triiodothyronine (reverse T3) in unextracted human serum. Gin Endocrinol 1976; S: 63141. Sterling K, Milch P O. Thermal inactivation of thyroxine-binding globulin for direct radioimmunoassay of triiodothyronine in serum. J Clin Endocrinol Metab 1974; 38: 866-75. Visser T J, Docter R, Hennemann G. Radiounmunoassay of reverse triiodothyronine. PY J Endocrinol 1977; 73: 395-6. Visser T J, Krieger-Quist L M, Doctor R, Hennemann G. Radioimmunoassay of 3,3'-diiodothyronine in unextiacted serum: the effect of endogenous triiodothyronine. J Endocrinol 1978; 79: 357-62.

V 52 II.3b. Triiodothyronine(7\).

Although triiodothyronine was identified in human plasma by Gross and Pitt-Rivers as long ago as 1952, it was, due to the cumbersone techniques I needed for measurement of T3, not until the late sixties that new interest in the fate of T3 was awakened by the introduction of less difficult techniques. T3 was measured by a competitive protein binding technique, after extraction of T3 from human serum and separation of T3 andT4 by chromatography (Nauman etal 1967,Sterling et al 1969). However, these techniques also were subject to severe artifacts (Fisher and Dussault 1971; Larsen 1971). The development of radioimmunoassays for the measurement of T3 begins in the early seventies, after the report of Ekins and co-workers (Brown et al 1970) that specific T3 antibodies could be obtained by immunization of rabbits with T3-protein conjugates. At present the T3-RIA is the accepted method for measurement of T3 in human plasma. We applied the technique of Larsen (1972), who employs sodium salicylate as TBP blocking agent. £- I 3b. 1. Ingredients. a. Buffer-solution. The composition of the buffer is: sodium barbital 0.075 M, merthiolate 10"4 M, bovine serum albumin 0.25% in distilled water, pH=8.6bylMHCl. b. T3Standard. 3,5,3'-triiodo-L-thyronine (free acid) was purchased from Sigma, St. Louis, Mo, USA. Standard preparations were made by dissolving the agent in 0.01 M NaOH/propylene-glycol (1:2), and further dilution in buffer. Storage was at -20°C. c. T3-Antiserum. Four rabbits ("Vlaamse reuzen") were immunized against a conjugate of T3 and bovine thyroglobulin (Sigma) as described by Chopra (1972). After three months all rabbits had developed antisera to T3 of good avidity. 12S d. T3-Tracer. I-T3 is purchased from Abbott Laboratories, North Chica- go, Illinois, USA. The specific activity of this preparation is ± 450 juCi//ig. Before use the tracer is purified by Sephadex G-25 chromatography (column size 50x1 cm; elution buffer 0.01 M NaOH in distilled water, pH=12). The elution pattern is similar to the profile shown in figure 11.11 (middle part). Storage of the tracer, diluted in buffer to ± 10.000 c.p.m./ 50/il,isat-20°C.

3b.2. Assay procedure. a. Milieu. Iodothyronine-free plasma prepared by the resin-technique (see section 3a.2a.) is employed in the tubes of the standard curve. Three agents were compared with respect to their effect on dissociating T3 from its binding to TBP: merthiolate (Thiomersal, BDH Chemicals Ltd, no. 30416; 250 jig/tube), salicylate (sodium salicylate cryst. GR, Merck, Art. 53 6601; 3000 /ng/tube), and ANS (8-anilino-l-naphtalene sulfonic acid, ffci Sigma, no. A-312S; 125 jug/tube). T3 concentrations in 23 human plasma samples were measured, the samples being assayed separately with each of the three agents. The results (values as mean ± SD) are: 2.60 ± 1.05 nmol/1 (merthiolate), 2.23 ± 0.99 nmol/1 (salicylate), 2.56 ± 1.03 nmol/1 (ANS). S There is no difference in T3 plasma concentration when merthiolate or ANS is used, but salicylate addition results in lower values as compared to $ addition of merthiolate (p<0.01) and of ANS (p<0.001) (paired t-test). b. Tracer concentration. Approximately 85 pg T3* per tube (± 10.000 c.p jn./50 Ml) is added. c. Antiserum dilution. Out of the four antisera obtained, antiserum 137 demonstrated the highest binding of T3 -tracer, and this antiserum was used exclusively. The tracer is bound for more than 90% by excess antiserum (figure 11.12, middle part). The optimal final dilution of this antiserum is 1=140.000. d. Incubation conditions. The test tubes are incubated for 18-24 hours at 4°C. e. Separation of bound and free hormone fraction. Although the dextran- coated charcoal technique was initially used, we subsequently changed to polyethylene-glycol, because PEG was easier to handle and less subject to nonspecific protein interference (for details see under section 3a.2e.). I' f. Assay procedure. The following reagents are consecutively pipetted in 4 ml %-• polystyrene tubes: x Ml buffer up to a final volume of 1 ml 200 Ml buffer containing 3 mg sodium salicylatc 100 Ml standard (containing 0-2 pg T3/jnl) in standard curve tubes only 50 nl iodothyronine-free plasma (in standard curve tubes only) SO Ml unknown plasma sample

50/il tracer (containing 10.000 c.p.m., ± 85 pgT3) 100 Ml AS 137, FD 1=140.000 The tubes are incubated for 18-24 hours at 4°C, whereafter the bound fraction is separated from the free by polyethylene-glycol (see under section 3a.2f.) Each sample is assayed in triplicate. The percent bound hormone is corrected for non-specific binding by simultaneously assaying samples in the absence of antiserum.

3b. 3. Results. •1 a. Sensitivity and precision. The detection limit of the assay is 0.1 nmol/1. The intra-assay variation is 12% at a plasma T3 concentration of 0.9 nmol/ I 1 and 5% at a level of 2.0 nmol/1. 1 b. Specificity. Cross-reactivity to related compounds is presented in table 11.15. Assuming that the T4 preparation used in these experiments is not contaminated with T3, the cross-reactivity of T4 in the T3 assay is at most

fc" 54 Table III5.

Cross-reactivity of thyroid hormone analogs in the T3 radioimmunoassay (AS 137, FD= 140.000) (see also legend table II.9).

thyroid hormone analog abbreviation % cross-reactivity 1U

3,5,3' - triiodo - L - thyronine T3 100 i § 3,5,3',5' - tetraiodo - L - thyronine T« <0.1 I 3,3',5' - triiodo - L - thyronine rT3 0.0031 3,5 - diiodo - L - thyronine 3,5-T, 0.18 3,3' - diiodo - L - thyronine 3,3'-Ta 0.63 3\5' - diiodo - L - thyronine 3',5'-T, < 0.0005 3,5,3!,5' - tetraiodo - L - thyroacetic acid TETRAC 0.077 3,5,3' - triiodo - L - thyroacetic acid TRIAC 8.3 thyroglobulin Tg 0.62

0.1%; this means an overestimation of T3 concentrations with 6% for the mean normal plasma T3 level of 1.8 nmol/1. No obvious deviation from parallelism between the standard curve and dilutions of unknown plasma samples was observed. The recovery of exogenously added T3 in ten experiments, was on average 101% with a range from 80 to 120%. c. Quality control Inter-assay variation is 13.7% at a plasma T3 concentration of 0.79 nmol/1, 9,6% at 2.08 nmol/1, and 5.6% at 4.42 nmol/1 (samples assayed in 26 consecutive experiments).

3b.4. Discussion.

Antibodies to T3 are easily obtained by immunization of rabbits with thyroglobulin, with thyroid microsomal fraction, or with conjugates of T3 coupled to serum albumin (Chopra et al 1971; Gharib et al 1971). Thyro- globulin injected in Freund's adjuvant was the most effective antigen in these studies; the antisera thus produced have sometimes a higher titer of "T3- specific" binding sites, sometimes the avidity to T4 is the highest (Chopra et al 1971). Our experience is similar: immunization of rabbits with Tg resulted in antisera with highest avidity to T4 which are used in the T4 radioimmunoassay (see figure 11.12), but also in antisera with highest avidity to T . How- f 3 ever, in the T3 assay we preferred antisera obtained by immunization with T3 I coupled to Tg (Chopra 1972), since these antisera demonstrated lower cross- k I reactivity to T4. In general, all these antibodies seem to recognize the thyronine structure and the number and location of the iodine atoms; the iodine atom attached at the 5' location is probably a major antigenic determinant. 55

The specific activity of the T3 tracer employed in most T3 assays is between 70 and 650 MCi/jug/tube, table 11.16). This is far below the theoretical maximum of 3500 juCi/^g. High specific activity tracers of about 3000 juCi/ jug can easily be produced by addition of one radioiodine atom to 3,5-T2, by which the sensitivity of the T3 assay increases to 0.0015 nmol/1 (Weeke and s Orskov 1973). The normal range of plasma T3 concentration is approximately 1.0-2.5 nmol/1, and consequently high specific activity tracers associated with great sensitivity are not needed. There is good agreement in the reported values of plasma T3 between measurements with low and high specific activity tracers k (Weeke and Orskov 1975). Tracers of high specific activity are applied in the direct determination of free triiodothyronine concentration in serum, which is in the order of 0.009 nmol/1 (Weeke and Orskov 1975). Much literature is available on the compounds used for the displacement of T3 from its binding to the serum proteins. Anilino-naphtalene-sulfonic acid (ANS) is reported to be most effective as blocking agent, although it also decreases the binding of the tracer to the antibody; this effect is partially abolished by addition of plasma (Hufner and Hesch 1973). Salicylate demonstrated similar effects, but the decrease in binding could be completely avoided by addition of plasma. As demonstrated by Larsen (1972), salicylate is more effective in displacing T3 than T4 from the binding proteins; theoretically this is of advantage, since cross-reactivity to T4 in the T3 assay thereby diminishes. We applied salicylate as blocking agent. Merthiolate, a deriv- ative of salicylate, is less frequently used in T3 assays. Hufner and Hesch (1973) recommend the use of merthiolate as blocking agent, since it would not influence the antigen-antibody reaction. However, in later studies merthiolate was also found to inhibit the antibody binding reaction (Sterling and Milch 1974; Kanagasabapathy and Wellby 1974).

Incubation conditions were not tested in detail for the T3 assay in our laboratory. Possibly, a much shorter incubation for a few hours at 37°C may also be sufficient to reach equilibrium (Standefer; Lenz et al 1976). Polyethylene glycol was chosen as separation method because of its low cost, reliability and rapidity, but other separation procedures can also be used (table 11.16).

Cross-reactivity to T4 in the T3 assay is important since plasma T4 concentrations are about 60 times higher than of T3 (on a molar base). Most assays demonstrate a T4-reactivity of 0.1% or less (table 11.16). The question is, if this represents real cross-reactivity or T3 contamination of the T4 preparation used. The latter possibility is suggested by reports that T4-reactivity in a particular T3 assay varies with different T4 batches (Patel and Burger 1973; Twomey and Sweat 1975; Seth et al 1976). it The T3 radioimmunoassay as performed in our laboratory is sensitive and precise, and meets the requirements for qualitative and quantitative specificity and quality control. The normal values obtained for the plasma T3 con- Main features of T, ladioimmunoassays, as repotted in the literature

REFERENCE TRACER ANT1SERUM T,TBP INCUBATION SENSI- % CROSS-REAC- NEUTRAL- TIVITY TIVITY' IZATION specact PgT3 immuni- FD incubation separation nmol/l MCI/Mg added zation B/F

s Gharib 1971, 90-120 SO T3-HSA 1=1000/ merthiolate 0.02% pre24hrs. 4°C DA 0.77 0.1-0.4 1972 1=4000 8 hrs. 4°C Mitsuma 1971 70-90 30 T3-BSA 1=400 0.2 Mg tetra- 24 hrs. 4°C charcoal 0.20 0.02 chlorothyronine Larsen 1972 300-500 <30 T3-BSA 1=106.000 salicylate pre 2-5 days 4°C charcoal 0.10 0.1 7 mg/tube 2-3 days 4° C Chopra 1972 90-100 100-150 T3-Tg 1=4000 ANS 250 Mg 24 hrs. 4°C DA 0.18-0.31 0.06 Surks 1972 70-90 75 T3-BSA 1=8000 Sephadex 18-20 hrs charcoal 0.09 <0.1 4°C Lieblich 1972 50 40-80 T3-BSA 1=800 DPH 50 Mg/tube 48 hrs. 5°C DA 0.39 0.2-1.6 Patel 1973 70-83 40-50 T3-BSA 1=125.000 ethanol pre 24 hrs. 4°C DA 0.09-0.22 0.2-1.4 extraction 48 hrs. 4° C Hüfnet 1973 80-110 40-60 T3-BSA 1=20.000 merthiolatc 24 hrs. 4° C charcoal 0.31 0.4-0.6 Sterling 1974 540 T3-BSA 1=5000 60°C2hrs. Ihr. 60°C PEG 0.15 "negligible" 18 hrs. 4°C Meinhold 1974 500 33 T3-BSA 1=180.000 ANS 200 Mg 1 hi. room DA ±0.18 0.05 temp. Twomey 1975 500 50 T3-BSA 1=5000 ANS 400 Mg 3 hrs. 37° C PEG 0.39 0.6-1.6 Burman 1975 600 25-50 1=150.000 ANS 250 Mg 1 hr. 37° C resin 0.39 <0.4 Burger 1975 100 T3-BSA ANS 80Mg 2 hrs. 40°C QAE 0.2-0.4 48 hrs. 4° C Sephadex A-25 Seth 1976 651 5 1=15.000 ANS 250 Mg 24 hrs. 4°C antibody- 0.29 0.02 wash solution3 Stepanas 1977» merthiolate 18 hrs. room resin <0.2 temp. Wiersinga 1979 450 85 T3-Tg 1=140.000 salicylate 18-24 hrs. PEG 0.1 <0.1 3 mg/tube 4°C V;. .'t,'" ,.•-:••

REFERENCE NORMALS (ntnol/l) HYPERTHYROIDISM HVPOTHYROIDISM mean ± SD range mean ± SD tange mean ± SD tange

Gharib 1971, 3.31 ± 0.85 1.85 - 4.80 11.70 ±4.45 5.67-22.68 1.59 ±0.66 0.62- 2.51 1972 Mitsuma 1971 2.13 ±0.35 1.48-2.65 82 7.61 ± 4.08 3.82 - 26.18 60 0.95 ±0.14 0.68- 1.23 45 10.96 ± 6.41 3.51 - 30.80 404 Laiscn 1972 1.69 ± 0.39 38 8.41 ± 6.81 24 0.60 ± 0.32 25 Chopra 1972 1.74 ± 0.50 0.69 - 3.33 96 7.56 ± 3.57 30 0.62 ±0.41 <0.18- 1.60 12 Sutks 1972 2.25 ± 0.37 1.54 - 3.02 38 10.24 ± 4.95 4.45 - 20.02 22 0.68 ±0.40 0.11 -<1.54 29 Lieblich 1972 2.23 ± 0..?9 79 6.61 ± 2.25 36 1.52 ± 0.37 45 Patel 1973 2.00 1.34-2.73 29<5 6.93 2.71 - 17.71 26 0.65 <0.09- 2.11 24 1.96 1.16 - 2.54 209 Hüfner 1973 1.85 ± 0.46 1.39 - 2.31 15 4.62 - 46.20 15 <0.31 - 0.77 7 Sterling 1974 2.91 ± 0.46 31 12.91 ± 6.13 14 0.72 ± 0.60 19 Meinhold 1974 1.77 ± 0.37 1.04 - 2.46 68 6.30 ±3.18 2.57 - 23.5 89 «C0.18- 2.63 62 3.69 ± 1.24 2.47- 7.684 764 Twomey 1975 2.17 1.02 - 3.31 49 Buiman 1975 1.91 ± 0.35 1.17 - 2.68 32 5.05 ± 1.74 2.80 - >9.24 23 0.63 ± 0.31 <0.39 - 1.25 17 Burger 1975 2.57 ± 0.59 99 3.39 - 16.02 31 0 - 1.23 10 Seth 1976 1.70 .1.12 - 2.21 52 Stepanas 19771 2.11 ±0.46 1.19 - 3.02 335 Wiersinga 1979 1.85 ± 0.37 1.00 - 2.85 63 4.94 ±2.51 2.50-13.35 55 0.94 ±0.49 0.19- 2.15 33 3.01 ± 0.56 2.53- 3.94 II4

' on a molar base 'T,-RIA kit, The Radiochcmical Centre, Amcrsham 'solid phase RIA with use of antibodies coupled to Scpharosc 4 T,-toxicosis 9 pre = preincubation - •._—i--.;

58 centrations are in the same Older as described in the literature, although at the lower end of the reported values (table 11.16).

3b.5. References.

Brown B L, Ekins R P, Ellis S M, Reith W S. Specific antibodies to T3. Nature 1970; 226: 359. Burger A, Sakaloff:C, Staeheli V, Valloton M B, Ingbar I H, Radioimmunoassays of 3,5,3'-triiodo-L-thyronine with and without a prior extraction step. Acta Endocrinol f 1975; 80: 58-69. Burman K D, Wright F D, Earll J M, Wartofsky L. Evaluation of a rapid and simple technique for the radioimmunoassay of triiodothyronine (T3). J Nucl Med 1975; 16: 662-5. Chopra I J, Nelson J C, Solomon D H, Beall G N. Production of antibodies specifically binding triiodothyronine and thyroxine. J Clin Endocrinol Metab 1971; 32: 299- 308. Chopra I J, Ho R S, Lam R. An unproved radioimmunoassay of triiodothyronir. ^ ,<) serum: its application to clinical and physiological studies. J Lab Clin Med 1971, -; J: 729-39. Fisher D A, Dussault J H. Contribution of methodological artifacts to the measurement of T3 concentration in serum. J Clin Endocrinol Metab 1971; 32: 675-9. Ghaiib H, Ryan R J, Mayberry W E, Hockert T. Radioimmunoassay for triiodothyronine (T3): I. Affinity and specificity of the antibody for T3. J Clin Endocrinol Metab 1971; 33: 509-16. Gharib H, Ryan R J, Mayberry W E. Triiodothyronine (T3) radioimmunoassay. A critical evaluation. Mayo CHnProc 1972; 47: 934-7. — Gross J, Pitt-Rivers R. The identification of 3,5,3'-L-triiodothyronine in human plasma. Lancet 1952; 1: 439-41. Hiifner M, Hesch R D. A comparison of different compounds for TBG-blocking used in radioimmunoassay for tri-iodothyronine. Clin Chim Acta 1973; 44:101-7. Hufner M, Hesch R D. Radioimmunoassay for T3 in human serum. Acta Endocrinol 1973: 72:464-74. Kanagasabapathy A I, Wellby A S. The use of merthiolate for TBG-blocking in the radioimmunoassay of triiodothyronine. Clin Chim Acta 1974; 55: 267-71. Larsen P R. Technical aspects of the estimation of triiodothyronine in human serum: evidence of conversion of thyroxine to triiodothyronine during assay. Metabolism 1971; 20: 609-24. Larsen P R. Direct immunoassay of triiodothyronine in human serum. J Clin Invest 1972:51:1939-49. Lieblich I, Utiger R D. Triiodothyronine radioimmunoassay. J Clin Invest 1972; 51: 157-66. Meinhold H, Wenzel K W. Studies on the direct radioimmunological determination of triiodothyronine in serum. J Clin Chem Gin Biochem 1974; 12:477-86. Mitsuma T, Nihei N, Gershengorn M C, Hollander C S. Serum triiodothyronine: measurements in human serum by radioimmunoassay with corroboration by gas-liquid chromatography. J Clin Invest 1971; 50: 2679-88. Nauman J A, Nauman A, Werner S C. Total and free triiodothyronine in human serum. J Clin Invest 1967; 46:1346-55. Patel Y C, Burger H G. A simplified radioimmunoassay for triiodothyronine. J Clin Endocrinol Metab 1973; 36:187-190. Seth J, Toft A D, Irvine W J. Simple solid-phase radioimmunoassays for total tri-iodo- 59

thyroninc and thyroxine in serum, and their clinical evaluation. Clin Chim Acta 1976; 68: 291-301. Standefer J C; Lenz P H, Wien G H, Toth J M. Triiodothyroninc radicimmunoassay: incubation time, temperature, and antiscra. Clin Chem 1976; 22: 396-7. Stepanas A V, Mashiter G, Maisey M N. Serum triiodothyronine: clinical experience with a new radioimmunoassay kit. Clin Endocrinol 1977; 6:171-83. Sterling K, Bellabarba B, Newman S I, Brenner M A. Determination of triiodothyroninc concentration in human serum. J Clin Invest 1969;48:1150-8. Sterling K, Milch P O. Thermal inactivation of thyroxine-binding globulin for direct radioimmunoassay of triiodothyronine in serum. J Clin Endocrinol Metab 1974; 38: 866-75. Surks M J, Schadlow A R, Oppenheimer J H. A new radioimmunoassay for plasma L- triiodothyronine: measurements in thyroid disease and in patients maintained on hormonal replacement. J Clin Invest 1972; 51: 3104-13. Twomey S L, Sweat R. Radioimmunoassay for triiodothyronine: use of polyethylene glycol to precipitate immune complexes. Clin Biochem 1975; 8: 169-76. Weeke J, Orskov H. Synthesis of I monolabcled 3,5,3'-triiodothyroninc and thyroxine of maximum specific activity for radioimmunoassay. Scand J Clin Lab Invest 1973; 32: 357-60. Weeke J, Orskov H. Ultrasensitive radioimmunoassay for direct determination of free triiodothyronine concentration in serum. Scand J CLin Lab Invest 1975; 35: 237-44.

r /v

'i « (' "v- 60

//. 3c. ThyroxinefTJ. V.. •rJ, Murphy and Pattee (1964) introduced a competitive protein binding method for measurement of T4 in human serum (referred to as T4-D,T4 assayed by displacement). Essentially, it is based on the competition between unlabeled T4 (derived from the unknown serum sample by ethanol extraction) and labeled T4 for binding sites on a fixed quantity of added TBG; separation of the free hormone fraction is effected by adsorption to resin or charcoal. Com- parison of the results of unknown samples with the standard curve gives the I hormone concentration of the unknown sample. The T4 radioimmunoassays which were soon developed after the success of the T3-RIA's, were found to m be superior to the Murphy-Pattee method.

We used the T4-D-method initially (employing the Tetrasorb-125 T4 diagnostic kit purchased from Abbott Laboratories, North Chicago), but changed to the T4-RIA during our studies. We found almost complete agreement in = the T4 values measured with both techniques (yx4.RIA •0-286 +1.064xx4 .£> • r=0.9825; n = 145). The T4-RIA, essentially the technique of Chopra (1972), will be described briefly. IS' 3c. 1. Ingredients a. Buffer-solution. The composition of the buffer is: sodium barbital 0.075 M, merthiolate 10~4M, bovine serum albumin 0.25% in distilled water, pH=8.6bylMHCl. b. T^Standard. 3,5,3',5'-tetraiodo-L-thyronine (free acid) was purchased from Sigma Chemical Company, St Louis, Mo. Standard preparations were I made by dissolving the agent in 0.01 M NaOH/propylene glycol(l:2), and n further dilution in buffer. Storage was at -20°C. c. Tn-Antiserum. Twelve rabbits ("Vlaamse reuzen") were immunized with bovine thyroglobulin (purchased from Sigma), emulsified in Freund's adjuvant. All rabbits developed antisera with good avidity to T4, although some antisera had higher T3 avidity. 125 d. Tn-Tracer. I-T4 is purchased from Abbott Laboratories, North Chica- go, 111., USA. The specific activity of this preparation is appr. 100 fiCi/ng. Before use the tracer is purified by Sephadex G-25 chromatography in the same manner as described for the T3-tracer. The purified tracer is diluted in buffer to ± 20.000 c.pjn./50/Ltl, and stored at -20°C. 3c.2. Assay procedure. a. Milieu. Iodothyronine-free plasma prepared by the resin technique (see section 3a.2a) is employed in the tubes of the standard curve. The tracer honnone was (mainly non-specifically) bound for 47% in the presence of salicylate (3mg/tube) as blocking agent, but for 6% when ANS (300 Mg/ tube) was used. ANS was therefore selected as blocking agent. b. Tracer concentration. Approximately 120 pg T4* per tube (± 20.000 c,p.m./50juI) is added. c. Antiserum dilution. Antiserum 3045 demonstrated the greatest avidity to T4, and was used exclusively. Excess antiserum binds the tracer for more than 95% (figure 11.12, upper part). The optimal final dilution of this anti- ' | serum is 1=2500. % d. Incubation conditions. The test tubes are incubated for 18-24 hours at |

e. Separation of bound and free liormone fraction. Separation by dextran- j g ' I coated charcoal resulted in a non-specific binding of 20%, and was there- ; I fore abandoned. Polyethylene glycol proved to be very effective as separa- % ting agent. f. Assay procedure. Reagents are pipetted in 4 ml polystyrene tubes in the )\ following order: \ x n\ buffer up to a final volume of 1 ml I 50 p\ ANS (300 fig, tube), freshly made ^ 50 nl standard (containing 0-5 ng T4 /tube) in standard curve tubes only) L 20 nl iodothyronine-free plasma (in standard cuve tubes only) f 20 nl unknown plasma If 50 Ml tracer (containing 20.000 c.p.m., ± 120 pg T4) < | 50 jul antisenim 3045, FD 1=2500. s | The tubes are incubated for 18-24 hours at 4°C. Separation is performed • |; % by PEG, as described under 3a. 2f. Each sample is assayed in triplicate. ,| The percentage bound hormone is corrected for non-specific binding by 1 simultaneously assaying samples in the absence of antiserum.

3c. 3. Results. ':. a. Sensitivity and precision. The detection limit of the assay is 50 pg/tube, or about 3.2 nmol/1. The intra-assay coefficient of variation is 5%at 62 I nmol/1,6.3% at 83 nmol/1,3.4% at 115 nmol/1, and 4% at 209 nmol/1 ;' (each sample assayed in tenfold). '• lv b. Specificity. Cross-reactivity to related compounds is presented in table : | 11.17. Dilution curves of unknown plasmas paralleled the standard curve. » { |; The percentage recovery of exogenously added T4 is on average 100.9% as ? determined in 16 experiments ranging from 98 to 104%. It was equal to I the recovery of 100.2% in thirteen experiments employing ethanol ex- iI traction of T4. Non-specific binding varies from 4 to 7%, and is rather con- : | stant in the controls of standards and unknowns in one assay. Correction | for non-specific binding is consequently calculated from the mean of 5 f controls in one assay. -; 1 c. Quality control. The inter-assay coefficient of variation is 11.8% at 38 i r nmol/1 (n=28), 6.0% at 86 nmol/1 (n=32), 5.3% at 105 nmol/1 (n=29), | | 3.5% at 148 runol/1 (n=30), and 4.4% at 159 ranol/1 (n=32). } | Cross-reactivity of thyroid hormone analogs in the T4 radioimmunoassay (AS 3045, FD 1=2500) (see also legend table H.9).

thyroid hormone analog abbreviation % cross-reac- tvity

- tetraiodo - L - thyroninc 100 3,5,3',5' T4 I- 3,5,3',5' - tetraiodo - D - thyionine D-T4 100

3,5,3' - triiodo - L - thyronine T3 0.2

3,3' - diiodo - L - thyronine 3,3'-Ta <0.01 3,5,3',5' - tetraiodo - L - thyroacetic acid TETRAC 6.5 3,5,3' - triiodo - L - thyroacetic acid TRIAC 0.3 3,5 - diiodo - L - tyrosine DIT <0.01 3 - monoiodo - L - tyrosine MIT <0.01 thyroglobulin Tg 0.26

3c 4. Discussion.

The T4 radioimmunoassay is rather easy as compared with T3 and rT3-RIA's, due to the relatively high plasma concentration of T4. Consequently, there is no need for great sensitivity and high specific activity tracers, and even considerable cross-reactivity to T3 does not result in detectable changes in measured T4-values. The robust character of the assay is reflected in the low intra- and inter-assay variation, and in the close agreement between the reported normal values (see table 11.18). Antiserum 3045 is peculiar in demonstrating a very low cross-reactivity to T3, but this is of no practical advantage. Most assays reported in the literature employ ANS to displace T4 from TBG. It should however be realized that in plasmas with high TBG concentrations ANS may not completely inhibit the T4-TBG binding, resulting in high values when polyethylene glycol and double-antibody methods are used for separation of the bound hormone fraction (Evans et al 1977). Recently, the effectiveness of ANS as a competitor for thyroid hormone binding sites has been explained by structural similarities between the molecular conformation of ANS and thyroid hormones (Cody and Hazel 1977).

The T4-RIA has been found to be superior to the T4-D technique in many respects. It obviates the extraction step, and due to its greater sensitivity much less plasma is needed than in the T4-D method. The linear range of the standard curve is from 1S-S18 nmol/1 in the T4 -RIA, and from 39-259 nmol/1 in the T4-D (Chopra 1972), permitting more samples to be assayed undiluted by T4-RIA. The T4-RIA is also more specific, since halofenate (Karen et al 63 1976), unesterified fatty acids (Rootwelt 1975; Shaw et al 1976) and orphen- J1.-'' adrine (Wiersinga and Touber 1977) all interfere in vitro in T4-D assays but not in T4-RIA's. These agents or its metabolites are soluble in ethanol, and I the ethanol extracts will contain appreciable quantities of these ethanol- soluble compounds, which compete with labeled thyroxine for the binding sites on the fixed quantity of TBG added in the T4-D assay, thus leading to spuriously elevated 'thyroxine" values. Applying the T4-D method in serum extracts obtained by Sephadex chromatography results in normal values, just as in the T4-RIA.

There is a very good agreement in T4 values as measured by T4-RIA and by the T4-D assay for the hyperthyroid and euthyroid range, but many authors report occasionally higher values by RIA than by D-assay for samples in the hyperthyroid range. The discrepancy disappears when the RIA is ap* plied to serum extracts (Chopra 1972). According to Chopra (1972) these inconsistences are due to the presence in serum of T4 covalently linked to I, serum proteins; the nature of this nonextractable T4 is unlikely to be thyroglobulin. Herrmann (1973) and Kubasik (1973) point to the. decrease in recovery found at high T4 concentrations in the T4-D method but not in the T4-RIA. Seth (1973) found the same T4-values in unextracted serum and in Sephadex-extracted serum, which also negates the presence of significant quantities of T4 covalently linked to protein. Different separation techniques in the T4-RIA were applied by Premachandra (1976); although mean T4- values did not differ with the various separation techniques used, individual I samples demonstrated considerable variation. The occasionally found dis- u crepancy between T4-values of hyperthyroid patients measured by RIA or the D-method, seems to be a methodological problem, probably resulting from incomplete extraction of T4 in the T4-D assay and from incomplete displacement of T4 from TBG by ANS in the T4-RIA.

3c. 5. References.

Beckers C, Cornette C, Thalasso M. Evaluation of serum thyroxine by radioimmunoassay. J Nud Med 1973; 14: 317-20. Chopra 1 J, Solomon D H, Ho R S. A radioimmunoassay of thyroxine. J Clin Endo- crinol Metab 1971; 33: 865-8. Chopra I J. A radioimmunoassay for measurement of thyroxine in unextracted serum. J Clin Endocrinol Metab 1972; 34: 938-47. Cody V, Hazel J. Molecular confirmation of ammonium 8-aniIino-l-naphthalene-sulfo- nate hemihydrate. A fluorescent probe for thyroxine binding to thyroxine binding globulin. J Med Chem 1977; 20:12-6. Dunn R T, Foster L B. Radioimmunoassay of thyroxine in unextracted serum, by a single-antibody technique. Clin Chem 1973; 19:1063-6. Evans S E, Burr W A, Hogan T C. A reassessment of 8-anilino-l-naphthalene sulphonic acid as thyroxine binding inhibitor in the radioimmunoassay of thyroxine. Ann Clin Biochem 1977; 14: 330-4. Table 11-18

Main features of T4 radioimmunoassays, as reported in the literature

REFERENCE TRACER ANTISERUM T4-TBP INCUBATION SENSI- % CROSS-REAC- NEUTRAL- TIVITY TIVITY' IZATION specact PgT4 immuni- FD incubation separation nmol/I ' !• T, '„ ''' _', '•'',;

MC/pg added zation B/F '•>•• :';• ••'• '.•

Chopra 1972 50-75 200 hTg 1=300 ANS 150pg 1 hr room DA 15 8.4 temp. 5 min. 4°C Mitsuma 1972 50 T4-BSA 1=1000 ANS 175ug 90' 37°C charcoal 0 10' 40°C Larsen 1973 bTg 1=2000 salicylate 18 mg/ml overnight 4°C, charcoal 10.9 or:90'37QC + 20' 4°C Beckers 1973 50-80 200 bTg 1=500 ANS 300pg pre-incub: DA 13 i 1 hr room temp. l'hr room temp. Dunn 1974 30-50 700 T4-BSA heat 1 hr room temp. charcoal 3.4 •:. - denaturation = boiling 1 hr. Meinhold 1974 80-100 50 T4-HSA 1=1800 I ANS300pg 20hrs. 4°C DA 5 Werner 1974 100 1=3000 [ ethanol extraction 30' 40° C PEG 1 ; <10.0 Herrmann 1974 100-170 10-30 T4-BSA 1=2000 ANS600pg 90' 37° C charcoal 3; • 2.1 10' 4°C Ratcliffe 1974 20-50 500 T4-HSA 1=5000 ANS 400jug 2 hr 37°C DA 5''.' , '"' 3.5 ' • ••,•'• 400-800 or overnight 4°C ! (1.1-5.2) Seth 1975 T4-BSA 1=4500 ANS 24 hrs. 4°C antibody 10 ! 1.3 : •;••• 1 1=6000 wash solu- 1 '< ' ' ' ' ; '

1 tion •' ' ' . " Visser 1975 500 bTg 1=4000 ANS 350/ig 16 hrs. 4°C DA 6 '' : '' 0.6 ' ' ''• "i Marsden 1975' 25-50 25 T4-BSA 1=500 ANS 200jug IV, hr 37° C charcoal ! 1.6 ,, !.''!•! 8,'- Prema- 1976 335 bTg 1=157 TCA-NaOH 1 hr 37°C resin 8 4 chandra 1 15' room temp. '. ' Farid 1977 ANS 300Mg 2 hrs room temp.PEG <13 0.6 Wiersinga 1979 100 120 bTg 1=2500 ANS 300Mg 18-24 hrs 4° C PEG 3 V) ! 0.2 :: : • ,;

.>.•• ..- .y \ i • I; I ,..-..*

REFERENCE NORMAL VALUES HYPERTHYROIDISM HYPOTHYROIDISM x± SD range n x + SD range n x±SD range n

Chopra 1972 108 ± 31 58- 171 40 323t 113 153- 596 40 45 ±12 30-62 7 Mitsuma 1972 96 ±25 51- 127 15 210± 47 156- 285 11 17 ± 10 5-26 14 Larsen 1973 108 ± 21 59 264± 84 40 26 ±17 32 Beckers 1973 124 ± 26 79 327 ± 104 30 34 ±20 21 Dunn 1974 58- 153 54 Meinhold 1974 104 ± 21 71- 152 25 179- 443 15 5-75 15 Werner 1974 Herrmann 1974 235 * 48 62 Ratcliffe 1974 100 ± 21 56- 145 61 249± 48 171- 332 23 18 ± 9 < 5-34 17 Seth 1975 Visser 1975 110 ± 25 36 242± 54 20 17 ±18 13 Marsden 19751 81 ±14 30 183 ± 55 18 8± 9 14 overt 52 ±25 4 borderline Prema- chandra 1976 104 ± 21 71- 161 no 242± 62 165- 368 58 32 t 17 0-71 43 Farid 1977 114 ± 19 96 235± 60 9 49 ±25 11 Wiersinga 1979 102 ± 18 62- 151 63 197± 50 123 - 338 55 26 ±20 < 5-68 33 123± 13 / 107- 145 11'

1 no iodothyronine-frce serum in standard curve 'Tj -toxicosis 3 on a molar base 66

Farid N R, Kennedy C. Assessment of a method for measuring serum thyroxine by radioimmunoassay, with use of polyethylene glycol precipitation. Clin Cheui 1977; 23: 1333-S. Herrmann J, Rusche H J, Kriiskemper H L. Rapid radioimmunoassay for the evaluation of thyroxine in unextracted serum. Clin Chim Acta 1974; 54:69-79. Karch F E, Morgan J P, Kubasik N P, Sine H E. Effect of halofenate on serum thyroid hormone determinations in vitro. J Clin Endocrinol Metab 1976; 43: 26-31. Kubasik N P, Sine H E, Murray M H. Serum thyroxine: results compared for a commercial radioimmunoassay and a commercial competitive protein binding analysisrCJin Chem 1973; 19:1307-08. Larsen P R, Dockalova J, Sipula D, Wu F M. Immunoassay of thyroxine in unextracted human serum. J Clin Endocrinol Metab 1973; 37:177-182. Marsden P, Facer P, Acosta M, Howorth P J N. A comparative study of serum total thyroxine estimation on unextracted serum by radioimmunoassay and by competitive protein binding. J Clin Path 1975; 28:608-12. Meinhold H, Wenzel K W. Radioimmunoassay of thyroxine in unextiacted serum. Horm Metab Res 1974;6:169-170. Mitsuma T, Colucci J, Shenkman L, Hollander C S. Rapid simultaneous radioimmunoassay for triiodothyronine and thyroxine in unextracted serum. Biochem Biophys Res Commun 1972; 46: 2107-13. Murphy B E P, Pattee C J. Determination of thyroxine utilizing the property of protein- binding. J Clin Endocrinol Metab 1964; 24:187-96.

Ptemachandra B N, Ibrahim II. A simple and rapid thyroxine radioimmunoassay (T4- RIA) in unextracted human serum; a comparison of T4-RIA and T4 displacement assay, (T4-D), in normal and pathologic sera. Clin Chim Acta 1976; 70: 43-60. Ratcliffe W A, Ratcliffe J G, McBride A D, Harland W A, Randall T W. The radioimmunoassay of thyroxine in unextracted human serum. Clin Endocrinol 1974; 3: 481-8. Rootwelt K. The influence of fatty acids on serum thyroxine determination* by competitive protein-binding radioassay. Scand J Clin Lab Invest 1975; 35: 649-53. Seth J, Rutherford F J, McKenzie I. Solid-phase radioimmunoassay of thyroxine in unextracted serum. Clin Chem 1975; 21:1406-13. Shaw W, Hubert I L, Spiertoo F W. Interference of fatty acids in the competitive protein-binding assay for serum thyroxine. Clin Chem 1976; 22: 673-8. Visser T J, van den Hout-Goemaat N L, Docter R, Hennemann G. Radioimmunoassay of thyroxine in unextracted serum. Neth J Med 1975; 18:111-115.

Werner S C, Acebedo G, Radichevich I. Rapid radioimmunoassay for both T4 and Ts in the same sample of human serum. J Clin Endocrinol Metab 1974; 38: 493-5. Wiersinga W M, Fabius A J M,Touber J L. Orphenadrine (Disipal), serum thyroxine and thyroid function. Acta Endocrinol 1977- 86: 522-32.

t ?;•• 67

Sc H.4. Traodothyronme uptake ratio (T3 U)

Y-r- In the blood thyroid hormones are bound to plasma proteins, predominantly ••- • to thyroxine-binding globulin (TBG), but also to albumin and a thyroxine- ; ^ binding prealbumin (TBPA). T3 is bound to TBG and albumin, but T4 and 5 v I rT3 also to TBPA (Snyder et al 1976; Chopra 1978). TBG has the highest j affinity for T4 and the lowest for T3, and serves as a high-affinity, low i capacity binding protein, whereas albumin and TBPA have a low-affinity high-capacity binding for iodothyronines. Therefore, under normal circum- 1 < stances TBG is the most important thyroid hormone binding protein in the blood. The proportion of unbound thyroid hormone in the blood is 0.03% for T4, and 0.3% for T3 and rT3. It is generally accepted that only unbound thyroid hormones can enter the target tissues, and it is the free thyroid hor- ' s; mone concentration in the blood which is of major importance in deter- ; mining the metabolic state of the peripheral tissues. Due to the very low con- , centration of free thyroid hormone it is not easy to measure it directly. An ; indirect estimation of the free thyroid hormone concentration can however . be easily obtained. The dynamic equilibrium between free and bound thyroid hormone in t plasma can be expressed as (T4 is chosen as the prototype of iodothyronines, and TBG as the most important binding protein) (Woeber 1978): ; T4 +TBG ^ T4 . TBG According to the law of mass action it follows:

[T4. TBG] K [T4].[TBG]~

in which [T4. TBG] represents the concentration of T4 bound to TBG, [T4] the free T4 concentration, [TBG] the concentration of unoccupied •; binding sites on TBG, and K the equilibrium constant. This expression can be rearranged as:

£ K. [T4] = [T4. TBG] . —— ': I [TBG] I The radioimmunoassay of T4 described under section 3c of this chapter, I measures the total T4 concentration in plasma, and this can be considered • I equivalent to [T4.TBG] since the unbound fraction of T4 is very small. '; I Measurement of——can easily be done with the triiodothyronine uptake : I [TBG] \ i ratio (T3U), a procedure originally introduced by Sterling and Tabachnick [\ t (1961). Radiolabeled T3 (preferred over labeled T4 due to its lower affinity ;| I for TBG) is incubated with a plasma sample and an insoluble, particulate % I material (usually an anion exchange resin). The uptake by the resin of any 3 I labeled T3 not bound to the plasma proteins, is determined (hence the term \

•••••*—r.y. "• u.

'resin T3 uptake' or T3U). Since the labeled T3 is bound to the plasma -a" proteins and to the resin in the same proportion as the endogenous hormone, the resin T3 uptake varies inversely with the binding affinity and concentra- vv. tion of unoccupied binding sites on the plasma proteins. It thus represents the proportion of free thyroid hormone in plasma. I From the product of total T4 concentration and the T3U in a plasma sample a free thyroxine index (FT4 index) can be calculated (Clark and Horn ¥ 1965). The FT4 index is in good correlation with the free T4 concentration (measured by equilibrium dialysis) in normal euthyroid adults, in hyperthyroid patients and in hypothyroid patients, but not in acutely or severely ill euthyroid patients ("sick euthyroid patients") (Anderson 1969). In principle, by the same procedure FT3 and FrT3 indices can be obtained, and a good correlation between the FT3 index and the free T3 concentration has been found in normal euthyroid subjects and in hypothyroid or hyperthyroid patients (Olsen and Andersen 1979). As described above, the free thyroxine concentration is positively related to the total T4 concentration, and negatively to the binding affinity and il concentration of unoccupied binding sites of TBG for T4. It thus is subject to changes in TBG concentration and to changes in T4 supply (Woeber 1978). An increased T4 supply (e.g. in hyperthyroidism) results in a greater total T4 concentration and a greater proportion of unbound T4, i.e. in a greater free thyroxine concentration; the reverse is true for a decreased T4 supply as in hypothyroidism. An increase in TBG concentration (e.g. induced by estrogens) results in a greater total T4 concentration and in a smaller proportion of unbound T4, i.e. the free thyroxine concentration does not change. In general , concordant changes in T4 and T3U tests indicate a change in hormone supply, whereas discordant changes indicate changes in hormone binding to the plasma proteins (see table 11.19).

The T3U method used in our laboratory is the Triosorb M-125 kit purchased from Abbott Laboratories, North Chicago. The inter-assay coefficient of variation of T3U is 4.4% at a ratio of 0.90 (n=10), 3.0% at a ratio of 1.01 (n=9), and 3.6% at a ratio of 1.11 (n=8). The results are expressed as the triiodothyronine uptake ratio, that is the T3U of the unknown plasma sample divided by the T3U of a pool of normal standard reference serum included in the same assay run. This form is strongly preferred over expression of the results as percentage uptake of the labels on the resin, and this form should be used exclusively in calculating a free thyroxine index (The Committee on Nomenclature of the American Thyroid Association 1976). By doing so the normal range of T3U is centered on 1.0, and the calculated FT4 index is of the same order as the total T4 concentration. Whereas many physicians have already difficulties in distinguishing total T3 concentration measured by RIA from the T3U (both are frequently referred to as T3 test), confusion has been complete by the use of the inverted fraction of T3U in 69

•*•'.

Factors that influence free thyroxine concentration, according to Woeber (1978). (t = in- I?'' creased; i = decreased; N = normal).

total T4 concentration % free T4 concentration of free T4 or or or i factor T4-RIA, T4-D T3U FT4 index increased T4 supply decreased T4 supply increased T4 binding N decreased T4 binding N

some methods, resulting in low ratios in hyperthyroidism and high ratios in hypothyroidism. This should be abandoned, and the notation of theT3U should always be in such a manner that hyperthyroidism is associated with increased ratios and hypothyroidism with decreased ratios.

References

Anderson B G. Free tliyroxine in serum in relation to thyroid function. JAMA 1968; 203:135-40. Chopra I J. Nature, source and biologic significance of thyroid hormones in blood. In: Werner S C, Ingbar S H, eds. The thyroid, 4th edition. Hagerstown: Harper and Row, 1978: 100-14. Clark F, Horn D B. Assessment of thyroid function by the combined use of the serum protein-bound iodine and resin uptake of I-triiodothyronine. J Clin Endocrinol Metab 1965; 25: 39-45. The Committee on Nomenclature of the American Thyroid Association. Revised nomenclature for tests of thyroid hormones in scrum. J Clin Endodrinol Metab 1976; 42: 595-8. Olsen K J, Anderson C J. Determination of free thyroid hormone fractions using a

simple T3-uptake test. Scand J Clin Lab Invest 1979; 39:119-23. Snydcr S M, Cavalieri R R, Goldfine ID, Ingbar S H, Jorgensen E C. Binding of thyroid hormones and their analogs to thyroxine-binding globulin in human serum. J Biol Chem 1976; 251: 6489-94. Sterling K, Tabachnik M. Resin uptake of I131-triiodothyronine as a test of thyroid function. J Clin Endocrinol Metab 1961; 21: 456-64. Woeber K A. Tests of thyroid hormone transport. In: Werner S C, Ingbar S H, eds. The thyroid, 4th edition. Hagerstown: Harper and Row, 1978: 33844. ^

if 70 II. 5. Sample handling and storage.

a. Plasma versus serum samples. Assays were done in plasma and in serum samples obtained simultaneously from the same subject (table 11.20). The results are the same in all assays whether plasma or serum is used; this is in accordance with other studies (Kubasik 1978). For reasons of convenience it fc. • was therefore decided to routinely collect blood in heparinized tubes (Searle LH/10, lithium heparin) and to use plasma exclusively. b. Time of sampling. Blood was collected in the morning between 8.30 and 10.30 pjn. in semirecumbent position, unless stated otherwise. This procedure avoids interference of diurnal variation in comparing hormone levels (De Costre et al 1970; Weeke 1973). c. Technique of blood collection. Thyroid hormones in the blood are bound to proteins for more than 99%. Thus, venous stasis is likely to induce increased values. The effect of various techniques of blood sampling was evaluated in 11 patients. Samples were withdrawn (a) without venous stasis by an indwelling venous catheter (b) with venous stasis by ordinary venapuncture in the contralateral arm (c) after 5 minutes of venous stasis through the indwelling venous catheter. The three samples in each patient were obtained within 7 minutes. As presented in table II.21, venous stasis is associated with an increase in the values of iodothyronine plasma concentrations; FT4 index does not change due to the concomitant decrease in T3U. These changes are related to the increase in protein concentration. The same observations have been made by several authors (Lalla and Grasbeck 1975; Judd et al 1975; Jacobsen et al 1977). Our experience is that in practice venous stasis can pose problems, since not infrequently elevated plasma T3 concentrations in patients referred under the diagnosis of T3-toxicosis, were found to be exclusively caused by longstanding venous stasis due to a difficult venapuncture. d. Storage of plasma samples. Plasma samples are stored at -20°C until assay. The samples are frozen within 4-6 hours after the blood collection. We

Table 11-20

Comparison of plasma and serum in the assays of TSH, rT3, T3, T4 and T3 U (values as meantSD).

assay n plasma serum

TSH mU/1 25 1.65 ± 0.9 1.59 ± 0.8 rT, nmol/l 14 0.23 ± 0.09 0.24 ± 0.09 i T3 nmol/l 24 1.61 ± 0.35 1.62 ± 0.34 T« nmol/l 25 105 ± 18 105 ± 16 T3U 20 0.973 ± 0.10 0.977 ± 0.10 Table 11-21

Effect of venous stasis on the plasma concentration of TSH, rT3, T, and T4, and on T3 U and FT4 index in 11 patients (values as mean ± SD, or as geometric mean with x-SD and x+SD between brackets; statistical significance of (b) vs (a) and of (c) vs (a) by the paired t test).

assay without venous stasis, with venous stasis after 5 min. venous stasis. by indwelling catheter by venapuncture by indwelling catheter (a) (b) p(b vs.a) (c) p(c vs.a) Hb mmol/1 8.3 ± 1.1 8.6 ± 1.1 <0.01 8.7 ± 1.1 <0.02 Htl/1 0.407 ± 0.06 0.417 ± 0.05 <0.02 0.422 ± 0.06 <0.05 total protein g/1 71.8 ± 5.4 73.1 ± 3.7 N.S. 76.6 ± 4.7 <0.01 albumin g/1 42.4 ± 3.3 42.5 ± 3.1 N.S. 44.2 ± 4.0 <0.01 TSH mU/1 1.2 ( 0.3 - 7.8 ) 1.1 ( 0.3. - 7.4) N.S. 1.1 ( 0.3 - 7.2 ) N.S. rT3 nmol/1 0.28( 0.14- 0.57) 0.30( 0.14- 0.63) N.S. 0.32( 0.16- 0.63) <0.05 T3 nmol/1 2.12( 1.11- 4.05) 2.13( 1.13- 4.01) N.S. 2.24( 1.18- 4.26) <0.05 T4 nmol/1 80 (46 -137 ) 84 (49 -145 ) <0.05 87 (49 -154 ) <0.05 T3U 1.06( 0.81- 1.38) 1.04( 0.81- 1.33) <0.05 1.01( 0.79- 1.29) <0.01 FT4 index 85 (41 -177 ) 87 (42 -180 ) N.S. 88 (41 -185 ) N.S. Seth J, Toft A u, Irvine w J. sunpie sguu-piiasc iauiui

72 have not encountered appreciable changes in hormone concentrations of I TSH, T4 andT3 stored at -20°C for 3 years. References.

jy: De Costre Ph, Buhler U, De Groot L J, Refetoff I. Diurnal rhythm in total serum thyroxine levels. Metabolism 1971; 20: 782-91. Jacobsen B B, Petersen B, Andersen H. Changes in serum level of thyroxine and thyroxine-binding proteins (TBG, TBPA, albumin) induced by venous stasis. Acta Endo- crinoll977; 85: 39-43. i Judd, SJ, Carter JN, Corcoran JM, Lazarus L. Circulating thyroid hormone concentrations and posture and venous compression. Brit Med J 1975; 4: 735-6. 5-1 Kubasik NP, Sine H E. Results for serum and plasma compared in 15 selected radio- assays. Clin Chem 1978; 24:137-9. Lalla M, GrSsbeck R. Effect of tourniquet application, muscle work and vacuum tube blood collection on serum concentrations of Na, K, creatinine, total protein, ASAT, ALAT, cholesterol, PBI, T3U and triiodothyronine. Scand J Clin Lab Invest 1975; Suppl. 143: 101. Weeke J. Circadian variation of the serum thyrotropin level in normal subjects. Scand J Clin Lab Invest 1973; 31: 337-42.

h CHAPTER HI.

STUDIES IN EUTHYROID VOLUNTEERS

///. 1. Introduction.

Normal values for the assays of TSH, T4, T3, rT3 and T3 U were determined in euthyroid volunteers and the effect of physiological factors as age, sex and weight on these parameters was evaluated. Stimulation with thyrotropin- releasing-hormone (TRH) was performed in all volunteers, and the normal response was determined. The TRH-test was included in our studies because of its great value in the diagnosis of thyroid function disorders (a positive TSH response to TRH virtually excludes the diagnosis hyperthyroidism), and because of its sensitivity to very small changes in thyroid hormone concentration in the plasma(Snyder and Utiger 1972; Vagenakis et al 1974).

///. 2. Subjects and methods.

In all subjects an indwelling venous catheter was inserted in an antecubital vein, and blood sampling was delayed until IS minutes after the insertion of the catheter. T4 was determined initially by the displacement method of Murphy-Pattee, and subsequently by radioimmunoassay. The TRH preparation used was kindly provided by Dr. W J. van Rijn (Relefact TRH, Hoechst Holland); in all cases 200 y% TRH was rapidly injected intravenously. In six fasting volunteers (3 women, 3 men, mean age 32 years, range 2248 year) control tests were performed: these subjects were injected intravenously at two occassiCiis with 200 jug TRH resp. physiological saline; the interval between the two tests was at'least one week. Sixty-three subjects were tested in order to determine the range of normal basal values and the normal response after TRH stimulation. These subjects were considered to be euthyroid on clinical grounds. Subjects were excluded for one or more of the following reasons: history of thyroid disease, goitre on physical examination, grossly abnormal body weight, medication known to influence thyroid function tests, acute or chronic disease. No attention was given to the time of testing with respect to the menstrual cycle. Healthy people of 60 years and older all resided in an home for the elderly in Amster- dam. The data for sex, age, weight and height in these normal adults are presented in table III. 1. From the 63 subjects 25 were tested after an overnight fast, and 38 were tested after breakfast. Table III-l Age, weight and height (given as mean ± SO, with range between brackets) of the normal adult population, distributed to sex, age and season in which blood sampling was performed. number age (years) weight (kg) height (cm) 9 d total sex 9 32 0 32 42.7 ± 15.0(22-88) 61.9 ± 6.2(44.5-73.0) 166 ± 6(154.5-177.5) 6 0 31 31 49.6 ± 17.4(24-86) 75.01 7.5(56.5-89.0) 175 ±6(161 -189 ) all subjects 32 31 63 46.1 ± 16.4(22-88) 68.4 ± 9.5(44.5-89.0) 171*8(154.5-189 ) age 20 - 29yr 7 3 10 25.8 ± 2.4(22-29) • 66.0 ± 12.8(44.5-88.5) 30 - 39yr 7 7 14 34.7 ± 2.8(30-39) 66.6 ± 8.7(47.5-77.0) 40 - 49yr 9 9 18 45.1 ± 2.6(40-48) 69.5 ± 9.5(54.0-89.0) 50 - 59yr 6 5 11 53.4 ± 2.8(50-57) 69.3 ± 9.8(58.7-86.0) > 60yr 3 7 10 76.0 ± 10.1(60-88) 70.2 ± 7.0(62.0-82.0) season winter 5 13 18 44.7 ± 11.3(23-64) 72.5 ± 9.4(56.5-86.0) spring 3 6 9 42.0 ± 11.9(27-63) 68.3 ±. 6.4(60.0-77.0) summer 22 8 30 47.4 ± 21.4(22-88) 66.2 ± 9.9(44.5-89.0) fall 2 4 6 44.7 ± 2.6(41-48) 66.6 ± 9.1(54.0-78.0) 75 Nineteen cord blood samples from healthy newborns were also analysed. The basal value of the thyroid function tests in each individual was calculated as the mean of the two samples taken before the injection of TRH. In the statistical analyses undetectable hormone concentrations were taken as SO percent of the detection limit; the level of significance was taken as a = 0.05. Differences between subgroups of the normal population are evaluated by Willcoxon's two sample test. In the determination of normal limits preference is given to warning limits over descriptive limits, since the risk that unusual values remain undetected should be kept small in the diagnostic process. The inner limits for the percentiles P2.5 and P97.5 with a reliability y = 90% are therefore determined according to a distributionfree method (Riimke and Bezemer 1972), and these warning limits constitute the nonnal limits. The hormone response to TRH stimulation was calculated as the difference between the observed peak value after TRH and the mean of the two basal values. This approach is more practical than calculating the surface area.

227.3. Results.

a. Basil values 1. General. There was a non-parametric distribution of all parameters under test (figure III.l); after logarithmic transformation the distribution of the plasma iodothyronine concentrations appeared to be nonnal.

n 20 16 TSH rT,

0 03 1-0 2.0 00 4.0 60 0 0.1 0.2 03 0.4 0.5 1.0 1.4 18 22 2.6 3.0 20 nmol/l ,_. nmol/l

16 T,U FT4 index 12

0 60 80 WO 120 140 160 0.75 O85 095 1.05 1.15 125 60 80 100 120 140 160 nmol/l

Figure III, 1. Distribution of TSH, rT8, T,, T4, T,U and FT4 index values in 63 normal euthyr jid adults. Effect of sex on the values of thyroid function tests in £3 normal adults (32 females, 31 males).

geometric mean x±SD median range inner limits (x-SD, x + SD) P2.5 and P97.5

TSHmU/1 9 0.7 (<0.3 - 2.3 ) 1.2 ± 1.0 1.1 <0.3 -4.3 d 0.7 «0.3 - 2.3 ) 1.2 ± 1.2 1.0 <0.3 -4.5 total 0.7 «0.3 - 2.3 ) 1.2 ± 1.1 1.1 <0.3 -4.5 <0.3 -3.1 rT,nmoI/l 9 0.22( 0.16 - 0.30 ) 0.23 ±0.08 0.22 0.11 - 0.44 d 0.23( 0.17 -0.31 ) 0.24 ± 0.07 0.23 0.14 - 0.40 total 0.22( 0.16 - 0.31 ) 0.24 ± 0.07 0.22 0.11 - 0.44 0.14 -0.38 T,nmol/1 9 1.74( 1.41 - 2.16 ) 1.78 ±0.38 1.72 1.00 - 2.78 d 1.88( 1.58 - 2.25 ) 1.91 ±0.35 1.81 1.30 - 2.85 total 1.81( 1.48 - 2.21 ) 1.85 ±0.37 1.75 1.00 - 2.85 1.30 -2.45

T4 nmol/1 9 98( 82 - 118) 100± 18 99 62 - 141 d 103( 88 - 121) 104 ± 17 105 70- 151 total 101( 85 - 119) 102 ± 18 101 62 - 151 75 - 135

TSU 9 0.93( 0.86 - 1.02 ) 0.94 ±0.08 0.93 0.75 - 1.09 d 0.95( 0.87 - 1.04 ) 0.95 ±0.09 0.95 0.81 - 1.19 total 0.94( 0.86 - 1.02 ) 0.94 ±0.09 0.93 0.75 - 1.19 0.81 -1.07

FT4 index 9 92( 77 - 109) 93 ± 16 92 66 - 135 d 98( 85 - 113) 99 ± 14 96 70 - 128 total 95( 80 - 111) 96 ± 15 95 66 - 135 70 - 122 3 rT3/TjXl0- 9 126{ 86 - 184) 135 ± 56 119 60 - 314 d 122( 85 - 174) 129± 47 128 60 - 277 total 124( 86 - 179) 133 ± 51 121 60 - 314 69 -- 209 s rT,/T4xl0- 9 2.24( 1.60 - 3.13 ) 2.40 ±0.80 2.13 1.02-4.29 d 2.22( 1.63 - 3.04 ) 2.30 ±0.70 2.30 1.23 - 3.53 total 2.23( 1.62 - 3.08 ) 2.35 ±0.76 2.20 1.02 - 4.29 1.38 -3.53 3 Tj/T4xl0" 9 17.8 (14.5 -21.9 ) 18.2 ± 4.0 18.0 11.7 -32.2 d 18.3 (15.3 -21.9 ) 18.6 ± 3.2 18.9 11.6 -24.7 all 18.0 (14.9 - 21.9 ) 18.4 ± 3.6 18.2 11.6 -32.2 12.8 -23.4 Table III-3 Effect of age on the values of thyroid function tests in 63 normal adults (20-29 year, n=10; 30-39 year, n=14; 40-49 year, n=18; 50-59 year, n=ll; >60 year, n=10).

age (years) geometric mean x± SD median range fel" (x-SD.x + SD) TSH mU/1 20- 29 0.6 «0.3 - 1.8 ) 0.9 ±0.6 1.15 <0.3 - 1.7 30- 39 0.4 «0.3 - 1.3 ) 0.8 ± 1.0 0.4 <0.3 - 3.5 40- 49 1.3 «0.5 - 3.8 ) 1.9 ±1.3 2.0 <0.3 - 4.5 £•- 50- 59 0.7 «0.3 - 2.5 ) 1.2 ± 1.1 0.9 <0.3 - 2.8 > - 60 0.6 «0.3 - 1.8 ) 0.9 ±0.7 1.0 <0.3 - 1.8

tT3nmol/l 20- 29 0.21(0.14 - 0.32) 0.23 ± 0.10 0.22 0.11 - 0.44 30- 39 0.23(0.17 - 0.30) 0.23 ± 0.07 0.22 0.15 - 0.40 40- 49 0.22(0.16 - 0.30) 0.23 ± 0.07 0.23 0.14 - 0.38 50- 59 0.22(0.16 - 0.31) 0.23 ± 0.08 0.21 0.14 - 0.39 > 60 0.25(0.20 - 0.31) 0.25 ± 0.06 0.24 0.17 - 0.36

T3 nmol/1 20- 29 1.96(1.62 - 2.37) 1.99 ± 0.40 1.84 1.60 - 2.45 30- 39 1.78(1.50 - 2.12) 1.81 ± 0.31 1.76 1.30 - 2.35 40- 49 1.77(1.40 - 2.24) 1.82 ± 0.43 1.71 1.00 - 2.85 50- 59 1.98(1.75 - 2.24) 1.99 ± 0.24 2.08 1.68 - 2.34 > 60 1.61(1.31 - 1.98) 1.64 ± 0.32 1.67 1.05 - 2.13

T4nmol/1 20- 29 100( 83- 122) 102 ± 20 100.5 75- 141 30- 39 104( 91- 120) 105 ± 14 111 77- 121 40- 49 96( 79- 118) 98 ± 20 96 62- 151 50- 59 103( 90- 117) 103 ± 14 100 85- 135 > 60 101( 83- 123) 103 ± 19 103 70- 137

T3U 20- 29 0.92(0.84-1.01) 0.92 ± 0.08 0.925 0.81 - 1.09 30- 39 0.96(0.91 - 1.02) 0.97 ± 0.07 0.96 0.84 - 1.07 40- 49 0.96(0.86 - 1.07) 0.96 ±0.10 0.96 0.75 - 1.19 50- 59 0.93(0.85 - 1.02) 0.94 ± 0.08 0.92 0.81 - 1.08 > 60 0.90(0.85 - 0.95) 0.90 ±0.06 0.90 0.80 - 1.00

FT4 index 20-29 93( 75- 114) 94 ± 20 91 69- 135 30- 39 101( 88- 116) 101 ± 13 104.5 77- 122 40- 49 93( 77- 111) 94 ± 17 92.5 66- 127 50- 59 96( 85- 108) 96 ± 13 92 83- 128 > 60 91( 78- 105) 91 ± 13 93 70- 109 3 rT3/T3xl0- 20- 29 109( 79- 152) 115 ± 38 107 69- 175 30- 39 126( 88- 180) 134 ± 46 125 64- 209 40- 49 124( 83- 184) 133± 51 127 60- 250 50- 59 111( 81 - 152) 116± 40 106 76- 189 > 60 154( 105 - 225 ) 130 ± 70 140 99- 314 3 rT3/T4xl0" 20- 29 2.14(1.45 - 3.14) 2.27 ± 0.79 2.20 1.02 - 3.53 30- 39 2.16(1.58 - 2.97) 2.27 ± 0.78 2.27 1.30 - 4.29 40- 49 2.28(1.61 - 3.24) 2.41 ± 0.82 2.30 1.23 - 4.07 50- 59 2.14(1.61 - 2.83) 2.22 ± 0.69 1.92 1.59 - 3.48 > 60 2.46(1.86 - 3.26) 2.55 ± 0.74 2.44 1.72 - 4.02 3 T3/T4xl0" 20- 29 19.5(15.9 - 24.0) 19.9 ± 4.7 19.6 14.8 - 32.2 30- 39 17.1(14.1 - 20.1) 17.4 ± 3.4 16.9 14.1 - 24.7 40- 49 18.4(15.8 - 21.5) 18.6 ± 2.9 18.3 13.4 - 25.4 " 50 -59 19.3(16.8 - 22.1) 19.4 ±2.6 18.4 15.5 - 23.1 > 60 16.0(12.6 - 20.3) 16.4 ± 3.9 16.9 11.6-23.4 78 £

The ratio's ofrT3/T3,rT3/T4 and T3/T4 were also log-normally distri- -,|£, buted. No differences were found in the test results between the 25 |£ fasting and the 38 non-fasting subjects. fe, 2. Effect of sex. No differences were found between the values for women \ ft and for men (table IE. 2). However, a tendency to somewhat higher ; , || values of plasma iodothyronines in men was noted. ' §v 3. Effect of age. No differences were found between the values in differ- ; || ent age groups (table III. 3). However, a tendency to lower plasma T3 f| and higher plasma rT3 concentrations was observed in subjects of 60 j §3 years and older. |y 4. Effect of weight. No correlation was found between the weight and the s |; plasma concentrations of rT3, T3 and T4 respectively. | 5. Effect of season. There is an uneven distribution of blood sampling over ! the four seasons (table III. 1), since the possibility of seasonal variation ; was not appreciated at the outset of the study. Therefore seasonal varia- | tion in hormone levels was not further investigated, although plasma j | T3 and T4 concentrations in summertime were lower than in the i winter. 6. Interrelationships of hormone concentrations. Basal TSH concentration was not related to either rT3, T3, T4 or FT4 index. There was no •• effect of sex and age on the ratio's of plasma rT3/*T3, rT3/T4 and T3/ T4, although a tendency to higher rT3/T3 and rT3/T4 and to lower | rl T3/T4 ratio's in old age was observed. r,

b. TRH-test. 1. Side effects. The observed side effects of a rapid intravenous injection of 200 jig TRH are listed in table III. 4. The side effects are experienced within seconds to a few minutes after the TRH injection, and have dis- rate///-* Nature and frequency of side effects of 200 fig TRH i.v. as a bolus injection in 63 normal adults.

no side effects 13 urinary urgency 38 nausea, epigastric discomfort 25 strange, bitter or metallic taste 14 dizziness 12 feeling of flushing 12 headache 4 oppressive feeling in neck or chest 3 heavy feeling in legs 2 palpitations 1 agitation 1 Table III-S

The response of TSH, rT,, T,, and T4 to 200 jig TRH i.v. and to 3 ml. physiological saline Lv. in 6 normal adults.

minutes: -15 0 20 60 120 180 240

TSH mU/1 :TRH 2.2 ± 1.3 2.2 ± 1.5 10.7 ± 3.3 8.4 ± 3.4 5.0 ±2.3 3.9 ± 2.0 2.7 ± 1.5 saline i.7 ± 1.4 1.7 ± 1.1 1.7 ± 1.2 1.5 ± 1.0 1.4 ±1.1 1.4 ± 1.3 1.4 ± 1.1 rT,nmol/l :TRH 0.26 ± 0.06 0.23 ± 0.07 0.22 ± 0.07 0.25 ± 0.06 0.31 ±0.08 0.31 ± 0.07 0.31 ± 0.07 saline 0.23 ± 0.04 0.25 1 0.05 0.25 ± 0.04 0.25 ±0.04 0.27 ±a.O4 0.28 ± 0.05 0.33 ± 0.07

T3nmol/1 :TRH 1.91 ± 0.41 1.89 ± 0.29 1.84i 0.27 2.34 ± 0.76 2.57 ±0.41 2.52 ± 0.51 2.56 ± 0.43 saline 1.681 0.42 1.66 ± 0.44 1.66 ± 0.38 1.65± 0.44 1.73 ±-0.42 1.76 ± 0.52 1.75 ± 0.47

T4 nmol/1 :TRH 104 ± 22 109 ± 24 126 ± 29 saline 1001 19 99 i 16 104 ± 22

,ti5 M^f-K*-'-1' 80

Table III-6 Effect of sex and age on the TSH response to TRH ( 200 Mg i-v.) in 63 normal adults (values in mU/1). M geometric mean me- range inner limits yf-j-• - (x-SD,x + SD) dian P2.S and P97.5

basal TSH : 9 0.7 (<0.3- 2.3) 1.1 <0.3- 4.3 <5 0.7 (<0.3- 2.3) 1.0 <0.3- 4.5 all 0.7 (<0.3- 2.3) 1.1 <0.3- 4.5 <0.3- 3.1 if;. peak TSH : 9 .9.7 ( 5.1 - 18.6) 11.4 0.8- 23.0 5.0- 22.0 6 6.4 ( 3.0 - 13.7) 7.0 0.8- 24.3 2.8- 15.5 all 7.9 ( 3.8- 16.5) 8.5 0.8- 24.3 2.8- 22.0

ATSH : 9 8.7 ( 4.5 - 16.8) 9.8 0.65- 20.4 4.7- 19.4 a 5.5 ( 2.6- 11.6) 5.5 0.65- 19.8 2.3- 13.0 all 6.9 ( 3.3 - 14.5) 7.85 0.65- 20.4 2.3- 19.4

peak TSH : 20 -29 yr. 11.0( 7.3- 16.7) 11.0 6.3 -22.0 30 -39 6.8( 4.0- 11.6) 6.7 3.0- 14.0 : 40 -49 10.9( 6.4- 18.4) 13.05 4.« -24.3 50 -59 8.9( 4.7- 16.8) 9.9 2.8 -23.0 > 60 3.5( 1.3- 9.5) 6.05 0.8* - 8.5 ATSH : : 20 -29 yr. 10.2( 6.6- 15.6) 10.6 5.3 -20.4 30 -39 6.2( 3.8- 10.2) 6.5 2.7- 13.9 40 -49 9.3( 5.6- 15.3) 10.0 4.3 -19.8 ! 50 -59 7.7( 4.0- 14.9) 9.3 2.5 -20.3 \ > 60 2.8( 1.0- 7.7) 4.8 0.65 - 7.9

appeared nearly always after ten minutes. No serious side effects were encountered.

2. Control tests. There is no change in plasma levels of TSH, T3 and T4 after administration of physiological saline, in contrast to the increase in these hormone levels after injection of TRH (table III.5). An increase of plasma rT3 was observed both after saline and after TRH injection. Plasma rT3 also increased in non-fasting subjects after TRH. 3. TSH-response. After injection of TRH plasma TSH increased in all subjects. The peak TSH was observed at 20 minutes in 58 subjects, and at 60 minutes in S subjects. Peak TSH values at 60 minutes were never more than 0.3 mU/1 above the 20 min. value. The TSH response to TRH in women is greater than in men (p < 0.05; table III. 6), and therefore normal response ranges are given separately for both sexes (figure HI. 2). The TSH response is not related to basal T3 or T4 values. Basal and peak TSH values are correlated (figure HI. 3); the correlation- regression equations are: I 81 1 20 16

j 12

- 8

-15 0 20 60 120 180 240 -15 0 20 60 120 180 240 min min Figure III. 2. The TSH response to 200 fig TRH i.v. in 32 female and in 31 male normal adults. Median values arc presented (solid lines), and the inner limits for the percentiles P2.5 and P97.5 (interrupted lines).

32 females: y = 3.05 x + 7.65 (r =0.57;p < 0.001) 31 males: y = 3.75 x +3:44 (r =0.82; p< 0.001) 63 adults: y = 3.42 x +5.55 (r =0.66; p< 0.001) The TSH response to TRH is lower in subjects of 60 years and older (p < 0.02; table III. 6). Peak TSH values of < 1.0 mu/1 were observed in one female and two males of old age.

4. 7^-response. Plasma T3 concentrations increased in all subjects after TRH injection. The peak value is reached at 20 minutes in 1 subject, at 60'in 1 subject, at 120'in 26 subjects, at 180'in 27 subjects and at 240' in 8 subjects. No effect of sex or age on the T3 response to TRH was found (table III. 7). Basal T3 values are correlated both with peak T3 values and with the relative increments of T3 (figure III. 4), but not with the absolute increments of T3. The correlation-regression equations are: basalT3 -peakT3 : y= 0.82x+ 1.22 (r =0.77;p<0.001) basal T3 - AT3 absolute : y = - 0.18 x + 1.22 (r =0.22; N.S.) basal T3 - AT3 relative : y = - 42 x + 128.52 (r = 0.61; p < 0.001) The median T3 response to TRH and the inner limits for the percentiles P2.5 and P97.5 are depicted in figure III. 5. The inner limit for the percentile P2.5 is 0.54 nmol/1 for the absolute increase in T3, and 28% for the relative increase in T3. There is an almost complete overlap between these two criteria, i.e. the subjects with an absolute increase of < 0.54 82 25 I I 20 ti

1R IS 1

3 MD 0.3 0.5 1.0 1.5 2X) 25 3.0 35 4O 45 BASAL TSH mU/l 1

Figure III. 3. The correlation between basal TSH and peak TSH after TRH (200 *tg Lv.) + in 63 normal adults (ypeak TSH = 3.42xbasalTSH 5.55;r= 0.66; p<0.001).

nmol/1 T3 all except one had a relative increase of < 28%. The lower limit for the normal response of T3 to TRH was found to be an absolute increment of > 0.55 nmol/1, and a relative increment of>30%.The lowest responses observed are an absolute increase of 0.43 nmol/1 T3, and a relative increase of 21%. The three subjects with peakTSH values of < 1.0 mU/1 all had an absolute T3 increase of > 0.55 nmol/1 and a relative T3 increase of > 30%. 5. T4-response. No effect of sex or age on the T4 response to TRH was found (table UI.8), although a tendency to a lower response in old age was observed. Basal T4 values are correlated to the peakT4 values and to the absolute increments, but not to the relative increments of T4. The correlation-regression equations are:' - basal T4 - peak T4 : y = 1.18 x-620 (r = 0.91; p<0.001) basal T4 - AT4absolute : y = 0.18 x - 0.20 (r = 0.32; p < 0.02 ) basal T4 - AT4relative : y = 0.03 x +14.73 (r = 0.06; N.S.). Table III-7 Effect of sex and age on the T, response to TRH (200 ng Lv.) in 63 normal adults (values as mean ± SD, range between brackets).

; basal T, peakT, AT, AT, hi (nmol/1) (nmol/1) (nmoj/l) (%) \ i , — . 5 * 9 1.78 ±0.37 2.60 ±0.34 0.82*0.27(0.43-1.60) 50(21-160) co !' 6 1.91 ±0.35 2.87 it 0.44 0.96 * 0.30(0.54 - 2.11) 52(28-140) w J all 1.85 ± 0.37 2.73 ± 0.41 0.89 ± 0.27(0.43 - 2.11) 51(21 - 160)

20-29yr 1.99 ±0.40 2.88 ±0.49 0.89 ± 0.29(0.43 -1.31) 46(21-69) ' 30-39 1.81 ±0.31 2.64 ±0.35 0.83 ± 0.30(0.48 -1.45) 48(21-112) 40 - 49 1.82 ± 0.43 2.81 ± 0.44 0.99 ± 0.38(0.63 - 2.11) 60(28 - 160) 50 -59 1.99 ± 0.24 2.81 ± 0.32 0.81 ± 0.14(0.61 -1.10) 41(33 - 55 ) \ f . > 60 1.64 ± 0.32 2.50 ± 0.40 0.86 ± 0.23(0.54 - 1.28) 55(33 - 105) 84

•wo

120

100 i

• $

40 ••':• i

1.0 1.2 14 1.6 1.8 20 22 24 2.6 23 BASAL T3 nmol/l

Figure III. 4. The correlation between basal T, and the relative increment of T3 after 42x + TRH (200 Mg i.v.) in 63 normal adults (y Ax % " basal T 128.52; r=0.61; p< 0.001).

The inner limit for the percentile P2.5 of the T4 response is 3 nmol/1 for the absolute increase in T4, and 4% for the relative increase in T4. The I increment was less than S nmol/1 in five subjects, and less than 10 nmol/1 in three other subjects. I

c. Cord blood. The results of the thyroid function tests in 19 cord blood samples are presented in table III. 9. The values of TSH, rT3 and T4 are higher, but T3 values are lower than in the normal adults, and consequently the ratio's of rT3/T3 and rT3/T4 are increased and of T3/T4 decreased relative to the values in nonnal adults. 85

220

2.9 190

-2S _160

130

1.7 100

70 -15 0 20 60 120 180 240 -15 0 20 60 120 180 240 min min

Figure III. S. The Ts and T4 response to 200 ng TRH i.v. in 63 normal adults. Median values are presented (solid lines), and the inner limits for the percentiles P2.S and P97.S (interrupted lines).

Table 7/7-8

Effect of sex and age on the T4 response to TRH (200 MS i.v.) in 63 normal adults (values as mean ± SD, range between brackets).

basal T4 peakT4 AT4 (nmol/1) (nmol/1) 4 (nmol/1) 9 100 ±18 118 ±24 18(-8 -42 ) 18(-7 -41) 6 104 ± 17 122 ± 21 .. 18( 2.5 - 32.5) 17( 2 - 29) all 102 ±17 120 ±22 18(-8 -42 ) 18(-7 -41)

20 - 29 yr 102 ± 20 125 ± 25 23( 10.5 -42 ) 23( 13 -39) 30-39 105 ± 14 128 ± 21 22(-8 - 37.5) 21(-7 -38) 40-49 98 ±20 115 ± 24 17( 3.5 -42 ) 17( 6 -41} 50-59 103 ± 14 118 ±17 14( 1 -28 ) 14( 1 -31) > 60 103 ± 19 118 ± 24 15( 2.5 -28 ) 14( 2 -24) Table III-9 Thyroid function tests in 19 cord blood samples from healthy newborns.

geometric mean x± SD median range (x-SD, x + SD)

TSH mU/1 9.2 ( 5.1 - 16.8 ) 10.6 ± 5.2 10.8 2.3 - 23.0 rT3 nmol/1 2.98 ( 2.24- 3.98) 3.11 ± 0.92 2.80 1.95- 5.40 T3mno!/1 1.03 ( 0.73- 1.46) 1.09 ± 0.38 1.10 0.50- 2.05 T4 nmol/1 137 ( 113 - 166 ) 139 ± 24 140 85 - 180 TaU 0.86 ( 0.80- 0.92) 0.86 ± 0.06 0.86 0.79- 1.09 FT4 index 118 ( 100 - 138 ) 119 ± 18 125 77 - 153 s rT,/T3xl0- 2887 (2063 - 4041 ) 3047 ± 1045 2792 1693 - 5200 3 rT,/T4xl0- 21.82 ( 17.82 - 26.72) 22.25 ± 4.56 22.76 15.60 - 33.75 5 T,/T4xl0- 7.54 ( 5.63- 10.11) 7.85 ± 2.29 7.59 4.00- 14.14 I 87 rv III.4. Discussion. *;.: 4a. Thyroid function tests: basal values. f:; The plasma concentration of iodothyronines in the normal adults are not ': normally distributed, but have a log-normal distribution. Consequently, the 1 geometric mean was calculated for each test including plasma TSH (the TSH [ concentrations are also non-parametrically distributed). The arithmetic mean | is also given, but only to facilitate comparison with the figures in the litera- I; ture which are nearly always reported as the arithmetic mean (table II. 5, II. 1 14, II. 16 and II. 18). Nevertheless, no great difference is seen between the geometric andarithmetic means of the plasma concentrations of iodothyronines (in contrast to the calculated ratio's), suggesting a near normal distribution. Indeed a normal distribution of plasma iodothyronines has been demonstrated by Evered et al (1978) in a very large sample (2779 subjects were investigated) of the population. It should be realized that the normal limits are warning limits, i.e. at least 2.5% of the values in a normal population is outside each of the two percentiles. Furthermore, the normal values were obtained from blood samples taken without venous stasis; in practice, a considerable increase in the iodothyronine plasma concentration sometimes occurs because of a difficult venapuncture (chapter U.S.). No effect of sex on the basal values of the thyroid function tests is observed, and this is in agreement with the literature (Ormstcn et al 1971; Mitsuma et al 1971; Snyder and Utiger 1972; Sanchez-Franco et al 1973; Meinhold and Wenzel 1974; Birk Lauridsen et al 1974; Blichert-Toft et al 1975; Stepanas et al 1977); only Wenzel et al (1974) report higher TSH values in young females than in males. The absence of a sex difference in man is in contrast to the findings of some studies in animals. Female rats have lower plasma T3 and higher plasma T4 values than males. In this respect it is !. of considerable interest that the peripheral conversion of T4 into T3, as studied very recently by Harris et al (1979) in fresh liver homogenates, is \. much lower in adult female rats than in males. This difference is caused by a ;• decreased 5 -deiodinating enzyme concentration in the females rather than a j decreased concentration: of co-factoi(s), since addition of dithiothreitol had {• no effect on the sex difference. The sex difference in the peripheral conver- | sion of T4 is most probably related to sex steroids: the difference is abolished |; by prepubertal castration and also after peripubertal castration. Testosteron | administration to castrated male and female rats in vivo increases the conver- I sion of T4 into T3, whereas 0-oestradiol decreases the T4->T3 conversion in I the castrated male (although no effect in vitro of estrogens is observed in the • castrated female rat). At the present time there is no information on a possi- L We sex-related difference in the peripheral conversion of T4 in man. | No effect of age on the basal values of the thyroid function tests could be | demonstrated in the adult normal population, although a tendency to lower 88

plasma T3 values in old age was observed. Low T3 and high rT3 plasma concentrations in old people have frequently been reported (Snyder and Utiger 1972; Rubenstein et al 1973; Herrmann et al 1974; Burger et al 1975; Mol- holm Hansen et al 1975; Hesch et al 1976; Westgren et al 1976; Nicod et al 1976). The explanation has usually been that there is a change in the peripheral conversion of T4 with increasing age (Wenzel and Horn 1976). How- • ever, in an elegant study Olsen et al (1978) demonstrated that low T3 and ; high rT3 plasma levels in old age are an effect of disease and not of age. They found no difference in T4, T3 and rT3 values between healthy young volun- : teers (aged 18-29 years) and carefully selected healthy elderly people (aged 70-90 years) living at home, but T3 levels were lower and rT3 levels were slightly, but not significantly, higher in elderly people living at a nursing home. The inference is clear: elderly subjects living at home seem to be healthier than those institutionalized in an home for the elderly. Recently, Smeulers et al (1979) reported a decrease in T3 with advancing age, while levels of T4, rT3 and TSH remained unchanged; they found no difference in serum T3 between elderly people living at home or at an elderly people centre. The problem whether impaired T4-*T3 conversion in old age is to be • regarded as a physiological or a pathological phenomenon, is akin to the problem of classification of an elevated blood pressure or an impaired glucose '. tolerance in elderly subjects. It may be concluded that the healthier and j the more active an elderly subject is, the more likely he is to have plasma ; iodothyronine values in the same range as young adults. The debate about the physiological or pathological nature of an impaired conversion of T4 into T3 in old age, is largely semantical and thereby philosophical of character.

In our study body weight and plasma concentrations of iodothyronines are not correlated. However, a positive correlation between body weight and ; plasma T3 (but not with plasma T4) is described by Bray et al (1976). How- ever, in this study grossly obese subjects were included, in contrast to our •! control subjects.

The effect of circadian and circannual rhythms on the plasma levels of ' thyrotropin and thyroid hormones cannot be evaluated in our study, since I blood sampling for basal values was always performed between 08.30 and ! ; 10.30 a.m. and seasonal variation was not studied prospectively. A circadian rhythm of TSH secretion in man is now firmly established, with the highest | levels during the night and a gradual decline to a nadir around 11.00 a.m. : (Nicoloff et al 1970; Patel et al 1972; Weeke 1973; Azukizawa et al 1976; | - Lucke et al 1977; Chan et al 1978). Plasma-TSH begins to rise in the evening j i before the onset of sleep, and in fact evidence for an inhibitory influence of ' | sleep has been found (Parker et al 1976). The neuroendocrine mechanisms < I underlying the circadian TSH rhythm are unknown; they are not related to a '•: 89 reduced dopaminergic inhibition of the TSH secretion at night (Scanlon et al 1979). The synchronous variation in TSH and free thyroid hormones excludes feedback control by thyroid hormones as a cause of the circadian TSH variation as concluded by Weeke and Gundersen (1978). These authors describe two rhythms for TSH as well as for FT3 and FT4 in man: synchronous diurnal rhythms with lower levels in the day-time and higher levels at night superimposed by short-term variations with a cyclelength of about 30 minutes; these data suggest a pulsatile release of hormones from the thyroid gland governed by a pulsatile TSH secretion, although other explana- tions (e.g. rapid changes of the distribution space of intracellular thyroid hormones) are equally possible. The diurnal variation in the plasma concentrations of total thyroxine and total triiodothyronine is non-synchronous with the TSH rhythm (De Costre et al 1971; O'Connor et al 1974; Balsam et al 1975; Azukizawa et al 1976; Lucke et al 1977; Chan et al 1978; Weeke and Gundersen 1978). This pattern is best explained by changes in hormonal binding by plasma proteins caused by haemodilution and haemoconcentra- tion due to postural changes (Judd et al 1975) (cf. chapter II. 5). Although of considerable physiological interest, the fluctuations in plasma hormone concentrations due to diurnal variation are minor, and consequently these changes are of little relevance in routine thyroid testing. In contrast, the variation in plasma concentrations of T3 and T4 due to circannual rhythms is much greater. Smals et al (1977) describe a seasonal variation in circulating T3 and T4 levels in 13 healthy male adults in the Netherlands. Lowest T4 and T3 levels were found in the summer, the mean values being appr. 13 nmol/1 resp. 0.23 nmol/1 lower than in winter (in contrast to a study of Postmes et al (1974) who could not demonstrate a seasonal variation). The findings of Smals et al however are in agreement with other reported studies in man. Ex- posure to the Antarctic cold results in a rise of serum T3 and T4 with a return to normal within 2 days of return to a normal environment (Eastman et al 1974). Although serum T4, T3 and TSH did not change throughout the year in a subtropical region, the mean urinary T3 and T4 excretion (reflecting the free plasma concentrations of these hormones) were higher during the coldest months than during the hottest months (Rastogi and Sawhney 1976). No seasonal variation was found in university employees who lived in air-con- ditioned rooms, but plasma T3 concentration increased significantly in winter (T4 and TSH did not change) in adult men living in a cold environment (Nagata et al 1976). The changes in plasma thyroid hormone concentrations in response to cold exposure are probably TSH mediated (Tumisto et al 1976), although there is no unanimous opinion on this point. Nevertheless, cold exposure in the rat stimulates the release of TSH, which can be suppressed by pretreatment with antibodies to TRH (Mori et a!1978) indicating a physiological role for TRH in the TSH response to cold. A shift in the conversion of 90

T4 towards a preferential generation of T3 might be another mechanism in the adaptation to cold, but so far this has not been studied.

4b. The TRH test. it' The side effects of an intravenous injection of TRH occur within seconds to minutes, and prior to the rise in TSH; the side effects are therefore not TSH mediated. It has been demonstrated that TRH is also present in the mammalian brain outside the hypothalamus (Wilber et al 1976), and its presence in the digestive tract of the rat has also been reported (Morley et al 1977). Recent studies indicate that TRH has a neurotransmittor function in the central nervous system outside the hypothalmus (Guillemin 1977). In this respect it is of interest that the urge to micturate in response to TRH was also observed in a patient with complete spinal cord transection at the level of the eight thoracic vertebra (Rosenthal 1974), and that a decrease in gastric motility after TRH has been seen endoscopically in human subjects (Dolva et al 1978). Thus an action of TRH as a neurotransmittor either cen- trally or locally in the smooth muscle of the gastrointestinal and genitourinary tracts may explain most of the side effects (cf. Ormston et al 1971). Exten- sive clinical experience over the last eight years has proven that the TRH test is a safe and reliable procedure.

In our study, we rather arbitrarily selected a dosage of 200 /ig TRH. As judged from dose-response curves 200 jig TRH i.v. elicits a submaximal response of TSH, and maximal responses are obtained at dosages of 400-500 Mg TRH i.v. (Haigler et al 1971; Snyder and Utiger 1972;Otsuki et al 1973). Other dosages and/or routes of administration (intramuscularly, orally) have been recommended in the literature with or without measurement of theT3 and T4 responses (Haigler et al 1972; Noel et al 1974; Azizi et al 1975; Van Kersen et al 1975; Staub et al 1976; Bremner et al 1977; Vogt et al 1978). We have sofar not been convinced of the advantages claimed in these studies, especially since in our hands the T3 response to TRH proved to be a very reliable criterion. Therefore we continue to use 200 /utg TRH intravenously, in accordance with the majority of the studies reported in the literature.

As judged from the control tests, a significant increase in plasma values of TSH, T3 and T4 is observed after TRH but not after physiological saline. No difference between TRH and saline is found in the plasma levels of rT3. The slight increase in plasma rT3 during both tests is not due to fasting, since the same finding is reported by Bremner et al (1977) in non-fasting normal subjects. Reverse T3 is present in the human thyroid gland (Chopra 1974; Burman et al 1977) and is secreted into blood especially when the gland is stimulated with TSH (Chopra et al 1975; Westgren et al 1976; Hooper et al 1978), but clearly intensive TSH stimulation is needed to obtain a measurable 91 response. Burman et al (1974) report a mean increase of 0.04 nmol/1 after 500 ng TRH i.v. in 10 normal human subjects and claim that this increase is r significant, but no control tests were performed.

We did not study the rT3 response any further, no explanation can be offered for the rise of rT3. Also, the T4 response to TRH was not studied further, since 8 of the 63 normal adults showed an absolute increment of less than 10 nmol/1 implying that the T4 response discriminates poorly between the euthyroid and the hyper- or hypothyroid state. A similar minor T4 response to 100-500 /ig TRH i.v. peaking around 4-6 hrs after TRH has been common experience (Hollander et al 1972; Lawton et al 1973; Uller et al 1973). In case no TSH or T3 assay is available and one has to rely on T4 determinations only, higher doses of TRH are needed (Lawton et al 1973; Azizi et al 1975; Van Kersen et al 1975; Vogt et al 1978).

Although peak TSH was < 1.0 mU/1 in three elderly subjects, a measurable TSH response to TRHv/as observed in all 63 normal subjects. This is in agreement with Andersen et al (1971), and with Birk Lauridsen et al (1974) who found a rise in TSH of < 2.0 mU/1 after TRH in 2 out of 60 controls. It is likely that no more infonnation is obtained by extending the measurement of TSH beyond 20 minutes. Also, the use of A TSH instead of the peak TSH fl value has no clear advantage (Sawin and Hershman 1976). Sometimes the TSH value at 60 minutes is somewhat higher than the 20 min. value, and this pattern seems to be a normal variant (Birk Lauridsen et al 1974; Sawin and Hershman 1976). We found the peak TSH to be correlated with basal TSH, indicating a proportional magnification of basal TSH by TRH; this is also described by Weeke (1974), Birk Lauridsen et al (1975) and Sawin and Hersh- man (1976). We, nor Wenzel et al (1974) could relate the peak TSH values to basal T3 or T4, but Sawin and Hershman (1976) report a moderate negative correlation of peak TSH with basal T4 (the coefficient of correlation between peak TSH and basal T3 or T4 in our study is indeed higher for T4 than T3, but failed to reach statistical significance in both). This would indicate that plasma T4 is one of the determinants for the secretion of TSH which is in agreement with the findings in hypothyroid patients (see chapter IV) and the interesting studies of Larsen and co-workers (Silva and Larsen 1977; see further chapter V. 4). Our results indicate that the TSH response is greater in females than in males. This has been reported by many others (Bowers et al 1971; Ormston et al 1971; Haigler et al 1971; Sanchez-Franco et al 1972; Noel et al 1974; Wen- zel et al 1974; McNeilly et al 1974; Blichert-Toft et al 1975; Sawin et al 1978); in the studies who fail to show this sex difference, the tendency to a greater response in women is nearly always observed (Snyder and Utiger 5 1972; Weeke 1974; Birk Lauridsen et al 1974; Vogt et al 1978). The sex '•3. difference cannot be caused by testing at different times of the menstrual 92 cycle, since the TSH response to TRH is the same at the follicular phase and at the luteal phase (Weeke 1972; McNeilly et al 1974; Sawin et al 1978), despite the initial findings of Sanchez-Franco et al (1972) describing a greater TSH response before ovulation than afterwards. Nor can the greater increase in TSH in women be explained by a relatively higher dose of TRH administered to women because of their lower body weight, since in subjects matched for height and weight women still demonstrated a tendency to higher TSH responses than men: peak values 16.8 vs. 14.3 mU/1 (Snyder and Utiger 1972). The common explanation has been that the greater TSH response in females is due to higher levels of plasma estrogens. The absence of a difference between the TSH responses in elderly women and men favours this explanation (Wenzel et al 1974) although this finding has been disputed (Snyder and Utiger 1972). The enhanced TSH response seen in males after the administration of estrogens also favours a causative role of estrogens (Faglia et al 1973; Mortimer et al 1974; Smyth et al 1977) although again this has not been a universal finding (Woolf et al 1973; Carlson et al 1973). Sawin et al (1978) conclude that the higher average estradiol levels in normally menstruating women tend to enhance the TSH response to TRH, but that cyclical changes in estradiol and progesterone have no effect. The mechanism by which estrogens enhance the TSH response to TRH is unclear; in this respect it is of interest that preincubation of rat anterior pituitary cells with estradiol led to increased TSH responsiveness to TRH (Labrie et al 1978), that estrogens decrease the metabolic clearance rate of TSH in the laboratory animal (Sawhney et al 1978), that the thyroid gland in normal females is larger than in normal males (Boyd 1961), and that the peripheral conversion of T4-*T3 in female rats is lower than in males (Harris et al 1979). Similarly, conflicting results are reported on the TSH response to TRH as a function of age. Snyder and Utiger (1972) and Birk Lauridsen et al (1974) find no effect of age in females, but a decrease in elderly males. In contrast, Wenzel et al (1974) report a lower response in elderly females but not in males, and Sakoda et al (1976) finds a decrease when applying 400 MgTRH i.v. but not at doses of 50-100 jug. Again, the general finding is that the lowest responses are found in elderly subjects (Otsuki et al 1973; Cuttelod et al 1974; Blichert-Toft et al 1975) like in our study. Some authors conclude that it is necessary to determine normal limits of the TSH response to TRH for males and females seperately, and also for each decade (Snyder and Utiger 1972; Wenzel et al 1974). We are of the opinion that, given the barely significant or insignificant differences between males and females and between young and old people, it is justified to apply one normal range for the whole population. Common sense then dictates that a low TSH response in a young female is more likely to be abnormal than in an octagenarian. 93

The T3 response to TRH in our study is not influenced by sex or age, in accordance with studies of Wenzel et al(1974) and Vogt et al (1978). In fact, the subjects with peak TSH values of < 1 mU/1 all had a clear increase in T3 levels; thus measuring the T3 response to TRH can be advantageous. In this respect we disagree with other authors who find the T3 increase too small or too inconsistent to be of diagnostic value (Wenzel et al 1974; Vogt et al 1978). The reason for this is no doubt that, whereas others have only determined T3 at 120 or 180 minutes, we measure the T3 response for 4 hours, and it has been clearly documented that plasma T3 peaks after 2-4 hours (Shenkman et al 1972; Lawton et al 1973; Uller et al 1973). The fact that the maximal increase in T4 is seen 4-6 hours after 200 jug TRH i.v. (Law- ton et al 1973; Uller et al 1973) again illustrates the preferential secretion of T3 over T4 in response to TSH stimulation.

4c. Cord blood. The results of the determination of hormone levels in the cord blood samples are most remarkable (very high rT3 and low T3 levels, in combination with elevated T4 and TSH levels), and are in agreement with previous I- reports (Larsen 1972; Chopra et al 1973, 1975; Erenberg et al 1974). The findings strongly suggest that impairment of the peripheral conversion of T4 is involved; indeed Chopra et al (1975) conclude from kinetic studies in fetal and adult sheep that peripheral production of T3 from T4 is reduced in the fetus, while the rate of conversion of T4 to rT3 is comparable to that in the adult; the very high rT3 concentrations are the result of an increased production of the substrate T4, and of a lower metabolic clearance rate as compared to adult sheeps. The relevance of these findings are discussed in chapter DC. The high TSH values are explained by thesuddenTSH surge occurring at birth as a result of cooling; the plasma TSH concentration in the neonate reaches a maximum level within 30 min of delivery, and decreases thereafter.

III. 5. References.

Anderson MS, Bowers CY, Kastin AJ et al. Synthetic thyrotropin-releasing hormone. New Engl J Med 1971; 285:1279-83. Azizi F, Vagenakis AG, Portnay GI, Rapopcit B, Ingbar SH, Braverman LE. Pituitary- thyroid responsiveness to intramuscular thyrotropin-releasing hormone based on analyses of serum thyroxine, tri-iodothyronine and thyrotropin concentrations. New Engl J Med 1975; 292: 273-7. Azukizawa M, Pekary AE, Hershman JM, Parker DC. Plasma thyrotropin, thyroxine and triiodothyronine relationships in man. J. din Endocrinol Metab 1976; 43: 53342. Balsam A, Dobbs CR, Seppo LE. Circadian variations in concentrations of plasma thyroxine and triiodothyronine in man. J Appl Physiol 1975; 39: 297-9. Birk Lauridsen U, Deckert T, Friis Th, Kirkegaard C, M0lholm Hansen J, Siersbaek-Niel- sen K. Estimation of serum thyrotropin (TSH) and stimulation with thyrotropin- releasing hormone (TRH) in thyroid diseases. Acta Med Scand 1974; 196: 171-6. Blichert-Toft M, Hummer L., Dige-Petersen H. Human serum thyrotrophin level and 94 response to thyiotrophin-relaeasing hormone in the aged. Geront Clin 1975; 17: 191-203. Bowers CY, Friesen HG, Hwang P, Guyda HJ, Folkers K. Prolactin and thyrotropin release in man by synthetic pyroglutamyl-histidyl-prolinamide. J Clin Endocrinol Metab 1971; 45:1033-41. Boyd W, The Thyroid. In: Boyd W. Textbook of Pathology. 7th edition. Philadelphia, Lea and Febiger, 1961: 1000-25. Bray GA, Fisher DA, Chopra IJ. Relation of thyroid hormones to body weight. Lan- . cet 1976; 1: 1206-8. Bremner WJ, Abreu R, Stockigt JR, de Kretser DM, Burger HG. Pituitary-thyroid responses to 4-hour constant infusions of thyrotropin-releasing hormone in man. J. Clin Endocrinol Metab 1977; 45: 981-7. Burger A, Sakoloff C, Staeheli V, Valloton MB, Ingbar SH. Radioimmunoassays of 3,5,3 '-triiodothyronine with and without a prior extraction step. Acta Endocrinol 1975; 80: 58-69. Burman KD, Diamond RC, Wright FD, Earll JM, Burton J, Wartofsky L. A radioimmunoassay for 3,3 ',5 -L-triiodothyronine (reverse T3): assessment of thyroid gland content and serum measurements in conditions of nonnal and altered thyroidal economy and following administration of thyrotropin releasing hormone (TRH) and thyrotropin (TSH). J Clin Endocrinol Metab 1977; 44: 660-72. Carlson HE, Jacobs LS, Daughaday WH. Growth hormone, thyrotropin and prolactin responses to thyrotropin-releasing hormone following diethystilboesterol pretreatment. J Clin Endocrinol Metab 1973; 37:488-90. Chan V, Jones A, Liendo ChP, MacNeilly A, Landon J, Besser GM. The relationship between circadian variations in circulating thyrotrophin, thyroid hormones and prolactin. Clin Endocrinol 1978; 9: 337-49. Chopra II. A radioimrnunoassay for measurement of 3,3 ',5 -triiodothyronine (reverse T,). J Clin Invest 1974; 54: 583-92. Chopra IJ, Sack J, Fisher DA. Circulating 3,3 ',5 '-triiodothyronine (reverse T3) in the human newborn. J Clin Invest 1975; 55: 1137-41. Chopra IJ, Sack J, Fisher DA. 3,3',5-triiodothyronine (reverse T3) in fetal and adult sheep: studies of metabolic clearance rate, production rate, serum binding and thyroidal content relative to thyroxine. Endocrinology 1975; 97: 1080-8. Chopra IJ, Solomon DH, Chua Teco GN. Thyroxine: just a pro-hormone or a hormone too? J Clin Endocrinol Metab 1973; 36:1050-7. Cuttelod S, Lemarchand-Beraud Th, Magnenat P, Percet C, Poli S, Vannotti A. Effect of age and role of kidneys and liver on thyrotropin turnover in man. Metabolism 1974: 23: 101-13. De Costre P, Buhler U, De Groot LJ, Refetoff S. Diurnal rhythm in total serum thyroxine levels. Metabolism 1971; 20: 782-91. Dolva 0, Hanssen KF, Stadaas J, Berstad A. Thyrotropin releasing hormone inhibits the pentagastrin stimulated gastric secretion and gastric motility in man. Scand J Gastroenterol 1978; 13: suppl49,49. Eastman CJ, Ekins RP, Leith SM, Williams ES. Thyroid hormone response to prolonged cold exposure in man. J. Physiol 1974; 241: 175-81. Erenberg A, Phelps DL, Lam R, Fisher DA. Total and free thyroid hormone concentrations in the neonatal period. Pediatrics 1974; 53: 211-16. Evered DC, Tunbridge WMG, Hall R, et al. Thyroid hormone concentrations in a large scale community survey. Effect of age, sex, illness and medication. Clin Chim Acta 1978; 83: 223-9. Faglia G, Seck-Perroz P, Ferrari C, Ambrosi B, Spada A, Travaglini P. Enhanced plasma thyrotrophin response to thyrotrophin-releasing hormone following oestradiol ad- 95 ministration in man. Clin Endocrinol 1973; 2: 207-10. Guillemin R. The expanding significance of hypothalamic peptides, or, is endocrinology a branch of neuroendocrinology? Recent Progr Horm Res 1977; 33:1-28. m Haigler Jr ED, Pittman Jr JA, Hershman JM, Baugh CM. Direct evaluation of pituitary thyrotropin reserve utilizing synthetic thyrotropin releasing hormone. J Clin Endo- crinol Metab 1971; 33: 573-81. Haigler Jr ED, Hershman JM, Pittman Jr JA. Response to orally administered synthetic thyrotropin-releasing hormone in man. J Clin Endocrinol Metab 1972; 35: 631-5. Harris ARC, Vagenakis AG, Braverman LE. Sex-related differences in outer ring monodeiodination of thyroxinc and 3,3 ',5 -triiodothyronine by rat liver homogenates. Endocrinology 1979; 104: 645-52.

Herrmann J, Rusch HJ, Kroll HJ, Hilger P, Kriiskemper HL. Free triiodothyronine (T3)- and thyroxine (T4) serum levels in old age. Horm Metab Res 1974; 6: 23940. Hesch R-D, Gatz J, Pape J, Schmidt E, von zur Miihlen A. Total and free triiodothyronine and thyroid-binding globulin concentration in elderly human persons. Eur J Clin In- vest 1976; 6: 139-45. Hollander CS, Mitsuma T, Shenkman L.Woolf P, Gershengorn MC. Thyrotropin-releasing hormone: evidence for thyroid response to intravenous injection in man. Science 1972;175:209-10. Hooper MJ, Ratcliffe JG, Ratcliffe WA, et aL Evidence for thyroidal secretion of 3,3 ',5 - triiothyronine in man and its control by TSH. Clin Endocrinol 1978; 8: 267-73. Judd SJ,Carter JN, Corcoran JM, Lazarus L. Circulating thyroid hormone concentrations and posture and venous compression. Brit Med J 1975; 4: 735-6. Kersen F van, Sluiter WJ, Doorenbos H, Woldring MG. Infusion of synthetic TRH in the diagnosis of thyroid and pituitary dysfunction. Neth J Med 1974; 17: 184-95. Labrie F, Drouin J, Ferland L, et al. Mechanism of action of hypothalamic hormones in the anterior pituitary gland and specific modulation of their activity by sex steroids and thyroid hormones. Recent Progr Horm Res 1978; 34: 25-94. Larsen PR. Direct immunoassay of triiodothyronine in human serum. J Clin Invest 1972; 51: 1939-49. Lawton NF, Ellis SM, Sufi S. The triiodothyronine and thyroxine response tothyro- trophin-releasing hormone in the assessment of the pituitary-thyroid axis. Clin Endo- crinol 1973; 2: 57-63. Lucke C, Hehrmann R, von Mayersbach K, von zur Miihlen A. Studies on circadian variations of plasma TSH, thyroxine and triiodothyronine in man. A eta Endocrinol 1977; 86: 81-8. McNeilly AS, Hagen C. Prolactin, TSH, LH and FSH responses to a combined LHRH/ TRH test at different stages of the menstrual cycle. Clin Endocrinol 1974; 3: 427-35. Meinhold K, Wenzel KW. Studies on the direct radioimmunological determination of triiodothyronine in serum. J Clin Chem Clin Biochem 1974; 12: 477-85. Mitsuma T, Nihei N, Gershengorn MC, Hollander CS. Serum triiodothyronine: measurements in human serum by radioimmunoassay with corroboration by gas-liquid chromatography. J Clin Invest 1971; 50: 2679-88. Malholm Hansen J, Skovsted L, Siersbaek-Nielsen K. Age dependent changes in iodine metabolism and thyroid function. Acta endocrinol 1975; 79: 60-5. Mori M, Kobayashi I, Wakabayashi K. Suppression of serum thyrotropin (TSH) concentrations following thyroidectomy and cold exposure by passive immunization with 1 antiserum to thyrotropin-releasing hormone (TRH) in rats. Metabolism 1978; 27: 1485-90. Moriey JE, Garvin TJ, Pekary AE, Hershman JM. Thyrotropin-releasing hormone in the gastrointestinal tract. Biochem Biophys Pes Commun 1977; 79: 314-8.

.-__,. r

96 Mortimer CH, Besser GM, Goldie DJ, Hook J, McNeilly AS. The TSH, FSH and prolactin responses to continuous infusions of TRH and the effects of oestrogen administration in normal males, Clin Endocrinol 1974; 3: 97-103. Nagata K, Izumiyama T, Kamata K et al. An increase of plasma triiodothyronine concentration in man in a cold environment J Clin Endocrinol Metab 1976; 43: 1153-6. Nicod P, Burger A, Staeheli V, Valloton MB. A radioimmunoassay for 3,3 ',5 -triiodo-L- thyronine in unextractcd serum: method and clinical results. J Clin Endocrinol Metab 1976; 42: 823-9. Nicoloff JT, Fisher DA, Appleman MD. The role of glucocorticoids in the regulation of thyroid function in man. J Clin Invest 1970; 49: 1922-9. Noel GL, Dimond RC, Wartosfsky L, Earll JM, Frantz AG. Studies of prolactin and TSH secretion by continuous infusion of small amounts of thyrotropin-rcleasing hormone (TRH). J din Endocrinol Metab 1974; 39: 6-17. O'Connor JF, Wu GY, Gallagher TF, Hellman J. The 24-hour plasma thyroxin profile in normal man. J CJin Endocrinol Metab 1974; 39: 765-71. Olsen Th, Laurberg P, Weeke J. Low serum triiodothyronine and high scrum reverse triiodothyronine in old age: an effect of disease not age. J Clin Endocrinol Metab 1978; 47:1111-5. Ormston BJ, Garry R, Cryer RJ, Besser GM, HallR. Thyrotrophin-releasing hormone as a thyroid-function test. Lancet 1971; 2:10-14. OtsukiM.DakodaM, Baba S, Influence of glucocorticoids on TRP-induced TSH-response in man. J Clin Endocrinol. Metab 1973; 36: 95-102. Parker C, Pekary AE, Hershman JM. Effect of normal and reversed sleepwake cycle upon nyctohemeral rhythmicity of plasma thyrotropin: evidence suggestive of an inhibitory influence in sleep. J din Endocrinol Metab 1976; 43: 318-29. Patel YC, Alford FP, Burger HG. The 24-hour plasma thyrotropin profile. Cli Sci 1972; 43: 71-7. Postmes TJ, van Hout JC, Saat G, Willems P, Coenegracht J. A radioimmunoassay study and comparison of seasonal variation in plasma T3 and T4 concentrations in normal healthy persons. Clin Chim Acta 1974; 50: 189-95. Rastogi GK, Sawhney RC. Thyroid function in changing weather in a subtropical region. Metabolism 1976; 25: 903-8. Rosenthal MB. TRH and urinary urgency. New Eng J Med 1974; 291: 1308-9. Rubenstein HA, Butler VP, Werner SC. Progressive decrease in serum triiodothyronine concentrations with human aging: radioimmunoassay following extraction of serum. J Clin Endocrinol Metab 1973; 37: 247-53. Riimke ChrL, Bezemer PD. Methods for the determination of normal values. II. New methods. Ned T Gen 1972; 116:1559-68. Sakoda M. Mori H, Tatziwa M, Kusaka T, Fukuda T, Baba S. Variation of pituitary thyrotropin reserve with different dosages of thyrotropin-releasing hormone (TRH) - Mode of failure of pituitary thyrotropin reserve. Kobe J Med Sci 1976; 22: 81-9. Sanchez-Franco F, Garcia MD, Cariredo L, Martin-Zurro A, Escobar del Rey F. Influence of sex phase of the menstruale cycle on thyrotropin (TSH) response to thyrotropin- releasing hormone (TRH). J Clin Endocrinol Metab 1973; 37: 736-40. Sawhney RC, Rastogi S, Rastogi GK. Effect of estrogens on thyroid function. I. Altera- tions in rhesus plasma thyrotropin and its kinetics. Endocrinology 1978; 102: 1310-6. Sawin CT, Hershman JM. The TSH response to thyrotropin-releasing hormone (TRH) in young adult men: intra-individual variation and relation to basal TSH and thyroid hormones. J Clin Endocrinol Metab 1976; 42: 809-16. Sawin CT, Hershman JM, Boyd HI AE, Longcope C, Bacharach P. The relationship of changes in serum estradiol and progesterone during the menstrual cycle to the thyro- 97 tropin and prolactin responses to thyrotropin-releasing hormone. J Clin Endocrinol Metabl978;47: 1296-1302. Scanlon MF, Weetman AP, Pourmand M et al, Dopaminergic regulation of circadian . II TSH rhythms and thyroid hormone levels in euthyroid subjects (abstract). Ann d' Endocrinol 1979; 40:14A. Shenkman L, Mitsuma T, Suphavai A, Hollander CS. Triiodothyronine and thyroid- stimulating hormone response to thyrotrophin-releasing hormone. A new test of f £ thyroidal and pituitary reserve. Lancet 1972; 1: 111-3. Sirva JE, Larsen PR. Pituitary nuclear 3,5,3 -triidothyronine and thyrotropin secretion: an explanation for the effect of thyroxine. Science 1977; 617-20. Smals AGH, Ross HA, Kloppenborg PWC. Seasonal variation in serum T3 and T4 levels in man. J Clin Endocrinol Metab 1977; 44: 998-1000. Smeulers J, Visser TJ, Burger AKC, Docter R, Henneman G. Decreased production of triiodothyronine (T3) and constant production of reverse T3 at advanced ages. Ned TGen 1979; 123:12-5. Smyth PPA, Turner JJ, O'Donovan DK. Effect of stilboesterol therapy on thyrotrophin- releasing hormone (TRH) responsiveness in males. Clin Endocrinol 1977; 6: 13944. Snyder PJ, Utiger RD. Thyrotropin response to thyrotropin releasing hormone in normal females over forty. J Clin Endocrinol Metab 1972; 34:1096-8. Snyder PJ, Utiger RD. Response to thyrotropin releasing hormone in man. J Clin Endo- crinol Metab 1972; 34: 380-5. Snyder PJ, Utiger RD. Inhibition of thyrotropin response to thyrotropin-releasing hormone by small quantities of thyroid hormones. J Clin Invest 1972; 51: 2077-84. Staub JJ, Girard J, Gemsenjäger E. Entwicklung eines einfachen oralen Kurztests mit dem TSH-Releasing-Hormon (TRH) und dessen Anwendung in der Schilddrüsen-diagnostik. Schweiz med Wschr 1976; 106: 1839-42. Stepanas AV, Mashiter G, Maisey MN. Serum trüodothyronine: clinical experience with a new radioimmunoassay kit. Clin Endocrinol 1977;6: 171-83. Tuomisto J, Männistö P, Lamberg BA, Linnoila M. Effect of cold exposure on serum thyrotrophin levels in man. Acta Endocrinol 1976; 83: 522-7. UHer RP, Van Herle AJ, Chopra IJ. Comparison of alterations in circulating thyroglobulin, triiodothyronine and thyroxine in response to exogenous (bovine) and endogenous (human) thyrotropin. J Clin Endocrinol Metab 1973; 37: 741-5. Vagenakis AG, Rapoport B, Azizi F, Portnay GI, Braverman LE, Ingbar SH. Hyper- response to thyrotropin-releasing hormone accompanying small decreases in serum thyroid hormone concentrations. J Clin Invest 1974; 54: 913-8. Vogt P, Girard J, Staub JJ. Thyroid-stimulating hormone (TSH), triiodothyronine (Ts) and thyroxine (T4) responses to intravenous and oral stimulation with synthetic thyrotropin-releasing hormone (TRH) in young healthy adults. Klin Wschr 1978; 56: 31-5. Weeke J. Circadian variation of the serum thyrotropin level in normal subjects. Scand J Clin Lab Invest 1973 ; 31: 337-42. Weeke J. Thyrotropin response to thyrotropin releasing hormone in normal subjects. Europ J Clin Invest 1974;4: 29-32. Weeke J, Gundersen HJG. Circadian and 30 minutes variation in serum TSH and thyroid hormones in normal subjects. Acta Endocrinol 1978; 89: 659-72. Wenzel KW, Meinhold H, Herpich M, Adlkofer F, Schleusener H. TRH-Stimulationtest mit alters- und geschlechtsabhängigem TSH - Anstieg bei Normalpersonen. Klin Wschr 1974; 52: 722-7. Wenzel KW, Horn WR. Triiodothyronine (T3) and thyroxine (T4) kinetics in aged men. In: Robbins J, Braverman LE, eds. Thyroid Research. Amsterdam: Excerpta Medica, 1976: 270-3. 98 i I Westgren U, Burger A, Ingemansson S, Melander A, Tibblin S, Wfihlin E. Peroperative [ |',;< studies on thyroid activity: secretion of T4, T3 and reverse T, in normal man. In: - ,fp Robbins J, Braverman LE, eds. Thyroid research. Amsterdam-Oxford: Excerpta ' fc Medica, 1976:226-8. f£ Westgren U, Burger A, Ingemansson S, Melander A, Tibblin S,WdhlinE. Blood levels of >f 3,5,3 -triiodothyronine and thyroxine: differences between chflderen, adults and : •: I? elderly subjects. Acta Med Scand 1976; 200:493-5. \ j> Wilber JF, Montoya E, Plotnikoff NP et al. Gonadotropin-releasing hormone and thyro- ,; !-:': tropin-releasing hormone: distribution and effects in the central nervous system. ^ Recent ProgrHorm Res 1976; 32:117-60. , J £i Woolf PD, Gonzales-Baicena D, Schlach DS et al. Lack of effect of steroids on thyro- 1- trophin-releasing hormone (TRH) mediated thyrotrophin (TSH) release in man. ;• Neuroendocrinology 1973; 13: 56-62.

5 I

CHAPTER IV.

STUDIES IN PATIENTS WITH THYROID DISEASES ¥.' IV.I. Introduction.

Thyroid function was evaluated in a series of patients referred to our clinic. The aim of this study is to obtain a spectrum of thyroid function disorders ranging from severe hyperthyroidism to severe hypothyroidism, and to evaluate the usefullness of the various tests in the diagnosis of increased or decreased thyroid function as an indicator of thyroid disease.

IV.2. Patients and methods.

Patients who for the first time were referred to our clinic under the suspicion of thyroid disease, entered the study. Patients who had been treated with thyroid and / or antithyrpid drugs in the six months before the date of referral, and patients with hypothalamic-pituitary disorders were excluded. Patients medicated with beta-adrenoceptor blocking agents, or with drugs which influence thyroid hormone binding to the plasma proteins, and patients with co-existent acute or chronic non-thyroidal diseases were not excluded. The nature of the thyroid disease was determined by clinical examination, by the presence or absence of thyrc A antibodies, and by thyroid scintigraphy. Ophthalmic Graves' disease is the term we used to describe patients with the ophthalmic manifestations of Graves' disease in the absence of clinical hyperthyroidism or a past history of hyperthyroidism (Hall et al 1970). Blood sampling was performed through an indwelling venous catheter, and after obtaining two basal samples 200 Kg TRH was rapidly injected intravenously (see chapter III. 2). The basal value of each test was calculated as the mean of the -IS' and 0' samples. In calculating ratio's etc. SO percent of the detection limit of the assay was taken when the value was below the detection limit of the assay. Patients with abnormal thyroid function tests were classified according to the FT4 index and T3 values. If the FT4 index and the T3 values were both normal, the TSH-response to TRH was used for further classification. I IV.3. Results. I One hundred forty-eight patients were tested, and abnormal thyroid Classification of 125 patients with abnormal results of thyroid function tests (t = increased; I = decreased; N = normal; GO >» Graves1 disease, s TA = toxic adenoma, TMG = Toxic multinodular goitre, SAT = subacute thyioiditis, AIT autoimmune thyroiditis, STT = subtotal thyroidectomy, TT = total thyroidectomy).

classification number of pat. age, yrs. weight, kg clinical diagnosis laboratory diagnosis group FT4 index T, ATSH total (females) (xtSD) (x ± SD)

At t i 55 (43) 53.8 ±16.9 62.6 i 12.5 hyperthyroidism due to: overt hyperthyroidism GD (33) TA( 9) TMG (12) SAT( 1) B N t j, 11 ( 9) 60.5 ±11.1 60.3 ±11.9 hyperthyroidism due to: mild hyperlhyroidism GD( 4) ("Ts-toxicosis") TA( 1) TMG( 3) ophthalmic GD ( 3) C N N i 15 (11) 58.5 ±13.0 66.5 ±11.8 equivocal hyperthyroidism subclinical hyperthy- due to: roidism GD( 3) TA( 3) TMG( 4) ophthalmic GD ( S) D N N t 6(4) 53.3 ±19.1 66.1 ± 17.9 equivocal hypothyroidism subclinical hypothy- due to: roidism AIT( 2) STT( 2) ophthalmic GD( 2) E i N t 6(6) 62.2 ± 12.0 67.11 7.0 hypothyroidism due to: mild hypothyroidism A1T( 4) STT( 1) SAT( 1) it 27 (21) 54.5 ±16.7 74.7 ± 10.5 hypothyroidism due to: overt hypothyroidism AIT(18) STT( 6) TT ( 3)

G t N I 5(3) 66.0 ± 5.1 61.51 10.4 equivocal hypcrthyroidism ("T4-toxicosis") due to: GD( 3) TA( 1) TMG( t)

K'^!!HSSc^?sC?55':rr-'™;'K'-";'"':!"'''''"."•'"~--•^•*'*?iteff^»i>Wjsii*^^ -,,...,,. ..,„,,„, 101 HYPERTHYROIDISM EUTHYROIDISM HYPOTHYROIDISM § overt mild subclinical subclinical mild overt A B C NORMALS D E F GROUP (n=55) (n=ni ln=15) (n=63) (n=6) (n=6) (n=27) _ 60 i | 20 <0.3

| 0.4 t 0 _ 4 \ I 2 c & 0 _ 200 \ § 100 c ? 0 1.4

& 0.6 240 I 120 £ 0

Figure IV. 1. Median values of thyroid function tests in 63 normal adults and in 120 patients with thyroid function disorders (groups A-F, for classification see table IV. I). The horizontal linesdepict the inner limits for the percentiles P2.5 and P97.5 of the normal values.

function tests were found in 125. These were divided in the groups A-G (see table IV. 1) Four patients in group A and one in group B were on oral contraceptive pills, two in group A and one in group G were on B-blockers. One patient had a normal FT4 index and an increased plasma T3, but also an increased TSH-response to TRH; therefore she was aUocated to group D. Median values and the observed range of the thyroid function tests in 102 the groups A-F are presented in table IV.2 (see also figure IV. 1). I, All unequivocally hyperthyroid patients were allocated in the groups A and B. Of these 66 patients, Graves' disease was diagnosed in 40 (60,5%), and toxic nodular goitre in 25 (38%). The patients with toxic nodular

Table IV-2 Thyroid function tests in 63 normal adults and in 120 patients with thyroid function disorders (groups A-F, for classification see table IV-1). The median values are presented; in brackets the range in the groups A — F, and the inner limits for the percentiles P2.5 and P97.5 in the normal subjects.

TSH mU/1 rT3nmol/l T3 nmol/1 group A <0.3 0.67 3.85 (n = 55) « 0.3 - 1.35) ( 0.26-4.05) (2.50 - 13.35) group B <0.3 0.40 2.70 « 0.3 - 1.1 ) ( 0.22-0.69) (2.53- 3.94) group C <0.3 0.28 2.00 (n=15) (< 0.3 - 1.1 ) ( 0.17 - 0.57) (1.63- 2.45) normals 1.1 0.22 1.75 (n = 63) « 0.3 - 3.1 ) ( 0.14-0.38) (1.30- 2.45) group D 4.9 0.29 2.19 (n = 6) ( 2.6- 46 ) ( 0.18- 0.36) (1.75- 2.62) group E 17.2 0.13 1.66 (n = 6) ( 5.7- 60 ) ( 0.08 - 0.17) (1.44- 2.15) group F 61 <0.03 0.88 ( 17 -343 ) «0.03- 0.11) (0.19- 1.25)

T4nmol/1 T,U FT4 index

group A 193 1.21 219 (n = 55) ( 123-338) (0.81 - 1.66) (123 - 476) group B 120 0.94 113 ( 107-145) (0.59 - 0.99) ( 85 - 122) group C 102 0.93 88 (n= 15) ( 75 - 135) (0.75 - 1.18) ( 73 - 122) 1 normals 101 0.93 95 (n = 63) ( 75 - 135) (0.81 - 1.07) ( 70-122) group D 84 0.84 72 ( 80 - 107) (0.71 - 0.91) ( 70- 80) group E 60 0.88 51 ( 42- 68) (0.74 - 0.95) ( 37 - 57) group F 17 0.71 10 « 5- 53) (0.55 - 0.87) ( 2-44) 103

3 3 &• rT3/T3xl0" rT3/T4xl0- T3/T4XIO -

group A 170 3.66 22.2 ; I (n=55) (58 - 387) (1.80-14.11) (12.4- 46.0) I *\ group B 130 3.18 24.1 j (76 - 218) (1.85- 5.17) (17.8- 36.8) j group C 142 2.83 20.2 1 (n=15) (90 - 285) (2.05- 5.56) (15.5 - 28.5) ; normals 121 2.20 18.2 ! (n = 63) (69 - 209) (1.38- 3.53) (12.8- 23.4) ; group D 118 3.03 23.9 \ (n = 6) (93 - 195) (2.19- 4.23) (20.3- 28.9) \ group E 77 2.15 29.9 (n=6) (45 - 111) (1.90- 3.33) (23.9- 41.9) group F 43 2.50 55.8 (n=27) (13 - 219) (0.43 - 16.00) (11.4 - 350. )

goitre are older and have lower plasma T3 and T4 values than the patients with Graves' disease (table IV.3). First-onset thyrotoxicosis was seen in 47 of 65 patients (72%), and recurrent thyrotoxicosis occurred in 18 (28%): 13 out of 54 patients in group A (24%), and 5 out of 11 patients in group B (45%). The incidence of Graves' disease and toxic nodular goitre was the same in patients with or without recurrence (table IV.4). Recurrent thyrotoxicosis was associated with lower plasma T3 and T4 values and with older age in comparison to first-onset thyrotoxicosis. Eleven out of the 66 patients of the groups A and B had increased plasma T3 levels (and FT3 index values) but a normal FT4 index. The incidence of this so-called "T3-toxicosis" is thus 17%. Plasma T3 values in T3-toxicosis were lower than in hyperthyroid patients with an increased FT4 index (group A), but plasma T4 values in T3-toxicosis (although by definition within the normal range) were higher than in the normal adult population. All unequivocally hypothyroid patients were allocated to the groups E and F. The cause of the hypothyroid state in these 33 patients was autoimmune thyroiditis in 22 (67%), subtotal thyroidectomy in 7 (21%), total thyroidectomy in 3 (9%) and subacute thyroiditis in 1 (3%).- A normal plasma T3, but a decreased FT4 index was found in 6 of these patients (18%; group E). The decrease in FT4 index in these patients is less than in the patients of group F who also demonstrate decreased plasma T3 values. In between the hyperthyroid patients of the groups A and B and the hypothyroid patients of the groups E and F are 21 patients (group C and D) with normal plasma T3 and FT4 index values; they were considered abnormal because of an abnormal TSH response to TRH stimulation. 104 Table IV-3 I Thyroid function tests (given as geometric mean, between brackets x - SD and x + SD) in 65 hyperthyroid patients with Graves' disease or nodular goitre.

Graves' disease toxic adenoma toxic multinodular goitre ¥(•-• number of patients 40 10 15 females 29 8 14 males 11 2 1 age yrs ( x ± SD) 47.8 ± 15.1 64.0 ± 10.7 69.4 ± 8.6

rT3 nmol/1 0.66 ( 0.39- 1.13) 0.56( 0.33 - 0.96) 0.68( 0.45 -- 1.03) T3 nmol/1 4.27 ( 2.25 - 8.10) 3.53( 2.67 - 4.66) 3.78( 2.86 -- 5.01) T4nmol/1 186 (138 - 249 ) 161 (127 -205 ) 168 (130 --218 ) T3U 1.15 ( 0.93- 1.42) 1.16( 0.95 - 1.41) 1.13( 0.98-- 1.31) FT4 index 213 (134 - 338 ) 187 (130 -269 ) 190 (131 --276 )

Fifteen had a decreased or absent TSH response (group C), and six showed an increased TSH response (group D). The FT4 index values in the patients of group D are lower, but plasma T3 values are higher than in the normal subjects. All patients of the groups A and G responded favourably to appropriate treatment. Almost all patients of the groups B and F were also treated, and the majority improved clinically. Several patients of group C and D were also treated specifically, however the response to treatment of these individuals was much less consistent.

Table IV-4 Thyroid function tests (given as geometric mean, between brackets x - SD and x + SD) in 47 patients with fust-onset thyrotoxicosis and in 18 patients with recurrent thyrotoxicosis.

first-onset recurrence

number of patients 47 18 Graves' disease 29 11 toxic adenoma 8 2 toxic multinodular goitre 10 5 age yxs (x + SD) 52.3 ± 17.3 61.4 ± 10.8

rT9 nmol/1 0.66 ( 0.40- 1.10) 0.6K 0.37 - 1.02) T9 nmol/1 4.40 ( 2.87- 6.74) 3.72( 2.53 - 5.47) T4 nmol/1 182 (136 - 244 ) 166 (130 -212 ) T3U 1.18 ( 0.98- 1.42) 1.07( 0.89 - 1.30) FT4 index 215 (140 - 329 ) 179 (119 -268 ) 105 JF-J

The TSH- and T3-response to 200 Mg TRH lv. in 63 normal adults and in 120 patients with thyroid function disorders (groups A-F, for classification see table IV-1). The median values are presented; in brackets the range in the groups A-F, and the inner limits for the percentiles P2,5 and P97.5 in the normal subjects.

basal TSH mU/1 peak TSH mU/1 ATSHmU/1 ATSH %

group A <0.3 <0.3 0 1 (n = 55) (< 0.3 - 1.35) (< 0.3- 1.9) (- 0.4- 1.05) group B <0.3 <0.3 0 « 0.3 - 1.1 ) « O.i- 1.3) (- 0.13- 0.4 ) group C <0.3 0.5 0.1 (n=15) « 0.3 - 1.1 ) « 0.3- 2.0) ( 0 - 1.85) normals 1.1 8.5 7.85 (n = 63) « 0.3 - 3.1 ) ( 2.8- 22.0) ( 2.3- 19.4 ) group D 4.9 39.4 34.5 713 (n = 6) ( 2.6- 46 ) ( 25.8-103 )( 19.5- 57.1 ) (124 - 1444) group E 17.2 62.4 47.5 246 (n = 6) ( 5.7- 60 ) (23-155 ) ( 17.4 - 95 ) (144- 400) group F 61 136 57 87 § (n = 27) ( 17 -343 ) ( 33 -708 )( 10 - 470 ) ( 24 - 555) $

basal T3nmol/1 peakT3 nmol/1 AT3nmol/l AT,

group A 3.85 4.40 0.30 (n = 55) (2.50- 13.35) (2.96 - 13.60) (-0.35 - 1.31) (-4- 29) group B 2.70 3.25 0.25 9 (2.53 - 3.94) (2.60- 4.23) (-0.05 - 0.68) (-2- 26) group C 2.00 2.35 0.37 18 (N=15) (1-63 - 2.45) (1.70- 3.08) (-0.18 - 0.93) (-9- 47) normals 1.75 2.70 0.83 45 (n = 63) (1.30- 2.45) (2.15- 3.40) ( 0.55-1.31) (30-105) group D 2.19 2.92 0.66 33 (n = 6) (1.75 - 2.62) (2.30- 3.70) ( 0.43-1.41) (18- 77) group E 1.66 2.00 0.42 26 (n=6) (1.44 - 2.15) (1.26- 3.05) (-0.50 - 0.90) (-28 - 42) group F 0.88 0.88 0.10 20 (n = 27) (0.19- 1.25) (0.24- 1.52) (-0.15 - 0.55) (-16 - 78)

The TSH- and T3-responses to TRH in the patients of the groups A-F are presented in table IV.S. All patients of the groups A, B, and C demonstrated a severely impaired or absent TSH - response. The patients of the groups D, E, and F demonstrated a greater than normal TSH - 106 200 r -I

100 I 50

CO normal

0.5

detection limit 0.3

V-=_=.__f=. _^c 0.1 .15 0 20 60 120 180 240 I Figure IV. 2. The TSH-rcsponse to 200 wg TRH i.v. in 63 normal adults and in 120 patients with thyroid function disorders (groups A-F, for classification see table IV. 1; the median values of each groups are presented, note the logarithmic scale). V- 107 HYPERTHYROIDISM EUTHYROIDISM HYPOTHYROIDISM overt mild subclinical subclinical mild overt B C NORMALS D E F if GROUP (n=55) (n=ni (n=15) (n=63) (n=6) (n=6) (n=27)

too

*40

-20

1.4

1.1

_0.8 I 0.5 c

.0.1

-0.4 Figure IV. 3. The T,-response to 200 wg TRH i.v. (upper panel: relative increment; lower panel: absolute increment) in 63 normal adults and in 120 patients with thyroid function disorders (groups A-F, for classification see table IV. 1). The horizontal lines depict the inner limit for the percentile P2.S of the normal response.

response, i.e. in all subjects the response was above the inner limit for the percentile P97.S of the TSH - response in normal subjects. The relative increment of TSH after TRH decreased progressively from groups D to F, t as can be seen in fig. IV.2. The individual values for the absolute and relative increment of T3 in 1 response to TRH, are depicted in figure IV. 3. None of the hyperthyroid patients of the groups A and B was observed to have a relative increment greater than 30%, although 18 patients had an absolute increment greater 108 than O.SS nmol/1. Two patients of group C, four patients of group D and three patients of group E had a normal relative and absolute increment of T3. Seven of the severely hypothyroid patients of group F had a normal relative increment of T3, but only one had a normal absolute increment of 1-'. 0.55 nmol/1. r As judged from the inner limits for the percentiles P2.5 of both the normal absolute and relative increments of T3 in response to TRH, all i hyperthyroid patients with an increased plasma -T3 and all hypothyroid patients with a decreased plasma -T3 (except one) have an impaired T3- response. Five patients with equivocal hyperthyroidism (group G) had elevated FT4 index values but a normal plasma T3; the TSH- and T3-response to TRH was impaired in each (table IV. 6); they are considered to have "T4- V' toxicosis". Their case histories are presented below.

Patient I, a 59-years old female with sickle cell anaemia, was admitted because of abdominal complaints and a decrease in renal function. A diagnosis was made of necrotising papillitis; her clinical condition'improvcd without specific treatment. She was on digoxin for four years because of atrial fibrillation, and had been treated ten years before admission with carbimazole because of thyrotoxic Graves' disease. Al- though clinically euthyroid, a TRH test was performed. In the two weeks prior to the test, oral and intravenous cholecystography and intravenous urography was performed. She was discharged without anttthyroid treatment, and died some months later from unknown cause. Patient 2. a 64-years old male schizophrenic, experienced a considerable loss of weight of 16 kg in three months. He was otherwise healthy, and no goitre was palpable. Hyperthyroidism was suspected, and thyroid function tests were compatible with the diagnosis of "T4-toxicosis". However, at the time the results were known, his condition

Table IV-6

Thyroid function tests (basal values, and the TSH- and T3-response to 200 ng TRH Lv.) in five patients with an elevated FT4 index but normal T3 values (group G; "T4 -toxicosis").

patl pat. 2 pat. 3 pat. 4 pat. 5 x±SD

TSH mU/1 <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 rT3 nmol/1 0.81 1.13 0.79 0.52 0.39 0.73 ± 0.29 T3 nmol/1 2.11 2.33 1.83 1.73 2.00 2.00 ± 0.24 T4 nmol/1 176 163 150 150 143 156 ± 13 T3U 1.02 1.27 1.27 0.99 1.04 1.12 ± 0.14 FT4 index 180 207 191 149 149 175 ± 26 3 rT3/T3xl0- 384 485 432 301 195 359 ± 114 3 rT,/T4xl0" 4.60 6.93 5.27 3.47 2.73 4.60 ± 1.63 3 T3/T4xl0" 12.0 14.3 12.2 11.5 14.0 12.8 ± 1.3 ATSH mU/1 0 0 0 1.15 0 0.2 ± 0.5 AT3 nmol/1 -0.37 0.45 0.33 0.25 0.55 0.24 ± 0.36 AT3% -17 19 18 15 28 13 ± 17 109

240 min

Figure IV. 4. TheTSH-and T3-responseto200wgTRH i.v. in an elderly female patient while asymptomatic (solid line), and during an hysterical twilight state (broken line).

had improved and no antithyroid treatment was instituted. Plasma T4 gradually decreased into the normal range, and has remained normal for two years. During the observation period the patient was on medication with various psychotropic drugs. Patient 3, a 73-years old female, was admitted because of fever from unknown origin. Atrial fibrillation and ischemic heart disease were diagnosed, but no pulmonary embolism could be demonstrated. She was treated with digoxin, furosemide and acenocoumarol. Two weeks prior to the TRH test plasma T, was 3.00 nmol/land plasma T4 14S nmol/1, one week later oral cholecystography was performed. At the time of the TRH test plasma T3 was 1.83 nmol/1 and T4 ISO nmol/1. Thyroid scintigraphy demonstrated a small multinodular goitre. The fever did not respond to antibiotics, but disappeared during entithyroid medication. Patient 4, a 67-years old male, was referred because of atrial fibrillation with a high ventricular response, which could not be suppressed satisfactorily with digoxin. The patient had no other signs or symptoms of thyrotoxicosis, and no goitre was palpable. A first blood sample revealed increased values of T} (3.1 nmol/1) and T4 (145 nmol/1). Propranolol was added to the medication, and this had a favourable effect: the ventricular rate decreased. The TRH test was performed under propranolol medication. Patient 5, a 67-years old female, was admitted to the psychiatric c!:.nic because of melancholy. She was treated with chlorodiazepoxide, perphenazine, amitryptiline and B-pyridyl-carbinol (Ronicol). Hyperthyroidism was suspected clinically, and thyroid scintigraphy revealed a toxic adenoma. The patient improved under carbimazole.

IV. 4. Discussion.

4a. The spectrum of thyroid function disorders. The plasma iodothyronine concentrations found in subjects with I increased, normal or decreased thyroid function, are depicted in figure IV. 1. The figure suggests that in the development of the hyperthyroid state the production of T3 is increased first, to be followed by an increased secretion -—-?•"

110

I yv> of T4. In this view T3 toxicosis is just a forerunner of the classical thyrotoxic state in which both T4 and T3 are produced in excess. In I contrast, in the development of hypothyroidism production of T4 is first of all decreased, later followed by a decrease in the production of T3. Several longitudinal studies have indicated that this proposed sequence can indeed occur (Hollander et al 1971; Patel and Burger 1973; Marsden and McKerron 1975; Marsden et al 1975; Toft et al 1975). It is thus plausible that patients with T3-toxicosis are less thyrotoxic than patients in whom both T3 and T4 are increased, and patients with a decrease in plasma T4 only have a milder form of hypothyroidism than patients in whom plasma T3 is also decreased. The relation between the plasma T3 concentration and the severity of the thyroid function disorder is supported in the literature (Surks et al 1972). The plasma T3 concentration in "classical thy rotoxicosis" where plasma T4 is also increased, is higher than in T3-toxicosis in which a clinically less severe hyperthyroidism is reported (Ivy et al 1971; Marsden and McKerron 1975). Mild hypothyroidism or clinical euthyroidism with a decrease in plasma T4 only is described by Patel and Burger(1973). Evered etal(1973)andStepanasetal(1977). In none of these studies (including ours) the clinical state has been evaluated quantitatively, and consequently the relation between the plasma T3 concentration and the clinical state cannot be expressed numerically. In general, however the level of the plasma T3 concentration reflects the severity of the thyroid function disorder, although exceptions to this rule are not rare.

Figure IV. 1 and the T3/T4 ratio's show that there seems to be an overproduction of T3 relative to T4 in both hyperthyroidism and hypothyroidism, and this has been verified in kinetic studies (Larsen 1972; Nicoloff et al 1974). The overproduction of T3 can be explained by the development of iodine deficiency in the thyroid gland, which then leads to a preferential synthesis of T3. Such a mechanism is Ideologically sound, at least in the development of hypothyroidism: T3 is metabolically four times more active than T4, the preferential synthesis of T3 saves 25% iodine, so that switching from T4 to T3 synthesis results in an approximately sixfold increase of the hormonal effect at the same expenditure of iodine (Havard 1974). Clinical and experimental studies support this explanation (Greer et al 1968; Pineda etal 1970). The increased incidence of T3-toxicosis in areas with endemic goitre also favours such a pathogenesis (Hollander et al 1972). The increased TSH stimulation in hypothyroidism may be responsible for a preferential secretion of T3 by the thyroid gland, in order to maintain euthyroidism (Sterling 1970; Patel and Burger 1973; Evered et al 1973). In this respect it is of interest that plasma T3 values in the patients with subclinical hypothyroidism (group D) are higher than the average plasma T3 values of the normal subjects, and in fact an abnormally increased plasma T3 was observed in some of these patients. The median value of plasma rT3 in these patients is also higher than in normal subjects HI although the median plasma T4 is subnormal. This may again indicate a preferential secretion of rT3 by the thyroid gland; a preferential secretion of rT3 and T3 over T4 in response to TSH stimulation has recently been reported in the thyroid of the dog (Laurberg 1978). It has been suggested that the overproduction of T} in hyperthyroidism is caused by a thyroid stimulating factor (immunoglobulins?) present in plasma (Izumi and Larsen 1977). Obviously, an increase in plasma T3 (and rT3) may also be caused by an increased rate of peripheral T4 conversion. However, thyroid tissue concentration of T3 is relatively more increased than of T4 in hyperthyroid patients (Larsen 1975), and the increase of plasma T3 thus seems primarily caused by increased T3 secretion by the thyroid gland. On the other hand, it is clear that in hyperthyroidism more T4 is available for peripheral conversion, and consequently extra-thyroidal production of T3 could be increased as compared to normal subjects, but this does not necessarily imply that the conversion rate of T4 into T3 is increased (Schimmel and Utiger 1977). Production rates and metabolic clearance rates of both T3 and rT3 are increased in hyperthyroidism and decreased in hypothyroidism in man (Smallridge et al 1978). These authors have also estimated the relative production rates (PR) of rT3 and T4, and reported a trend to higher rT3/T4PR ratio's in hyperthyroidism and to lower rT3/T4 PR ratio's in hypothyroidism. Inada et al (1975) have found that the rate of extrathyroidal conversion of T4 to T3 averages 3.00 ± 0.68% of the extrathyroidal T4 pool per day in normal subjects, but is increased to 4.16 ± 0.72% in untreated hypothyroid patients; thus, although the absolute extrathyroidal production of T3 in normal subjects is greater than in hypothyroid patients, in hypothyroidism relatively more T3 is derived from the extrathyroidal deiodination of T4. This is in agreement with the study of Maguire et al (1976) who report an increased conversion rate of T4 to T3 in hypothyroidism, and also in hyperthyroidism, with a fall to normal after treatment. Therefore, it can be concluded that plasma concentrations of T3 and rT3 are modulated in hyper- and hypothyroidism by both thyroidal and extrathyroidal factors, as they are in the normal subjects. In our patients the plasma ratio's of T3/T4 are much more increased in hypothyroidism than in hyperthyroidistn; since T3 and T4 change in the same direction in these states, this indicates a relative overproduction of T3 to T4, which is greater in hypothyroidism than in hyperthyroidism. In contrast, the plasma ratio's of rT3/T4 are very much increased in hyperthyroidism and about normal in hypothyroidism; since in these states rT3 and T4 also change in the same direction, this may be interpreted as indicating a relative overproduction of rT3 in hyperthyroidism as compared to hypothyroidism.

Finally, we observed a steady decline in the plasma ratio's of rT3/T3 from severe hyperthyroidism to severe hypothyroidism, indicating that in 112 a l hyperthyroidism a relative overproduction of rT3 occurs, while in ; ^= j hypothyroidism there is a relative overproduction of T3. One can therefore , |?, speculate that in hyperthyroidism more rT3 and less T3 is generated from fg! 7 T4, and that in hypothyroidism less rT3 and more T3 is generated from T4. || The homeostatic significance of such a mechanism is clear: by modulating i * fcc. e the conversion of T4 into either active T3 or the metabolically inactive rT3 If the peripheral tissues may counteract to some extent the deleterious effects ' ij of excessive or insufficient thyroid hormone secretion. We are well aware c that our patient data do not allow any firm conclusions on the extent or the 0 side of the overproduction of the various iodothyronines in different states of thyroid function. Clearly more studies are needed.

4b. Hypothyroidism. . =| Plasma T3 was normal in 18% of the hypothyroid patients of the groups • E and F, and this limits the value of plasma T3 determinations foi the , : diagnosis of hypothyroidism. Earlier studies of Mitsuma et al (1971) and j Surks et al (1972) reported decreased plasma T3 values in all hypothyroid patients (probably because only severely hypothyroid patients were studied), but in all later studies normal plasma T3 values are found in a portion of the hypothyroid patients: in 25% by Patel and Burger (1973), in 26% by Chopra (1974), in 12% by Burman et al (1975), in 43% by Seth et al (1976), and in 48% by Stepanas et al (1977). These studies include 217 patients, and 83 of them (38%) had a normal plasma T3 concentration. Our incidence is lower (18%), and this may be caused by the warning limits that we apply for the normal range. It is obvious that plasma T4 (or the FT4 index) is a more sensitive indicator for hypothyroidism than plasma T3, although both tests are in no way specific for the diagnosis hypothyroidism: decreased values can be caused by impairment of T4-»T3 conversion and abnormalities in the protein binding of thyroid hormones. The basal TSH value and the TSH response to TRH appear to be better indicators for the hypothyroid state, since 6 patients were found with normal T3 and T4 values, but increased TSH responses to TRH; three of these patients had normal basal TSH values. Such patiens have been labeled as •;.]| having subclinical or preclinical hypothyroidism (Evered and Hall 1972; \ Evered et al 1973), a term used to define an asymptomatic state in which a f i reduction in thyroid activity has been compensated for by an increased TSH !] i output to maintain a euthyroid state. Neither term is entirely satisfactory, U j because some patients of this group with minor complaints may improve by I ! L-thy roxine therapy and because the natural history of the thyroid function || disorder in this group of patients is not necessarily progressive: in some plas- \ < ma TSH reverts into the normal range, in some plasma TSH remains slightly .j - elevated and in others a further increase in plasma TSH is seen together with j { the development of clinical hypothyroidism (Tunbridge et al 1978). Others •' I have used the term preclinical hypothyroidism to indicate states with circu- 113 lating thyroid antibodies (Bastenie et al 1971) or with hypercholesterol- V'<' aemia (Fowler et al 1970) claiming an association with ischemic heart dis- t ease. However, subsequent studies have not always supported the concept of hypercholesterolaemia as a premonitory sign of hypothyroidism (Nilsson et al 1976; Tunbridge et al 1977; Gordin et al 1979), and also mean TSH levels in subjects with antibodies to thyroid tissue are higher than in controls in some studies (Bastenie 1977; Gordin et al 1979), but not in others (Tunbridge et al 1977). Wilkin et al (1977) and Sluiter (1979) have presented a kinetic model which explains the increased release of TSH in subclinical hypothyroidism by a reduced functioning thyroid mass. Tests of peripheral tissue function may be normal in subclinical hypothyroidism, as demonstrated by a normal Achilles tendon reflex time in clinically euthyroid subjects with high TSH levels living in an area of severe endemic goitre (Goslings et al 1977). Thus, it may well be that the sensitivity for thyroid hormone of different tissues can vary. However this may be, there are subjects judged clinically to be euthyroid and with normal plasma concentrations of T3, T4 and TSH, who have an increased TSH response to TRH (Alaghband-Zadeh et al 1977); these subjects may be at risk for developing hypothyroidism. It can be concluded that the TSH response to TRH is the most sensitive indicator presently available for detecting early hypothyroidism, and further that hypothyroidism is a graded phenomenon. The pattern of thyroid function tests in our patients of group D to F closely resembles the pattern described by Evered et al (1973), Wenzel et al (1974), Bigos et al (1978) and Kahn (1978). Detection and classification of hypothyroidism by means of laboratory tests has proven to be superior to clinical examination, but it remains to be proven if patients with subclinical hypothyroidism benefit from treatment. In earlier studies on the TSH response to TRH it is often stated that hypothyroid patients have an exaggerated TSH response (Ormston et al 1971), and indeed the absolute increment of TSH progressively increases when hypothyroidism is becoming more severe (table IV. S) However, the relative increment of TSH progressively decreases at the same time (figure IV. 2). The same trend is found in other studies (Evered et al 1973; Wenzel et al 1974; Bigos et al 1978). This indicates a progressive decrease in the release of TSH by TRH with the advancement of more severe hypothyroidism. In this respect it is of interest that in the first weeks after instituting replacement therapy in hypothyroid patients an enhanced TSH response to TRH relative to the pre-treatment response is observed (Fischer et al 1978; Ridgway et al 1979). Both phenomena may be explained by a decrease of the TSH content in the pituitary thyrotrophic cells, which is re- stored during substitution therapy. Indeed, a low content of stored hormone I is suggested by the paucity of secretory granules observed inTSH-produc- ing pituitary adenomas secondary to primary hypothyroidism (Samaan et 114 4v al 1977). Sixteen of our 27 hypothyroid patients had the peak TSH value at 60 minutes after TRH (in contrast to five of the 63 normal subjects), and this delay may also indicate a decrease in the directly releasable pool of stored i TSH in the pituitary thyrotrophs. •O7 The T3 response to TRH in hypothyroidism must be analysed by the absolute increment of T3, because the relative increments are not reliable due to the low plasma concentrations of T3. Although the T3 response to TRH is impaired in all our patients with overt hypothyroidism except one (group F) which is in accordance with other studies (Shenkman et al 1972; Squire and Gimlette 1977; Bigos 1978), measuring the T3 response to TRH has little to offer in the diagnosis of hypothyroidism. In fact, the only indication for TRH stimulation in the process of diagnosing hypothyroidism, is a normal or borderline elevated plasma TSH value when T3 and T4 are also normal.

4c. Hyperthyroidism. The preponderance of females in the hyperthyroid patients of group A and B (with a female/ male ratio of about 4:1) and the older age of patients with toxic nodular goitre as compared to patients with thyrotoxic Graves' disease, is in agreement with the literature (Werner 1978). In general thyrotoxic Graves' disease is associated with higher plasma T3 concentrations than toxic nodular goitre, and this has also been described by others (Patel and Burger 1973; Zurcher et al 1973; Shalet et al 1975; Marsden and McKerron 1975; Sobrinho etal 1977). Plasma T3 is increased \\ in all patients with unequivocal hyperthyroidism in contrast to plasma T4; thus T3 is a more sensitive test for detecting hyperthyroidism than T4, but again the specificity of the T3 assay in this respect is decreased by the occurrence of increased T3 values due to other causes (e.g. increased binding to plasma proteins, preferential secretion of T3 in subclinical hypothyroidism).

The incidence of 17% of T3-toxicosis in our series is greater than - reported initially in the literature: 5% by Mitsuma et al (1971), 4% by Hollander et al (1972), 11% by Bellabarba (1972), all in the U.S.A.; 8% by Patel and Burger (1973) in Australia, 9% by Zurcher et al (1973) in Switzerland, and 12.5% by Hollander et al (1972) in an area with endemic goitre (Chile). Subsequently an ever higher incidence was reported, also from areas without iodine deficiency: 8.2% by Turner et al (1975) in Ireland, 10% by Carter et al (1975) in Australia, 17% by Marsden and McKerron (1975) in England, 30% by Shalet et al (1975) in England, 14% at first-onset thyrotoxicosis and 25% if recurrent thyrotoxicosis is included by Stepanas et al (1977), also in the U.K., and lastly 31% by Sobrinho etal (1977) in an area with endemicgoitre (Portugal). One may conclude from all these studies that the real incidence of T3-toxicosis is unknown, but also that the incidence even in areas without iodine deficiency is considerable in the order of about 20% of all hyperthyroid patients. The incidence of T3 toxico- 115 sis in the above-mentioned studies (as far as this could be ascertained), is -i - V' 11% if patients with recurrent thyrotoxicosis are excluded and 21% if these , !.. patients are included (the corresponding figures in our study are 13% resp. ] |: 17%). Thus, T3-toxicosis occurs relatively often in recurrent thyrotoxicosis. S . j| This could be due to the development ofintrathyroidal iodine deficiency as a 1 •; ' % result of the thyrostatic treatment and by the likelihood that recurrent > I- thyrotoxicosis will be diagnosed earlier than first-onset hyperthyroidism. | | T3 toxicosis occurs both in thyrotoxic Graves' disease and in toxic nodular •< goitre. In the studies mentioned previously the nature of the hyperthyroid : state is given in 133 patients. 55% had Graves' disease, and 43% toxic nodular goitre. The corresponding figures in our study are 50% and 50% (excluding the patients with ophthalmic Graves' disease), as opposed to 60% resp. 38% in the patients of group A. T3 toxicosis thus occurs more often in toxic nodular goitre than in thyrotoxic Graves' disease.

In theory, T3 toxicosis might be caused by an enhanced peripheral conversion of T4 into T3 Herrmann et al (1975) described two patients in j whom excessive peripheral conversion of T4 to T3 supposedly contributed t to the development of T3 toxicosis. However, these patients were studied ' during antithyroid drug treatment. Also the observed ratio's of plasma rT3, T3 and T4 in our study are not in favour of such a mechanism. In this i respect our results are in disagreement with other reports demonstrating | normal or subnormal plasma rT3 values (Griffiths et al 1976) and j decreased rT3/ rT3 ratio's (Wenzel and Meinhold 1975; Ratcliffe et al 1976) • in patients with T3 toxicosis as compared to normal subjects. Finally f] Snarski and Snarska (1979) by measuring the daily degradation rate and the j metabolic clearance rate of T4 found no indication for an increased peripheral conversion of T4 to T3 in two patients with T3 toxicosis. It must ! be concluded that so far no patient has been described in whom T3 toxicosis is caused by an enhanced peripheral conversion of T4 to T3.

All patients of the groups A and B demonstrate a severely impaired or \ absent TSH response TRH. It is of interest that not infrequently patients \ are found with normal plasma T3 and T4 values but an absent TSH ,;' response to TRH (group C). This has been a common finding by many | authors, especially in patients with ophthalmic Graves' disease, 8 hyperfunctioning thyroid adenoma or multinodular goitre, who are >; judged clinically to be euthyroid (Lawton et al 1971; Ridgway et al 1973; 1 Chopra etal 1973; Franco etal 1973; Ormstonetal 1973; Clifton-Blighet | al 1974; Gemsenjager et al 1976; Dige-Petersen and Hummer 1977; Elte et | al 1977; Smeulers et al 1977; Kirkegaard et al 1977; Morgans et al 1978; | Blichert-Toft etal 1978; Emrich and Bahre 1978). A possible explanation | for this phenomenon is, that thyroid hormone secretion in these patients is ft unduly high and sufficiently elevated (due to an expanded functioning '$, thyroid mass) to inhibit the release of TSH by the negative feedback of 3 116

»V- thyroid hormones on the pituitary thyrotrophs. This situation can be 1:' considered as subclinical hyperthyroidism. The recurrence of a normal TSH response to TRH after partial thyroidectomy in such patients, which in some studies is accompanied by a significant decrease in plasma T4 and T3 to within the normal range, also favours this view (Gemsenjager et al 1976; Morgans etal 1978; Blichert-Toft et al 1978). The suppression of TSH release in these patients is not absolute, since oral administration of high doses of TRH (40 mg) may result in TSH release in some of them (Staub et al 1979; Kirkegaard et al 1979). This makes an abnormality in the process of TSH synthesis or release in these patients less likely. Results of peripheral tissue function tests (systolic time interval, sex-hormone binding globulin and Achilles tendon reflex time) in patients with subclinical hyperthyroidism are reported to be within the euthyroid range (BirkhSuser et al 1979). The T3 response to TRH as measured by the relative increment of T3 is below normal in all hyperthyroid patients (groups A and B), and this is in agreement with the findings of Murie et al (197S) and Squire and Gimlette (1977). Clearly the expression of the T3 response to TRH by the relative increment of T3 is superior to the absolute increment in hyperthyroidism due to the wide fluctuations and more imprecise measurements of plasma T3 concentrations in the very high range. As can be seen in figure IV. 3., there are two patients in group C with a normal increment of T3 (both absolute and relatively) despite an absent TSH response. This suggests that biologically active TSH is released in response to TRH, which is not detected by the TSH radioimmunoassay (alternatively, it is possible that TRH has a direct influence on the thyroid). Indeed, in a single case a TRH- induced rise in TSH as assessed by an increased thyroidal iodine secretion has been demonstrated when no substantial increase in immunoassayable TSH could be detected (Boehm et al 1976). If so, the T3 response to TRH in our hands is even more sensitive for measuring the negative feedback of thyroid hormones at the pituitary level than the TSH response. This is further illustrated by the following case history.

Patient X, a 69-years old female, was admitted to another hospital because of depression and bizarre behaviour. Extensive investigation did not reveal any organic cerebral disease or endocrine abnormality, with the exception of an absent TSH response to TRH. She was referred to our clinic. Her behaviour had improved I spontaneously, and a tentative diagnosis of hysteria was made. On clinical i examination she was euthyroid, and the FT4 index and the TSH- and T3-response to TRH were all normal (figure IV. 4). Nine months later the patient returned from a journey to the U.S.A. with the same bizarre behaviour. She was still clinically euthyroid and the FT4 index was again normal; a hysterical twilight state was diagnosed. The TSH-response to TRH was now absent, but the T, response was normal. Thus, an absent TSH-response to TRH can be caused by other states than "•*:

117 hyperthyroidism, especially by psychiatric disorders with a return to a normal TSH-response after recovery (Loosen et al 1977; Gregoire et al 1977). Nevertheless, the maintained T3 response in our case indicates a normal set-point of the pituitary-thyroid axis, and therefore euthyroidism.

.*-.: u The patients of group G, tentatively labeled as T4-toxicosis" are peculiar in that they do not fit in the described categories. These patients are also peculiar from a clinical point of view: they were certainly not cases of overt hyperthyroidism, and all five patients either suffered from a co- existent non-thyroidal disease, or were or had been treated with drugs known to influence thyroid hormone metabolism (propranolol, and especially radiographic contrast agents). All patients had an abnormal TSH and T3 response to TRH, compatible with hyperthyroidism. In retrospect, hyperthyroidism at the time of testing seems probable in patients 3 and S, possible in patients 1 and 4, and unlikely in patient 2. Considerable debate has arisen in the literature if T4 toxicosis really exists as a separate entity. Some consider the finding of raised FT4 index values together with normal T3 and FT3 index values as a normal albeit unusual finding in elderly especially female subjects (Britton et al 1975), or merely as the borderline area of the upper normal range and the lower hyperthyroid range (Hadden et al 1975), while others suggest that T4 toxicosis really exists (Turner et al 1975; Kirkegaard et al 1975; Turner et al 1975; Engler et al 1978; Dorfman and Young 1978). The patients described by these authors all have more or less severe co-existent non- thyroidal disease, and very frequently have been exposed to iodine- containing radiographic contrast agents; the only exception to this is the one patient described by Dorfman and Young. More light upon this problem has been shed by Birkhauser et al (1977). They describe 31 patients with various non-thyroidal diseases and a tentative laboratory diagnosis of T4 toxicosis; none of the patients was clinically hyperthyroid After recovering from their disease classic hyperthyroidism with a rise of plasma T3 into the hyperthyroid range developed in 14 patients, a rapid return of T4 values to normal was observed in 11 patients and in 6 patients T4 remained elevated (in two of these transient iodine-induced hyperthyroidism was diagnosed with spontaneous return of T4 to normal as the iodine was eliminated). Exposure to excess iodine was eventually traced in 16 of the 31 patients. In this respect a study of Sobrinho et al (1977) from I Portugal is of interest. These authors describe the preferential secretion of it T3 in 59 hyperthyroid subjects with Graves' disease or with nodular goitre; I' however in 5 out of 15 patients with iodine-induced thyrotoxicosis no preferential secretion of T3 was noted, i.e. these patients had T4 toxicosis. Most cases of T4 toxicosis therefore are due to impairment of the extrathyroidal conversion of T4 into T3 caused by intercurrent non- 118 thyroidal disease or by a particular medication (as shown by Biirgi et al (1976) and Wu et al (1978) some iodine-containing radiographic contrast -a? I agents inhibit the T3 generation from T4, but iodine itself has no effect on the conversion of T4 as demonstrated by Burger et al in 1976). There are only a few cases of T4 toxicosis in which T4 secretion by the thyroid gland is really excessive in relation to T3 secretion, usually associated with exposure to excess iodine. The transient increase in T4 observed in some patients remains unexplained; it may well be caused by a decreased MCR of T4 and/ or by abnormalities in the affinity or capacity of thyroxine f binding to the plasma proteins due to the intercurrent disease (recently i'i' Hennemann et al (1979) described elevated T4 and FT4 index but normal FT4 values in euthyroid subjects, caused by an inherited increased affinity of iodothyronines to the plasma proteins). The incidence of "T4 toxicosis" in our series (groups A,B and G) is 7%, of T3-toxicosis 15.5%, and of "classical" thyrotoxicosis 77.5% (two other patients with T4 toxicosis are described in the chapters V and VII). Caplan et al (1978) found T4 toxicosis in 12% of 83 elderly hyperthyroid patients. This limits the value of the plasma T3 determination for the diagnosis of hyperthyroidism. It is our policy to start with the assay of T4 and T3U, and to proceed with a T3 determination if the FT4 index is normal and hyperthyroidism is suspected; in all borderline cases a TRH stimulation test is performed. A normal TSH response to TRH excludes the existence of primary hyperthyroidism - this is the most important diagnostic aspect of the TRH test. It should be realized however, that an absent or impaired TSH response to TRH is not specific for hyperthyroidism, since blunted responses are occasionally seen in normal subjects and psychiatric patients.

IV. 5. References.

Alaghband-Zadeh J, Carter GD, Daly JR, Fowler PBS, Greenwood TW. Association between exaggerated responsiveness to thyrotrophin-releasing hormone and hypercholesterolaemia. Lancet 1977; 2: 998-1000. Bastenie PA, Vanhaelst L, Bonnyns M, Neve P, Starquet, M. Preclinical hypothyroidism: a risk for coronary heart-disease. Lancet 1971; 1: 203-4. Bastenie PA. Diagnostic de l'hypothyreoidie fruste. Sem Hop Paris 1977; S3: 1447-9. Bellabarba D. Further observations on T3 thyrotoxicosis (abstract). Clin Res 1972; 20:421. Bigos ST, Ridgway EC, Kourides I A, Maloof F. Spectrum of pituitary alterations with mild and severe thyroid impairment. J Clin Endocrinol Metab 1978; 46: 317-25. Birkhattser M, Burer Th, Busset R, Burger A. Diagnosis of hyperthyroidism when serum- thyroxine alone is raised. Lancet 1977; 2: 53-6. Birkhattser MH, Staub JJ, Burckhardt D et al. Oral TRH and metabolic parameters in "preclinical hyperthyroidism" (abstract). Acta Endocrinol 1979; suppl. 225:33. I Blichert-Toft M, Christiansen C, Axelsson CK, Egedorf J, Ibsen H, Ibsen J. Effect of selective goitre resection on absent thyrotropin response to thyrotrophin releasing hormone in idiopathic euthyroid goitres. Clin Endocrinol 1978; 8: 95-100. Boehm TM, Dimond RC, Wartofsky L. Isolated thyrotropin deficiency with thyrotropin-

.<&:• 119 releasing hormone induced TSH secretion and thyroidal release. J Clin Endoci inol Metab 1976; 43: 1041-6. I Britton KE, Quinn V, Ellis SM et al. Is "T4-toxicosis" a normal biochemical finding in elderly women? Lancet 1975; 2: 141-2. ..w Burger A, Dinichert D, Nicod P, Jenny M, Lemarchand-Beraud T, Vallotton MB. Effect of amiodarone on serum triiodothyronine, reverse triiodothyronine, thyroxin and thyrotropin. J Clin Invest 1976; 58: 255-9. Burgi H, Wimpfheimer C, Burger A, Zaunbauer W, Rosier H, Lemarchand-Beraud T. Changes of circulating thyroxine, triiodothyronine and reverse triiodothyronine after radiographic contrast agents. J Clin Endocrinol Metab 1976; 43: 1203-10. Burman K D, Wright F D, Earll J M, Wartofsky L. Evaluation of a rapid and simple technique for the radioimmunoassay of triiodothyronine (T3). J Nud Med 1975; 16: 662-5. Caplan RH, Glasser JE, Davis K et al. Thyroid function tests in elderly hyperthyroid patients. J Am Geriatr Soc 1978; 26: 116-20. 7 Carter JN. Eastman CJ, Casey JH, Farrell JC, Lazarus L. T3 toxicosis Med J Aust 1975; 1: 229. Chopra U, Solomon DH, Chua Teco GN. Thyroxine: just a pro-hormone or a hormone too? J Clin Endocrinol Metab 1973; 36:1050-7. Chopra IJ, Solomon DH, Chopra U, Young RT, Chua Teco GN. Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhoris: evidence for euthyroidism despite subnormal serum triiodothyronine. J Clin Endocrinol Metab 1974; 39: 501-11. Clifton-Bligh P, Silverstein GE, Burke G. Unresponsiveness to thyrotropin-releasing hormone (TRH) in treated Graves' hyperthyroidism and in euthyroid Graves' disease J Clin Endocrinol Metab 1974; 38:531-7. Dige-Petecsen H, Hummer L. Serum thyrotropin concentrations under basal conditions and after stimulation with thyrotropin-releasing hormone in idiopathic non-toxic goitre. J Clin Endocrinol Metab 1977; 44: 1115-20. Dorfman SG, Young RL. T4- thyrotoxicosis. Arch Int Med 1978; 138:1016-7. Editorial. Problems with serum-thyroxine. Lancet 1977; 2:74-5. Elte JWF, Haak A, Frdlich M, Wiarda KS, Wermeskerken RKA. Autonomously functioning euthyroid multinodular goitre. Neth J Med 1977; 20:1-4. Emrich D, Bahre M. Autonomy in euthyroid goitre: maladaptation to iodine deficiency. Clin Endocrinol 1978; 8:257-65. Engler DM, Stockigt JR, Feller V, Roy JA, Taft HP. T3 thyrotoxicosis in Melbourne. Proc EndocrSoc Austr 1975; 18-22. Engler D, Donaldson EB, Stockigt JR et al. Hyperthyroidism without triiodothyronine excess: an effect of severe nonthyroidal illness. J Clin Endocrinol Metab 1978; 46:77-82. Evered D, Hall R. Hypothyroidism. Brit Med J1972; 1:290-3. Evered DC, Ormston BJ, Smith PA, Hall R, Bird T. Grades of hypothyroidism. Brit Med J 1973:1:657-62; Fischer HRA, Hackeng WHL, Schopman J, Silberbusch J. Effects of substitution with thyroxine on the thyrotrophin (TSH) response to thyrotrophin-releasing hormone (TRH) in severe primary myxoedema and in mild hypothyroidism following prolonged thyrostatic therapy. Acta Endocrinol 1978; 89: 303-15. Fowler PBS, Swale J, Andrews H. Hypercholesterolaemia in borderline hypothyroidism (stage of premyxoedema). Lancet 1970; 2:488-91. Franco PS, Hershman JM, Haigler JrED, Pittman Jr J A. Response to thyrotropin-releasing hormone compared with thyroid suppression tests in euthyroid Graves' disease. Meta- bolism 1973; 22:1357-65. Gemsenjager E, Staub JJ, Girard J, Heitz Ph. Preclinical hyperthyroidism in multinodular goitre. J Clin Endocrinol Metab 1976; 43:810-6. Gemsenjager E, Staub JJ, Girard J, Heitz Ph. Die hypophysSre TSH-Reserve in einem

3'' 120 chirurgischen Krankengut von biander Struma und Rezidivstruma. Schweiz Med Wschr 1976; 854-60. Gordin A, Mantela J, Miettinen A, Helenius T, Lamberg BA. Serum thyrotropin and circulating thyroglobulin and thyroid microsomal antibodies in a Finnish population. Acta Endocrinol 1979; 90:33-42. Goslings BM, Djokomoeljanto R, Docter R et al. Hypothyroidism in an area of endemic goitre and cretinism in Central Java, Indonesia. J Clin Endocrinol Metab 1977; 44: 481-90. Greer MA, Grimm Y, Studer H. Qualitative changes in the secretion of thyroid hormones induced by iodine deficiency. Endocrinology 1968; 83:1193-8. I Grégoire F, Brauman H, de Buck R, Corvilain J. Hormone release in depressed patients before and after recovery. Psychoneuroendocrinology 1977; 2:303-12. Griffiths RS, Black EG, Hoffenberg R. Measurement of 3,3',5'-triiodothyronine-(reverse) Tj, with comments on its derivation. Clin Endocrinol 1976; 5:679-8S. Hadden DR, McMaster A, Bell TK, Weaver JA, Montgomery DAD. Does T4-toxicosis Iv - exist? Lancet 1975; 1:754-5. Hall R, Doniach D, Kirkham K, el Kabir D. Ophthalmic Graves' disease. Lancet 1970; 1: 375-8. L Havard CWH. Which test of thyroid function? Brit Med J 1974; 1:553-6. Hennemann G, Docter R, Krenning EP, Bos G, Otten M, Visser TJ. RAised total thyroxine and free thyroxine index but normal free thyroxine. A serum abnormality due to inherited increased affinity of iodothyronines for serum binding protein. Lancet 1979; 1:639-42. Herrmann J, Lehr HJ, Kröll HJ, Rusche HJ, Rudorff KH, Krttskemper HL. Excessive peripheral conversion of thyroxine (T4) to triiodothyronine (T3) in the pathogenesis of Tj-hyperthyroidism. Dtsch med Wschr 1975; 100:2319-23. Hollander CS, Shenkman L, Mitsuma T, Blum M, Kastin AJ, Anderson DG. Hyper- triiodoti yroninaemia as a premonitory manifestation of thyrotoxicosis. Lancet 1971; 2: 731-3. Hollander CS, Mitsuma T, Nihei N, Shenkman L, Burday SL, Blum M. Clinical and laboratory observations in cases of triiodothyronine toxicosis confirmed by radioimmunoassay. Lancet 1972; 1: 609-11. Hollander CS, Stevenson C, Mitsuma T, Pineda G, Shenkman L, Silva E. T3 toxicosis in an iodine-deficient area. Lancet 1972; 2:1276-8. Hooper PL, Caplan RH. Thyroid uptake of radioactive iodine in hyperthyroidism. JAMA 1977; 238:411-3. Inada M, Kasagi K, Kurata S, et al. Estimation of thyroxine and triiodothyronine distribution and of the conversion rate of thyroxine to triiothyronine in man. J Clin Invest 1975; 55: 1337-48. Ivy HK, Wahner HW, Gorman CA. "Triiodothyronine (T3) toxicosis". Its role in Graves' disease. Arch Int Med 1971; 128:529-34. Izumi M, Larsen PR. Triiodothyronine, thyroxine, and iodine in purified thyroglobulin from patients with Graves' disease. J Clin Invest 1977; 59:1105-121. Joasso A. Tt-thyrotoxicosis with normal or low serum T3 concentration. Aust NZ J Med 1975; 5: 432-4. Kahn A. Patterns of primary thyroid failure. Israel J Med Sei 1978; 14:719-24. Kirkegaard C, Siersbaek-Nielsen K, Friis Th, Rogowski P. Does T4 toxicosis exist? Lancet 1975; 1:868. Kirkegaard C, Faber. J, Friis T, Lauridsen ÜB, Rogowski P, Siersbaek-Nielsen K. Intrave- nous and peroral TRH stimulation in sporadic atoxic goitre. Acta Endocrinol 1977; 85: 508-14. Kirkegaard C, 01gaard K, Faber J, Siersbaek-Nielsen K, Friis Th. Thyrotropin-releasing- hormone tests. Lancet 1979; 1:556. Kumar MS, Safa AM, Deodhar SD, Schumacher OP. The relationship of thyroid- \r 121 f ' I f>. stimulating hormone (TSH), thyroxine (T4), and triiodothyronine (T3) in primary thyroid •] ' • fj '& failure. AmJ Clin Pathol 1977; 68:747-51. \ £ |i Larsen PR. Triiodothyronine: review of recent studies of its physiology and pathophysiology | -^ |S in man. Metabolism 1972; 21:1073-92. j | Larsen PR. Thyroidal triiodothyronine and thyroxine in Graves* disease: correlation with • • ' t; presurgical treatment, thyroid status, and iodine content. J Clin Endocrinol Metab ''•--. . '• ['f 1975; 41:1098-1104. 1 f, Laurberg P. Non-parallel variations in the preferential secretion of 3,5,3'-triiodothyronine ;,5 t; (T3) and 3,3',5'-triiodothyronine (rT,) from dog thyroid. Endocrinology 1978; 102: j& 757-66. ' < ' J $ Lawton NF, Ekins RP, Nabarro JDR. Failure of pituitary response to thyrotropin-releasing • t hormone in euthyroid Graves' disease. Lancet 1971; 2:14-6. , ;' Loosen PT, Prange Jr AJ, Wilson JC, Lara PP, Pettus C. Thyroid stimulating hormone response after thyrotropin releasing hormone in depressed, schizophrenic and normal women. Psychoneuroendocrinology 1977; 2:137-48. Maquire SB, Dennehy A, Cullen MJ. The effect of thyrotoxicosis and hypothyroidism on the extrathyroidal conversion of thyroxine to triiodothyronine in man. In: Robbins J, Braverman LE, eds. The Thyroid. Amsterdam: Excerpta Medica, 1976:259-62. ; Marsden P, McKerron CG. Serum triiodothyronine concentration in the diagnosis of . ; | hyperthyroidism. Clin Endocrinol 1975; 4:183-9. ••> Marsden P, Howorth PJN, Chalkley S, Acosta M, Leatherdale B, McKerron CG. Hormonal < pattern of relapse in hyperthyroidism. Lancet 1975; 1:944-7. : Mitsuma T, Nihei N, Gershengorn MC, Hollander CS. Serum triiodothyronine: meas- .-• urements in human serum by radioimmunoassay with corroboration by gas-liquid :•• chromatography. J Clin Invest 1971; 50:2679-88. | Morgans ME, Thompson BD, Whitehouse GS. Sporadic non-toxic goitre: an investigation | of the hypothalamic-pituitary-thyroid axis. Clin Endocrinol 1978; 8:101-8. •.'?, Murie N, Auvray E, Delaborde F, Herry JY. Interet diagnostique du dosage radio-immuno- \ • logique de la triiodothyronine avant et apres TRH (151 patients). Arch Med de l'Ouest : ' 1975:7:287-97. ; Nicoloff JT, Low JC. Dussault JH, Fisher DA. Simultaneous measurement of thyroxine • and triiodothyronine peripheral turnover kinetics in man. J Clin Invest 1972; 51:473-83. ; Nilsson G, Nordlander S, Levin K. Studies on subclinical hypothyroidism with special j, reference to the serum lipid pattern. Acta med Scand 1976; 200:63-7. '• Ormston BJ, Garry R, Cryer RJ, Besser GM, Hall R. Thyrotrophin-releasing hormone as a ; thyroid function test. Lancet 1971; 2:10-4. j '. Ormston BJ, Alexander L, Evered DC et al. Thyrotropin response to thyrotrophin-releasing I \ • hormone in ophthalmic Gravss' disease: correlation with other aspects of thyroid I | function, thyroid suppressibility and activity of eye signs. Clin Endocrinol 1973; 2: ~!\ p 369-76. I ' | Patel YC, Burger HG. Serum triiodothyronine in health and disease. Clin Endocrinol 1973; 2:

123 i ,v • L |l* Wenzel KW, Meinhold H, Raffenberg M, Adikofer F, Schleusener H. Classification of hypo- ' I< 'Cz. thyroidisminevaluatingpatientsafterradioiodinetherapybyserumcholesterol,T3-uptake, -. fe |$ total T4, FT4-index, total T,, basal TSH and TRH-test. Europ J Clin Invest 1974: 4: £ II 141-148. ' ; § r'v Wenzel KW, Meinhold H.Triiodothyronine/reverse-triiodothyronine balanceandthyroxine ft •?,' metabolism. Lancet 1975; 2:413. v p I":' Werner SC. Hyperthyroidism. Introduction. In: Werner SC, Ingbar SH, eds. The Thyroid, a : |? i> fundamental and clinical text. 4th edition. Hagerstown: Harper and Row, 1978; 591-603. '':} |. Wilkin TJ, Storey BE, Isles TE, Crooks J, Swanson Beck J. High TSH concentrations in 'fi. y: "euthyroidism": explanation based on control-loop theory. Brit Med J 1977; 1:993-6. I f P Wu SY, Chopra IJ, Solomon DH, Bennett LR. Changes in circulating iodothyronines in ' H; i euthyroid and hyperthyroid subjects given ipodate (Oragrafin), an agent for oral . |' cholecystography. J Clin Endocrinol Metab 1978; 46:691 -7. j & Zurcher H, Sakoloff C, Burger A, Vallotton MB. Importance clinique de la triiodothyronine "t serique. Schweiz Med Wschr 1973; 103:1710-3. < I

1 'I

i ! • •'' f. &'• IK I ?;<•• CHAPTER V. i%t %-: STUDIES IN PATIENTS WITH ACUTE NON-THYROIDAL II- DISEASES §: V.I. Introduction. [< 7

In a comatose patient, a case of self-intoxication with dextromoramide (Palfium), plasma T3 was found to be below the detection limit of the assay, but plasma T4 and TSH values were normal. After recovery the plasma T3 level returned into the normal range. In another patient suffering from acute leukemia with a rapid downhill course, plasma T3 decreased progressively; it was speculated that the time of death would coincide with the time at which plasma T3 would be undetectable. The date of death was estimated as april 12th, the patient died on april 1 lth; plasma T3 was undetectable in a blood sample drawn on april 10th (figure V.I). These observations raised the following questions: a. What is the incidence and what is the cause of a low plasma T3 concentration in patients with acute non-thyroidal diseases? b. Is there any relation between the plasma T3 concentration and the severity of the disease? c. Is a low plasma T3 level in these patients of any relevance in the set- point of the pituitary-thyroid axis? In an attempt to answer these questions a prospective clinical study was set up.

V.2. Patients and methods.

Patients with an acute non-thyroidal illness in whom a complete recovery was expected within two weeks, were studied. From the 28 initial patients eight were excluded from the study: 4 patients could not be observed for 2 weeks, and 4 were treated during the study period with drugs that may interfere with thyroid function tests (radiographic contrast agents, systemic heparin, diphenlhydantoin, heroin and methadone). The remaining 20 patients were evaluated clinically after two weeks, and | divided into four groups; group A (n=13) - full recovery, group B (n= 4) — improvement, but no full recovery, group C (n = 1) — no change in the clinical condition, group D (n = 2) — deceased (see table V.I.). T.:ST*

125

5 10 Marctl

Figure V.I. Plasma T3 concentrations (• +) in a 63-years old male with acute leukemia, who died at april 1 lth; the estimated time of death (by extrapolation of the two first plasma

T3 values, (•—-•) was april 12th.

Table V-l Sex, age and diagnosis of 20 patients with an acute non-thyroidal disease. group nr. sex age(yis) diagnosis

- full recovery in 2 weeks (n = 13) 1 F 15 infectious mononucleosis 2 M 15 meningitis 3 F 16 ihinophaiyngitis, meningism 4 M 21 chickenpox 5 M 33 toxicodermia 6 M 39 fever, polyarthralgia and rash of unknown origin 7 M 40 self-intoxication 8 M 45 psittacosis 9 F 53 pyelonephritis 10 M 58 tonsillitis, pharyngitis 11 M 62 biliary colics 12 F 65 pneumonia 13 M 67 pneumonia

- improvement, but no full recovery in 2 weeks (n = 4) 14 M 27 meningitis 15 M 30 amebic abcess of liver 16 F 35 meningitis 17 M 38 active pulmonary tuberculosis

— no change in 2 weeks (n = 1) 1 18 F 47 sarcoidosis

- deceased within 2 weeks ( n = 2) 19 M 81 acute myocardial infarction 20 M 72 progressive obstructive jaundice due to carcinoma of pancreas Table V-2 Thyroid function tests in 16 patients admitted because of acute non-thyroidal disease: group A (n = 12) - full clinical recovery in 2 weeks, group B (n = 4) - clinical improvement but no full recovery in 2 weeks. (Values as mean ± SD, except plasma TSH and iodothyronine ratio's which are presented as the geometric mean ± SD).

group day 1 2 3 4 5 6 i TSH mU/1 A 0.9 (0.4 - 2.1 ) 0.9 (0.3 - 2.8 ) 1.4 (0.6 - 3.1 ) 1.5 (0.9 - 2.4 ) 1.2 (0.7 - 2.1 ) 1.2 (0.7 - 2.0) B 2.6 (1.5 7- 4.5 ) 1.3 (0.6 - 2.8 ) 1.2 (0.5 - 3.1 ) 1.1 (0.5 - 2.8 ) 1.2 (0.5 - 3.2 ) 1.1.(0.3 - 4.1 ) ft} rTjiimol/1 A 0.42 ± 0.18 0.43 ± 0.17 0.42 ± 0.20 0.39 ± 0.17 0.34 ±0.12 0.33 ± 0.07 B 0.56 ± 0.20 0.54 ± 0.21 0.54 ± 0.30 0.58 ± 0.36 0.54 ±0.31 0.57 ± 0.33 Ts nmol/1 A 1.01 ± 0.33 1.06 ± 0.38 1.31 ± 0.54 1.48 ± 0.43 1.44 ± 0.37 1.42 ± 0.42 B 0.85 ± 0.46 0.94 ± 0.66 0.84 ± 0.48 1.02 ±0.54 0.91 ± 0.34 0.78 ± 0.41 4

T4nmol/1 A 87 ±22 86 ±22 981 18 106 ±19 102 ± 16 105 ± 20 B 85 ±28 96 ±39 88 ±26 96 ±36 94 ±26 90 ±22

TSU A 0.97 ±0.13 0.94 ±0.12 0.92 ±0.12 0.93 ±0.11 0.92 ± 0.10 0.93 ±0.11 B 1.04 ±0.19 1.02 ± 0.16 0.99 ±0.12 0.98 ±0.13 1.00 ±0.12 0.99 ± 0.12

FT4 index A 84 ±24 80 t 18 90 ± 17 99 ±21 94 ±17 99 ±26 B 85 ±23 96 ±41 86 ±26 94 ±42 93 ±28 88 ±25

rT,/Taxl0-' A 412(225 - 753 ) 403(208 - 781 ) 330(146 -•745 ) 261(150 - 452 ) 233(137 - 396 ) 236(147 - 378 ) B 788(326 - 1903 ) 731(247 - 2168) 696(251 - 1931) 540(204 - 1426) 564(250 - 1271) 742(276 - 1997) J rT,/T4xl0- A 4.61(3.01 - 7.05 4.76( 3.15 - 7.18) 4.01( 2.61 - 6.15) 3.53( 2.50- 4.99) 3.22( 2.27 - 4.58) 3.11( 2.41- 4.02) B 6.24(3.73 - 10.45 5.62( 3.30- 9.57) 5.59( 3.27 - 9.53) 5.40( 2.92 - 9.97) 5.30( 2.96 - 9.50) 5.70( 3.69 - 8.81)

Ts/T4xl0-' A 11.2 (7.9 - 15.8 ,> 11.8 ( 8.4 - 16.6 ) 12.2 ( 7.5 - 19.7 ) 13.6 ( 9.7 - 19.0 ) 13.8 (10.4 - 18.4 ) 13.2 ( 9.9 - 17.5 ) B 8.4 (5.0 - 14.0 ' 7.7 ( 4.1 - 14.4 ) 8.0 ( 4.5 - 14,2 ) 10.0 ( 7.0 - 14.3 ) 9.4 ( 7.4 - 11.9 ) 7.7 ( 4.2 - 14.1 )

g?^^^ 7 10 14 ±20 ±30 ±40 normal values 1.3 (0.8 - 2.0) 1.2 (0.8 - 2.0) 1.3 (0.8 - 2.1) 1.8 (1.0 - 3.2) 2.0 (1.0 - 3.9) 1.9 (0.7 - 5.6) 2.1 (1.0 - 4.6) 1.3 (0.3 - 6.0) 1.5 (0.5 - 4.8) 0.7 (<0.3 - 2.3) 0.29 ± 0.08 0.27 ± 0.07 0.27 ± 0.10 0.51 ± 0.30 0.50 ± 0.33 0.46 ± 0.23 0.30 ± 0.15 0.30 ± 0.11 0.34 ± 0.18 0.24 ± 0.07 1.50 ±0.45 1.56*0.47 1.63 ± 0.39 0.91 ± 0.37 0.84 ± 0.39 1.15 ± 0.45 1.21 ±0.27 1.95 ± 0.76 1.96 ±0.52 1.85 ± 0.37 104 ± 22 103 ± 20 106 ± 17 98 ±34 85*20 104 ±29 108 ± 32 106 ± 31 110 ±33 102 ±18 0.95 ± 0.10 0.92 ± 0.13 0.94 ±0.12 0.99 ±0.13 1.02 ± 0.15 0.99 ± 0.14 0.9S ± 0.10 0.95 ± 0.16 0.95 ± 0.14 0.94 ± 0.09 99 ±25 94 ±23 99 ±20 97 ±37 87 ±28 102 ± 27 104 ±39 101 ± 33 106 ±46 96 ±15 196(113 - 340) 172(104 - 285) 157( 95 - 258) 529(242 - 1155) 571(223 -1464) 391(193 - 793) 2*3(110 - 453) 157( 77 - 317) 162( 92 - 286) 124( 86 - 179) 2.74(2.00- 3.75) 2.54(1.78 - 3.61) 2.38(1.70 - 3.33) i 3 4.82(3.39 - 6.84) 5.29(2.95 - 9.50) 4.15(2.68 - 6.42) 2.55(1.94 - 3.34) 2.79(2.08 - 3.73) 2.93(2.33 - 3.69) 2.23(1.62 - 3.08) 14.0 (10.1 - 19.3) 14.8 (11.0 -19.8) 15.1( 11.8 - 19.5) 9.1 ( 5.9 - 14.2) 9.3 ( 6.3 - 13.7) 10.6( 8.0 - 14.1) 11.4 (7.1 - 18.4) 17.8 (11.1 - 28.4) 18.1 (12.1 - 27.1) 18.0 (14.9 - 21.9) 128 Blood samples were taken on days 1,2,3,4,5,6,7,10 and 14; day 1 was the day of admission. In the patients of group B several blood samples were also taken after two weeks. In each plasma sample the following $ I determinations were done: TSH, rT3, T3, T4 and T3U. The competitive protein binding method of Murphy and Pattee was used for the assay of T4. Stimulation tests with 200 jug TRH i.v. were performed as described in chapter III.2 on days 1 and 14, if feasible. All samples of one patient were assayed in the same run in order to avoid inter-assay variation. Statistical evaluation of the trend in hormone levels and of the TSH- and I: T3-responses to TRH stimulation was performed by analysis of variance (Bennett and Franklin 19S4). Because of their skewed distribution, plasma TSH levels and the ratio's of plasma iodothyronine concentrations were analyzed after logarithmic transformation. In the statistical analyses the level of significance was taken as a= 0.0S; undetectable hormone concentrations were taken as SO percent of the detection limit.

V.3. Results.

The sequence of plasma T4, T3 and rT3 concentrations in patient nr 12 is depicted in figure V.2. Plasma TSH was undetectable in this patient, and the TSH- and T3-response to TRH were absent resp. subnormal both at day 1 and at day 14. Subsequently hyperthyroidism was diagnosed, and she responded favourably to antithyroid drug treatment. The data of this patient were excluded from the further statistical evaluation. On admission plasma T3 was subnormal (i.e. below the percentile P2.s of the normal values) in 16 out of 19 patients (84%), and plasma rT3 was supranormal (i.e. above the percentile P97.5) in 12/19 patients (63%);

14 days

Figure V.2. Plasma T4,T3 (solid lines) and rT3 (dashed line) concentrations in a 65-years old female admitted because of pneumonia. The biochemical pattern of T4-toxicosis (day 1) 1 evolved to 'classical Tr and T4-thyrotoxicosis at the time the intercurrent pneumonia had subsided (days 10 and 14). 129

Table V-3 I Thyroid function tests in two acutely ill patients, who died within two weeks after admission (group D).

patient D, patient Da on admission - last sample on admission - last sample

IS TSH mU/1 1.3 0.5 0.4 < 0.3 rT3 nmol/1 0.68 1.10 1.44 3.15 in T3 nmol/1 1.11 0.59 0.86 0.68 T4 nmol/1 104 89 143 137 M T,U 0.91 0.96 0.99 1.26 >1 FT« index 95 85 142 173 s rT3/T,xl0- 613 1864 1674 4632 3 rT,/T4xl0- 6.54 12.36 10.07 22.99 T,/T4xl0-» 10.67 6.63 6.01 4.96 peak TSH mU/1 11.6 - 4.4 - peak T3 nmol/1 1.68 — 1.42 — i;

plasma ,T4 and the FT4 index were subnormal in 6/19 patients (32%). The results of the thyroid function tests in the patients of groups A and B are presented in table V.2. In the patients of group A plasma T3 (p< 0.001), T4 (p < 0.001) and the FT4 index (p < 0.01) increased, whereas rT3 (p < 0.01) and T3U (p < O.OS) decreased; plasma TSH did not change (p > 0.1). In contrast, no change in these parameters was found in the first two

3.0 r

10 2.0

1.0

01-

-30 0 20 60 120 180 240 -30 0 20 60 120 180 240 min min

Figure V.3. The TSH- and T,-response to 200//g TRH i.v. in 16 patients with acute non- thyroidal diseases on admission (solid lines) and two weeks later (interrupted lines) (values as mean ± SD).

t'7 130

The TSH- and Ta-response to 200 fig TRH i.v. in 15 patients with acute non-thyroidal disease on admission (~ day 1) and appr. two weeks later (~ day 14) after full clinical recovery (group A, n = 11) or when clinically improved but not fully recovered (group B, n = 4). ii-' (TSH values are presented as the geometric mean ± SD, T3 values as x ± SD with in brackets the range of the relative increment). •:•'• i group day of basal value peak value absolute relative L~; • 'if'' TRH- increment increment test I TSH mU/1 A 1 0.9(0.4 - 2.2) 5.5(3.1 - 10.1) 4.5(2.5- 8.1) 14 1.3(0.8 - 2.1) 6.2(4.3- 9.0) 4.8(3.2- 7.2)

B 1 2.0(0.7 - 5.7) 8.5(4.3 - 16.6) 6.3(3.4 - 11.6) I', 14 1.9(0.7 - 5.6) 10.7(5.6 - 20.7) 8.5(4.4 - 16.1)

T3nmol/1 A 1 1.02 ± 0.34 1.55 ± 0.58 0.53 ±0.28 52(21- 80) 14 1.59 ± 0.40 2.29 ± 0.65 0.71 ± 0.29 44(20- 62) B 1 1.01 ± 0.66 1.68 ± 0.79 0.68 ± 0.32 111(29 - 247; 14 1.15 ± 0.45 1.97 ± 0.53 0.82 ±0.22 83(41 - 140)

weeks in group B. However, when the assay results of the subsequent weeks were also analyzed, plasma T3 (p < 0.01), T4 (p < 0.01) and the FT4 index (p< 0.01) increased, whereas rT3 decreased (p< 0.02); T3U and TSH did not change. Also, the ratio's rT3/T3 and rT3/T4 decreased (p < 0.01) and T3/T4 increased (p< 0.001) in group A, but did not change in the first two weeks in group B. The results of the thyroid function tests in the patient whose clinical condition remained unchanged (group C), were almost the same at day 1 and 14: TSH 1.3 vs. 1.2 mU/1, rT3 0.28 vs 0.24 nmol/1, T3 1.29 vs. 1.23 nmol/1, T4 68 vs. 67 nmol/1, T3U 1.02 vs. 1.13, FT4 index 69 vs. 75. In the s two patients who died (group D) a progressive decrease in plasma T3 and increase in plasma rT3 was observed (table V.3). The TSH- and T3-response to TRH on admission and appr. two weeks later of 16 patients (nrs. 1-5, 7-11, 13-18) are depicted in figure V.3. The TSH- and T3-responses were normal on both occasions, and no difference was observed between day 1 and day 14 (although T3 levels were higher on day 14). Evaluation of the responses in group A and B separately also failed to reveal a difference between day 1 and 14 (table V.4.). In the one patient of group C the peak TSH values after TRH were 6.9 and 6.3 mU/1, and the peakT3 values 1.60 and 1.63 nmol/1 on days 1 and 14 respectively.

I- ," -»'••'

131 D V.4. Discussion.

|f The high incidence of a subnormal plasmaTy (84%) and a supranormal p rT3 level (63%) in our patients with non-thyroidal diseases is in agreement ft with a reported incidence of 71,59, 80 and 77% for low T3 values and of ? % 53% for high revalues (Bermudez etal 1975, Burrows etal 1975, Lim etal I; 1977, and Chopra et al 1979). Thus it appears that subnormal T3 levels, jfe often in combination with supranormal revalues, are a common finding ; g; |1 in patients with acute or chronic non-thyrpidal disease (Carter etal 1974; , ;> ft | Chopra et al 1974; Bermudez et al 1975; Burrows et al 1975; Chopra et al p 1975, Green etal 1975; Hflfner and Gless 1975; Nomura etal 1975;Burgeret 5 al 1976; Gavin et al 1976; Jarnerot et al 1976; Ramirez et al 1976; \ Brinckmeyer et al 1977; Burrows et al 1977; Finucane et al 1977; Lim et al \ I977;Ljunggrenetall977;Rudmanetall977;Talwaretall977;Wartofsky : f et al 1977; Kolendorf et al 1978; Maharajan et al 1978; Naeye et al 1978; \ \ Ratcliffe etal 1978; Saunders etal 1978; Afrasiabi etal 1979; Helenius and i j Lieuwendahl 1979; Pittman et al 1979; Rose and Davis 1979; Weissel et al 1979).Thenatureofthediseaseseemstobeirrelevantforthedevelopmentof j the 'low T3 syndrome', since subnormal T3 values have been observed in ! almost every disease so far studied. The nonspecific nature of the reciprocal changes in plasma T3 and rT3 is further indicated by similar alterations in the plasma iodothyronines after surgical stress (Burr et al 1975; Adami et al 1978; Chan etal 1978; Hagenfeldtetal 1979; Kehletetal 1979; Prescott etal : 1979). !

The reciprocal changes in plasma T3 and rT3 strongly suggest that they I are caused by modulation of the peripheral conversion of TA. Kinetic ; studies in patients with hepatic cirrhosis (Nomura et al 1975; Chopra 1976; \ Lumholtz et al 1978), renal failure (Lim et al 1977), or diabetes (Pittman et al 1979) indicate impaired production rates (but normal metabolic ; clearance rates) of T3 and normal or slightly increased production rates ) I (but decreased metabolic clearance rates) ofrT3, whereas the production of | : T4 is equal to or slightly less than that found in control subjects.The fall in ; | [ plasma T3 and rise in plasma rT3 can thus be explained by impairment of i the T4-»T3 and rT3*3,3'-T2 deiodinative process due to inhibition of 5'- j • | deiodination. It has been speculated that the decrease in 5'-deiodinating \ f activity is related to the stress-induced rise in catecholamines and • \ especially in cortisol secretion, since corticosteroid medication in man also :': [ results in low T3 and high rT3 plasma concentrations (Chopra et al 1975). A [ However, no relation has been found between plasma cortisol and thyroid ^ I hormone levels after surgery (Kirby et al 1973; Prescott et al 1979), and | j inhibition of the glucose and cortisol response to hysterectomy by % blockade of afferent pathways with spinal anaesthesia (Madsen et al 1977) % 132 fcl. jV, 'IS 2.00 I 1.50

1.25 1.00 sr •,1.00 L0.75

0.75 0.50

0.50 0.25 L

14 14 days days

Figure V.4. Mean plasma T3 and rT3 concentrations in 18 patients with acute non-thyroidal diseases, who after admission (day 1) made a full (group A, n=12) or incomplete (group B, n=4) clinical recovery in 2 weeks, or who died (group D, n=2). I did not prevent the postoperative decrease in T3 (Brandt et al 1976). As to the catecholamines, it is difficult to reconcile a possible direct inhibition of the T4-»T3 conversion by catecholamines with the similar inhibitory effect of propranolol (cf. chapters VII and VIII). Glucose metabolism in the hexose-monophosphate shunt is decreased during illness, resulting in a lower concentration of NADPH, which is the cofactor in the reduction of glutathione. Visser (1978) has put forward the hypothesis that the modulation of the conversion of T4 during illness is related to a lower intracellular reduced glutathione concentration, since reduced glutathione probably serves as a cofactor for the S'-deiodinating enzyme. Fasting per se results in similar changes, however, not all acute or chronically ill patients reduce their dietary intake. The biological significance of the low T3 syndrome in non-thyroidal disease is unknown, but it may be speculated that the low T3 syndrome fits into the general adaptation syndrome in which the body seeks to diminish the catabolic effects of stress (see further general discussion, chapter IX).

In our patients, plasma T3 increased and rT3 decreased concomitantly with the improvement in clinical condition; in fact, these changes occurred also within the normal range, and were observed in every patient. The same sequence of events has been observed by others in (sometimes experimen- tally induced) non-thyroidal disease and after surgery (Burr et al 1975; Burger et al 1976; Talwar et al 1977; Wartofsky et al 1977; Adami et al 1978; Chan etal 1978; Naeyeetal 1978; Maharajanetal 1978; Saundersetal 1978; Hagenfeldt et al 1979; Kehlet et al 1979; Prescott et al 1979). A relation 133 v,'v. between the plasma T3 concentration and the prognosis in non-thyroidal disease is strongly suggested by the rather slow return of T3 and rT3 values into the normal range if clinical recovery was delayed, and by the further decrease in T3 and increase in rT3 in the patients who died (figure V.4). A pre- terminal fall in plasma T3 is also reported by McLarty et al in 1975 and by Maharajan et al in 1978. It has been our experience that patients are especially at risk for an unfavourable course of their disease if plasma T3 decreases below 1 nmol/1 and plasma rT, increases above 1 nmol/1.

A relation between the severity of a disease and plasma T3 is also suggested by other studies. The fall in plasma T3 is related to the degree of renal failure (Ramirez et al 1976; Finucane et al 1977; Kolendorfet al 1978; Weissel et al 1979), to the degree of liver damage as estimated by the plasma albumin concentration (Nomura et al 1975; Green et al 1977), and to the degree of fever in patients with infectious diseases (Ljunggren et al 1977; Maharajan et al 1978). Patients with malignant lymphoma stage IV B have lower plasma T, concentrations than patients in stage I-IV A, and the prognosis of patients with lung cancer is worse if they have a subnormal T3 (Ratcliffeetal 1978). A more quantitative approach relating plasma T3 and rT3 levels to the severity of a disease, is offered in the next chapter.

The set-point of the pituitary-thyroid axis in the 'low T3 syndrome' is of considerable interest. The low concentration of plasma T3 is usually associated with a low free T3 concentration (Chopra et al 1974; Chopra et al 1975; Talwar et al 1977; Chan et al 1978; Kehlet et al 1979; Prescott et al 1979), and since even small decreases in plasma thyroid hormone concentrations are followed by hyperresponsiveness of TSH to TRH stimulation (Vagenakis et al 1974) one would expect an increase in basal and stimulated TSH values in these patients. As our results indicate, this is not so: despite substantial decreases in plasma T3 concentration the TSH levels are not elevated, nor is the TS H- and T3-response different on day 1 and day 14, and in this respect the patients of group A and B behave similarly. Apparently the pituitary thy rotrophic cells do not react to the low plasma T3 levels: the pituitary-thyroid axis functions as if a euthyroid state exists. A normal increase in T3 levels in patients with non-thyroidal disease in response to either exogenous TSH or a TRH-induced rise in endogenous TSH, has been reported by others (Carter et al 1974; Chopra et al 1974; Carter et al 1976; Talwar et al 1977; Weissel et al 1979), and also a normal TSH-response to TRH is found by most authors. Therefore patients with the low T3 syndrome are often called 'sick euthyroid patients', also because they do not present with signs or symptoms of hypothyroidism. Occasion- ally however slightly elevated basal TSH values and increased TSH responses to TRH are reported, especially in liver disease (Chopra et al 1974; Nomura et al 1975; Bermudez et al 1975; Green et al 1977; Rose and Davis 1979), whereas a blunted TSH release has also been observed, especially in f:

St: 134 ? renal failure (Lim et al 1976; Ramirez et al 1976; Hooper 1976; Naeye et al •) 1978; Weissel et al 1979). It thus cannot be denied that in the low T3 syndrome adaptations at the hypothalamic/ pituitary level may occur. However, the magnitude of the adaptation of thyroid hormone metabolism ; p tostressfullconditionsseemstobemuchgreaterintheperipheraltissues.lt : , l£ is noteworthy that patients dying of a chronic wasting illness show a striking j fe; reduction of the concentration of T3 in the heart, liver and kidney, together j || with a relative increase of T4 in these tissues (Reichlin etal 1973). Thus, there E; seems to be a discrepancy between some sort of'hypothyroidism'at the level j of the peripheral tissues, and the 'euthyroidism' at the central hypothalamic- ' pituitary-thyroid level. Ideologically, the low T3 concentration could help to mitigate the excessive catabolism in the peripheral tissues; in this situation a stimulation of TSH release and thereby of thyroid hormone secretion would serve no useful purpose. It is unlikely that in the low T3 , | syndrome the absence of an increase in TSH release is effectuated by a | suppressive effect of the high rT3 concentrations (Nicod etal 1976). The ro/e | ! { ofthyroxine seems to be crucial in this respect. ! i : Plasma T4 was subnormal in one third of our acutely ill patients, which is in agreement with reports of others (Carter et al 1974; Chopra et al 1974, ^ 197S; Burger et al 1976; Adami et al 1978; Naeye et al 1978; Saunders et al | 1978; Ratcliffe et al 1978; Kehlet et al 1979; Chopra et al 1979). A major | determinant of the fall in plasma T4 induced by illness or surgery, is a •', decrease in binding capacity of the plasma proteins (especially of TBP A) for .; thyroid hormones, resulting in an increase of the T3U values (Surks and : Oppenheimer 1964; Bellabarba et al 1968; Lutz et al 1972; Kirby et al 1973; Bermudez et al 1975; Talwar et al 1977; Adami et al 1978; Naeye et al 1978; Chopra et al 1979; Helenius and Liewendahl 1979). It has been argued that : unknown substances circulate in the plasma of these patients that may ; inhibit the binding of T4 to the plasma proteins (Chopra et al 1979; Helenius • and Liewendahl 1979). This would explain why the FT4 concentrations (as : measured directly by dialyses) is normal or even elevated in 'sick euthyroid \ patients' (Burrows et al 1977; Brinckmeyer et al 1977; Talwar et al 1977; ; Chopra etal 1979; Kehlet etal 1979; Prescott et all 979). In this situation the h FT4 index—as calculated from T4 and T3U — is often low and is, therefore, JJ \ not a reliable parameter. ' - Moderately elevated T4 values have also been reported, especially in I elderly patients (Burrows et al 1975,1977; Wartofsky et al 1977; Chan et al >] 1978; Maharajan et al 1978; Ratcliffe et al 1978; Prescott et al 1979). One 2 then finds a pattern of T4-toxicosis\ The reason why in some elderly qj patients an increase in T4 is found is not clear, but may well be due to an j| impairment of the deiodination of T4. c( The elegant studies of Silva and Larsen have recently clarified some i I aspects of the feedback control of thyroid hormones at the pituitary level '\ I 135 t i. - 50 £ X 40 S

to*

Figure V.5. Plasma T4, T3 and rT3 in a 74-years old female, who died 7 days after admission because of progressive cardiac failure due to mitral insufficiency. Hypothyroidism was recognised after death (TSH values in all samples greater than 100 mU/1). Note the absence of a pre-terminal rise in rT3, whereas T3 continuously decreases.

(SilvaandLarsen 1977,1978;Silvaetal 1978,1978; Larsenetal 1979). They demonstrated in the rat: a) that there is an excellent correlation between the nuclear T3 content of the pituitary thyrotrophs and the release of TSH, b) that the anterior pituitary converts T4 into T3, c) that pituitary nuclear T3 is derived from intrapituitary T4 monodeiodination and also directly from plasma T3, d) that local T4 monodeiodination contributes relatively more to the nuclear T3 than plasma T3 in the pituitary, as compared with liver and kidney, and e) that intrapituitary T4-*T3 conversion is blocked by iopanoic acid but not by propylthiouracil, whereas both agents inhibit T4-iT3 conversion in liver and kidney. Thus the pituitary thyrotroph and the peripheral tissues can respond differently with regard to the conversion of T4, and it may well be that the capability of the pituitary thyrotroph to derive its nuclear T3 from local T4 conversion to a much greater extent than the peripheral tissues, answers the question of 'central euthyroidism' and 'peripheral hypothyroidism' in non-thyroidal illness. It underlines the uncertainty with regard to the real metabolic situation at the cellular level, which may not always be reflected correctly by the levels of the circulating thyroid hormones; the tissues may locally regulate the production of T3 according to their need (Obregon et al 1979).

The non-specific changes in plasma iodothy ronine concentrations caused by non-thyroidal disease compromise the usefulness of plasma assays in the diagnosis of thyroid disease in a sick patient. If T4 and FT4 index values are normal, thyroid disease is probably absent. If they are abnormally low, the finding of a high rT3 concentration in plasma indicates euthyroidism rather "•i than hypothyroidism (see fig. V.5). Finally, it can be concluded that even in sick patients the TRH test maintains its great diagnostic value as witnessed by the responses seen in all our acutely ill patients. 136 V.5. References

Adami H O, Johansson H, Thoren L, Wide L, Akerstrom G. Serum levels of TSH, rT3, T4 and T,-resin uptake in surgical trauma. Acta Endocrinol 1978; 88: 482-9. Afrasiabi M A, Vaziri N D, Gwinup G et al. Thyroid function studies in the nephrotic syndrome. Ann Int Med 1979; 90:335-8. Bellabarba D, Inada M, Varsano-Aharon N, Sterling K. Thyroxine transport and turnover in major nonthyroidal illness. J Clin Endocrinol Metab 1968; 28:1023-30. ft Bennett C A, Franklin N L. Statistical analysis in chemistry and the chemical industry. New York: John Wiley and Sons, 1954. Bermudez F, Surks M I, Oppenheimer J H. High incidence of decreased serum triiodothyro- I nine concentration in patients with nonthyroidal disease. J Clin Endocrinol Metab 1975;

41:27-40. .. J . Brandt M R, Kehlet H, Skovsted L, Hansen J M. Rapid decrease in plasma tniodothyronine during surgery and epidural analgesia independent of afferent neurogenic stimuli and of cortisol. Lancet 1976; 2:1333-6. Brinckmeijer L M, Woem A M, Nissen N I. Thyroid function in malignant lymphoma. Acta Med Scand 1977; 202: 275-80. . Burger A, NicodP, SuterP, VallottonMB, VagenakisA, BravermanL. Reduced active thyroid hormone levels in acute illness. Lancet 1976; 1: 653-5. Burr W A, Griffiths R S, Black E G, Hoffenberg R, Meinhold H, Wenzel K. W. Serum triiodothyronine and reverse triiodothyronine concentrations after surgical operation. Lancet

BurrowsYw, Shakespear R A, Hesch RD, Cooper E, AickinCM, BurkeCW. Thyroid

hormones in the elderly sick: "T4 euthyroidism". Brit Med J 1975; 437-9. Burrows AW, Cooper E, Shakespear R A, et al. Low serum L-T3 levels in the elderly sick: protein binding, thyroid and pituitary responsiveness and reverse T3 concentrations. Clin Endocrinol 1977; 7: 289-300. Carter J N, Eastman C J, Corcoran J M, Lazarus L. Effect of severe, chronic illness on thyroid function. Lancet 1974; 2: 971-4. Carter J N, Eastman C J, Corcoran J M, Lazarus L. Inhibition of conversion of thyroxine to triiodothyronine in patients with severe chronic illness. Clin Endocrinol 1976; 5:587-94. Chan V, Wang C, Yeung R TT. Pituitary-thyroid responses to surgical stress. Acta Endocri- nol 1978; 88: 490-8. Chopra IJ, Solomon D H. Chopra U, Young R T, Chua Teco GN. Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhosis: evidence for euthyroidism despite subnormal serum triiodothyronine. J Clin Endocrinol Metab 1974; 39: 510-11. Chopra I J, WilliamsDE, Orgiazzi J, SolomonDH. Opposite effects of dexamethasoneon serum concentrations of 3,3',5'-triiodothyronine (reverse T3) and 3,3',5-triiodothyronine (T3). J Clin Endocrinol Metab 1975; 41: 911-20. Chopra I J, Chopra U, Smith S R, Reza M, Solomon D H. Reciprocal changes in serum concentrations of 3,3',5'-triiodothyronine (reverse Tj) and 3,3',5-triiodothyronine (T3) in systemic illnesses. J Clin Endocrinol Metab 1975; 41: 1043-9. Chopra I J. An assessment of daily production and significance of thyroidal secretion of 3,3',5'-triiodothyronine (reverse T3) in man. J Clin Invest 1976; 58: 32-40. Chopra I J, SolomonDH, HepnerGW, MorgensteinAA. Misleading low free thyroxine index and usefulness of reverse triiodothyronine measurement in nonthyroidal illnesses. Ann Int Med 1979; 90: 915-12. FinucaneJF, GriffithsRS, BlackEG, HallCL. Effects of chronic renal disease on thyroid hormone metabolism. Acta Endocrinol 1977; 84: 750-8. Gavin L A, McMahon F A, Castle J N, Cavalieri R R. Alterations in serum thyroid hormones and thyroxine-bindingglobulin in patients with nephrosis. J Clin Endocrinol Metab 1978; 46: 125-30. 137 '; i GreenJRB, SnitcherEJ, MowatNAG, EkinsRP, ReesLH, DawsonAM. Thyroid { ,i •i.v. function and thyroid regulation in euthyroid men with chronic liver disease: evidence f. of multiple abnormalities. Clin Endocrinol 1977; 7: 453-61. , | %': Hagenfeldt I, Melander A, Thorell J, Tibblin S, Westgren U. Active and inactive thyroid hor- : • » ;f; mone levels in elective and acute surgery. Acta Chir Scand 1979; 145:77-82. .. % '•: Helenius T, Liewendahl K. Abnormal thyroid function tests in severe non-thyroidal illness: j ,, | f. Diagnosis and pathophysiologic aspects. Scand J Clin Invest 1979; 39:389-97. ", i %'} Hooper M J. Diminished TSH secretion during acute non-thyroidal illness in untreated I :| primary hypothyroidism. Lancet 1976; 1:48-9. ' I ^- Hiifner M, Gless K H. Die euthyreote hypotriiodothyroninamie. Schweiz Med Wschr 1975; f\ 105: 1698-1701. Jarnerot G, Kagedal B, Schenk H von, Truelove S C. The thyroid in ulcerative colitis and Crohns disease. V. Triiodothyronine. Effect of corticosteroids and influence of severe disease. Acta Med Scand 1976; 199: 229-32. Kehlet H, Klauber P V, Weeke J. Thyrotropin, free and total triiodothyronine, and thyroxine in serum during surgery. Clin Endocrinol 1979; 10: 131-6. if Kirby R, Clark F, Johnston I DA. The effect of surgical operation of moderate severity on i thyroid function. Clin Endocrinol 1973; 2: 89-99. I Kolendorf K, Moller B B, Rogowski P. The influence of chronic renal failure on serum and . ', | urinary thyroid hormone levels. Acta Endocrinol 1978; 89: 80-8. , I >! LarsenPR, Dick T E, Markovitz B P, Kaplan MM, GardTG. Inhibition of intrapituitary / \\ thyroxine to 3,5,3'-triiodothyronine conversion prevents the acute suppression of ' thyrotropin release by thyroxine in hypothyroid rats. J Clin Invest 1979; 64:117-28. Lim V S, Fang V S, Katz A S, Refetoff S. Thyroid dysfunction in chronic renal failure. A study of the pituitary-thyroid axis and peripheral turnover kinetics of thyroxine and triiodothyronine. J Clin Invest 1977; 60: 522-34. Ljunggren J G, Kallner G, Try selius M. The effect of body temperature on thyroid hormone- levels in patients with non-thyroidal illness. Acta Med Scand 1977; 202: 459-62. Lumholtz IB, Faber J, Buch Sarensen M, Kirkegaard C, Siersbaek-Nielsen K, FriisTh. Peri- pheral metabolism of T4, T3, reverse T3,3',5'-diiodothyronine and 3,3'-diiodothyroninein liver cirrhosis. Horm Metab Res 1978; 10: 566-7. Lutz J H, Gregerman R I, Spaulding S W, Hornick R B, Dawkins A T Jr. Thyroxine binding proteins, free thyroxine and thyroxine turnover interrelationships during acute infectious illness in man. J. Clin Endocrinol Metab 1972; 35: 230-49. Maharajan G, Etta K M, Singh A, Ahuja IS, Ahuja G K. Thyroxine, triiodothyronine and thyrotrophin levels in meningococcal meningitis, typhoid fever and other febrile conditions. Clin Endocrinol 1978; 9: 401-6. Madsen S N, Brandt M R, Engquist A, Badawi I, Kehlet H. Inhibition of plasma cyclic AMP, glucose and cortisol response to surgery by epidural analgesia. Brit J Surgery 1977; 64: 669-71. McLarty D G, Ratcliffe W A, McColl K, Stone D, Ratcliffe J G. Thyroid-hormone levels and prognosis in patients with serious non-thyroidal illness. Lancet 1975; 2: 275-6. Naeye, R, Golstein J, Chomeck N, Meinhold H, Wenzel K W, Vanhaelst L. A low T3 syndrome in diabetic ketoacidosis. Clin Endocrinol 1978; 8: 467-72. Nicod P, Burger A, Strauch G, Vagenakis A G, Braverman L E. The failure of physiologic doses of reverse T3 to effect thyroid-pituitary function in man. J Clin Endocrinol Metab 1976; 43: 478-81. Nomura S, Pittman C S, Chambers J B Jr, Buck MW, Shimizu T. Reduced peripheral conversion of thyroxine to triiodothyronine in patients with hepatic cirrhosis. J Clin Invest 1975; 56: 643-52. Obregon M J, Roelfsema F, Morreale de Escobar G, Escobar del Rey F, Querido A. Ex- change of triiodothyronine derived from thyroxine with circulating triiodothyronine as studied in the rat. Clin Endocrinol 1979; 10: 305-15. 138 "• Pittman CS, Suda A K, Chambers JBJr, Ray YG. Impaired 3,5,3'-triiodothyronine produc- ! | tion in diabetic patients. Metabolism 1979; 28: 333-8, j ^ Pittman C S, Suda A K, Chambers J B Jr, Me Daniel H G, Ray G Y, Preston B K. Abnormal- i $, ities of thyroid hormone turnover in patients with diabetes mellitus before and after S <* insulin therapy. J Clin Endocrinol Metab 1979; 48: 854-60. j f Prescott R W G, Yeo P P B, Watson M !, Johnston ID A, Ratcliffe J G, Evered D C. Total j ft and free thyroid hormone concentrations after elective surgery. J Clin Pathol 1979; 32: j j|

321-4. . mj i £ Ramirez G, O'Neill W Jr, Jubiz W, Bloomer A H. Thyroid dysfunction in uremia: evidence ,, £ for thyroid and hypophyseal abnormalities. Ann IntMed 1967; 84:672-6. ' | RatcliffeJG, Stack BHR.BurtRW et al. Thyroid function in lung cancer. Brit MedJ 1978; j | 1:210-2. " K ReichlinS, Bollinger J, Nejad I, Sullivan P. Tissue thyroid hormone concentration of rat and j~ man determined by radioimmunoassay: biologic significance. Mt Sinai J Med 1973; 40: i jj 502-10. (>: Rose D P, Davis T E. Plasma triiodothyronine concentrations in breast cancer. Cancer 1979; ' \- 43: 1434-8. ' Rudman D, Fleischer A S, Kutner M H, Raggiv J F. Suprahypophyseal hypogonadism and j hypothyroidism during prolonged coma after head trauma. J Clin Endocrinol Metab , ,. \- 1977;45:747-54. ; j [ SaundersJ, Hall S EH, Sonksen PH. Thyroid hormones in insulin requiring diabetes before r j and after treatment. Diabetologia 1978; 15: 29-32. : i Silva J E, Larsen P R. Pituitary nuclear 3,5,3'-triiodothyronine and thyrotropin secretion: an explanation for the effect of thyroxine. Science 1977; 198: 617-20. Silva J E, Kaplan M M, Cheron R G, Dick T E, Larsen P R. Thyroxine to 3,5,3'-triiodothy- ; ronine conversion by rat anterior pituitary and liver. Metabolism 1978; 27: 1601-7. A Silva J E, Larsen P R. Contributions of plasma triiodothyronine and local thyroxine mono- | deiodination to triiodothyronine to nuclear triiodothyronine receptor saturation in 3 pituitary, liver and kidney of hypothyroid rats. Further evidence relating saturation of ? pituitary nuclear triiodothyronine receptors and the acute inhibition of thyroid- ) stimulating hormone release. J Clin Invest 1978; 61: 1247-59. j Silva J E, Dick T E, Larsen P R. The contribution of local tissue thyroxine monodeiodination to the nuclear 3,5,3'-triiodothyronine in pituitary, liver and kidney of euthyroid rats. Endocrinology 1978; 103: 1196-1207. Surks M I, Oppenheimer J H. Postoperative changes in the concentration of thyroxine-binding prealbumin and serum free thyroxine. J Clin Endocrinol Metab 1964; 24: 794-802. • Talwar K K, Sawhney R C, Rastogi G K. Serum levels of thyrotropin, thyroid hormones and | their response to thyrotropin releasing hormone in infective febrile illnessess. J Clin Endocrinol Metab 1977; 44: 398-403. j Vagenakis A G, Rapoport B, Azizi F, Portnay G I, Braverman L E, Ingbar S H. Hyperre- . sponse to thyrotrophin-releasing hormone accompanying small decreases in serum v p thyroid hormone concentrations. J Clin Invest 1974; 54: 913-8. •.• • i Visser T J. A tentative review of recent in vitro observations of the enzymatic deiodination of ; Ij iodothyronines and possible physiological implications. Mol Cellul Endocrinol 1978; 10: 1 c| 241-6. I • || Wartofsky L, Martin D, Earll J M, Alterations in thyroid iodine release and the peripheral 4 | metabolism of thyroxine during acute felciparum malaria in man. J Clin Invest 1972; 51: j 'I? 2215-32. I Wartofsky L, Burman K D, Dimond R C, Noel G L, Frantz A G, Earll J M. Studies on the 3 nature of thyroidal suppression during acute falciparum malaria: integrity of pituitary y • j response to TRH and alterations in serum T3 and reverse T3. J Clin Endocrinol Metab ••] i_ 1977; 44:85-90. '*> f Weissel M, Sturmvoll H K, Kolbe H, Hflfer R. Basal and TRH-stimulated thyroid and pitui- i | tary hormones in various degrees of renal insufficiency. Acta Endocrinol 1979; 90:23-32. : | •fi

CHAPTER VI.

STUDIES IN PATIENTS WITH ACUTE MYOCARDIAL IN- FARCTION ^ _^^

VI. 1. Introduction.

In patients with non-thyroidal disease a qualitative relationship exists between the severity of the disease and the changes in the peripheral conversion of T4 (see chapter V). A quantitative relationship is more difficult to establish, since the severity of a disease cannot be graded easily. The entity of acute myocardial infarction (AMI) is exceptional in this respect, because the infarct size (which can be estimated by serial measurements of serum enzyme concentrations) is directly correlated to the severity of the condition: a greater infarcted area is associated with more complications and a worse prognosis (Chapman 1972). We therefore studied prospectively thyroid hormone metabolism in patients with an acute myocardial infarction, in order to detect a possible quantitative relation between the infarct size and the changes in the peripheral conversion of T4. The pituitary-thyroid axis was also evaluated by serial measurements of plasma TSH, since conflicting reports have been published on thyroid function in AMI: a normal, a decreased or an increased secretion of T4 and of TSH have been suggested (Ceremuzynski et al 1970; Saeed-Uz-Zafar et al1971; Harland et al 1972; Vanhaelst et al 1976; Kaplan et al 1977; Westgren et al 1977; Smith et al 1978; Ljunggren et all979).

VI. 2. Patients and methods.

A series of consecutive patients admitted to the coronary care unit because of a suspected AMI, was studied. Acute myocardial infarction was s diagnosed on a characteristic history correlated with typical electrocardio- graphic changes and a serial rise in the serum enzymes glutamic oxalacetic transaminase (SGOT), glutamic pyruvic transaminase (SGPT) and lactic dehydrogenase (LDH). Infarct location was defined according to the criteria proposed by the New York Heart Association (1973). The size of infarction was estimated by 6-hourly measurements of SGOT until the peak value was reached (West et al 1966; Killen and Tinsley 1966; Kibe and Nilsson 1967). From the initial 43 patients seven were excluded from the Data of 36 patients admitted because of acute myocardial infarction (AMI). nr sex age HAT peak SGOT localization infarct 1 complications hrs U/l pump rhythm conduction failure disturbances

1 uncomplicated AMI (group A, n = 17). 1 M 48 2.0 23 anterolateral i 2 M 59 15.5 38 anterolateral 1 A 3 F 71 3.0 40 inferolatcral 4 M 67 3.75 45 anterolateral i 5 M 57 1.0 49 lateral 6 F 59 4.5 50 anteroseptoinferolateral ) ' !' 7 M 66 4.5 62 anteroseptal 8 M 54 4.0 68 anterolateral ; ! 9 F 88 2.5 68 anteroseptolateral '"] 10 M 77 13.0 71 inferior * ( 11 M 65 -2.0 74 anteroseptal 12 M 67 4.0 85 inferoposterolateral 'J 13 M 51 3.5 115 inferoposterolateral ; 14 M 58 5.0 116 inferolateral 15 M 59 14.5 125 anteroseptolateral • •'•'••;''

16 M 72 5.0 139 anteroseptolateral '" ', ' : 17 M 86 31.0 168 anterolateral • ' '! x(range) 65(48 - 88) 6.8(-2 -31) 79(23 - 168) i , * ,' ' '

AMI ferouD B, n = 17) complicated AMI (group B, n = 17) 18 M 66 24.0 34 infcrolateral X 19 M 60 1.25 39 anteroiateral AF 20 F 76 3.0 61 anterolateral AF 21 F 62 2.5 81 inferolateral X AF CB 22 M 67 0.25 105 anterolatcral AF 23 M 60 1.0 119 inferoposterior X CB 24 M 65 14.5 122 infcropostcrolatcral X 25 M 78 5.5 143 anteroseptal X 26 M 72 1.0 149 inferoposterolateral X AF SA 27 M 45 18.5 150 anterolateral X 28 M 66 4.0 168 inferior X AV 29 M 51 4.0 172 anterolateral AF 30 M 45 4.5 196 inferoposterior AV 31 M 57 1.0 200 inferoposterolatcral X AF 32 M 54 1.75 225 inferior VF CB 33 M 49 2.0 251 anteroscptal VT 34 M 47 2.75 257 antcroseptolateral X VF x(range) 60(45 - 78) 5.4 (0.25 - 24) 145(34 - 257) fatal AMI (group C, n = 2) 35 F 72 2.5 64 anterolateral rupture of papillary muscle 36 M 66 4.0 167 inferoposterior cardiac tamponade all patients (n = 36) 30M x(range) 6F 63(45 - 88) 5.9 (-2 - 31) 112(23- 257) 9 antero/septal 12 antero/scptal 15 infero/posterior

Abbreviations: SGOT = serum glutamic oxalacetic transaminase; HAT = hospital arrival time; AF = atrial fibrillation; VF = ventricular fibrillation; VT = ventricular tachycardia; CB = complete (third degree) AV block; SA = sino-atrial conduction disturbances; AV = atrioventricular conduction disturbances other than CB. 142 V: study: 3 patients had no AMI, and 4 were treated during the study period ; % with drugs that interfere with thyroid function tests (B-adrenoceptor |<3 blocking agents, oral contraceptives, anabolic steroids). Thyroid disease <*S, I was absent in all patients, as judged from the history and the physical |1 ?• examination. • , ^< I From the remaining 36 patients (table VI.l) two died within the study p h period, and 17 of the 34 survivors had an uncomplicated AMI. A |'; •;', complicated AMI was characterized by the occurrence of pump failure | (grade> II according to Killip and Kimbal 1967), or disturbances of cardiac j | rhythm and conduction (excluding sinus bradycardia, sinus tachycardia, I and extrasystolic beats). The hospital arrival time, defined as the time interval between the acute onset of precordial pain and the time of admission, was assessed in all patients. In two patients (nr. 2 and 17) the peak SGOT value was found in the first blood sample withdrawn on ; admission. The treatment of the patients with AMI consisted in general of narcotic : , )- analgesics according to need, antiarrhythmic drugs (lidocaine for 36 hours ' ! by intravenous drip, unless contraindicated), and anticoagulants (either j continuation of coumarin derivatives in 22 patients, or institution of j subcutaneous low-dose heparin in 12 patients); digoxin and diuretics were prescribed, if necessary. Mobilization was begun on the third day after admission, if feasible, and the administration of subcutaneous heparin was discontinued some days later. Blood samples were taken on days 1,2,3,4,5,6,7,10 and 14, between 0830 and 1030 a.m. except on day 1, when blood was taken at the time of admission. In each plasma sample the following determinations were done: TSH, rT3, T3, T4 and T3U. The competitive protein binding method of Murphy and Pattee was used for the assay of T4. The 45 samples of 5 patients were also analyzed by T4-RIA, and the results were in close agreement with the T4-D values. All samples of one patient were assayed in the same run in order to avoid inter-assay variation. Hematocrit values (Ht) were determined on the day of admission and in the second week. Statistical evaluation of the trend in hormone levels was performed by analysis of variance (Bennett and Franklin 1954). Differences between . [ several points of time or groups of patients were analyzed by the (paired) t I test. Because of their skewed distribution, plasma-TSH levels and the | ratio's of plasma iodothyronine concentrations were analyzed after I logarithmic transformation. In the statistical analyses the level of ' significance was taken as a = 0.05; undetectable hormone concentrations ) \ were taken as 50% of the detection limit. I

I VI. 3. Results. \ t i I In the patients who died, a progressive decrease in plasma T3 and an ; 143

0.8

0.6

0.4

0.2

2.4 i.i G I: ao

1.6

1.2

150

S 130

140

1 120 - — 1

u. 100

3.0

•5 —-—. E 1.0 I CO / 1-

0.5

0.3 1

w days Figure VIA. Thyroid function tests (values as x ± SD but TSH as geometric mean ± SD) in 34patients with acute myocardiai infarction. Thyroid function tests in 34 patients admitted because of acute myocardial infarction (uncomplicated in 17 patients, group A; complicated in 17 patients, group B). Values as mean ± SD, except plasma TSH and iodothyronine ratio's which are presented as the geometric mean ± SD. group dayl 2 3 4 5 TSH mU/1 all 0.9( 0.3-2.6) 0.7(<0.3 - 2.0) 1.0(0.4 - 2.7) 1.3(0.6 - 3.0) 1.35(0.6 - 2.8) A 0.8«0.3 - 2.4) 0.7(<0.3 - 2.2) 1.1(0.5 - 2.7) 1.3(0.6 - 2.8) 1.3 (0.6 - 2.9) B 1.1( 0.4-2.9) 0.7(<0.3 - 1.8) 0.9(0.3 - 2.7) 1.4(0.6 - 3.2) 1.4 (0.7 - 2.8) rTa nmol/1 all 0.33 ± 0.08 0.47 ± 0.18 0.51 ±0.22 0.47 ± 0.16 0.44 ±0.15 A 0.32 ± 0.08 0.41 * 0.12 0.44 ± 0.14 0.41 ± 0.14 0.38 ± 0.13 B 0.33 ± 0.09 0.53 ± 0.21 0.58 ± 0.27 0.52 ± 0.17 0.50 ± 0.15

T3nmol/1 all 1.91 ± 0.36 1.62 ± 0.39 1.55 ± 0.39 1.65 ± 0.39 1.65 ± 0.37 A 1.90t 0.44 1.66 ± 0.42 1.67 ± 0.37 1.79 ±0.36 1.81 ± 0.33 B 1.91 ± 0.29 1.58 ± 0.36 1.43 ± 0.38 1.52 ± 0.38 1.49 ± 0.36 T« nmol/1 all 117 ± 19 117 ±22 119 ± 26 126 ± 29 128 ± 26 A 116 ±19 117 ± 24 123 ±28 129 ± 29 134 ±27 B 117 ±20 117 ±20 115 ±24 123 ± 31 121 ± 24

T3U all 0.94 ±0.11 0.96 ±0.12 0.95 ± 0.09 0.93 ± 0.08 0.93 ± 0.07 A 0.93 ± 0.08 0.95 ± 0.09 0.94 ± 0.06 0.91 ± 0.07 0.90 ± 0.06 B 0.96 ± 0.13 0.97 ± 0.14 0.96 ±0.12 0.95 ± 0.09 0.95 ± 0.08

FT4 index all 110 ±26 112 ±23 112 ± 24 117 ±28 118 ±23 A 108 ± 17 110 ±20 115 ±26 118 ±28 121 ±25 B 113 ±33 114±26 109 ± 22 116 ±30 115 ±21 3 rT3/T3xl0- all 168(121 - 234) 282(193-413) 317(205 - 491) 277(179-429) 261(168 - 405) A 167(115 - 244) 249(182 - 340) 259(180 - 373) 225(154 - 329) 205(141 - 299) B 170(128 - 224) 320(212 - 483) 388(256 - 588) 341(228 - 508) 331(228 - 481) s rT3/T4xl0- all 2.73(2.07 - 3.61) 3.85(2.78 - 5.35) 4.10(2.83 - 5.94) 3.62(2.57 - 5.10) 3.35(2.36 - 4.78) A 2.69(2.03 - 3.58) 3.48(2.72 - 4.44) 3.51(2.60-4.76) 3.12(2.46 - 3.97) 2.78(2.11 - 3.66) B 2.77(2.09 - 3.68) 4.27(2.94 - 6.21) 4.79(3.30 - 6.97) 4.20(2.89 - 6.09) 4.05(2.92 - 5.61) 3 T3/T4xl0" all 16.2(13.3 - 19.8) 13.7(11.3- 16.4) 12.9(10.7 - 15.6) 13.1(11.0 - 15.5) 12.9(10.9 - 15.2) A 16.1(13.2 - 19.7) 14.0(11.4-17.1) 13.6(11.3- 16.3) 13.9(11.4- 16.9) 13.5(11.2-16.3) B 16.4(13.4-20.0) 13.4(11.3- 15.8) 12.3(10.2 - 14.9) 12.3(11.0-13.8) 12.2(10.7 - 14.0) group day 6 7 10 14 normal values TSH mU/1 all 1.2(0.5 - 2.9) 1.1(0.5 - 2.6) 0.9( 0.3-2.2) 0.?( 0.3-2.5) 0.7(<0.3 - 2.3) A 1.0(0.3 - 3.0) 0.9(0.3 - 2.6) 0.6«0.3 - 1.7) 0.7(<0.3 - 2.3) B 1.4(0.7 - 2.6) 1.3(0.7 - 2.4) 1.2( 0.6-2.4) 1.2( 0.5-2.6) rT3 nmol/1 all 0.39 ±0.11 0.36 ± 0.10 0.33 ± 0.10 0.32 ±0.11 0.24 ± 0.07 A 0.36 ±0.11 0.33 ±0.11 1.30 ± 0.10 0.31 ± 0.09 B 0.42 ±0.10 0.38 ± 0.10 0.36 ±0.10 0.34 ± 0.12

T3 nmol/1 all 1.68 ± 0.39 1.62 ± 0.38 1.59 ± 0.28 1.63 ±0.32 1.85 ± 0.37 A 1.81 ± 0.37 1.76 ± 0.33 1.71 ± 0.19 1.69 ± 0.25 B 1.54 ± 0.36 1.49 ± 0.38 1.47 ± 0.31 1.57 + 0.37

T4 nmol/1 all 130 ±24 130 ±24 123 ±22 118 ±20 102 ± 18 A 134 ±25 130 ± 27 123 ± 19 117 ± 18 B 126 ± 22 129 ± 23 124 ± 25 119 ± 22

T3U all 0.93 ± 0.07 0.93 ± 0.07 0.93 ± 0.07 0.93 ± 0.08 0.94 ± 0.09 A 0.91 ± 0.05 0.92 ± 0.05 0.93 ± 0.05 0.92 ± 0.07 B 0.94 ± 0.09 0.94 ± 0.08 0.93 ± 0.09 0.94 ± 0.10 P\\ index all 120 ± 21 121 ±23 115 + 20 109 ± 18 96 ± 15 A 121 + 23 • 120 ± 25 114 ± 19 107 ± 17 B 118 ± 19 122 ± 22 115 + 21 111 ± 19 3 rT3/T3xl0- all 229(155 - 337) 224(142 - 352) 203(133- 309) 191(122 - 300) 124(86 - 179) A 193(133-280) 183(126 - 264) 170(115-252) 176(120 - 257) B 271(195 - 376) 274(176 - 428) 243(167 - 354) 209(126 - 347) 3 rT3/T4xl0" all 2.91(2.16 - 3.92) 2.69(2.03 - 2.56) 2.62(1.90 - 3.60) 2.62(1.90 - 3.62) 2.23(1.62 - 3.08) A 2.60(2.01- 3.37) 2.47(1.92 - 3.18) 2.38(1.76 - 3.22) 2.53(1.88 - 3.41) B 3.26(2.43 - 4.38) 2.93(2.19 - 3.90) 2.88(2.10 - 3.94) 2.71(1.91 - 3.85) 3 T3/T4xl0" all 12.7(10.6 - 15.4) 12.4( 9.7 - 15.7) 12.9(10.6 - 15.6) 13.7(11.0- 17.0) 18.0(14.9 - 21.9) A 13.5(11.0 - 16.5) 13.5(10.9 - 16.7) 14.0(11.8-16.6) 14.4(11.7 - 17.8) B 12.1(10.3 - 14.1) 11.3( 8.9 - 14.2) 11.9( 9.9 - 14.2) 13.0(10.4 - 16.2) 146

Table Vl-3

Plasma T3 and rT3 concentrations (values as mean ± SD) on admission as a function of ft the hospital arrival time (HAT) in 34 patients with acute myocardial infarction. f.

HAT n T3 *T3 hrs. nmol/1 nmol/1

< 2 8 1.94 ± 0.39 0.29 ± 0.05 2-3 S 1.98 ± 0.31 0.31 ± 0.05 3-4 4 1.89 ± 0.24 0.30 ± 0.05 4-S 7 1.82 ± 0.37 0.30 ± 0.12 5-6 3 1.89 ± 0.46 0.34 ± 0.06 6-24 5 2.11 ± 0.34 0.39 ± 0.07 > 24 2 1.45 ± 0.49 0.47 ± 0.12

increase in plasma rT3 levels was observed. The results of the thyroid function tests in the 34 survivors are presented in figure VI. 1. and table VI.2. A transient increase in the plasma levels of TSH, rT3 and T4 (p < 0.001) and in the FT4 index (p < 0.01) was found, whereas plasma T3 (p < 0.001) and T3U (p< 0.05) decreased. No difference was found in the plasma concentrations of TSH, rT3 and T4 and in the FT4 index between days 1 and 14, but plasma T3 and T3U were lower on day 14. The trend in hormone levels in the patients with an uncomplicated AMI (group A) was similar to the trend found in those with a complicated AMI (group B). The increase in plasma rT3 and the decrease in plasma T3 was greater in group A than in group B, but plasma TSH, T4 and T3U and the FT4 index did not differ between these two groups. Peak SGOT values were higher in complicated AMI than in uncomplicated AMI (p < 0.01). Concomitant with the changes in the plasma iodothyronine concentrations a transient increase in the ratio's of rT3/T3 and rT3/T4 (p. < 0.001) and a decrease in the ratio of T3/T4 was found (p < 0.001). The return to base-line values of the ratio's was delayed in group B as compared with group A. No difference was found between day 1 and day 14 in the ratio's of rT3 / T3 and of rT3/T4, but the ratio T3/T4 was lower on day 14 than on day 1 (p< 0.01). The 34 patients were further divided in three groups according to infarct size as estimated by the peak SGOT value (table VI.4). The increase in plasma rT3 and the decrease in plasma T3 concentrations (figure VI.2) were calculated as the difference between the highest or lowest observed value and the value on day 14, since the initial values of these hormones on the day of admission were clearly related to the hospital arrival time (table VI.3). No differences were found in any of the parameters tested between patients with small infarctions (peak SGOT < 80) and those with medium sized infarctions (peak SGOT 80-160). However, the values of A rT3, lowest T3 and A T3, and highest rT3/ T3 and rT3 / T4 of patients with large infarctions 147 Table VI-4 Some parameters of thyroid hormone metabolism (values as x ± SD, ratio's as geometric mean ± SD) in 34 patients with acute myocardial infarction, divided in three groups according to infarct size as estimated by the peak SGOT value.

peak SGOT value <80 80 - 160 >160 (n=14) (n=12) (n=8) peak SGOT 52 ±16 121 ± 23 205 ± 36 highest rT3 0.53+ 0.19 0.56+ 0.18' 0.65 ± 0.37' 3 ArT3° 0.17 ± 0.14 0.24+ 0.15' 0.39 ± 0.34 1 3 lowest T3 1.37+ 0.19 1.31 ± 0.35 1.16 ± 0.23 2 4 AT3° 0.16+ 0.30 0.36+ 0.28 0.60 ± 0.20 3 highest rT3/T3 339(237 -486) 357(231 - 552)' 478(346 - 660) 3 highest rT3/T4 4.12( 2.98 -5.71) 4.88( 3.41 -6.98)' 5.48( 4.14-7.25) lowest T3/T4 10.9 ( 9.3 -12.9) 11.1 ( 9.0 -13.6)' 10.2 ( 9.2 -11.4)'

°ArT3 andAT3: the difference between the highest or lowest observed value and the value on day 14. 1 not significant;2 0.05

(peak SGOT > 160) were different from those of patients with small infarctions. The peak values of T4 and TSH did not differ in these three groups, although a tendency to higher peak TSH values was observed in patients with greater infarctions. A linear correlation was found between the infarct size and all of the following parameters: A rT3, AT3, highest rT3/T3

Table VIS The correlation between infarct size as estimated by the peak SGOT value (variable x) and some parameters of thyroid hormone metabolism (variable y) in 34 patients with acute myocardial infarction.

variable y regression equation coefficient p-value of correlation

highest rT3 y = O.OOlx + 0.46 r=0.28 0.05

lowest T3 y=-O.OOlx + 1.43 r = 0.28 0.05

148 PEAK SGOT (U/l) <80 80-160 >160 0.75 r (n=14) (ru8)

0-50 So

0.25

S 0.25 o c 0.50

0.75 L

Figure VI.2. The increase in plasma rTj and the decrease in plasma T3 concentrations in 34 patients with acute myocardial infarction, divided according to infarct size as estimated by the peak SGOT value. ( AIT, and A T3: the difference between the highest or the lowest observed value and the value on day 14; values as x±SEM).

and highest rT3/T4 (table VI.5 and figure VI.3). As depicted in figure VI. 1, the peak values of TSH precede those of T4 in time. Inspection of thedata of each individual strongly suggested this sequence in most cases. Since the day on which the peak T4 value was reached, differed f rompatient to patient, we rearranged the T4 values according to that day. The TSH- and T3-values were similarly rearranged around the peak T4 value. Isolated high T4 values found on occasion during the first three days after AMI, were not considered '•4- as the "peak", since they apparently were due to the initial hemoconcentration: Ht values on admission were higherthan in thesecond week (0.45±0.04 ^ _^ ,„ .,„,,,, -,-s.r:7---v^fJ*«J2^S^lt?*#^?'"r .'?-S'f\'^s"lT''$'?t'?''.*l"'-» '.'

Table VT-6

Plasma T4, TSH and T, concentrations in 34 patients with acute myocardial infarction, arranged around the time of the observed "peak" T, value in the first two weeks after admission (T4 and T3 values as x ± SD with the range in brackets, TSH as geometric mean ± SD).

day -7 -6 -5 -4 -3 -2 n (6) (13) (22) (28) (33) (34)

f4 nmol/1 108 ± 21 101 ± 20 109 ± 19 113 ± 21 119 ± 22 125 ± 25 (86- 138) (79- 144) (72- 138) (72- 150) (80- 173) (89- 188) TSH mU/1 0.6(<0.3 - 2.1) 0.7(<0.3 - 3.0) 0.6«0.3 - 2.0) 0.7(<0.3 - 1.9) 1.3(0.6 - 3.0) 1.3(0.7 - 2.5)

T3 nmol/1 1.59 t:0.31 1.70:t0.52 1.54 ± 0.47 1.54:tO.36 1.62 ± 0.38 1.68 d:0.35 (1.03 - 1.94) (0.85 - 2.63) 0.89-- 2.36) (0.91 - 2.42) (0.99 - 2.53) (0.92 - 2.39) 4

- peakT4 day —1 0 +1 +2 +3 +4 +5 n (34) (34) (33) (28) (21) (12) (6)

T4 nmol/1 129 ± 24 139 ± 26 129 ± 22 123 ± 23 126 ± 20 123 ± 25 127± 26 (93 - 182) (97- 197) (97- 187) (85- 169) (92- 178) (94- 159) (90- 147) TSH mU/1 1.4(0.7 - 2.9) 1.2(0.5 - 2.7) 1.1(0.4 - 3.1) 0.9(0.4 - 2.2) 1.0(0.4 - 2.6) 0.9(0.4 -2.0) 1.0(0.4 - 2.5)

T3 nmol/1 1.68 ± 0.33 1.79 ± 0.42 1.67 ± 0.35 1.67 ± 0.27 1.67 ± 0.35 1.61 ± 0.29 1.53:t0.51 (0.99 - 2.22) (1.00 - 2.63) (1.14 - 2.43) (1.23 - 2.34) (0.80 - 2.16) (0.85 - 1.94) (0.71 - 2.20) 150 r=0.56 1.2 r=0.46 1.2 p<0.01 p<0.001

1.0 1.0

|0.8 0.8 "rii- 3 50.6 10.6

0.4

0.2 0.2

40 80 120 160 200 240 40 80 120 160 200 240 PEAK SGOT U/l PEAK SGOT U/l

g F/.5. The relation between the changes in the conversion of T4 (as indicated by A rT3 and &T3: the difference between the highest or lowest observed value and the value on day 14) and the infarct size (as estimated by the peak SGOT value) in 34 patients with acute myocardial infarction.

vs. 0.43±0.04, p < 0.001). The results are listed in table VI.6. Again, the ;• substantial increase in T4 is preceded by a rise in TSH, and is associated with a concomitant increase in plasma T3 values.

VI. 4, Discussion.

The reciprocal changes in plasma concentrations of T3 and r T3 found in our patients with acute myocardial infarction, are similar to those found in other acute non-thyroidal diseases (see chapter V). The decrease in plasma T3 and the increase in plasma rT3 thus seems to be caused by inhibition of the 5'-deiodinating process in the peripheral conversion of T4. The nadir of T3 and the peak of rT3 occurred on the third day, in agreement with the study of Westgren et al (1977). As in patients with other non-thyroidal diseases, the decrease in plasma T3 in patients with AMI is not related to the increase in plasma cortisol levels (Ljunggren et al 1979).

A qualitative relationship between the severity of AMI and the changes in the conversion of T4 was found: in patients with a complicated AMI the decrease in plasma T3 and the increase in plasma rT3 was greater than in patients with uncomplicated AMI, and the return to base-line levels was delayed in complicated AMI as compared to uncomplicated AMI. The greatest fall in plasma T3 in patients with the most severe course of AMI has also been observed by Naumann et al (1975). Our data also indicate that there is a quantitative relationship between the severity of AMI and &*.' 151

the changes in the conversion of T4, since the infarct size was greater in patients with a complicated AMI, as was expected (Chapman 1972). The existence of a quantitative relationship is further demonstrated by the linear correlation between peak SGOT values and various parameters of 8 thyroid hormone metabolism: A rT3, AT3,'highest rT3 / T3 and rT3 / T4 ratio's (but not with the lowest T3/T4 ratio). The only report in the literature relating infarct size to thyroid hormone metabolism is of Smith et al (1978). They found a quantitative relationship between the infarct size and the lowest T3/T4 ratio (but not with the lowest plasma-T3 value). The cause for the discrepancy in results is probably related to a different selection of patients: we excluded only those patients who were treated with drugs that interfere with thyroid function tests, whereas Smith et al (1977) applied more rigid selection criteria. A real quantitative relationship between the severity of a disease and the changes in thyroid hormone metabolism in other states than AMI has thus far only been reported in diabetes (Saunders et al 1978; Pittman et al 1979). A negative correlation between fasting blood glucose levels and the T3/T4 ratio has been demonstrated in diabetic patients, and plasma T3 was positively related to glucose utilization and glucose MCR but negatively to plasma ketone body concentration.

As in other non-thyroidal diseases, the fate ofthyroxine in patients with AMI is most intriguing. The occasionally found high T4 values on the first three days are best explained by hemoconcentration, as witnessed by the observed higher hematocrit values at the time of admission, and the reported high Ht values in the first days after the onset of AMI followed by a steady decrease up to day 10 (Jan et al 1975; Kaplan et al 1977; Smith 1977). Vanhaelst et al (1976) have suggested that the elevated plasma T4 levels at the onset of AMI are due to stimulation of thyroid secretion induced by the rise in catecholamines, which can stimulate thyroid secretion in animals (Ericson et al 1970; Melander 1970). This explanation seems, however, rather far-fetched in the light of the plausible explanation of hemoconcentration. Also, the thyroxine secretion rate determined immediately after AMI has been reported as normal (Harland et al 1972). One would expect a decrease in plasma T4 values concomitant with the decrease in Ht (Kaplan et al 1977). Smith et al (1978) found no changes in the T4 values (which therefore already indicates an increase in circulating T4), but after correction for the change in Ht an increase of 15% in T4 levels between day 2 and day 9 was calculated (Smith 1977). We observed a clear increase in uncorrected plasma T4 values of appr. 12%, peaking around day 6-7 (table VI.2). This increase is not caused by hemoconcentration, since the Ht decreased and T3U remained constant as of day 3. The initially higher T3U values - and not lower values as would be expected from the hemoconcentration in the initial phase (cf. table 11.21) - are most probably 152 caused by a decreased binding capacity of the plasma proteins for thyroid hormones, analogous to the situation found in other non-thyroidal diseases (see chapter VI), due to a decrease in TBP caused by the 'acute phase reaction' (Johansson et al 1972; Smith et al 1977; Kushner et al 1978). The concentration of TBG has been measured in patients with AMI, I and after an initial fall to a nadir on day 3 a return to the initial levels is observed (Westgren et al 1977). Thus the increase in T is not due to !-• 4 changes in the plasma volume or to changes in the concentration of thyroid hormone binding proteins. It is also highly unlikely that the increase in T4 is caused by a rise in the plasma concentration of free fatty acids (FFA) for several reasons: 1) FFA falsely increase T4-values as measured in the competitive-protein binding method of Murphy and Pattee (T4-D assay), but are without effect on the results in theT4-RI A (Rootwelt 1975; Liewendahl and Hele- nius 1976; Pannal et al 1977); the rise of T4 in our study was demonstrated byT4-DandbyT4-RIA. 2) intravenous heparin increases T4 (as measured in theT4-D assay) and FT4 (Saeed-Uz Zafar et al 1971), most probably via the induced rise in FFA; subcutaneous low-dose heparin does not produce a rise in FFA (Arnesen et al 1979), and furthermore the increase in T4 was also seen in our patients who were treated with oral anticoagulants, which are without effect on T4 and FT4 concentration in AMI (Saeed-Uz Zafar et al s 1971). 3) the catecholamine-mediated rise in FFA in AMI is an early phenomenon, the highest levels of FFA are found at the time of admission whereafter a decrease to near normal levels on day 3-5 occurs (Gupta et al 1969; Sharma 1977); however, plasma T4 peaks later, on day 6-7. It thus appears that there is a real increase in the amount of circulating T4 around day 6-7 - in fact, a laboratory pattern of "T4-toxicosis" is found in 16 of the 34 patients at the time of the T4 peak. In general, an increase in plasma T4 is caused by either a decrease in MCR or an increase in PR. The impairment in the deiodination of T4 (which is quantitatively an important pathway in the degradation of T4) may have lowered the M CR. A higher P R is caused by increased thyroidal secretion of T4. Since the T4 peak is preceded in time by a significant increase in TSH, this suggests that a TSH- mediated secretion of T4 is responsible (at least in part) for the rise in T4. Furthermore, indirect evidence for this sequence of events is found in the associated rise of plasma T3 (table VI.6). If the T4 rise is TSH mediated, the question arises why the release of TSH is stimulated. Again, the time sequence suggests that the low T3 levels play a causative role - mean lowest T3 levels are found on day 3, mean highest TSH levels on day 4 and 5. In the rat, TSH release by the anterior pituitary is a function of nuclear T3, which is derived half from 153 intrapituitary T4 -*T3 deiodination and half from plasma T3 (Silva and Larsen 1978). A normal concentration of FT4 is reported in the initial phase of AMI, whereas a substantial decrease in FT3 is observed starting 8 hours after the onset of AMI (Saeed-Uz Zafar et al 1971; Bugiardini et al 1979). Therefore, it seems reasonable to propose that the fall in T3 and FT3 concentrations provokes the release of TSH which, in a later phase, is switched off by the increase in T4.

In summary, our studies suggest the following sequence of events in thyroid hormone metabolism in patients with AMI: 1. the occasionally observed high T4 values in the initial phase are due to hemoconcentration. 2. the 5'-deiodination of T4 is inhibited, resulting in a peak of plasma rT3 and in a nadir of plasma T3 values on day 3, and in a lower metabolic clearance ofT4. 3. TSH secretion, provoked by the lower T3 levels, increases and JV-AICS on days 4 and 5. 4. a TSH-mediated increase in thyroidal secretion of T4 and T3 occurs mainly on days 6 and 7, and is switched off thereafter by the negative feedback of thyroid hormones on the pituitary. 5. subsequently, all parameters return to normal in about two weeks with the exception of the T4-*T3 deiodination which remains impaired up to appr. 3 months after AMI (Smith et al 1978). This proposition satisfactorily explains many of the seemingly contradictory results of thyroid function tests in AMI so far published.

From a teleological point of view, the low T3 levels in AMI may serve to prevent further catabolism (see chapter IX). In this respect the observed changes in thyroid hormone metabolism in patients with AMI fits in the general adaption syndrome of the body to stress. It remains to be demonstrated if it is wise to treat the increase in thyroidal secretion with antithyroid drugs as has been proposed (Ceremuzynsky et al 1970). On the other hand, recent studies on the allegedly beneficial effect of B- adrenoceptor blocking-agents in the reduction of infarct size (Julian 1976; Hjalmarson and Werko 1976) are of great interest, since one of the effects of these drugs is impairment of the S'-deiodination of T4 (cf. chapters VII and VIII).

VI.5. References

Arnesen H, Bjerkedal I, Skjaeggestad 0, Godal HC. Plasma free-fatty-acids and small doses of subcutaneous heparin in acute myocardial infarction. Thrombosis Research 1979; 14: 541-9. •WJ-

154 Bennett CA, Franklin NL. Statistical analysis in chemistry and the chemical industry. New York: John Wiley and Sons, 19S4. Bugiardini G, Miniero R, Mattioli L et al. Modificazioni di ormoni tiroidei in corso di infarcto miocardico acuto. In: The Proceedings of the International Symposium on free thyroid hormones (Venice 1978). Amsterdam: Excerpta Medica, 1979 (in press). Ceremuzynski L, Kuch J, Markiewicz L, Lawecki J, Taton J. Patterns of endocrine reactivity in patients with recent myocardiat infarction. Clinical and biochemical correlations: trial of endocrine therapy. Brit Heart J 1970; 32:603-10. '€•• Chapman BL. Relation of cardiac complications to SGOT level in acute myocardial infarction. Brit Heart J 1972; 34:890-6. The Criteria Committee of the New York Heart Association. Nomenclature and criteria for diagnosis of diseases of the heart and great vessels. 7th ed. Boston: Little, Brown and Compagny, 1973. Ericson LE, Melander A, Owman Ch, Sundler F. Endocytosis of thyroglobulin and release of I: thyroid hormone in mice by catecholamines and 5-hydroxy-tryptamine. Endocrinology 1970:87:915-21. Gupta DK, Young R, Jewitt DE, Hartog M, Opie LH. Increased plasma-free-fatty-acid concentrations and their significance in patients with acute myocardial infarction. Lancet 1969; 2:1209-13. Harland WA, Orr JS, Dunnigan MG, Sequeira RFC. Thyroxine secretion rate after myocardial infarction. Brit Heart J 1972; 34: 1072-4. Hjalmarson A, Werko L (editors). Experimental and clinical aspects on preservation of the ischetnic myocardium. Acta Med Scand 1976; suppl 587:177-220. Jan KM, Chien S, Bigger JT. Observations on blood viscosity changes after acute myocardial infarction. Circulation 1975; 51:1079-84. Johansson BG, Kindmark CO, Trell EY, Wollheim FA. Sequential changes of plasma proteins after myocardial infarction. Scand J Clin Lab Invest 1972; 29: suppl 124,117-26. Julian DG (chairman). Beta-blockade and myocardial infarction. Postgrad Med J 1976; 52: suppl 4:124-56. Kaplan MM, Schimmel M, Utiger RD. Changes in serum 3,3',5'-triiodothyronine (reverse T3) concentrations with altered thyroid hormone secretion and metabolism. J Clin Endo- crinol Metab 1977; 45:447-56. Kibe O, Nilsson NJ. Observations on the diagnostic and prognostic value of some enzyme tests in myocardial infarction. Acta Med Scand 1967; 182:597-610. Killen DA, Tinsley EA. Serum enzymes in experimental myocardial infarcts. Relation of blood levels of serum glutamic oxaloacetic transaminase, lactic dehydrogenase, and heat stable lactic dehydrogenase to size of experimental myocardial infarct. Arch Surg 1966; 92:418-22. Killip T III, Kimbal JT. Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients. Am J Cardiol 1967; 20:457-64. Kushner I, Broder ML, Karp D. Control of the acute phase response. Serum C-reactive protein kinetics after acute myocardial infarction. J Clin Invest 1978; 61:235-42. Ljunggren J-G, Falkenberg C, Savidge G. The influence of endogenous cortisol on the peripheral conversion of thyroxine in patients with acute myocardial infarction. Acta Med Scand 1979; 205:267-9. Liewendahl K, Helenius T, Effects of fatty acids on thyroid function tests in vitro and in vivo. Clin Chim Acta 1976; 72:301-13. Melander A. Amines and mouse thyroid activity: release of thyroid hormone by catecholamines and indoleamines and its inhibition by adrenergic blocking drugs. Acta Endo- crinol 1970; 65:376-80. Nauman J, Ceremuzynski L, Nauman A, Gunther-Krawezynska E, Jarczynska J. Serum triiodothyronine, thyroxine and thyroid stimulating hormone in relation to the clinical 155

course of recent myocardial infarction. Abstract 94, 9th annual meeting European I Society for Clinical Investigation. Rotterdam 1975. Pannall PR, Swanepoel E, Bennett JM, Pauw FH. Increased concentrations of serum free fatty acids falsely increase serum thyroxine as determined by competitive protein-binding. ClinChem 1977; 6:990-3. Pittmann CS, Suda AK, Chambers JB Jr, Ray GY. Impaired 3,5,3'-triiodothyronine (T3) production in diabetic patients. Metabolism 1979; 28:333-8. Rootwelt K. The influence of fatty acids on serum thyroxine determinations by competitive protein-binding assay. Scand J Clin Lab Invest 197S; 35:649-53. Saeed-Uz Zafar M, Miller JM, Breneman GM, Mansour J. Observations on the effect of heparin on free and total thyroxine. J Clin Endocrinol Metab 1971; 32:633-40. Saunders J, Hall SEH, SOnksen PH. Thyroid hormones in insulin requiring diabetes before and after treatment. Diabetologia 1978; 15:29-32. Sharma SC. Catecholamines and free fatty acids in myocardial infarction and angina. J Clin Pathol 1977; 30:1037-9. Silva JE, Larsen PR. Contributions of plasma triiodothyronine and local thyroxine monodeiodination to triiodothyronine to nuclear triiodothyronine receptor saturation in pituitary, liver and kidney of hypothyroid rats. Further evidence relating saturation of pituitary nuclear triiodothyronine receptors and the acute inhibition of thyroid-stimulating hormone release. J Clin Invest 1978; 61:1247-59. Smith SJ, Bos G, Esseveld MR, van Eyk HG, Gerbrandy J. Acute-phase proteins from the the liver and enzymes from myocardial infarction; a quantitative relationship. Clin Chim Ada 1977; 81:75-85. Smith SJ. Plasma-eiwitten en collold-osmotische druk na het acute myocard infarct. Thesis Rotterdam 1977. Smith SJ, Bos G, Gerbrandy J, Docter R, Visser TJ. Lowering of serum 3,3',5-triiodothyronine thyroxine ratio in patients with myocardial infarction: relationship with extent of tissue injury. Eur J Clin Invest 1978; 8:99-102. V&nhaelst L, Degaute JP, Golstein J, Bastenie PA. Pituitary-thyroid axis reaction after myocardial infarction. Horm Metab Res 1976; 8:42-6. West M, Eshchar J, Zimmerman HJ. Serum enzymology in the diagnosis of myocardial infarction and related cardiovascular conditions. Med Clin N Am 1966; 50: 171-92. Westgren U, Burger A, Levin K, Melander A, Nilsson G, Petterson U. Divergent changes of serum 3,5,3'-triiodothyronine and 3,3',5'-triiodothyronine in patients with acute myocardial infarction. Acta Med Scand 1977; 201:269-72. V

CHAPTER VII. STUDIES ON THE EFFECT OF PROPRANOLOL IN HUMAN SUBJECTS

VII. 1. Introduction. Although it is well known that in thyrotoxic patients treatment with a 0- adrenoceptor blocking agent will reduce many of the manifestations of the disease (Shanks et al 1969; Malcolm 1972), the precise mode of action of these agents in hyperthyroidism is not well understood (Levey 1971; Spauldingand Noth 1975; Levey 1975;Landsberg 1977). Since propranolol does not decrease plasma thyroxine (T4) or "'I-uptake in the thyroid (Krikler 1966; Hadden et al 1968,1969; Biran and Tal 1972; McLarty et al 1973) it is thought that 0-blocking agents do not act directly on the hyperactive thyroid, but rather decrease the — perhaps increased — sensitivity of the|3-adrenoceptors in the hyperthyroid patient. The possibility that0 -blocking agents influence the p!asma concentration of triiodothyronine (T3) suggested itself in a goitrous patient, who was under treatment with alprenolol (400 mg per day) because of vague cardiac

vulhyroid euthyraid

\, 1 ' 1

20 > day»

Figure VIM. Effect of alprenolol treatment on plasma levels of T4, T, (solid lines) and rT, (dotted line) in a patient with multinodular goitre. (TSH-Ievels were below the detection limit of the assay during the whole period depicted; the TSH-response to TRH, days 0,20 and 42, was absent). 157 £? complaints. Although this 59-year old woman was clinically euthyroid, the if results of thyroid function tests were compatible with a diagnosis of : i W - hyperthyroidism, except for the fact that plasma-T3 measured on several JT occasions was either marginally elevated or even normal. Discontinuing ; the alprenolol medication led to a rise in plasma-T3 to clearly hyperthyroid ;•£ levels (figure VII. 1) and the patient became clinically thyrotoxic. When the U treatment was resumed plasma-T3 decreased again to normal levels. ft We therefore decided to study the effect of therapy with a 0-blocking agent in hyperthyroidism on the plasma levels of T4, T3, rT3 and TSH. A group of patients with primary hypothyroidism on substitution therapy with L- thyroxine was included in the study in order to differentiate between a thyroidal or extrathyroidal effect, if any. Propranolol was chosen because il it is the best known0-antagonist and is widely used in the treatment of hyperthyroidism (Ramsay 1975). Moreover, propranolol has no intrinsic sympathomimetic activity. This study has been published in part elsewhere (Wiersinga and Touber 1977).

VII.2. Patients and methods.

Group A consisted of 11 hyperthyroid patients (10 females, one male, I varying in age from 22 to 73 years, with a mean of 54.8 years); six patients jr- had Graves' disease, three had a toxic multinodular goitre and two a toxic adenoma. In the hyperthyroid patients the mean T4 was 213 nmol/1, range 163-259, the mean T3 4.56 nmol/1, range 2.70-6.14. Of the hyperthyroid patients six were newly diagnosed cases and had never been treated. Five others had a relapse of thyrotoxicosis; in these patients no antithyroid medication had been given for at least two months prior to the study. Two of the five patients had been treated with radioactive iodine two and twelve years before the study. Group B consisted of six hypothyroid patients (all female, varying in age from 33 to 68 years with a mean of 58.5 years); three patients had primary myxoedema, one was hypothyroid after subtotal thyroidectomy and two had undergone total thyroidectomy because of thyroid carcinoma (follicular and papillary carcinoma, respectively). The last two patients had a neck uptake of less than 0.4% of the administered dose of I31I two I months prior to the study; no signs of functioning metastases could be I detected by whole-body scintigraphy at that time. All six patients were on a I constant substitution dose of L-thyroxine (200//g per day in five and 150 it Mg per day in one patient) for at least one month prior to the study and all were clinically euthyroid, their mean T4 was 114 nmol/1, range 89-146; mean T3 1.80 nmol/1, range 1.29-2.28. Both groups of patients received a fourteen-day course of propranolol, consisting of 80 mg per day in the first and 160 mg per day in the second Table VIM Thyroid function tests during propranolol treatment, 80 mg/day in the first and 160 mg/day in the second week (values as x ± SD, but plasma TSH and iodothyronine ratio's as geometric mean ± SD; day 0 represents the mean of the two days prior to treatment). days 0 1 2 3 5 group A. hyperthyroid patients (n = 11) "i TSH mU/1 <0.3 <0.3 <0.3 <0.3 <0.3 rT3 nmol/1 0.91 ± 0.43 0;97 ± 0.37 0.92 ± 0.37 1.03 ± 0.38 1.01 ± 0.42 Ts nmol/1 4.S6 ± 1.22 4.45 ± 1.29 4.05 ± 1.26 3.99 ± 1.26 3.91 ± 1.20 T4 nmol/1 213 ±32 213 ± 39 210 ± 39 212 ± 43 214 ± 40 T3U 1.16 ± 0.15 FT4 index* 2S0 ± 58 3 rT3/T3xl0- 189(126 - 284) 209(150 - 292) 221(152 - 322) 253(176 - 365) 248(153 - 403) 3 rT4/T4xl0" 3.95(2.76-5.64) 4.27(3.08-5.92) 4.13(2.97-5.74) 4.64(3.43-6.27) 4.41(3.03-6.40) 3 00 T3/T4xl0- 20,9(16.4-26.6) 20.4(16.6-25.1) 18.7(15.4-22.8) 18.3(15.0-22.3) 17.8(13.9-22.7)

group B. hypothyroid patients on L-T4 substitution (n = 6) TSH mU/1 1.3(<0.3 - 9.1) 1.3«0.3 - 8.9) 1.6«0.3- 11) 1.6«0.3 - 12) 1.7(<0.3 - 13) rT3 nmol/1 0.22 ± 0.14 0.30 ± 0.20 o!31 ± 0.15 0.29 ± 0.12 0.26 ±0.11 T3 nmol/1 1.80 ± 0.37 1.93 ± 0.55 1.79 ± 0.54 1.65 ± 0.54 1.45 ± 0.42 T4 nmol/1 114 ± 21 114 ±23 115 ± 22 114 ±23 119 ±26 T3U 0.96 ± 0.08 FT4 index 108 ± 27 3 rT3/T3xl0" 105( 50 - 221) 131( 67 - 258) 165( 92 - 298) 173(102 - 294) 167( 86 - 323) 3 rT3/T4xl0" 1.64(0.83-2.27) 2.14(1.17-3.89) 2.52(1.58-4.01) 2.42(1.61-3.64) 2.00(1.39-2.89) 3 T3/T4xl0" 15.6(12.0-20.4) 16.6(12.0-23.0) 15.2(10.7-21.6) 14.0(10.1-19.4) 12.0( 8.4-17.2) days 12 14 group A. hypeTthytoid patients (n = 11) TSH mU/1 <0.3 <0.3 <0.3 <0.3 rT3 nmol/1 1.02 ± 0.30 1.04 ± 0.39 1.09 ± 0.39 1.12 ±0.44 T3 nmol/1 3.76 ± 1.09 3.68 ± 1.16 3.57 ± 1.08 3.20 ±1.16 T4 nmol/1 215+31 215 ± 37 218 ±40 208 ± 37 T3U 1.19 ± 0.19 FT4 index 6 250 ± 75 l rT3/T3xl0 271(188 - 391) 276(176 - 432) 297(185 - 475) 343(216 - 543) 3 rT4/T4xl0" 4.59(3.48-6.04) 4.59(3.41-6.16) 4.74(3.28-6.84) 5.05(3.46-7.36) 3 T3/T4xl0- 16.9(13:4-21.4) 16.6(13.2-20.9) 16.0(12.6-20.0) 14.7(11.4-19.0) in group B. hypothyroid patients on L-T, substitution (n = 6)

TSH mU/1 1.9(0.3-13) 1.9(<0.3 - 17) 1.8(<0.3 - 17) 2.2(0.3 - 19) rT3 nmol/1 0.28 ±0.11 0.30 ±0.18 0.28 ±0.13 0.34 ± 0.16 T3 nmol/1 1.48 ± 0.31 1.51 ± 0.46 1.42 ± 0.34 1.29 ± 0.42 T4 nmol/1 119 ±23 123 ±26 123 ± 26 120 ± 19 T3U 0.97 ± 0.06 FT4 index 117 ±25 3 rT3/T3xl0" 174( 99 - 305) 175( 81 - 379) 176( 85 - 365) 245(118-509) 3 rT3/T4xl0" 2.18(1.40-3.39) 2.13(1.20-3.78) 2.04(1.09-3.80) 2.55(1.53-4.28) 3 T3/T4xl0" 12.5( 9.3-16.9) 12.2( 8.9-16.7) 11.5( 8.7-15.2) 10.4( 8.1-13.3) 160 week; other medication, if any, was kept constant during the study period. Propranolol was administered 8-hourly in oral doses of 40-20-20 mg and 80-40-40 mg, respectively. Blood was taken 90 min after the morning dose; the patients were not allowed to take any drug (including L-thyroxine) except propranolol, before blood sampling was completed. Blood samples t were collected two days prior to, and on days 1,2,3,5,7,9,12 and 14 of the propranolol medication. In each sample the following determinations were done in plasma: T4 by the competitive protein binding method of Murphy-Pattee, T3, rT3 and TSH by radioimmunoassay. Stimulation tests with thyrotropin releasing hormone were carried out before and on the last day of treatment (200 fig TRH through an indwelling venous catheter, in the second test 90 min after 80 mg propranolol orally; TSH- and T3- i ' responses were followed for 4 hours). The triiodothyronine uptake ratio (T3U) was measured before and on the last day of propranolol treatment. All samples of one patient were run in the same assay in order to avoid interassay variation. Pulse rate, blood pressure and weight of each patient were recorded on the same days as the blood samples were taken. Statistical evaluation of the trend in hormone levels and of the TSH- and 1 . I T3-responses after TRH stimulation were performed by analysis of variance (Bennett and Franklin 1954). Plasma-TSH levels and the ratio's of iodothyronines, because of skewed distribution, were analyzed after logarithmic transformation. Duncan's test (1955) was used to determine the day on which the mean T3 value was different from the pre-treatment value. T3U and the calculated FT4 index were evaluated by Student's paired t test. In all analyses the level of significance was taken as a =0.05; undetectable hormone concentrations were taken in the statistical analyses as 50 percent of the detection limit.

VII.3. Results.

The results are presented in tables VII. 1 and VII.2. Propranolol treatment of the hyperthyroid patients did not influence T4, T3U, FT4 index and TSH, but T3 decreased (p< 0.001) as of day 2, and rT3 increased (p<0.05). Plasma-T3 declined in all 11 patients, in four to within the normal range (figure VII.2). The ratio's of rT3/T3 (p<0.01) and rT3/T4 (p < 0.05) increased, the ratio T3/T4 decreased (p < 0.001). The TSH- response oh TRH stimulation remained absent; T3-levels during the second test were lower than during the first test (p < 0.001.) In all hypothyroid patients on L-thyroxine substitution plasma-T3 decreased during propranolol medication, in two to subnormal values (figure VII.2); T3-values were lower (p< 0.001) as of day 5. Reverse T3(p< 0.02) T4 (p < 0.001) and TSH (p < 0.001) increased, T3U did not change. The ratio's of rT3/T3 (p < 0.01) and rT3/T4 (p< 0.02) increased, the ratio T,/T4 decreased (p < 0.0\). The TSH-levels after TRH stimulation during ,, 7Wr^-.-~;.-ya^S3«£S^^

K//-2 TRH-stimulation test before propranolol and during 160 mg propranolol; TSH-values are presented as the geometric mean ± SD in mU/1, and T3-values as x ± SD in nmol/1.

minutes 0 20 60 120 180 240 group A. hyperthyroid patients (n = 11)

TSH before propranolol <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 TSH during 160 mg propranolol <0.3 <0.3 <0.3 <0.3 <0.3 <0.3 T3 before propranolol 4.40.± 1.23 4.16 ± 1.23 4.33 ± 1.23 4.36 ± 1.23 4.84 ± 1.48 4.31 ± 1.39 T3 during 160 mg propranolol 3.20 ± 1.16 3.34 ± 1.29 3.47 ± 1.17 3.54 ± 1.17 3.67 ± 1.20 3.68 ± 1.40 group B. hypothyroid patients ori L-T4 (n = 6)

TSH before propranolol 1.4 4.0 3.9 2.7 2.3 2.0 «0.3 -10) (0.3 -47) (0.3 - 47) (0.3 - 28) (0.3- 20) K0.3 -16) TSH during 160 mg 2.2 5.2 5.6 4.4 3.3 2.2 propranolol ( 0.3 -19) (0.3- 86) (0.5 - 67) (0.4 - 46) (0.4- 31) «0.3 -23) T3 before propranolol 1.56 ± 0.39 1.59 ± 0.34 1.68 ± 0.46 1.65 ±0.51 1.72 + 0.32 1.77 ± 0.39 T3 during 160 mg propranolol 1.29 ± 0.42 1.31 + 0.49 1.26 ± 0.43 1.29 ± 0.34 1.40 ± 0.43 1.37 ± 0.35

!!!^?R?K^ -r—.--. - -- •• • •••-• •-'-•" ••<•• '• " 162 GROUP A- HYPERTHYROIDISM GROUP B- HYPOTHYROIDISM ON L-T« j 1 6.15 2.30 5.70 2.15 525 200 4.80 1.85 4.35 1.70 3.90 1.55 3.45 1.40 3.00 1.25 2.55 1.10 2.10 0.95

1.65 • 0.80 \ 1.20 0.65

160 0 160 propranolol propranolol nig/day

Figure VII.2. Decrease in plasma T, during propranolol treatment in each individual hyperthyroid and hypothyroid patient. v ;

'i- I GROUP A- HYPERTHYROIDISM GROUP B- HYPOTHYROIDISM ON L-T«

230 T3 nmol/, 5.05 200 430 1.70 3.55 1.40 " -. 260 1.10 2.05 0B0

7. 10 • 10 change ' from 0 0 baseline -10 T3 -20 -20 -30 -30 -40 -40 -50 -50

0 80 160 0 80 160 propranolol mg/day propranolol mg/day

Figure VII.3. Absolute and relative decrease in plasma T, during propranolol treatment in the hyper- and hypothyroid patients (mean ± SD). 163 propranolol medication were not higher in comparison with the pretreatment levels (p > 0.1); T3-levels during this test were lower i (p<0.05) on propranolol. Figure VII.3 depicts the absolute and relative decrease in plasma-T3 of groups A and B. The mean per cent change from pretreatment plasma-T3 in the hyperthyroid patients is -17% (range: +6% to -38%) after one week and -31% (range: -16% to -55%) after the second week. In the hypothyroid patients these values are -17% (range: -8% to -34%) and -28% (range: -13% to -55%). Statistical analysis indicated no differences in per cent decrease % between the hyper- and hypothyroid groups. The corresponding figures I for the mean per cent change in plasma-rT3 are: group A, after one week +20% (range: -23% to +85%) and +30% (range: -20% to+118%) after the second week; group B, after one week +46% (range: -8% to +233%) and +77% (range: +8% to +300%) after the second week.

VII.4. Discussion.

The results of this study indicate that propranolol treatment lowers plasma-T3 in hyperthyroid patients and in hypothyroid patients on L- thyroxine substitution therapy. The decrease in plasma-T3 cannot be explained by an in vitro effect of propranolol on the T3- radioimmunoassay. Propranolol even in a concentration of 250//g/ml plasma did not influence the assays of T3 and TSH; it has been demonstrated that plasma levels of propranolol during treatment with 160 mg per day are in the order of 0.1/*g/ml (Evans and Shand 1973; Niesand Shand 1975). Interference of propranolol or its metabolites with plasma protein binding sites for thyroid hormones is highly unlikely; the T3U ratio remained constant, and plasma-T4 did not decrease concomitantly. Furthermore, the rise in TSH-and T4-levels in the hyothyroid group points very strongly to a real effect of propranolol on the secretion and/or metabolism of thyroid hormones. However, impairment of T3-secretion by the thyroid in the hyperthyroid patients seems not to be a major effect of the drug (given in this dosage), since a decrease in plasma-T3 of the same magnitude was found in the hypothyroid patients. Peripheral conversion of exogenous T4 was the main, if not the only source of plasma T3 and rT3 in these patients, because any residual thyroid function was suppressed by substitution therapy. The concomitant rise in plasma rT3 strongly suggests impairment of the 5'-deiodination of T4 as the cause for the reciprocal changes in the triiodothyronines. A similar decrease in plasma T3 and increase in plasma rT3 concentrations has been found in subjects treated with propylthiouracil (Abuid and Larsen 1974; Saberi et al 1975; Geffner et al 1975; Westgren et al 1977; Siersbaek-Nielsen et al 1978; Laurbergand Weeke 1978), dexamethasone (Chopra et al 1975; Burr et al 1976; Westgren et al 1977), amiodarone (Burgeret al 1976; Jonckheer et al 1978)! -•r—j-

164 iopanoic acid (Biirgi et al 1976), and ipodate (Wu et al 1978). Kinetic studies of Lumholtz et al (1978,1979) in hypothyroid patients maintained on L-T4 replacement therapy have demonstrated a decrease in the T4-T3 conversion rate (but no change in the T4*rT3 conversion rate), and a decrease in the MCR of rT3 during propranolol treatment. These findings constitute conclusive evidence that propranolol inhibits the 5'- deiodination of T4. The reciprocal changes in plasma T3 and rT3 under propranolol medication have been reported by many authors (Nauman et al 1974; Murchison et al 1976; Harrower et al 1977; Lotti et al 1977; Theilade et al 1977; Verhoeven et al 1977; Wiersinga and Touber 1977; Kallner et al 1978; Saunders et al 1978; Tevaarwerk et al 1978; Faber et al 1978; Feely et al 1979). In retrospect, Nauman et al were the first who described the decrease in plasma T3, and Verhoeven et al first reported the increase in plasma rT3. The mean per cent decrease in plasma T3 in these studies varies between 16 and 40%, according to the dose and duration of propranolol medication; the mean per cent increase in plasma rT3 is reported as 16- 43%. The decrease in plasma T3 is correlated to the plasma propranolol concentration (Feely et al 1979). This may explain the variable responses seen in individual patients, since plasma propranolol levels may vary considerably (Rubenfeld et al 1979). i We observed an increase in T4 and TSH during propranolol medication in the hypothyroid patients on L-T4 substitution therapy. A rise in TSH is also reported by Feely et al (1979) and — although not significantly — by Faber et al (1979); however, an increase in T4 is not observed in other studies (Lumholtz et al 1978, 1979; Feely et al 1979; Faber et al 1979). Similar changes in T4 and TSH are reported in L-T4 treated subjects during administration of other drugs that inhibit the deiodination of T4: 1) Propylthiouracil medication results in an increase of TSH but not of T4 levels in the studies of Saberi et al (1975) and Geffner et al (1975), although no change in TSH (Siersbaek-Nielsen et al 1978) and an increase in T4 (Westgren et al 1977) is also reported. 2) Treatment with amiodarone resulted in a slight but not significant rise ii in T4; TSH, however, remained undetectable possibly due to the high substitution dose of L-T4 (300 /Ag) (Burger et al 1976). 3) Iopanoic acid medication increased TSH but was without effect on T4 i (Bflrgi et al 1976). 4) Administration of ipodate increased T4, but did not alter TSH; however, in 2 of the 4 patients studied TSH was undetectable (Wu et al 1978). 1 5) The studies on dexamethasone are of no use in this respect, since corticosteroids interfere with the release of TSH (Wilber and Utiger 1969; Nicoloff et al 1970; Sowers et al 1977). These studies therefore indicate that inhibition of the 5'-deiodinationof T4 165 r is often associated with a rise in T4 and TSH. With regard to our own results, we suggest that the increase in TSH is caused by the fall in plasma T3 — a direct effect of propranolol on the release of TSH could not be demonstrated in several studies (Woolf et al 1972; Epstein et al 1975; Wartofsky et al 1975; Lauridsen et al 1976)—and that the increase in T4 is caused by a lower MCR of T4 due to the inhibition if of the 5'-deiodination of T4. For obvious reasons an increased thyroidal release of T4 in the L-T4 treated subjects can be ruled out; moreover, a i direct effect of propranolol on thyroidal release was absent in a study of Wartofsky et al (1975). The kinetic studies of Lumholtz et al (1978,1979) in L-T4 treated hypothyroid patients on propranolol medication indeed demonstrate a decreased MCR of T4; however, the MCR of T3 and rT3 are also lowered by the drug, so that no clear conclusion can be drawn on the contribution of the impaired deiodination to the decrease in MCR of T4. Studies in euthyroid subjects treated with one of the above-mentioned drugs further support the proposed explanation, although the picture is complicated by the capability of the normal thyroid to respond to increased TSH levels. Plasma T4 concentration increases during treatment with amiodarone, iopanoic acid or ipodate, and these drugs also induce a rise in plasma TSH (with the exception of ipodate). Propranolol medication in euthyroid subjects also decreases plasma T3 P and increases plasma rT3 concentrations (Kristensen et al 1978; Faber et al 1979). Plasma TSH does, however, not change (Faber et al 1979), and i plasma T4 either remains unaltered (Lotti et al 1977; Shenkman etal 1977; Faber et al 1979) or increases (Williams and Jacobs 1970; Kristensen and Weeke 1977). Again, chronic propranolol treatment in the intact or hypophysectomized rat increases plasma T4 (Tal et al 1972). More detailed studies in a greater number of subjects with more frequent blood sampling may further elicit the effects of pharmacological inhibition of 5'-deiodination on the plasma levels of T4 and of TSH. Nevertheless, the rise in TSH and the rise in T4 (probably due to both TSH-stimulated thyroidal release and to a decreased MCR) subsequent to impaired 5'- deiodination of T4, as observed in a relatively large number of patients with AMI (see chapter VI), strongly favours the mechanism suggested.

The nature of the inhibitory effect of propranolol on the 5'-deiodination of T4 is unclear. Dratman (1974) has hypothesized that T4, an amino acid analog, has a neurotransmitter function in the adrenergic nervous system, and that T4 is deiodinated to T3 by tyrosine hydroxylase, the initial enzyme in the synthesis of the catecholamines. However, administration of a- methyl-p-tyrosine (a-MPT), an inhibitor of tyrosine hydroxylase, to L-T4 treated human subjects did not change the plasma T4 or T3 concentrations (Dvorak etal 1978). Also, a -MPT failed to influence the T4 -T3 conversion in vivo or in vitro in animal studies (Chopra 1977; Hiifner et al 1977; •!: 166 t |i Kaplan and Utiger 1978; Pascual et al 1978). The hypothesis of Dratman . |; ?f' therefore seems unlikely. On the other hand, in-vitro studies on the effect |£ Fjj of propranolol (chapter VIII) indicate that intact{3-adrenergic receptors l|\ g are necessary for the inhibitory effect on the deiodination of T4. It is thus ff. | conceivable, that changes in the adenylate-cyclase/ cyclic AMP system , ft | secondary to stimulation or blockade of thefl-adrenergic receptors play a % g role in the modulation of the 5'-deiodinating activity. g

! ?;i The biological significance of the 'low T3 syndrome' induced by |: I propranolol is incompletely understood. The decrease in plasma T3 may |; explain in part the benificial effects of 0-adrenergic blockade in : hyperthyroid patients: the per cent reduction of plasma T3 is directly ' related to the fall in oxygen consumption (Saunders et al 1978) and to the i degree of continued weight loss or weight gain (Feelyetal 1979). However, : • 0 -adrenoceptor blocking agents do not suppress all peripheral mani- |. festations of thyroid hormone excess (McDevitt et al 1968; Grossman '. ]- et al 1971; Malcolm 1972; Zwillich et al 1978), and have many actions other ' [ than the S'-deiodination inhibition. The decrease in T3 thus seems to be [ only partially responsible for the clinical effects of propranolol in jjj hyperthyroidism. [I It has been suggested that the propranolol withdrawal syndrome, \{ characterized by tachycardia, angina pectoris, sweating or tremor \\ (resembling the signs and symptoms of thyrotoxicosis), is due to a rebound ' • increase of the suppressed plasma T3 (Kristensen et al 1978). However, it is more likely that the withdrawal syndrome is causally related to the increased number of 0-adrenergic receptors induced by propranolol treatment (Glaubiger and Lefkowitz 1977), with a consequent transient hypersensitivity to j3-agonists after sudden withdrawal of propranolol (Boudoulas et al 1977; Nattel et al 1979). Naturally, an increase in plasma T3 may well aggravate the hypersensitivity to (5-agonists.

I VII.5. References.

I Abuid J, Larsen P R. Triiodothyronine and thyroxine in hyperthyroidism. Comparison of : | the acute changes during therapy with antithyroid agents. J Clin Invest 1974; 54: 201-8. I Bennett C A, Franklin N L. Statistical analysis in chemistry and the chemical industry. New § York: John Wiley and Sons, 1954. ; i Biran S, Tal E. Effect of beta-blocking drug propranolol on mI utilization in euthyroid | patients. J Clin Pharmacol 1972; 12: 105-7. - J Boudoulas H, Lewis R P, Kates R E, Dalamangas G. Hypersensitivity to adrenergic stimula- : I ; tion after propranolol withdrawal. Ann Int Med 1977; 87: 433-6. i j Burger A, Dinichert D, Nicod P, Jenny M.Lemarchand-BeraudTh, Vallotton MB. Effect of I r amiodarone on serum triiodothyronine, reverse triiodothyronine, thyroxin, and ;| ' ; thyrotropin. A drug influencing peripheral metabolism of thyroid hormones. J Clin • | Invest 1976; 58:255-9. \ | Bflrgi H, Wimpfheimer C, Burger A, Zaunbauer W, Rdsler H, Lemarchand-Beraud Th. \ -?~, Changes of circulating thyroxine, triiodothyronine and reverse triiodothyronine after fc radiographic contrast agents. J Clin Endocrinol Metab 1976; 43: 1203-10. ;i;i Burr W A, Ramsden DB,Griffiths R S, Black EG, Hoffenberg R, Meinhold H, Wenzel K W. '$£' Effects of a single dose of dexamethasone on serum concentrations of thyroid hormones^ re Lancet 1976; 2: 58-61. 1} Chopra 1 J, Williams D E, Orgiazzi J, Solomon D H. Opposite effects of dexamethasone on [ serum concentrations of 3,3',5'-triiodothyronine (reverse T,) and 3,3',5-triiodothyronine fj (T3) in man. J Clin Endocrinol Metab 1975; 41: 911-20. i| Chopra I J. A study of extrathyroidal conversion of thyroxine (T4) to 3,3',5-triiodothyronine* (T3) in vitro. Endocrinology 1977; 101: 453-63. Dratman M B. On the mechanism of action of thyroxine, an amino acid analogue of tyrosine. J Theor Biol 1974; 46: 255-70. Duncan D B. Multiple range and multiple F tests. Biometrics 1955; 11: I. Dvorak J C, Engelman K, Utiger R D. Failure of a-methyltyrosine to inhibit peripheral triiodothyronine formation. J Cltn Endocrinol Metab 1978; 47: 442-4. Epstein S, Pimstone B L, Vinik A I, McLaren H, Failure of adrenergica and (3 receptor blockade to elevate the TSH and prolactin response to TRH in hyperthyroidism. Clin Endocrinol 1975; 4: 501-4. Evans G H, Shand D G. Disposition of propranolol. V. Drug accumulation and steady-state concentrations during chronic oral administration in man. Clin Pharmacol Ther 1973; 14: 487-93. Faber J, Friis T, Kirkegaard C et al. Serum T4, T3 and reverse T3 during treatment with propranolol in hyperthyroidism, L-T4 treated myxedema and in normal man. Horm Metab Res 1979; 11:34-6. Feely J, Isles T E, Ratcliffe W A, Crooks J. Propranolol, triiodothyronine, reverse triiodothyronine and thyroid disease. Clin Endocrinol 1979; 10: 531-8. Geffner D L, Azukizawa M, Hershman H M. Propylthiouracil blocks extrathyroidal conversion of thyroxine to triiodothyronine and augments thyrotropin secretion in man. J Clin Invest 1975; 55: 224-9. Glaubiger G, Lefkowitz R J. Elevated beta-adrenergic receptor number after chronic propranolol treatment. Biochem Biophys Research Commun 1977; 78: 720-5. Grossman W, Robin N I, Johnson L W, Brooks K, Selenkow H A. Effects of beta-blockade on the peripheral manifestations of thyrotoxicosis. Ann Int Med 1971; 74: 875-9. Hadden D R, Montgomery DAD, Shanks R G, Weaver J A. Propranolol and iodine-131 in the management of thyrotoxicosis. Lancet 1968; 2: 852-4. Hadden D R, Bell T K, McDevitt D G, Shanks R G, Montgomery DAD, Weaver J A. Pro- pronolol and the utilisation of radioiodine by the human thyroid gland. Acta Endocrinol 1969; 61: 393-9. Harrower A D B, Fyffe J A, Horn D B, Strong J A. Thyroxine and triiodothyronine levels in hyperthyroid patients during treatment with propranolol. Clin Endocrinol 1977; 7:41-4. Hiifner M, Grussendorf M, Ntokalou M. Properties of the thyroxine (T4) monodeiodinating system in rat liver homogenate. Clin Chim Acta 1977; 78: 251-9. Jonckheer M H, Blockx P, Broeckaert I, Cornette C, Beckers C. 'Low T3 syndrome' in patients chronically treated with an iodine-containing drug, amiodarone. Clin Endocrinol 1978; 9: 27-35. Kallner G, Ljunggren J-G, Tryselius M. The effect of propranolol on serum levels of T4, T3 and reverse-Tj in hyperthyroidism. Acta Med Scand 1978; 204: 35-7. Kaplan MM, Utiger RD. Iodothyronine metabolism in rat liver homogenates. J Clin Invest > 1978; 61: 459-71. j. Krikler D M. Adrenergic receptor blockade. Lancet 1966; 1: 268. | Kristensen B 0, Weeke J. Propranolol induced increments in total and free serum thyroxine in I patients with essential hypertension. Clin Pharmacol Ther 1977; 22: 864-7. 168 Krktensen B 0, Steiness E, Weeke J. Propranolol withdrawal and thyroid hormones in patients with essential hypertension. Clin Pharmacol Ther 1978; 23:624-9. Landsberg L. Catecholamines and hyperthyroidism. Clin Endocrinol Metab 1977; 6:697-718 #!:- Laurberg P, Weeke J. Opposite variations in serum T3 and reverse T3 during propylthiouracil treatment of thyrotoxicosis. Acta Endocrinol 1978; 87: 88-94. Lauridsen U B, Faber J, Friis Th, Kirkegaard C, Nerup J. Thyrotropin (TSH) release during altered adrenergic a and j8 receptor influence. Horm Metab Res 1976; 8: 406-7. Levey G S. Catecholamine sensitivity, thyroid hormone and the heart, a reevaluation. Am J Med 1971; 50: 413-20. Levey G S. The heart and hyperthyroidism: use of beta-adrenergic blocking drugs. Med Clin N Am 1975; 59: 1193-1201. r Lotti G, Delitala G, Devilla L, Alagna S, Masala A. Reduction of plasma triiodothyronine (T,) induced by propranolol. Clin Endocrinol 1977; 6: 405-10. Lumholtz IB, Siersbaek-Nielsen K, Faber J, Kirkegaard C, Friis Th. Effect of propranolol on extrathyroidal metabolism of thyroxine and 3,3',5-triiodothyronine evaluated by non- compartmental kinetics. J Clin Endocrinol Metab 1978; 47: 587-9. Lumholtz IB, Busch-Serensen M, Faber J.Friis Th, Kirkegaard C, Siersbaek-Nielsen K. The influence of propranolol on the extrathyroidal metabolism of 3,3',5'-triiodothyronine (reverse T3). Acta Med Scand 1979; suppl 624: 31-4. Malcolm J. Adrenergic beta receptor inhibition and hyperthyroidism. Acta Cardiol 1972; 27, suppl XV: 307-26. McDevitt D G, Shanks R G, Hadden D R, Montgomery DAD, Weaver J A. The role of the thyroid in the control of heart-rate. Lancet 1968; 1: 998-1000. McLarty D G,.Brownlie B E W, Alexander W D, Papapetrou P D, Horton P. Remission of thyrotoxicosis during treatment with propranolol. Brit Med J 1973; 2: 332-4. Murchison L E, Bewsher P D, Chesters M I, Ferrier W R. Comparison of propranolol and practolol in the management of hyperthyroidism. Brit J Clin Pharmacol 1976; 3:273-7. Nattel S, Rangno R E, Loon G van. Mechanism of propranolol withdrawal phenomena. Circulation 1979; 59: 1158-64. Nauman J, Nauman A, Roszkowska K. Influence of propranolol on levels of thyroxine and triiodothyronine in hyperthyreotic patients. Mat Med Pol 1974; 6: 178-82. Nicoloff J T, Fisher D A, Appleman M D. The role of glucocorticoids in the regulation of thyroid function in man. J Clin Invest 1970; 49: 1922-9. Nies A S, Shand D G. Clinical pharmacology of propranolol. Circulation 1975; 52: 6-15. Pascual A, Obregon M J, Morreale de Escobar G. Tyrosine hydroxylase and the conversion of L-thyroxine into 3',3,5-triiodo-L-thyronine in the rat. Endocrinology 1979; 104:1574- 9. Ramsay I. Adrenergic j3-receptor blockade in hyperthyroidism. Brit J C!in Pharmacol 1975; 2: 385-8. Rubenfeld S, Silverman V E, Welch K M A, Mallette L E, Kohler P O. Variable plasma propranolol levels in thyrotoxicosis. New Engl J Med 1979; 300: 353-4. Saberi M, Sterling F H, Utiger R D. Reduction in extrathyroidal triiodothyronine production by propylthiouracil in man. J Clin Invest 1975; 55: 218-23. Saunders J, Hall S E H, Crowther A, SSnksen P H. The effect of propranolol on thyroid hormones and oxygen consumption in thyrotoxicosis. Clin Endocrinol 1978; 9: 67-72. Shanks R G, Hadden D R, Lowe D C, McDevitt D G, Montgomery DAD. Controlled trial of propranolol in thyrotoxicosis. Lancet 1969; 1: 993-4. Shenkman L, Podrid P, Lowenstein J, Hyperthyroidism after propranolol withdrawal. JAMA 1977; 238: 237-9. Siersbaek-Nielsen K, Kirkegaard C, Rogowski P, Faber J, Lumholtz B, Friis Th. Extra- thyroidal effects of propylthiouracil and carbimazole on serum T4, T3, reverse T3 and TRH-induced TSH-release in man. Acta Endocrinol 1978; 87: 80-7. .-M

169 f^ Sowers J R, Carlson H E, Brantbar N, Hershman J N. Effect of dexamethasone on prolactin •£•; and TSH responses to TRH and metoclopramide in man. J Clin Endocrinol Metab 1977; «'i 44:237-41. @ Spaulding S W, Noth R H. Thyroid-catecholamine interactions Med Clin N Am 1975; 59: % 1123-31. S: TalE, BiranS, Sulman F G. Influence of propranolol on serum thyroxine in the rat. JEndo- ? %\ crinol 1972; 53: 503-4. % Tevaarwerk GJM, Malik M H, Boyd D. Preliminary report on the use of propranolol in ¡v thyrotoxicosis: I. Effect on serum thyroxine, triiodothyronine and reverse triiodothy- 'i? ronine concentrations. Can Med Ass J 1978; 119: 350-1 and 360. ,¡: [}? Theilade P, Hansen J M, Skovsted L et al. Propranolol influences serum T3 and reverse T3 in ! hyperthyroidism. Lancet 1977; 2:363. : ¡if Verhoeven R P, Visser T J, Docter R, Hennemann G, Schalekamp M A D H. Plasma thyroxine, 3,3',5-triiodothyronine and 3,3',5'-triiodothyronine duringjS-adrenergic blockade in hyperthyroidism. J Clin Endocrinol Metab 1977; 44: 1002-5. f Wartofsky L.DimondRC, Noel G L.Frantz A G, EarllJ M. Failure of propranolol to alter ; r thyroid iodine release, thyroxine turnover, or TSH and PRL response to thyrotropin- : ; releasing hormone in patients with thyrotoxicosis. J Clin Endocrinol Metab 1975; 41: ; ¡; 485-90. • j | Westgren U, Melander A, Wâhlin E, Lindgren J. Divergent effects of 6-propylthiouracil on Í! 3,5,3'-triiodothyronine (T3) and 3,3',5'-triiodothyronine (rT3) serum levels in man. Acta -" U Endocrinol 1977; 85: 345 -50. 4 Westgren U, Ahren B, Burger A, Ingemansson S, Melander A. Effects of dexamethasone, desoxycorticosterone, and ACTH on serum concentrations of thyroxine, 3,5,3'-triiodo- ',] thyronine and 3,3',5'-triiodothyronine. Acta Med Scand 1977; 202: 89-92. Wiersinga W M, Touber J L. The influence of jß-adrenoceptor blocking agents on plasma , thyroxine and triiodothyronine. J Clin Endocrinol Metab 1977; 45: 293-8. Wilber J F, Utiger R D. The effects of glucocorticoids on thyrotropin secretion. J Clin Invest 1969; 48: 2096-103. Williams E S, Jacobs H S. Effect of propranolol on serum-thyroxine. Lancet 1970; 2:829-30. Woolf P D, Lee L A, Schalch D S. Adrenergic manipulation and thyrotropin-releasing hormone (TRH)-induced thyrotropin (TSH) release. J Clin Endocrinol Metab 1972: 35: 616-8. Wu S-Y, Chopra: I J, Solomon D H, Bennett L R. Changes in circulating iodothyronines in euthyroid and hyperthyroid subjects given ipodate (Oragrafin), an agent for oral cholecystography. J Clin Endocrinol Metab 1978; 46: 691-7. Zwillich C W, Matthay M, Potts D, Adler R, Hofeldt F, Weil J V. Thyrotoxicosis: comparison of effects of thyroid ablation and beta-adrenergic blockade on metabolic rate and ventilatory control. J Clin Endocrinol Metab 1978; 46: 491-500. 'I

CHAPTER VIII.

STUDIES ON THE EFFECT OF PROPRANOLOL IN ISOLATED RAT LIVER PARENCHYMAL CELLS

VIII. 1. Introduction.

In man several drugs have been shown to influence the peripheral conversion of T4. Propylthiouracil (Abuid and Larsen 1974; Saberi et al 1975; Geffner et al 1975; Westgren et al 1977; Siersbaek-Nielsenetal 1978; Laurberg and Weeke 1978), dexamethasone (Chopra et al 1975; Burr et al 1976; Westgren et al 1977), amiodarone (Burger et al 1976; Jonckheer etal 1978), iopanoic acid (Biirgi et al 1976), and ipodate (Wu et al 1978), all decrease plasma levels of T3 with a concomitant rise in plasma rT3 due to inhibition of the 5'-deiodination process. These drugs also inhibit the T4-T3 (and when investigated the rT3-»3,3'-T2) conversion in vitro as studied in rat liver or kidney cell preparations; the only exception are the corticosteroids, which do not inhibit the conversion of T4 when added directly in vitro, but decrease the conversion after pre-treatment of the animal in vivo (Visser et al 1975; Hiifner and Kndpfle 1976; Chopra 1977; Balsam et al 1978; Chiraseveenuprapund et al 1978; Chopra et al 1978; Heyma et al 1978; HOffken et al 1978; Htifner and Grussendorf 1978; Kaplan and Utiger 1978; Leonard and Rosenberg 1978; Visser et al 1978). We have previously demonstrated that the B-adrenoceptor blocking agent propranolol lowers the plasma level of T3 in hyperthyroid patients and hypothyroid patients on L-thyroxine substitution therapy (Wiersinga and Touber 1977). Again, inhibition of the peripheral conversion seemed the most likely explanation for this finding. Since the liver is a major site of T4- deiodination the decrease in plasma T3 - and presumably, in T3 production - could be caused by a direct effect of propranolol on the parenchymal liver cells. However, a decrease in T3 production as a consequence of a decrease in liver perfusion caused by propranolol (Nies and Shand 1975) seemed equally possible. In order to further elucidate the action of propranolol, we studied the effect of the drug on liver cells in vitro. The isolated liver cell preparation - rather than liver cell homogenate - was chosen because the action of propranolol is mediated via blockade of the 13-adrenergic receptors of the liver cell plasma membrane (Wolfe et al 1976). 171 VIII. 2. Materials and methods.

Liver cells of two groups of male Wistar rats were used; mean weight ± SD are215± 11 gin group A (n=4) and 298 ± 18gingroupB(n=12). The rats were anaesthetized with nembutal and received intravenously 0.2 ml heparin (1000 I.U.) dissolved in 0.8 ml 1% NaNO2 to prevent blood coagulation and vascular contraction. Parenchymal liver cells were isolated by perfusion of the liver with calcium-free phosphate buffered Ringer- solution saturated with carbogen (pH 7.4, 37°C) containing 20 g/1 bovine serum albumin and 0.6 g/1 collagenase (Sigma) (Berry and Friend 1969; James 1977). The isolated cells were suspended in 3 ml of a Krebs-Ringer- bicarbonate buffer (pH 7.4,37°C, containing 2.5 g/1 bovine serum albumin but no collagenase) at a concentration of 10-20 mg protein per ml and incubated at 37°C in slowly rotating tubes. Cell suspensions of each individual rat of group A were incubated in duplicate with 1 .3,MM sodium- L-thyroxine (Sigma) in the presence or absence of 50/iM carbimazole, 50 //M 6-propyl-2-thiouracil or 580/*M D,L-propranolol (I.C.I.). In the same manner, cells of the rats of group B were incubated with 1.3//M sodium L- thyroxine with or without D,L-propranolol at concentrations of 290,580 and 1160/JM respectively.

T4 and the various drugs were added to the incubation mixture and 500/vl samples were taken from the incubation mixture at times 0,15, 30 and 60 minutes. The samples were immediately transferred to 1 ml ice-cold absolute ethanol. After centrifugation (2000 g, 15 min, 4°C) the clear supernatants were stored at -20°C. As controls, liver cells were incubated in the absence of T4, and T4 was incubated in the absence of liver cells. The protein concentration of the incubation mixture was determined by the biuret method (Itzhaki and Gill 1964); the concentration served as a measure of the number of cells present.

The T3 and T4 concentrations of the ethanol extracts (diluted if necessary) were measured in duplicate by specific radioimmunoassays (Visser et al 1975; Chopra 1977), using 300 g/1 polyethyleneglycol (Cheung and Haunwhite Jr. 1976) for the separation of the bound and free hormone fractions. Samples presenting the standard curve were incubated under identical conditions, i.e. with a final ethanol concentration of 6.7% (v/v). The detection limit of the T3-assay is 0.1 nM, of the T4-assay 3nM. Cross- reactivity of T4 in the T3-assay is< 0.1%; cross-reactivity of T3 in the T4-assay is 0.3%. The coefficient of intra-assay variation of the T3-assay is 13.5% at a level of 1.1 nM, 4.5% at 8.2 nM and 6.9% at 21.7 nM. Carbimazole, propylthiouracil or propranolol in concentrations used in these experiments, did not interfere in the T3- and T4-assays. Corrections were made for base-line T3 hormone levels and for the losses of T3 and T4 in the ethanol extraction procedure (the recovery of the thyroid hormones is 76.0 ± 4.5% (x ± SD). The amount of T3 generated by the liver cells from the 172 added T4 was expressed as the conversion ratio, i.e. (nmol T3 formed/ nmol T4 added) x 100%. Statistical evaluation was performed by two ways analysis of variance with replications, mixed model. The factors are rats and treatments and the observations constitute slopes of regression lines. Each slope is determined by a least square fit to the conversion ratio versus time.

VIII3. Results.

The endogenous T4 concentration of the liver cells (i.e. cells incubated in the absence of T4) was below the detection limit of the T4-assay. The T3 concentration of the incubation mixtures varied between 0.1 and 1.0 nM in the absence of T4, and between 5 and 10 nM immediately afterthe addition of 1300 nM T4. The increase in base-line T3 in the presence of T4 is due to contamination with T3 « 0.5%) of the T4 preparation used and to the immunological cross-reactivity of T4 in the T3-assay. T3 generation could

o I 1 s

100 l4nnwl'l 200 60 90 120 TIME IminulM) 0.75 ISO T,nmol/I

Figure VIII. 1. Demonstration of similarity between standard curves (•) and dilutions of ethanol extracts (•) from isolated rat liver cells in the radioimmunoassays of T3 and T4.

Figure VIII.2. Conversion of exogenous T4 (1.3.//M) into T3 by isolated rat liver cells (protein content 7.9 mg/ml). . 173

Table VIII-1 "$•• Effect of carbimazole, propylthiouracil and propranolol on the conversion of exogenous

T4 (1.3 MM) into T3 by isolated rat liver cells. Conversion ratios are given as the mean W ± SEM (group A, n = 4; mean protein content ± SD is 18.9 ± 3.8 mg/ml).

Incubation time conversion ratio (nmol T3 formed/nmol T, added x 100%) in minutes controls carbimazole PTU propranolol 50 nM SOjuM 580/iM

IS 0.49 ± 0.10 0.39 ± 0.13 -0.05 ± 0.02 0.16 ± 0.03 30 0.86 ± 0.20 0.78 ± 0.18 -0.03 ± 0.06 0.21 ± 0.10 60 1.06 ± 0.25 1.10 ± 0.06 -0.03 ± 0.06 0.26 ± 0.10

not be detected in liver cells incubated in the absence of T4, nor in T4 incubated in the absence of liver cells. Measurements of T4 and T3 in diluted ethanol extracts resulted in dose response curves similar to the standard curves (fig. VIII. 1). Figure VIII.2 depicts an example of T3 generation from exogenous T4 by isolated rat liver cells; in general approximately 10 nM T3 is formed after 30 min. incubation.

o I \! i10

o.s

TIME (mlnulsi) TIME (minules)

Figure VIII.3. Conversion of exogenous T4 (1.3//M) into T5 by isolated rat liver cells (group A, n = 4; mean protein content ± SD is 18.9 ± 3.8 mg/ml). Conversion ratios are given as mean ± SEM. • • controls, *—* carbimazole 50//m,o o proprylthiouracil 50//M, • • propranolol

Figure VIII.4. Conversion of exogenous T4 (1.3.^M) into T3 by isolated rat liver cells (group B, n = 12; mean protein content ± SD is 18.1 ± 1.1 mg/ml). Conversion ratios are given as mean ± SEM. • • controls,* * propranolol 290 pM. • • propranolol 580//M, o o propranolol 1160//M. .--J,—-

174 Table VUI-2

Effect of various concentrations of propranolol on the conversion of exogenous T4 (1.3 MM) into T3 by isolated rat liver cells. Conversion ratios are given as the mean ± SEM (group B, n = 12; mean protein content ± SD is 18.1 + 1.1 mg/ml).

Incubation time conversion ratio (nmol T3 formed/nmol T4 added x 100%) in minutes controls propranolol propranolol propranolol t 290 MM 580 juM 1160 MM

15 0,52 * 0.06 0.34 ± 0.05 0.21 ± 0.02 0.26 ± 0.05 30 0.82 ±0.11 0.69 ±0.12 0.55 ± 0.10 0.56 ± 0.09 60 1.47 ± 0.20 1.46 ±0.16 1.11 ±0.13 0.99 ± 0.16

The results of the experiments with the cells of rats of group A are presented in table VIII. 1 and figure VIII. 3. In the statistical analysis indications were found for the existence of interactions and treatments (p < 0.01). We used linear contrast (Armitage 1971) for the evaluation between ft control and treatments. No difference in conversion ratios was found between control and 50 pM carbimazole (p> 0.1). However, both 50//M propylthiouracil and 580 pM propranolol lowered conversion ratios (p< 0.01). No interactions are found in the experiments of group B (results are given in table VIII.2 and figure VIII.4). The interactions- and residual sum of squares were pooled. The analysis was continued by a stepwise multiple comparison procedure for ordered parameters (Spjotvoll 1977). The level of significance was taken as a = 0.05. Addition of 580 and 1160//M propranolol both resulted in lower conversion ratios, but no difference was found between the 580 and 1160 juM dose. No inhibitory effect on the conversion was found for 290/*M propranolol.

V1II.4. Discussion.

Our results indicate that isolated rat liver cells are suitable for the in vitro study of the conversion of T4 into T3; T3 generation cannot be detected in the absence of T4 nor is there a change in T3 level when T4 is incubated without cells. We have applied conversion ratios because the deiodinating capacity of the liver cells is not saturated at the T4 concentrations used (Hesch et al 1975; Chopra 1977).

Conversion of T4 into T3 by isolated liver cells has been reported previously by Hesch et al (1975). Using a comparable amount of cells, they found a lower conversion ratio (0.1%) than in the present study (0.8%). Rat liver homogenate has also been shown to generate T3 from T4 (Visser et al 1975; HuTner and Knfipfle 1976; Visseretal 1976; Chopra 1977;Hiifneretal 1977); conversion ratios in this model were somewhat higher dependent on the concentrations of substrate and (cell) protein. The results of the

•nrr-y I

175 experiments with the cells of the rats of group A show quite clearly that whereas the addition of carbimazole has no effect on the conversion process, propylthiouracil inhibits T,-generation. This is in accordance with the results of studies using rat liver homogenate (Visser et al 1975; Chopra 1977). An effect of propranolol on the in vitro conversion of T4 has thus far not been reported. Our data show that this j}-adrenoceptorblockingagentalso inhibits the conversion of T4 into T3 by isolated rat liver cells. The effect of propranolol is not an artefact, since the drug did not influence either the T3 I levels of cells incubated without T4 or the radioimmunoassays of T3 and T4. Chiraseveenuprapund et al (1978) could not demonstrate an effect of propranolol (added in a dose of 2000/4M) on the T4 -»T3 conversion in rat kidney homogenate; the discrepancy with our results is most likely explained by the destruction of the B-adrenergic receptors in their homogenization procedure. We have been unable to establish an unequivocal dose dependent effect of propranolol. Although both 580 and 1160 y/M propranolol inhibit the conversion process while 290juM d oes not, there is no difference between the effects of 580 and 1160 /*M.

Since propranolol is rapidly metabolised by the liver (Nies and Shand 1975) low levels of propranolol may not be sufficient to initiate or to I maintain an inhibitory effect. This possibility might be explored in experiments using 0-adrenoceptor blocking agents which are not metabolised by the liver, such as atenolol and practolol. The concentrations of propranolol used in our incubation mixture may have been much higher than in vivo portal plasma levels of subjects given the usual therapeutic dosage. The results of the studies of Hayes and Cooper (1971) indicate that, in the dog, propranolol concentrations in liver tissue are in the order of 0.5//g/ g (about ten times higher than the corresponding blood levels) after the administration of 1 mg/kg orally. It is therefore conceivable that in our experiments other than the (3-blocking properties of the drug - such as the membrane stabilising activity - may have influenced the results. Studies with B-blocking agents which lack this activity (e.g. atenolol and metoprolol) are required to further elucidate the mechanism of action in this respect. In the rats of group A propranolol seems to inhibit T4 conversion more effectively than in the rats of group B. This difference may well be due to the difference in age: the rats of group B are older than those of group A. In conjunction with our earlier finding (Wiersinga and Touber 1977) of the in vivo lowering of plasma levels of T3 by propranolol, the I demonstration of the in vitro inhibitory effect of this drug underlines the importance of B-adrenoceptor agonists and antagonists in the metabolism and possibly ultimate biological effect of thyroid hormones. 176 VIII. 5. References.

Abuid J, Larsen PR. Triiodothyronine and thyroxine in hyperthyroidism. Comparison of the acute changes during therapy with antithyroid agents. J Clin Invest 1974; 54:201-8. Armitage P. Statistical methods in medical research. Blackwell Scientific publications, Oxford 1971. Balsam A, Ingbar SH. The influence of fasting, diabetes, and several pharmacological agents on the pathways of thyroxine metabolism in rat liver. J Clin Invest 1978; 62:415-24. Berry MN, Friend DS. High-yield preparation of isolated rat liver parenchymal cells. A biochemical and fine structural study. J Cellul Biol 1969; 43:506-20. Burger A, Dinichert D, Nicod P, Jenny M, Lemarchand-Beraud Th, Vallotton MB. Effect of amiodarone on serum triiodothyronine, reverse triiodothyronine, thyroxin, and thyrotropin. A drug influencing peripheral metabolism of thyroid hormones. J Clin Invest 1976; 58:255-9. Btirgi H, Wimpfheimer C, Burger A, Zaunbauer W, R6sler H, Lemarchand-Beraud Th. Changes of circulating thyroxine, triiodothyronine and reverse triiodothyronine after radiographic contrast agents. J Clin Endocrinoi Metab 1976; 43:1203-10. Burr WA, Ramsden DB, Griffiths RS, Black EG, Hoffenberg R, Meinhold H, Wenzel KW. Effect of a single dose of dexamethasone on serum concentrations of thyroid hormones. Lancet 1976; 2:58-61. Cheung MC, Haunwhite Jr WR. Use of polyethylene glycol in separating bound from unbound ligand in radioimmunoassay of thyroxine. Clin Chem 1976; 22:299-304. Chiraseveenuprapund P, Buergi U, Goswatni A, Rosenberg JN. Conversion of L-thyroxine to triiodothyronine in rat kidney homogenate. Endocrinology 1978; 102:612-22. Chopra IJ, Williams DE, Orgiazzi J, Solomon DH. Opposite effects of dexamethasone on serum concentrations of 3,3',5'-triiodothyronine (reverse T,) and 3,3',5-triiodothyronine (T3) in man. J Clin Endocrinoi Metab 1975; 41:911-20. Chopra IJ. A study of extrathyroidal conversion of thyroxine (T4) to 3,3',5-triiodothyronine (T3) in vitro. Endocrinology 1977; 101:453-63. Chopra IJ, Wu S-Y, Nakamura Y, Solomon DH. Monodeiodination of 3,5,3'-triiodothyronine and 3,3',5'-triiodothyronine to 3,3'-diiodothyronine in vitro. Endocrinology 1978; 102:1099-1106. Geffner DL, Azukizawa M, Hershman HM. Propylthiouracil blocks extrathyroidal conversion of thyroxine to triiodothyronine and augments thyrotropin secretion in man. J Clin Invest 1975; 55:224-9. Hayes A, Cooper RG. Studies on the absorption, distribution and excretion of propranolol in rat, dog and monkey. J Pharmacol ExpTher 1971; 176:302-11. Heyma P, Larkins RG; Stockigt JR, Campbell DG. The formation of tri-iodothyronine and reverse tri-iodothyronine from thyroxine in isolated rat renal tubules. Clin Sci Mol Med 1978; 55:567-72. Hesch RD, Brunner G, SOling HD. Conversion of thyroxine (T4) and triiodothyronine (T}) and the subcellular localisation of the converting enzyme. Clin Chim Acta 1975; 59: 209-13. Hoffken B, Kodding R, Muhlen A von zur, Hehrmann T, Juppner H, Hesch RD. Regulation of thyroid hormone metabolism in rat liver. Biochim Biophys Acta 1978; 539:114-24. Httfner M, Knopfle M. Pharmacological influences on T4 to T, conversion in rat liver. Clin Chim Acta 1976; 72:337-41. Hiifner M, Grussendorf M, Ntokalou M, Properties of the thyroxine (T4) monodeiodinating system in rat liver homogenate. Clin Chim Acta 1977; 78:251-9. Hiifner M, Grussendorf M. Investigations on the deiodination of thyroxine (T4) to 3,3'- diiodothyronine (3,3'-T2) in rat liver homogenate. Clin Chim Acta 1978; 85:243-51. L 177 -j ltzhaki RF, Gill DM. A micro-biuret method for estimating proteins. Anal Biochem 1964; fl 9:401-10. 'd. James J. The genesis of polypoidy in rat liver parenchymal cells. Cytobiology 1977; 15: » 410-9. :?- Jonckheer MH, Blockx P, Broeckaert I, Cornette C, Beckers C. 'Low T3 syndrome' in •: patients chronically treated with an iodine-containing drug, amiodarone. Clin Endocrinol •'.a I: 1978:9:27-35. I? Kaplan MM, Utiger RD. Iodothyronine metabolism in rat liver homogenates. J Clin Invest % 1978; 61:459-71. '•3 |.; Laurberg P, Weeke J. Opposite variations in serum T3 and reverse T3 during propylthiouracil |: treatment of thyrotoxicosis. Acta Endocrinol 1978; 87:88-94. ? Leonard JL, Rosenberg IN. Thyroxine 5'-deiodinase activity of rat kidney: observations I on activation by thiols and inhibition by propylthiouracil. Endocrinology 1978; 103: I 2137-44. i Nies AS, Shand DG. Clinical pharmacology of propranolol. Circulation 1975; 52; 6-15. ; Saberi M, Sterling FH, Utiger RD. Reduction in extrathyroidal triiodothyronine production i by propylthiouracil in man. J Clin Invest 1975; 55:218-23. <. Siersbaek-Nielsen K. Kirkegaard C, Rogowski P, Faber J, Lumholtz B, Friis Th. Extra- thyroidal effects of propylthiouracil and carbimazole on serum T4, T3, reverse T3 and TRH-induced TSH-release in man. Acta Endocrinol 1978; 87:80-7. 1, Spjotvoll E. Ordering ordered parameters. Biometrica 1977; 64:327-34. Visser TJ, van der Does-Tobe I, Docter R, Hennemann G. Conversion of thyroxine into tri-iodothyronine by rat liver homogenate. Biochem J 1975; 150:489-93. Visser TJ, van der Does-Tobe I, Docter R, Hennemann G. Subcetlular localization of a rat liver enzyme converting thyroxine into tri-iodothyronine and possible involvement of essential thiol groups. Biochem J 1976; 157:479-82. Visser TJ, Fekkes D, Docter R, Hennemann G. Sequential deiodination of thyroxine in rat liver homogenate. Biochem J 1978; 174:221-9. Westgren U, Melander A, Wahlin E, Lindgren J. Divergent effects of 6-propylthiouracil on 3,5,3'-triiodothyronine (T3) and 3,3',5'-triiodothyronine (rT3) serum levels in man. Acta Endocrinol 1977; 85: 345-50. Westgren U, Ahren B, Burger A, Ingemansson S, Melander A. Effects of dexamathasone, desoxycorticos^erone, and ACTH on serum concentrations of thyroxine, 3,5,3'-triiodothyronine and 3,3',5'-triiodothyronine. Acta Med Scand 1977; 202:89-92. Wiersinga WM, Touber JL. The influence of B-adrenoceptor blocking agents on plasma thyroxine and triiodothyronine. J Clin Endocrinol Metab 1977; 45:293-8. Wolfe BB, Harden TK, Molinoff PB. B-adrenergic receptors in rat liver: effects of adrena- lectomy. Proc Nat Acad Sci USA 1976; 73:1343-7. Wu S-Y, Chopra IJ, Solomon DH, Bennett LR. Changes in circulating iodothyronincs in euthyroid and hyperthyroid subjects given ipodate (Oragrafin), an agent for oral cholecystography. J Clin Endocrinol Metab 1978; 46:691-7.

:€> I

CHAPTER IX.

GENERAL DISCUSSION g

IX. 1. Physiological modulation of the peripheral conversion of T4.

The effect of physiological factors as sex, age and environment on the conversion of T4 have been discussed in chapter III.4a. In man, sex- or season-related differences have not been reported, but in the female rat the T4~T3 conversion is lower than in the male, due to a decrease in the concentration of the 5'-deiodinating activity. The effect of age in man is debated; the frequently observed trend towards lower plasma T3 and higher plasma rT3 concentrations in elderly subjects (whether or not caused by co-existent non-thyroidal disease) is, I however, almost certainly due to inhibition of the S'-deiodination process (Wenzel and Horn 1976). More unanimity exists on the fate of the iodothyronines in the fetal and neonatal period. There is no doubt that the low T3 and high rT3 levels in the fetus are caused by inhibition of the S'- deiodination of T4 (Harris et al 1978; see also chapter III.4c). The preferential pathway of monodeiodination of T4 to rT3 has also been demonstrated in the human fetus (Lightner et al 1977; Klein et al 1978). Interestingly, the S'-deiodinating pathway is activated soon after birth resulting in a rise of T3 and a afall of rT3 plasma levels (Chopra et al 1975) and in fact, the increase in T4-»T3 conversion already begins some days before labor in association with a rise in plasma cortisol, as demonstrated in the sheep and the rat (Klein et al 1978; Wu et al 1978). With regard to the mechanism of the 5'-deiodination inhibition during the fetal and neonatal period, conflicting data have been published. Chopra (1978) reported a marked dithiothreitol (DDT)-induced increase in the T4-»T3 conversion by fetal sheep liver homogenates, suggesting that the low conversion is related more to the availability of sulfhydryl groups than to a deficiency of the S'- deiodinating enzyme. In contrast, Harris et al (1979) concluded that the decreased T3 generation in the fetal rat is due to a decrease in the enzyme 1 concentration per se, while in the neonatal rat it is secondary to both a I reduced availability of sulfhydryl groups and to a deficiency of the I deiodinating enzyme. Finally, Woeber (1978) found no difference in T3- and rT3-generating activities between human polymorphonuclear leucocytes taken from normal adults and from cord blood. Differences in technical procedures and in species may explain some of the discrepancies. -r-vr-J-

179 ';;). The nutritional state of the body is another prominent factor in the ?1 peripheral conversion of T4. It is appreciated that body weight is one of the t: determinants of the plasma T3 concentration, high levels being found in £ obesity (Bray et al 1976; Danforth 1978) and low levels—associated with a !> high plasma rT3 — in malnutrition or simple weight loss (Chopra and • 'p Smith 1975; Ingenbleek and Beckers 1975; Vigersky et al 1977; Thomson et li al 1977) and in anorexia nervosa (Miyai et al 1975; Moshang et al 1975; § Bradlow et al 1976; Boyar et al 1977; Croxson et al 1977; Leslie et al 1978). j If Experimental obesity is associated with an increase of plasma T3 and Fl sometimes with a decrease of plasma rT3 concentrations; the PR of T3 is y increased in the overfed state, irrespective of the dietary composition, but l{ T4 kinetics do not change (Danforth 1978). In contrast, starvation results in a fall of plasma T3 and in a rise of plasma rT3 concentrations (Portnay et : al 1974; Vagenakis et al 1975; Merimee and Fineberg 1976; Carlson et al \ i 1977; Croxson etal 1977; Palmblad etal 1977; Azizi 1978; Visseretal 1978; I Burman et al 1979). Kinetic studies in fasting human subjects have i [, demonstrated a decrease in the PR ofT3 and in the MCR of rT3, but also an -t, | increased PR of rT3, indicating a lower T4^T3 and a higher T4*rT3 i conversion rate (Vagenakis et al 1977; Eisenstein et al 1978; Suda et al \. 1978). The impairment of the 5'-deiodination of T4 has been confirmed in '1 in vitro studies using liver homogenates from fasting animals (Balsam and | Ingbar 1979; Gavin etal 1979; Harris etal 1979; Kaplan 1979). The nature '': h of the defect in the 5'-deiodination process during fasting seems to be a i' decrease in the concentration of sulfhydry 1 groups (leading to less reduced v glutathione, the possible co-factor of the putative 5'-deiodinase); however, according to Gavin et al (1979) and Kaplan (1979) a change in the enzyme concentration is also involved. The above-mentioned studies raise the possibility that some specific dietary constituent rather than the total amount of ingested calories might be responsible for the observed changes in the conversion of T4. Several studies suggest that carbohydrates (CHO) are of prime importance in this respect. Refeeding fasting human subjects with glucose results in a return to normal of the plasma T3 and rT3 levels; refeeding with protein is without I effect (Croxson et al 1977; Westgren et al 1977; Azizi 1978; Burman et al | 1979). However, Davidson and Chopra (1979) recently suggested that the I effect of non-CHO calories on the 5'-deiodination of T4 is more f pronounced, if at least a normal amount of CHO (~200 g) is ingested. CKO : l and proteins, but not lipids, are modulators of the 5'-deiodination of T4 as * ? studied in rat liver homogenates (Harris et al 1978).

IX.2. Pathophysiological modulation of the peripheral conversion of T4. $

L; The effect of thyroidal disease upon the peripheral conversion of T4 in \ l man cannot be properly evaluated at the present time. The overproduction 180 of T3 relative to T4 in hyperthyroidism is most likely caused by both preferential secretion of T3 and increased T4-T3 conversion, since the conversion rate of T4 to T3 in untreated hyperthyroid patients is higher than after treatment. This increased conversion may well be induced by a higher concentration of the substrate, T4. The conversion rate of T4 to T3 is also higher in the untreated hypothyroid patient, despite the lower substrate concentration. It is therefore speculated that the extrathyroidal tissues have the capability to modulate the conversion of T4 in such a way that, relative to rT3, less T3 is produced in hyperthyroidism but more in hypothyroidism (cf. the ratio's of plasma iodothyronine concentrations found in patients with thyroid function disorders, chapter IV). No data are available on the generation of rT3 from T4 in states of thyroid hormone excess or deficiency, neither in man nor in animals. In vitro studies with liver and kidney homogenates from the rat and the dwarf mouse, have shown that the 5'-deiodinating activity is stimulated in hyperthyroidism, and is inhibited in hypothyroidism: the production of T3 from T4 and the degradation of rT3 are enhanced in homogenates from hyperthyroid animals but inhibited in hypothyroid animals (Grussendorf and Hufner 1977; Balsam et al 1978; Kaplan and Utiger 1978; Harris et al 1979). This modulation of the peripheral conversion of T4 is most likely due to changes in the concentration of the 5'- deiodinase since the 5'-deiodination process is normalized after pretreatment of the hypothyroid animal with thyroid hormones, but not after the in vitro addition of DTT (Harris et al 1979). It thus appears that T4 and T3 serve as enzyme inductors. These in vitro observations do not necessarily negate the speculation mentioned earlier: in this respect the ratio's of the T4-*T3 and T4*rT3 conversion rates are more important than the absolute values of the conversion rates. Many non-thyroidal diseases exert a profound effect on the peripheral conversion of T4. One wonders if not all non-thyroidal diseases, traumas, surgical procedures etc. are associated with this phenomenon. As discussed in chapter V, kinetic studies in patients with non-thyroidal disease indicate an inhibition of the 5'-deiodination of T4; the low plasma T3 is thus caused by a decreased conversion of T4 to T3, and the high plasma rT3 mainly by a decreased metabolic clearance rate of rT3.

IX.3. Pharmacological modulation of the peripheral conversion of T4.

Several pharmacological agents widely used in clinical medicine inhibit the S'-deiodination of T4 and are listed in table IX. 1. However, carbimazole (Abuid and Larsen 1974; Saberi et al 1975; Laurberg and Weeke 1978; Siersbaek-Nielsen et al 1978), lithium (Burman et al 1976; Blomqvist et al 1977; Chiraseveenuprapund et al 1978) and iodide (Burger et al 1976; Chiraseveenuprapund et al 1978) do not inhibit the peripheral

.-_,. V conversion of T4. Conflicting reports have been published on the effect of & diphenylhydantoin (Cullen et al 1973; Hflfnerand Knopfle 1976; Tsukuiet •$ al 1978), and the influence of sex steroids depends apparently on the I' experimental circumstances (Chen and Walfish 1978; Harris et al 1979) (cf. i chapter III.4a). Polychlorinated biphenyls, which are widespread I environmental pollutants, enhance the S'-deiodination process in animals | (Bastomsky et al 1976). i i Most agents listed in table IX, 1 have been tested in vitro where similar t\ effects as in vivo have been demonstrated, with the exception of the

Table IX-1 The effect of physiological, pathological and pharmacological factors on the 5'-deiodination process in the peripheral conversion of T4 (inhibition -, no effect 0, stimulation +; * animal data only).

PHYSIOLOGICAL FACTORS 1. age : fetus — neonate + elderly subjects (-) , :i 2. sex* : females (relative to males) - '~\ males (relative to females) + • = 3. weight : weight loss - , rj obesity + 4. diet : fasting - carbohydrates +

PATHOLOGICAL FACTORS 1. thyroid disease: hyperthyroidism + hypothyroidism — 2. non-thyroidal disease, traumas, surgical procedures

PHARMACOLOGICAL FACTORS 1. propylthiouracil 2. carbimazole 3. corticosteroids 4. amiodaxone (Cordarone) 5. iodide 6. iopanoic acid (Bilijodon, Telepaque) 7. amidotrizoic acid (Urografin) o a 8. ioglycamic acid (Bilivistan) 9. ipodate (Oragrafin) o I 10. propranolol 11. lithium 12. growth hormone 13. estrogens* 14. androgens* • i 182 corticosteroids (chapter VIII. 1). The mechanism of action of these agents has, in most cases, not satisfactorily been elucidated. The enhancing effect of growth hormone is probably related to the anabolic action of the hormone wHh its attendant need for more energy (Sato et al 1977). With respect to amiodarone and the radiographic contrast agents, it has been suggested that some similarity in the structure of these agents with the thyroxine molecule may be of significance (Burger et al 1976; Biirgi et al 1976; Wu et al 1978), as has been demonstrated for other iodinated benzene derivatives (Escobar del Rey and Morreale de Escobar 1962) and thyroid hormone analogs (Larson and Albright 1961; Pittman et al 1964). Visser (1979) has recently clarified the intriguing phenomenon that, in contrast to PTU, carbimazole has no effect on the conversion of T4. It is proposed (cf. chapter I) that the intermediate sulphenyl iodide of the enzyme (E-SI) is reduced by the cofactor to the native enzyme (E-SH), or reacts with thioureylenes yielding mixed disulfides; it was found that the methimazole-enzyme complex is more easily reduced by the cofactor. The study of the pharmacological modulation of the peripheral conversion of T4 has already provided a new agent for the treatment of hyperthyroid patients: ipodate (Wu et al 1978).

IX.4. The peripheral conversion of TA and the pituitary-thyroid axis: regulatory mechanisms.

The deiodination of T4 is governed by two putative enzymes, 5'- deiodinase and 5-deiodinase (chapter I). Physiological, pathophysiological or pharmacological modulation of the peripheral conversion of T4 very often results in the 'low T3, high rT3 syndrome', which is caused by a decreased T4-»T3 conversion rate and a decreased MCR of rT3, sometimes associated with a small increase in the T4^rT3 conversion rate. Thus, inhibition of the S'-deiodinating activity, rather than stimulation of the 5- deiodinating activity is of great importance, and outer- and inner-ring monodeiodination of T4 is not a random process as initially proposed by Surks and Oppenheimer (1971). The factors that regulate the 5- deiodination process are still obscure. The substrate and enzyme concentration and the availability of non-protein sulfhydryl (SH) groups in the tissues have been recognized as important regulators of the 5'- deiodination — i.e. the conversion of T4-»T3 and rT3-»3,3'-T2. The decrease in the S'-deiodinating activity has, in general, been explained by a lower substrate concentration in hypothyroidism, by a lower concentration of the enzyme in the fetus and in females, and by a lower concentration of SH- groups during starvation and non-thyroidal illness (cf. chapter V.4). It is assumed that the 5-deiodinase is le??. sensitive to lower concentrations of reduced glutathione than the S'-deiodinase. Undoubtedly more regulatory pathways for the deiodination of T4 will be recognized in the near future. It 183 is intriguing that oral glucose, in contrast to intravenous glucose, is much more effective in the restoration of the 'low T3, high rT3 syndrome' (Westgren et al 1977; Burr et al 1979). Factors from the gut (gastrointestinal hormones?) may therefore participate in the regulation of the deiodination of T4. t The 'low T3, high rT3 syndrome' is not without consequences for the setting of the pituitary-thyroid axis. This is best demonstrated by the rise in plasma TSH secondary to pharmacologically induced S'-deiodination inhibition in L-T4 treated subjects (cf. chapter VII). A slight transient increase in TSH was also observed in the patients with acute myocardial infarction (chapter VI), but in a more heterogeneous group of patients with various non-thyroidal diseases no changes in the TSH levels or in the TSH response to TRH could be detected (chapter V). It is therefore surprising that a more pronounced increase in plasma TSH is not seen in patients with a non-thyroidal illness, in whom the plasma T3 concentration decreases to a very low or even undetectable level. Apparently there is an adaptation in the set point of the pituitary thyrotrophic cell for the release of TSH. In theory, this could be done by modulation of the release of TRH, or of the receptors on the thyrotroph, and also by a more efficient intrapituitary monodeiodination of T4 to T3 (cf. the studies of Larsen and co-workers, described in chapter V.4). An altered sensitivity of the pituitary thyrotroph has been reported in iodine deficiency (Chopra et al 1978), and also in the fasting state, where a blunted TSH response to TRH is observed by most authors (Portnay et al 1974; Croxson et al 1977; Carlson et al 1977; Azizi 1978; Burman et al 1979).

There are problems with thyroxine when the conversion of T4 is changed. In some patients with non-thyroidal disease, plasma T4 is found to be low, which is probably due to a decreased binding capacity of the plasma proteins for thyroid hormones (cf. chapter V). However, in a homogeneous group of patients with acute myocardial infarction we observed a transient rise in plasma T4, presumably caused by some TSH- induced thyroidal release of T4 and by impaired degradation of T4 (chapter VI). A lowered MCR of T4 due to impaired deiodination is again strongly suggested by the rise in T4 in the L-T4 treated subjects with pharmacologically-induced inhibition of the S'-deiodination of T4 (chapter VII). Although the data in the literature are rather confusing, higher plasma T4 levels have been reported when the S'-deiodination is inhibited (by fasting — Ingbar and Galton 1975, by PTU — Escobar del Rey and 1 Morreale de Escobar 1961; Slingerland and Burrows 1962; Hershman 1964; Frumess and Larsen 1975; for other agents see chapter VII.4) and lower plasma T4 levels have been demonstrated when the 5'-deiodination is activated (by growth hormone — Sato et al 1977, by polychlorinated 184 biphenyls — Bastomsky et al 1976, or by estradiol benzoate — Chen and Walfish 1978). Also, the clearance of radioactive PBI or T4 is reported to be delayed by PTU (Oppenheimer et al 1972) and during fasting (Ingbar and Galton 1975; Grant et al 1978), while an increased MCR of T4 has been noted in obesity (Felt et al 1976). In our opinion many cases of the so-called "T4-toxicosis" (a condition not infrequently encountered in elderly subjects) are due to co-existent non-thyroidal disease, the relatively low T3 and the high T4 plasma levels being caused by inhibition of the 5'-deiodination of T4 (cf. chapter IV. 4c).

IX.5. The biological significance of the peripheral conversion of T4.

It is clear from the fore-going discussion that the extrathyroidal deiodination of T4 into either T3 or rT3'is not a random process, but that two independent although related mechanisms are involved. It is clear that modulation of the peripheral conversion of T4 is a common feature of non- thyroidal disease, and is qualitatively and quantitatively related to the severity of the disease (chapters V and VI); it constitutes a basic response of the body to stress. The existence of the 'low T3, high rT3 syndrome' in 'sick euthyroid patients' suggests that the shift in the conversion of T4 (resulting in less metabolically active T3 and more metabolically inactive rT3) is an adaptive mechanism, by which the body seeks to protect itself against excessive catabolism at a time when anabolism is challenged. The question then arises whether, under these circumstances, some sort of 'hypothyroidism' exists at the level of the peripheral tissues, not withstanding the fact that the functional state of the hypothalamic- pituitary-thyroid axis is indicative of euthyroidism. If the teleological points raised above are valid, rT3 should have less catabolic activity than T3. Chopra et al (1978) have demonstrated in the rat, that rT3 administration is without effect on body growth, urinary urea nitrogen and plasma levels of T4, T3 and TSH. In contrast, T3 medication resulted in an increase of urinary urea nitrogen and in a decrease of body growth; thus, T3 exerts a catabblic effect as opposed to rT3. In another series of experiments in rats, Cho.pra et al (1978) compared the effect of T4 treatment to that of treatment with T4 together with sodium ipodate (ipodate strongly inhibits the T4-*T3 conversion). T4 alone increased plasma T3 and urinary urea nitrogen and decreased body grov/th as compared to controls, but the concomitant administration of ipodate completely abolished the effects of T4, although plasma T4 had increased dramatically because of the decrease in deiodinative degradation. These results constitute strong evidence that in these experiments T4 in itself has minimal catabolic action, and that the catabolic effects are mainly due to the generation of T3 from T4. Further evidence that modulation of the conversion of T4 has homeostatic significance comes from starvation

-r~r 185 experiments. Gardner et al (1979) studied seven volunteers who fasted for 80 hours at two occasions: first without T3 replacement, and subsequently with a 'physiological' replacement dose of T3. Serum T4 did not change and serum rT3 increased in the two fasting periods, but the decrease in serum T3 was prevented in the T3 fast. The excretion of urea was 9.1% higher during the T3 fast. They conclude that the decrease in serum T3 during fasting spares muscle protein. A similar conclusion is reached by Carter et al (1975). Studies in patients with anorexia nervosa have also demonstrated hypothyroid-like alterations in the peripheral tissues. The Achilles reflex half-relaxation time in anorectic patients with a subnormal plasma T3 is longer than in those with a normal T3 (Croxson and Ibbertson 1977). The observed lowering of the androsterone/etiocholanolone ratio — due to decreased 5 a reductase — (Bradlow et al 1976) and the elevated tetrahydrocortisol/tetrahydrocortisone ratio — due to decreased 11- hydroxysteroid dehydrogenase — (Boyar et al 1977) in anorexia nervosa, are abnormalities in the androgen and cortisol metabolism also found in hypothyroid patients (Hellman et al 1959; 1961). In patients with anorexia nervosa the abnormal ratio's revert to normal when the patient is treated with T3, which suggests that the abnormalities are due to the 'low T3 syndrome' caused by the weight loss per se. Interestingly, a reduced androsterone/etiocholanolone ratio associated with the 'low T3 syndrome' is also observed in lung cancer—where it is known to be a poor prognostic sign — (Ratcliffe et al 1978), and an increase in the tetrahydrocortisol/ tetrahydrocortisone ratio is one of the nonspecific consequences of illness (Zumoff et al 1973). The situation with respect to the presumably enhanced 5'-deiodination of T4 in obesity is less clear. It has been reported that the body can adapt to caloric restriction by decreasing its energy needs and by becoming metabolically more efficient; these changes partially offset the expected weight loss in obese subjects (Bray 1969; Grant et al 1978). On the other hand, it is difficult for normal subjects to gain much weight by overeating. The debate on increased thermogenesis in overnutrition is continuing since 1900. The generation of extra heat in the face of increased food intake (the old concept of'Luxuskonsumption') may play a role in the maintenance of a stable body weight, and is possibly induced by an increase in plasma T3 levels (Sims 1976; Davidson and Chopra 1979). Subjects with morbid obesity may be limited in their ability to produce heat; their reduced thermogenesis seems to be constitutive rather than the result either of developmental or other environmental factors (Bray et al 1976; James and Trayhurn 1976; Bray 1977; Jung et al 1979).

The significance of the low T3 and high rT3 plasma concentrations in fetal life is unknown, but could be related to the energy metabolism of the fetus. In this respect it is of interest that in monkey hepatocarcinoma cells 186 the T4-»rT3 pathway is very much enhanced and the T4-iT3 pathway strongly inhibited (Sorimachi and Robbins 1977, 1978). It has been speculated that this represents dedifferentiation to a more immature cell line resembling fetal hepatocytes. The striking resemblance between the fate of rT,and that of a M-fetoprotein (an acute phase protein that inhibits the inflammatory reaction, found in high concentrations in the plasma of the fetus, in cord blood, after injury and during liver carcinogenesis) further underlines the basic significance of the conversion of T4. It thus appears that modulation of the conversion of T4 is an important homeostatic mechanism of the body in the regulation of its energy needs. But, as Tweedledee said: "If it was so it might be; and if it were so, it would be; but as it isn't, it ain't. That's logic". It has been suggested that the monodeiodination of T4 to T3 is essential for the action of T4, and that T4 itself possesses little or no biological activity, T3 being the active thyroid hormone (among others, Oppenheimer et al 1972; Frumess and Larsen 1975). Thus viewed, the function of T4 is merely a prohormone of T3. Chopra et al (1973) have argued against this proposition because 1) a normal plasma T3 concentration in the presence of a low plasma T4 is not sufficient to sustain euthyroidism, 2) a supranormal plasma T3 concentration is required in T3-treated hypothyroid patients to maintain euthyroidism, and 3) despite a low plasma T3 concentration, the fetus is assumed to be euthyroid. However, in the light of the recent knowledge on the intracellular T4-»T3 deiodination and the binding of T3 to nuclear receptors the validity of these arguments is now questionable. In this respect, the studies of Larsen and co-workers (cf. chapter V.4) are especially important, as they have shown that the intrapituitary T4-»T3 monodeiodination generates half of the nuclear T3 in the thyrotropic cell, the other half being derived from plasma T3. Studies from the Escobar's have also repeatedly demonstrated the importance of the T4-»T3 conversion in vivo in the expression of T4 activity (Hervas et al 1976; Obregon et al 1979). Although it cannot be denied that T4 may have some intrinsic activity (du Breuil and Galton 1978; Chopra et al 1978; Surks and Oppenheimer 1978) it seems to us that thyroid hormone activity in the tissues is expressed mainly by T3. This is not to say that T4 has no functions other than that of a prohormone. The significance of its binding tc nuclear receptors and its occupancy of receptor sites has so far not been elucidated.

IX.6. Epilogue.

The rapid development of endocrinology has deepened our insight in hormonal mechanisms, but also has brought confusion because classical concepts have been undermined. For instance, the very concept of a 187 hormone is now unclear since agents that transmit signals by endocrine, paracrine or neurocrine mechanisms are all referred to as hormones. And what is synthesized within an endocrine cell, is not necessarily the same product that leaves the cell (e.g. preproparathormone, proparathormone and parathormone), and what is secreted by an endocrine cell into the V: bloodstream is not necessarily the same product that ultimately provokes the hormonal effect in the peripheral tissues (e.g. testosterone and 6. dihydrotestosterone). And the hormonal effects in the target cells may be modulated by the receptor state (e.g. 'up' and 'down' regulation of insulin receptors), not to mention the possibiliy of post-receptor alterations (Hoffenberg 1979). The fate of thyroxine is an excellent illustration of these points. Firstly, thyroxine synthesized in the thyroid gland may be converted intrathyroidally to T3 as has been demonstrated in animals (Bidey et al 1976; Mindroiu et al 1976; Green 1978; Laurberg 1978, 1979). Secondly, thyroxine is deiodinated in the peripheral tissues into either the potent hormone T3 or the metabolically inactive rT3. And thirdly, the receptors for thyroid hormones can be altered - fasting induces a decrease in nuclear I-" T3 receptors (DeGroot et al 1977; Schussler and Orlando 1978) —, but thyroid hormones can also modulate the receptors of other hormones: thyroid hormone excess induces an increase of glucagon receptors (Madsen and Sonne 1976), and the number of /J-adrenergic receptors of skeletal muscle, heart, fat cells and erythrocytes is increased in hyperthyroidism and decreased in hypothyroidism (Kunos et al 1974; Banerjee and Kung 1977; Ciaraldi et al 1977; Kunos 1977; Williams et al 1977; Giudicelli 1978; Malbon et al 1978; Sharma and Banerjee 1978; Bilczikian et al 1979; Popovic et al 1979). It thus appears that the peripheral tissues have the capability to modulate the expression of the hormonal effect to a great extent, by the regulation of the amount of T3 generated from T4 and by the regulation of the affinity and capacity of the thyroid hormone receptors. Further study of these mechanisms in the target cells is therefore needed to establish the real significance of the hypothyroid-like alterations observed in the 'low T3, high rT3 syndrome' (Wiersinga and Touber 1978). It may even be necessary to redefine the concepts of hyperthyroidism and of hypothyroidism. % IX. 7. References. a Abuid J, Larsen PR, Triiodothyronine and thyroxine in hyperthyroidism. Comparison of 1-3 acute changes during therapy with antithyroid agents. J Clin Invest 1974; 54:201-8. Azizi F. Effect of dietary composition on fasting-induced changes in serum thyroid hormones and thyrotropin. Metabolism 1978; 27:935-42. Balsam A, Sexton F, Ingbar SH. The effect of thyroidectomy, hypophysectomy, and hormone 188 : | replacement on the generation of triiodothyronine from thyroxine in rat liver and kidney. Vv*; Endocrinology 1978; 103:1759-67. , <«« Balsam A, Ingbar SH. Observations on the factors that control the generation of triiodo- •; ,«,•. thyronine from thyroxine in rat liver and the nature of the defect induced by fasting. J Clin .. 'pjZ Invest 1979; 63:1145-56. , >%' Banerjee IP, Kung LS. B-Adrenergic receptors in rat heart: effects of thyroidectomy. Europ J ''• f# Pharmacol 1977; 43:207-8. : f;i Bastomsky CH, Murthy PVN. Banovac K. Alterations in thyroxine metabolism produced by |, v cutaneous application of microscopic immersion oil: effects due to polychlorinated • ; f;' biphenyls. Endocrinology 1976; 98: 1309-14. v --,; ||, Bidey SP, Anderson J, Marsden P, McKerron CG. Post-secretory deiodination of iodothyronines released from normal human thyroid cells in vitro. Biochem Biophys Research Commun 1976; 72:67-73. Bilczikian JP, LoebJR, Gammon DE. The influence of hyperthyroidismandhypothyroidism on theB-adrenergic responsiveness of the turkey erythrocyte. J Clin Invest 1979; 63:184-92. Blomqvist N, Lindstedt G, Lundberg P-A, Walinder J. No inhibition by Li+ of thyroxine monodeiodination to 3,5,3'-triiodothyronine and 3,3',5'-triiodothyronine (reverse triiodothyronine). Clin Chim Acta 1977; 79:457-64. i Boyar RM, Lem D, Roffware H et al. Cortisol secretion and metabolism in anorexia nervosa. j New Engl J Med 1977; 296:190-3. ; Bradlow HL, Boyar RM, O'Connor J, Zumoff B, Hellman L. Hypothyroid-like alterations : in testosterone metabolism in anorexia nervosa. J Clin Endocrinol Metab 1976; 43:571-4. Bray GA. Effect of caloric restriction on energy expenditure in obese patients. Lancet 1969; 2: , 397-8. I Bray G A, Fisher DA, Chopra IJ. Relation of thy roid hormones to body weight. Lancet 1976; 1: | 1206-8. 1 Bray GA. Obesity and thyroid hormones. Pharmac Ther C 1977; 2:205-14. Breuil A du, Galton VA. Thyroxine: studies concerning its intrinsic physiological activity. Acta ' Endocrinol 1978; 88:87-93. • Burger A, Dinicheit D, Nicod P, Jenny M, Lemarchand-Beraud T, Vallotton MB. Effect of amiodarone on serum triiodothyronine, reverse triiodothyronine, thyroxine, and thyrotropin. A drug influencing peripheral metabolism of thyroid hormones. J Clin Invest 1976; 58:255-9. Burman KD, Dimond RC, Earll JM, Wright FD, Wartofsky L. Sensitivity to lithium in treated Graves' disease: effects on serum T4, T3 and reverse T,. J Clin Endocrinol Metab 1976; 43:606-13. Burman KD, Dimond RC, Harvey GS et al. Glucose modulation of alterations in serum iodo- { thyronine concentrations induced by fasting. Metabolism 1979; 28:291-9. i Burr W, Beer F de, Black E, Harrison S. Serum TSH increase after oral carbohydrate j (abstract). Ann d'Endocrinol 1979; 40:47A. .{. Carlson HE, Drenick EJ, Chopra IJ, Hershman JM. Alterations in basal and TRH-. >i stimulated serum levels of thyrotropin, prolactin, and thyroid hormones in starved \[ obese men. J Clin Endocrinol Metab 1977; 45:707-13. : • Carter WJ, Shakir KM, Hodges S, Faas FH, Wynn JO. Effect of thyroid hormone on meta- -j] bolic adaptation to fasting. Metabolism 1975; 24:1177-83. | Chen HJ, Walfish PG. Effects of estradiol benzoate on thyroid-pituitary function in female • | rats. Endocrinology 1978; 103:1023-30. | Chiraseveenuprapund P, Buergi U, Goswami A, Rosenberg JN. Conversion of L-thyroxine 1 to triiodothyronine in rat kidney homogenate. Endocrinology 1978; 102: 612-22. f Chopra IJ, Solomon DH, Chua Teco GN. Thyroxine: just a prohormone or a hormone too? .*•; J Clin Endocrinol Metab 1973; 36:1050-7. 1 Chopra IJ, Sack J, Fisher DA. Circulating 3,3',5'-trisodothyronine (reverse T3) in the human ~ newborn. J Clin Invest 1975; 55:1137-41. - r

189 Chopra IJ, Smith SR. Circulating thyroid hormones and thyrotropin in adult patients with protein-calorie malnutrition. J Clin Endocrinol Metab 1975; 40:221-7. Chopra IJ, Carlson HE, Solomon DH. Comparison of inhibitory effects of 3,5,3'-triiodo- Si thyronine (T3), thyroxine (T4), 3,3',5'-triiodothyronine (rT3), and 3,3'-diiodothyronine (T2) on thyrotropin-releasing hormone-induced release of thyrotropin in the rat in vitro. Endocrinology 1978; 103:393-402. Chopra IJ. Sulfhydryl groups and the monodeiodination of thyroxine to triiodothyronine. Science 1978; 199:904-6. Chopra IJ, Solomon DH, Chopra U, Wu S-Y, Fisher DA, Nakamura Y. Pathways of metabolism of thyroid hormones. Recent Progr Horm Research 1978; 34: 521-67. Ciaraldi T, Marinetti GV. Thyroxine and propylthiouracil effects in vivo on alpha and beta adrenergic receptors in rat heart. Biochem Biophys Research Commun 1977; 74:984-91. Croxson MS, Hall TD, Kletzky OA, Jaramillo JE, Nicoloff JT. Decreased serum thyrotropin induced by fasting. J Clin Endocrinol Metab 1977; 45:560-8. Croxson MS, Ibbertson HK. Low serum triiodothyronine (T3) and hypothyroidism in anorexia nervosa. J Clin Endocrinol Metab 1977; 44:167- 74. Cullen MJ, Burger AG, Ingbar SH. Effects of diphenylhydantoin on peripheral thyroid hormone economy and the conversion of T4 to T3 (abstract). Program Vth Annual Meeting of the European Thyroid Association, 1973; 46. Danforth E. Discussion. In: Chopra IJ, Solomon DH, Chopra U, Wu S-Y, Fisher DA, Nakamura Y. Pathways of metabolism of thyroid hormones. Recent Progr Horm Research 1978; 34:521-67. Davidson MB, Chopra IJ. Effect of carbohydrate and noncarbohydrate sources of calories on plasma 3,5,3'-triiodothyronine concentrations in man. J Clin Endocrinol Metab 1979; 48:577-81. DeGroot LJ, Coleoni AH, Rue PA, Seo H, Martino E, Refetoff S. Reduced nuclear triiodothyronine receptors in starvation-induced hypothyroidism. Biochem Biophys Research Commun 1977; 79:173-8. Eisenstein Z, Hagg S, Vagenakis AG et al. Effect of starvation on the production and peripheral metabolism of 3,3',5'-triiodothyronine in euthyroid obese subjects. J Clin Endocri- nol Metab 1978; 47: 889-93. Escobar del Rey F, Morreale de Escobar G. The effect of propylthiouracil, methylthiouracil and thiouracil on the peripheral metabolism of L-thyroxine in thyroidectomized L- thyroxine maintained rats. Endocrinology 1961; 69: 456-65. Escobar del Rey F, Morreale de Escobar G. The effect of butyl-4-hydroxy-3:5-diiodoben- zoate on the availability of thyroid hormones to peripheral tissues in isotopically equilibrated rats. Acta Endocrinol 1962; suppl. 40: 1-15. Felt V, Strasek J, Bednar J. Thyroid hormones and obesity. Rev Czechoslovak Med 1976; 22: 216-24.

Frumess RD, Larsen PR. Correlation of serum triiodothyronine (T3) and thyroxine (T4) with biologic effects of thyroid hormone replacement in propylthiouracil-treated rats. Meta- bolism 1975; 24: 547-54. Gardner DF, Kaplan MM, Stanley CA, Utiger RD. Effect of tri-iodothyronine replacement on the metabolic and pituitary responses to starvation. New Engl J Med 1979; 300:579-84. Gavin LA, McMahon FA, Livermore B, Cavalieri RR. The effect of fasting on T4-deiodination, involves changes in enzyme concentration and thiol levels (abstract). Ann d'Endocri- nol 1979; 40: 23A. Ginsberg J, Chen HJ, Walfish PG. Multiple sites of action of diphenylhydantoin on rat thyroid regulation (abstractX Ann d'Endocrinol 1979; 40: 49A. •a Giudicelli Y, Thyroid-hormone modulation of the number of B-adrenergic receptors in fat- cell membranes. Biochem J 1978; 176: 1007-10. Grant AM, Edwards OM, Howard AN, Challand GS, Wraight EP, Mills IH. Thyroidal t hormone metabolism in obesity during semi-starvation. Clin Endocrinol 1978; 9:227-31. 190 Green WL. Metabolism of thyroid hormones by rat thyroid tissue in vitro. Endocrinology 1978; 103: 826-37. Grussendorf M, Hiifner M. Induction of the thyroxine (T4) to triiodothyronine (T3) converting enzyme in rat liver by thyroid hormones and analogs. Clin Chim Acta 1977; 80: 61-6.

Harris ARC, Fang S -L, Vagenakis AG, Braver man LE. Effect of starvation, nutriment re- •"••' placement, and hypothyroidism on in vitro hepatic T4 to T, conversion in the rat. Meta- bolism 1978; 27: 1680-9. Harris ARC, Fang S-L, Prosky J, Braverman LE, Vagenakis AG. Decreased outer ring Ji""1. monodeiodination of thyroxine and reverse triiodothyronine in the fetal and neonatal rat. Endocrinology 1978; 103: 2216-22. Harris ARC, Vagenakis AG, Braverman LE. Sex-related differences in outer ring monodeiodination of thyroxine and 3,3',5'-triiodothyronine by rat liver homogenates. Endocri- ft: nology 1979; 104: 645-52. Harris ARC, Fang S-L, Hinerfeld L, Braverman LE, Vagenakis AG. The role of sulfhydryl groups on the impaired hepatic 3',3,5-triiodothyronine generation from thyroxine in the hypothyroid, starved, fetal and neonatal rodent. J Clin Invest 1979; 63: 516-24. Hellman L, Bradlow HL, Zumoff B et al. Thyroid-androgen interrelations and the hypo- cholesteremic effect of androsterone. J Clin Endocrinol Metab 1959; 19: 936-48. Hellman L, Bradlow HL, Zumoff B et al. The influence of thyroid hormone on hydrocorti- sone production and metabolism. J Clin Endocrinol Metab 1961; 21: 1231-47. Hershman JM. Effect of 5- and 6-propylthiouracil on the metabolism of L-thyroxinein man. J Clin Endocrinol Metab 1964; 24: 173-9. Hervas F, Morreale de Escobar G, Escobar del Rey F. Conversion of L-thyroxine to triiodo- L-thyronine and biological activity of L-thyroxine, as measured by changes in growth hormone. Endocrinology 1976; 98: 77-83. Hoffenberg R. Peripheral metabolism of hormones: clinical implications. J Roy Soc Med 1979; 72: 400-8. Hiifner M, Kn8pfle M. Pharmacological influences on T4 to T3 conversion in rat liver. Clin Chim Acta 1976; 72: 337-41. Ingbar DH, Galton VA. The effect of food deprivation on the peripheral metabolism of thyroxine in rats. Endocrinology 1975; 96: 1525-32. Ingenbleek Y, Beckers C. Triiodothyronine and thyroid-stimulating hormone in protein- calorie malnutrition in infams. Lancet 1975; 2: 845-8. James WPT, Trayhurn P. An integrated view of the metabolic and genetic basis for obesity. Lancet 1976; 2: 770-3. Jung RT, Shetty PS, James WPT, Barrand MA, Callingham BA. Reduced thermogenesis in obesity. Nature 1979; 279: 322-3. Kaplan MM, Utiger RD. Iodothyronine metabolism in liver and kidney homogenates from hyperthyroid and hypothyroid rats. Endocrinology 1978; 103: 156-61. Kaplan MM. Subcellular alterations causing reduced hepatic thyroxine-5'-monodeiodinase activity in fasted rats. Endocrinology 1979; 104: 58-64. Klein AH, Oddie TH, Fisher DA. Effect of parturition on serum iodothyronine concentrations in fetal sheep. Endocrinology 1978; 103: 1453-7. Klein AH, Kobel CJ, Sack J, Fisher DA. Effect of intraamniotic fluid thyroxine injection on fetal serum and amniotic fluid iodothyronine concentrations. J Clin Endocrinol Metab 1978; 47: 1034-7. Kunos G, Vermes-Kunos I, Nickerson M. Effects of thyroid state on adrenoceptor properties. Nature 1974; 250: 779-81. Kunos G. Thyroid hormone-dependent interconversion of myocardiala- and 0-adrenoceptors in the rat. Brit J Pharmacol 1977; 59: 177-89. Larson FC, Albright EC. Inhibition of L-thyroxine monodeiodination by thyroxine analogs. J Clin Invest 1961; 40: 1132-8. 191 Laurberg P, Weeke J. Opposite variations in serum T, and reverse T, during propylthiouracil treatment of thyrotoxicosis. Acta Endocrinol 1978; 87: 88-94. •4- Laurberg P. Selective inhibition of the secretion of triiodothyronines from the perfused canine thyroid by propylthiouracil. Endocrinology 1978; 103: 900-5. Laurberg P. The-effect of radiographic contrast agents on iodothyronine release from the H perfused canine thyroid isolated in situ (abstract). Ann d'Endocrinol 1979; 40: 39A. Leslie RDG, Isaacs AJ, Gomez J, Raggatt PR, Bayliss R. Hypothalamo-pituitary-thyroid function in anorexia nervosa: influence of weight gain. Brit Med J 1978; 2: S26-8. Lightner ES, Fisher, DA, Giles H, Woolfenden J. Intra-amniotic injection of thyroxine (T4) to a human fetus. Evidence for conversion of T4 to reverse Tv Am J Obstet Gynecol 1977; 127: 487-90. Madsen SN, Sonne O. Increase of glucagon receptors in hyperthyroidism. Nature 1976; 262: 793-5. Malbon CC, Moreno FJ, Cabelli RJ, Fain JN. Fat cell adenylate cyclase and 0-adrenergic receptors in altered thyroid states. J Biol Chem 1978; 253: 671-8. Merimee TJ, Fineberg ES. Starvation-induced alterations of circulating thyroid hormone concentrations in man. Metabolism 1976; 25: 79-83. Mindroiu Th, Serban M, Ghiea D. Conversion of thyroxine (T4) to triiodothyronine (T3) by the rabbit thyroid gland. Rev Roum Biochim 1976; 13: 197-202. Miyai K, Yamamoto T, Azukizawa M, Ishitashi K, Kumahara Y. Serum thyroid hormones and thyrotropin in anorexia nervosa. J Clin Endocrinol Metab 1975; 40: 334-8. Moshang Jr J, Parks JS, Baker L et al. Low serum triiodothyronine in patients with anorexia nervosa. J Clin Endocrinol Metab 1975; 40: 470-3. Obregon MJ, Pascual A, Mallol J, Morreale de Escobar G, Escobar del Rey F. Marked decrease of the effectiveness of a T4 dose in iopanoic acid-treated rats (abstract). Ann d'Endocrinol 1979; 40: 40A. Oppenheimer JH, Schwartz HL, Surks MI. Prophylthiouracil inhibits the conversion of L- thyroxine to L-triiodothyronine. An explanation of the antithyroxine effect of propylthiouracil and evidence supporting the concept that triiodothyronine is the active thyroid V: hormone. J Clin Invest 1972; 51: 2493-7. Palmblad J, Levi L, Burger A et al. Effects of total energy withdrawal (fasting) on the levels of growth hormone, thyrotropin, cortisol, adrenaline, noradrenaline, T4, T,, and rT3 in healthy males. Acta Med Scand 1977; 201: 15-22. Pittman CS, Barker SB. Ar.tithyroxine effects of some thyroxine analogues. Am J Physiol 1959; 197: 1271-4. Pittman CS, Barker SB. Inhibition of thyroxine action by 3,3',5'-triiodothyronine. Endocri- nology 1959; 64: 466-8. Pittman CS, Shinohara M, Thrasher H, McCraw EF. Effect of thyroxine analogues on the peripheral metabolism of thyroxine: the half-life and pattern of elimination. Endocri- nology 1964; 74: 611-6. Pittman JA, Tingley JO, Nickerson JF, Hill Jr SR. Antimetabolic activity of 3,3',5'-triiodo- DL-thyronine in man. Metabolism 1960; 9: 293-5. Popovic WJ, Brown JE, Adamson JW. Modulation of in vivo erythropoiesis. Studies with euthyroid and hypothyroid dogs. J Clin Invest 1979; 64: 56-61. Portnay GI, O'Brian JT, Bush J et al. The effect of starvation on the concentration and binding of thyroxine and triiodothyronine in serum and on the response to TRH. J Clin Endocrinol Metab 1974; 39: 191-4. Ratcliffe JG, Stack BHR, Burt RWetal. Thyroid function in lung cancer. Brit Med J1978; I: 210-2. Saberi M, Sterling FH, Utiger RD. Reduction in extrathyroidal triiodothyronine production by propylthiouracil in man. J Clin Invest 1975; 55: 218-23. Sato T, Suzuki Y, Taketani Tet al. Enhanced peripheral conversion of thyroxine to triiodo- 192 thyronine during hGH therapy in GH deficient children. J Clin Endocrinol Metab 1977; 45: 324-9. Schussler GC. Orlando J. Fasting decreases triiodothyronine receptor capacity. Science 1978; 199: 686-8. ' Sharma VK, Banerjee SP, 0-Adrenergic receptors in rat skeletal muscle: effects of thyroidectomy, Biochim Biophys Acta 1978; 539: 538-42. Siersbaek-Nielsen K, Kirkegaard C, Rogowski P, Faber J, Lumholtz B, Friis Th. Extrathy- roidal effects of propylthiouracil and carbimazole on serum T4, T3, reverse T3 and TRH- induced TSH-release in man. Acta Endocrinol 1978; 80-7. Sims E AH. Experimental obesity, dietary-induced thermogenesis, and their clinical implications. Clin Endocrinol Metab 1976; 5: 377-95. Slingerland DW, Burrowt BA. Inhibition by propylthiouracil of the peripheral metabolism of radiothyroxine. J Clin Endocrinol Metab 1962; 22: 511-7. Sorimachi K, Robbins J. Metabolism of thyroid hormones by cultured monkey hepatocarcinoma cells. Nonphenolic ring deiodinr lion and sulfation. J Biol Chem 1977; 252:4458-63. Sorimachi K, Robbins J. Thyroxine inactivation by starvation in cultured hepatocarcinoma cells. Formation of reverse triiodothyronine. Horm Metab Res 1978; 10: 528-31. Suda AK, Pittman CS, Shimizu T, Chambers Jr JB. The production of metabolism of 3,5,3'- triiodothyronine and 3,3',5'-triiodothyronine in normal and fasting subjects. J Clin Endo- crinol Metab 1978; 47: 1311-9. Surks MI, Oppenheimer JH. Metabolism ofphenolic and tyrosyl-ring labeled L-thyroxine in human beings and rats. J Clin Endocrinol Metab 1971; 33: 612-8. Surks MI, Oppenheimer JH. Concentration of L-thyroxine and L-triiodothyronine specifically bound 'to nuclear receptors in rat liver and kidney; quantitative evidence favoring a ••,' major role of T3 in thyroid hcrmone action. J Clin Invest 1978; 60: 555-62. ;¡ Thomson JE, Baird SG, Thomson JA. Thyroid function in dietary amenorrhoea. Clin Endo- [; crinol 1977; 7: 383-8. f Tsukui T, Aizawa T, Yamada T, Kawabe T. Studies on the mechanism of goitrogenic action of diphenylthiohydantoin. Endocrinology 1978; 102: 1662-9. ; Vagenakis AG, Burger A, Portnay GI et al. Diversion of peripheral thyroxine metabolism from activating to inactivating pathways during complete fasting. J din Endocrinol Metab 1975; 41: 191-4. Vagenakis AG, Portnay GI, O'Brian JT et al. Effect of starvation on the production and metabolism of thyroxine and triiodothyronine in euthyroid obese patients. J Clin Endo- crino! Metab 1977; 45: 1305-8. Vigersky RA, Andersen AE, Thompson RH, Loriaux DL. Hypothalamic-dysfunction in secondary amenorrhoea associated with simple weight loss. New Eng J Med 1977; 297: ; 1141-5. Visser TJ, Lamberts SWJ, Wilson JHP, Docter R, Hennemann G. Serum thyroid hormone concentrations during prolonged reduction of dietary intake. Metabolism 1978; 27:405-9. . Visser TJ. Mechanism of enzymatic 5'-deiodination; site of inhibition by thioureylenes and < sulfite (abstract). Ann d'Endocrinol 1979; 40: 23A. Wenzel KW, Horn WR. Triiodothyronine (T3) and thyroxine (T4) kinetics in aged men. In: Robbins J, Braverman LE, eds. Thyroid Research. Amsterdam: Excerpta Medica, 1976; :.' 270-3. f s Westgren U, Ahren B, Burger A, Melander A. Stimulation of peripheral T3 formation by oral but not by intravenous glucose administration in fasted subjects. Acta Endocrinol 1977; 'ft 85:526-30. :] Wiersinga WM, Touber JL. Het perifere regulatiemechanisme van het biologisch effect van & schildklierhormonen. Werkgroep 'De schildklier', Amsterdam 1978. f Williams LF, Lefkowitz RJ. Watanabe AM, Hathaway DR, Besch HR. Thyroid hormone ï regulation of 0-adrenergic receptor number. J Biol Chem 1977; 252: 2787-9. ! 193 Woeber KA. L-Triiodothyronine and L-reverse triiodothyronine generation in the human polymorphonuclear leucocyte. J Clin Invest 1978; 62: 577-84. Wu S-Y, Chopra IJ, Solomon DH, Johnson DE. The effect of repeated administration of I ipodate (Oragrafin) in hyperthyroidism. J Clin Endocrinol Metab 1978; 47: 1358-62. fa Wu S-Y, Klein AH, Chopra IJ, Fisher DA. Alterations in tissue thyroxine-5'-monodeiodinating activity in perinatal period. Endocrinology 1978; 103: 235-9. Zumoff B, Bradlow HL, Fukushima DK, Hellman L. Increase in the tetrahydrocortisol/ tetrahydrocortisone ratio from cortisol 4-14C: a nonspecific consequence of illness. J Clin k Endocrinol Metab 1973; 39: 1120-4.

I k1

SUMMARY The peripheral conversion of the thyroid hormone tetraiodothyronine or thyroxine (T4) into triiodothyronine (T3) and reverse triiodothyronine (rTj) is defined as the extrathyroidal deiodination of T4, by which - "I depending on the site of removal of an iodine atom- either the S i metabolically very potent hormone T3 or the metabolically inactive rT3 arises {chapter I). This deiodination process generates the bulk of the daily production of T3 and rT3, and is a major pathway in the degradation of T4. The conversion of T4 has been demonstrated in a wide variety of tissues; the greatest deiodination capacity is found in the liver and kidneys. The deiodination of T4 is enzymic in nature, and sulfhydryl groups serve as cofactor; the putative enzyme 5'-deiodinase promotes the T4 -• T3 conversion, and the 5-deiodinase the T4 -• rT3 conversion. The main goal of this study was to delineate several physiological, pathological and pharmacological factors involved in the modulation of the peripheral conversion of T4.

The assays used in this study are described in chapter II. Thyrotropin (TSH) and the iodothyronines rT3, T3 and T4 were determined in plasma by radioimmunoassay (the method of Murphy and Pattee was initially applied for the assay of T4). The triiodothyronine uptake ratio (T U) was measured by the Triosorb-M 125 kit. The normal values of these t. sts, as determined in 63 euthyroid volunteers, are presented in chapter HI. Median values (in brackets the inner limits for the percentiles P2.5 and P97.5) are: TSH 1.1 (< 0.3-3.l)mU/l, rT3 0.22(0.14-0.38)nmol/l,T31.75 (1.30-2.45) nmol/l,T4101 (75-135) nmol/l,T3U 0.93 (0.81-1.07), FT4 index 95 (70-122). No effect of sex or age was found, although a tendency to lower T3 and higher rT3 values was observed in old age. The TSH response to 200//g TRH i.v. was lower in males and in subjects of 60 years and older as compared to females and younger subjects. The median peak TSH value was 8.5 mU/1 (inner limits 2.8-22.0 mU/1). The normal T3 response to TRH was defined as an absolute increment of ^>0.55 nmol/1 and a relative increment of > 30%.

In 19 cord blood samples from healthy newborns plasma T3 was low and plasma rT3 very high.

Abnormal thyroid function tests were found in 125 patients suspected of thyroid disease {chapter IV). According to the test results, they were 195

classified into 7 groups: overt hyperthyroidism (supranormal FT4 index and T3, subnormal TSH response; n=55), mild hyperthyroidism or T3- toxicosis (normal FT4 index, supranormal T3, subnormal TSH response; n=l 1), subclinical hyperthyroidism (normal FT4 index and T3, subnormal TSH response; n= 15), subclinical hypothyroidism (normal FT index and r 4 f; T3, supranormal TSH response; n=6), mild hypothyroidism (subnormal FT4 index, normal T3, supranormal TSH response; n=6), overt hypothyroidism (subnormal FT, index and T3, supranormal TSH response; n=27), and T4-toxicosis (supranormal FT4 index, normal T3, subnormal TSH response; n=S). It is concluded that hyperthyroidism and hypothyroidism are graded phenomena, and that the TSH response to YRH is the most sensitive test in the detection of abnormal thyroid function. The finding of a normal TSH response virtually excludes hyperthyroidism, but a sub-normal TSH response is also found in other (especially psychiatric) disorders, sometimes associated with a normal T3 response. The biochemical pattern of T4-toxicosis is in most cases caused by impaired S'-deiodination of T4. From the observed ratio's rT3/T3, rT3/T4 and T3/T4 it is speculated that there is an overproduction of rT3 relative to T3 in hyperthyroidism, and of T3 relative to rT3 in hypothyroidism. li In 20 patients with various acute non-thyroidal diseases, a subnormal T3 . was found in 84%, a supranormal rT3 in 63%, and a subnormal T4 in 32% (chapter V). The plasma T3 and rT3 values slowly returned into the normal range if clinical recovery was delayed, whereas a further decrease of T3 and increase of rT3 was observed in the patients who died; this suggests that the severity of a disease is qualitatively related to the changes in the conversion of T4. The subnormal T4 values are caused by a decreased binding capacity of the plasma proteins for thyroid hormones, possibly with involvement of an inhibitor for thyroid hormone binding. Despite the substantial decrease • in plasma T3 concentrations the TS H levels were not elevated, nor were the ; TS H- and T3-responses to TRH different on the day of admission and two f; weeks later. The discrepancy between some sort of 'hypothyroidism' in the £'. peripheral tissues and the 'euthyroidism' at the central pituitary-thyroid [1 level may be explained by the capability of the pituitary thyrotroph to '4 derive its nuclear T3 from local T4 conversion to a much greater extent than < the peripheral tissues.

i;< A quantitative relationship between the severity of a disease and the changes in the conversion of T4 was found in 34 patients with acute f. myocardial infarction (AMI) (chapter VI). The fall in T3 and the rise in rT3 | were greater in patients with a complicated AMI than in those with an | uncomplicated AMI, and a linear correlation was found between the V infarct size (as estimated by the peak SGOT value) and the increment of

------••-••.-.>--;,r 196 rT3, the decrement of T3, and the highest observed rT3/T3 and rT3/T4 ratio's. Plasma T4 levels peaked approximately one week after the onset of AMI, probably due to a decreased degradation of T4 and to a slight TSH- mediated increase in thyroidal secretion provoked by the inhibited 5'- deiodination of T4. p..

The effect of the B-adrenoceptor blocking agent propranolol on the conversion of T4 was studied in 11 hyperthyroid patients and in 6 hypothyroid patients on L-T4 substitution therapy (chapter VII). Plasma T3 decreased and plasma rT3 increased in both groups under propranolol medication, whereas T4 and TSH increased in the hypothyroid patients. These changes can be explained by an inhibitory effect of propranolol on the 5'-deiodination of T4. The effect of propranolol on the in vitro conversion of T4 into T3 using isolated rat liver parenchymal cella is described in chapter VIII. Propranolol lowers the conversion ratio when added in concentrations of 580 and 1160//M, but has no inhibitory effect when 290//M is added.

It thus appears that physiological, pathophysiological, or pharmacological modulation of the peripheral conversion of T4 very often results in the 'low T3, high rT3 syndrome' (chapter IX). As demonstrated in kinetic studies by others, the syndrome is caused mainly by a decreased PR of T3 from T4 and by a decreased MCR of rT3, both due to a decrease in the 5'- deiodinating activity. This decrease has, in general, been explained by a lower substrate concentration in hypothyroidism, by a lower concentration of the enzyme in the fetus and in female rats, and by a lower concentration of sulfhydryl groups during starvation and in non-thyroidal illness. Inhibition of the S'-deiodination of T4 not infrequently results in a decreased MCR of T4, and in an increased release of TSH. In general, however, the pituitary-thyroid axis functions as if an euthyroid state exists, despite hypothyroid-like alterations in the peripheral tissues. The biological significance of the iow T3, high rT3 syndrome' is presumably related to the energy metabolism of the organism: the shift in the conversion of T4 can be interpreted as an adaptive mechanism, by which the body seeks to protect itself against excessive catabolism at a time when anabolism is challenged. The peripheral tissues thus have the capability to modulate the expression of the hormonal effect to a great extent by the regulation of the amount of T3 locally generated from T4. k I

SAMENVATTING % D' De perifere conversie van het schildklierhormoon tetrajodothyronine of j thyroxine (T4) in trijodothyronine (T3) en 'reverse' trijodothyronine (rT3) wordt gedefinieerd als de dejodering van T4 buiten de schildklier, waarbij -afhankelijk van de plaats waar een jodiumatoom wordt verwijderd - het metabool sterk werkzame hormoon T3 of het metabool inactieve rT3 ontstaat (hoofdstuk I). Deze dejodering van T4 levert verreweg het grootste aandeel in de dagelijkse productie van T3 and rT3, en is een kwantitatief : belangrijke weg in de afbraak van T4. De omzetting van T4 is aangetoond i in uiteenlopende weefsels; de lever en de nieren hebben het grootste ; I dejoderend vermogen. De dejodering van T4 is een enzymatisch proces, • waarbij SH-groepen fungeren als co-factor; het hypothetische enzym 5'- i \ dejodinase bevordert de omzetting van T4 in T3, het S-dej odinase van T4 in í : rT3. Het belangrijkste doel van deze studie was om factoren van : fysiologische, pathologische en farmacologische aard die van invloed zijn • op de conversie van T4, nader te leren kennen. De gebruikte bepalingsmethodieken worden beschreven in hoofdstuk IL Het schildklier-stimulerend-hormoon TSH en de jodothyronines rT3, T3 en T4 zijn langs radioimmunologische weg bepaald in plasma (voor de bepaling van T4 is aanvankelijk de methode van Murphy-Pattee toegepast). De T3-(hars)opname (T3U) is bepaald met behulp van de : Triosorb-M 125 kit. De normaalwaarden voor deze testen, vastgesteld bij 63 euthyreote vrijwilligers, zijn vermeld in hoofdstuk III. De médiane :. waarden, met tussen haakjes de (waarschuwende) grenzen voor de ! percentielen P2.5 en P97.5, zijn: TSH 1.1 (< 0.3-3.1) mU/1, rT3 0.22 (0.14- i | 0.38) nmol/l,T31.75 (1.30-2.45) nmol/l,T4101 (75-135) nmol/l,T3U 0.93 \ \ \- (0.81-1.07), FT4 index 95 (70-122). Leeftijd en geslacht waren niet van '1 j [i invloed op de uitslagen, hoewel er een tendens bestond tot lagere T, en : ; \\ hogere rT3 waarden bij mensen van 60 jaar en ouder. j ) {! De TSH response op 200 //g TRH i.v. was bij mannen lager dan bij | vrouwen, en bij personen van 60 jaar en ouder lager dan bij jongeren. De 'f médiane waarde van de TSH-piek was 8.5 mU /1 (waarschuwende grenzen fi i 2.8-22.0 mU /1). De normale T3 response op TRH werd gedefinieerd als een | | absolute toename van > 0.55 nmoï /1 en een relatieve toename van > 30%. j i In het navelstrengbloed van 19 gezonde pasgeborenen was het plasma T3 -\ I laag en het plasma rT3 erg hoog. j |; l < f •tr-j-

198 Afwijkende uitslagen van de schildklierfunctietesten werden gevonden bij 125 patiënten die verdacht werden van een gestoorde schildklierfunctie {hoofdstuk IV). Op grond van de testresultaten waren 7 groepen te onderscheiden: manifeste hyperthyreoïdie (supranormale FT4 index en T3, subnormale TSH response; n=55), lichte hyperthyreoïdie of T3-toxicosis (normale FT4 index, supranormaal T3, subnormale TSH response; n=l 1), subklinische hyperthyreoïdie (normale FT4 index en T3, subnormale TSH response; n=15), subklinische hypothyreoïdie (normale FT4 index en T3. supranormale TSH response; n=6), lichte hypothyreoïdie (subnormale FT4 index, normaal T3, supranormale TSH response; n=6), manifeste hypothyreoïdie (subnormale FT4 index en T3, supranormale TSH response; n=27), en T4-toxicosis (supranormale FT4 index, normaal T3, subnormale TSH response; n=5). Geconcludeerd wordt, dat hyper- thyreoïdie evenals hypothyreoïdie een gegradueerd verschijnsel is, en dat de TSH response op TRH de gevoeligste indicator is voor het opsporen van een afwijkende schildklierfunctie. Een normale TSH response op TRH sluit het bestaan van een hyperthyreoïdie vrijwel uit, maar een subnormale tot afwezige TSH response komt ook voor bij andere (vooral psychiatrische) ziektebeelden, waarbij de T3 response dan nog normaal kan zijn. Het biochemische patroon van een T4-toxicosis is in de meeste gevallen te verklaren uit een remming van de 5'-dejodering van T4. Op grond van de waargenomen ratio's rT3/T3, rT3/T4 en T3/T4 lijkt het niet onmogelijk, dat bij hyperthyreoïdie relatief meer rT3 dan T3 wordt geproduceerd, en bij hypothyreoïdie relatief meer T3 dan rT3.

Bij 20 patiënten met verschillende acute niet aan de schildklier gerelateerde ziektebeelden werd een subnormaal T3 gevonden in 84%, een supranormaal rT3 in 63%, en een subnormaal T4 in 32% (hoofdstuk V). De plasma T3 en rT3 waarden keerden vrij langzaam tot in het normale gebied terug indien het klinisch herstel traag verliep, terwijl een verdere afname van het T3 en een toename van het rT3 werd gezien bij patiënten die overleden; dit suggereert een kwalitatief verband tussen de ernst van een ziekte en de veranderingen in de conversie van T4. De subnormale T4 waarden zijn te verklaren uit een afgenomen bindingscapaciteit van de plasma-eiwitten voor schildklierhormoon, waarbij mogelijk ook een remmer van deze binding betrokken is. Ondanks de aanzienlijke verlaging van het plasma T3 gehalte zijn de TSH spiegels niet verhoogd; de TSH- en '•% Tj-responsies op TRH op de dag van opname verschillen niet van de

responsies twee weken later. De discrepantie tussen 'hypothyreoïdie' in de 'ii perifere weefsels en 'euthyreoïdie' in de hypofyse-schildklier as is wellicht te verklaren uit het vermogen van de hypofysaire TSH-producerende cel om zijn aan de kern gebonden T3 in een veel sterkere mate te ontlenen aan de plaatselijke omzetting van T4 dan de perifere weefsels. 19S Een kwantitatief verband tussen de ernst van een ziekte en de veranderingen in de conversie van T4 kon worden aangetoond bij 34 s patiënten met een acuut hartinfarct (hoofdstuk VI). De daling van T3 en de stijging van rT3 waren bij patiënten met een gecompliceerd verlopend hartinfarct groter dan bij patiënten met een ongecompliceerd verloop. Er bestond een lineair verband tussen de infarctgrootte (v ¿chat aan de hand van de hoogste SGOT waarde) en de toename van rT3, de afname van T3, en de hoogst waargenomen rT3/T3 en rT3/T4 ratio's. De plasma T4 waarden stegen tot een maximum ongeveer 7 dagen na ontstaan van het infarct; deze stijging is waarschijnlijk het gevolg van een verminderde afbraak van T4 en een geringe door TSH gestimuleerde toename in de schildkliersecretie, beide veroorzaakt door remming van de S'-dejodering van T4.

Het effect van de ß-blokker propranolol op de conversie van T4 werd bestudeerd bij 11 thyreotoxische patiënten en bij 6 hypothyreotische patiënten op L-T4 substitutie (hoofdstuk VII). Het plasma T3 daalde en het plasma rT3 steeg in beide groepen tijdens toediening van propranolol, terwijl het T4 en het TSH stegen in de patiënten met hypothyreoïdie. De veranderingen zijn te verklaren uit een remmend effect van propranolol op de S'-dejodering van T4. Het effect van propranolol op de in vitro omzetting van T4 in T3 (met gebruikmaking van geïsoleerde I leverparenchymcellen van ratten) wordt beschreven in hoofdstuk VIII. Propranolol toegevoegd in een concentratie van 580 en 1160^/M verlaagt de conversie, maar toevoeging van 290 f*M heeft geen invloed op de conversie.

Fysiologische, pathophysiologische of farmacologische beïnvloeding van de perifere conversie van T4 resulteert dus dikwijls in het 'laag T3, hoog rT3 syndroom' (hoofdstuk IX). Zoals uit kine- tisch onderzoek door anderen verricht blijkt, wordt dit syndroom hoofdzakelijk veroorzaakt door een verminderde productie van T3 uit T4 en door een verminderde metabole klaring van rT3, beide als gevolg van een afname in de S'-dejodering. Deze afname is - in grote lijnen - te verklaren uit een lagere substraatconcentratie bij hypothyreoïdie, uit een lagere concentratie van het enzym bij de foetus en bij vrouwtjesratten, en uit een lagere concentratie aan sulfhydryl-groepen tijdens vasten en bij ziekten in het algemeen. Remming van de S'-dejodering van T4 leidt niet zelden tot een verminderde metabole klaring van T4 en een toegenomen TSH secretie. Doorgaans functioneert de hypofyse-schildklier as echter alsof er sprake is van euthyreoïdie, ondanks de hypothyreoïdie-achtige veranderingen in de perifere weefsels. De biologische betekenis van het 'laag T3, hoog rT3 syndroom' houdt vermoedelijk verband met de energiehuishouding van het lichaam: de veranderingen in de conversie van

ï é kr. 200 T4 zijn te interpreteren als een aanpassingsmechanisme, waarmee het lichaam zichzelf kan beschermen tegen overmatig katabolisme ten tijde van een toch al moeizaam anabolisme. De perifere weefsels lijken derhalve over het vermogen te beschikken het hormonale effect in aanzienlijke mate te beinvloeden door regulatie van de hoeveelheid ter plaatse uit T4 I gevormd T3.

i-.

I,- f •

4 SAMENVATTING VOOR DE LEEK ;

De Nobelprijs voor geneeskunde is in 1977 toegekend aan Rosalyn ¡ Yalow voor haar baanbrekend werk bij de ontwikkeling van de ; radioimmunologische bepaling van hormonen, en aan Roger Guillemin en Andrew Schally voor de isolering van stoffen uit de hersenen die de afgifte van hormonen door het hersenaanhangsel (de hypofyse) reguleren (met name voor de isolering van het TRH of het TSH-releasing-hormone, een hormoon dat de afgifte van TSH door de hypofyse stimuleert; TSH is het, \ schildklierstimulerend hormoon uit de hypofyse). Zonder hun pioniers- ; , werk zouden de studies in dit proefschrift niet mogelijk zijn geweest. De schildklier produceert twee hormonen, T4 en T3, waarvan het T3 ; eigenlijk niet goed meetbaar was door de bijzonder lage concentratie waar- \ in het in het menselijk bloed voorkomt. Met behulp van de door Yalow en Berson ontwikkelde radioimmunologische methodiek is de bepaling van '•'• | T3 sinds het begin het begin van de zeventiger jaren veel eenvoudiger en . | betrouwbaarder geworden. Schildklierziekten uiten zich vaak doordat teveel of te weinig schildklierhormoon wordt geproduceerd; men spreekt dan van hyperthyreoïdie of hypothyreoïdie, waarbij de T4 en T3 concentraties in het bloed verhoogd of verlaagd zijn. Na de invoering van de T3 bepaling bleken er patiënten te zijn bij wie het ziektebeeld van de hyperthyreoïdie : uitsluitend veroorzaakt werd door een overmatige T3 productie (de zgn. T3-toxicosis). Dit beeld kon in het verleden, toen de T3 bepaling nog niet : beschikbaar was, en doordat het T4 gehalte in het bloed hierbij normaal is, I vaak moeilijk herkend worden. Uit ons onderzoek blij kt dat de Tj-toxicosis • S niet zeldzaam is: het kwam voor in 17% van de patiënten met ! > SÏ hyperthyreoïdie. ] jj_ g Ook de invoering van TRH-test in het begin van de zeventiger jaren |' [| heeft de diagnostiek van schildklierziektes sterk verbeterd. Men geeft |: ;| hierbij een injectie van TRH in de ader, en meet vervolgens de reactie van • '• ! ! | TSH in het bloed; normaal stijgt het TSH gehalte na toediening van TRH. '| , b | De TSH afgifte wordt echter ook beïnvloed door T3 en T4: een te lage '} 1 1 concentratie schildklierhormoon stimuleert de TSH afgifte, een te hoge i t\ \ concentratie schildklierhormoon remt de TSH afgifte. Bij hyperthyreoïdie | ^ blijkt de TSH response op TRH verminderd tot afwezig te zijn, bij f i hypothyreoïdie verhoogd. Het grote belang van de TRH test ligt in het feit, ,'j ' |> dat een normale TSH response het bestaan van een hyperthyreoïdie vrijwel i ¿; uitsluit. Zoals ook uit ons onderzoek blijkt, is de gevoeligheid van deze test \ ffe -BV

• 202 V; i : it v in de diagnostiek van een gestoorde schildklierfunctie zelfs zo groot, dat '*£• £ niet zelden reeds een afwijkende TSH response wordt gevonden wanneer 'j£. f de T3 en T4 gehaltes in het bloed (nog) normaal zijn en de patiënten geen |% £ klinische verschijnselen van een teveel of een tekort aan schildklierhor- fe | moon tonen. Of een dergelijke'subklinische'hyper-of hypothyreoïdie ook ; ? |¿: l behandeld moet worden, is overigens de vraag. Wel lijken deze patiënten ' 'ix een grotere kans te hebben op de ontwikkeling van een manifeste hyper- of ii hypothyreoïdie. , |Ï; ; Na het beschikbaar komen van de T3 bepaling vond men lage tot zeer p lage T3 gehaltes in het bloed van patiënten met uiteenlopende |) ziektebeelden die op zich niets met de schildklier te maken hebben. Het E opvallende was, dat deze patiënten niet de indruk maakten een tekort aan f. schildklierhormoon te hebben. Na genezing van de ziekte bleek het T3 t gehalte in het bloed weer normaal te zijn. De verklaring voor dit í intrigerende verschijnsel is als volgt. Het is de laatste jaren duidelijk i: geworden dat het grootste deel van het dagelijks geproduceerde T3 niet - ; !: • zoals men zou denken - gevormd wordt in de schildklier, maar afkomstig is : 'il ? van de omzetting van T4 in T3 buiten de schildklier. Bij deze 'perifere conversie' wordt van T4, dat vier jodiumatomen heeft, één jodiumatoom afgesplitst, waardoor T3 ontstaat. Uit studies van Chopra is voorts gebleken, dat door afsplitsing van een jodiumatoom op een andere plaats in het T4 molecuul een verwante verbinding met eveneens drie jodiumatomen ontstaat (het spiegelbeeldige of 'reverse' T3, afgekort als rT3), dat echter in tegenstelling tot T3 geen hormonaal effect heeft in de zin van stimulatie der stofwisseling. Bij ziek zijn in het algemeen daalt nu het T3 gehalte en stijgt het rT3 gehalte in het bloed door veranderingen in de perifere conversie van T4. Herkenning van dit patroon is belangrijk, daar het waarschijnlijk schadelijk is patiënten met een aldus ontstaan laag T3 met schildklierhormoon te behandelen. Een tweede opvallend fenomeen bij dit 'laag T3, hoog rT3' beeld is, dat het TSH gehalte in het bloed hierbij normaal is, en niet stijgt zoals bij patiënten met een laag T3 t.g.v. hypothyreoïdie wel het geval is. Wij vonden dat de TRH-test onder deze : omstandigheden bij acuut zieke patiënten zijn diagnostische waarde blijft behouden. Verder namen wij waar, dat bij patiënten die ernstig ziek zijn en |? zich slechts langzaam herstellen van hun ziekte, de T3 gehaltes lager zijn en ¡ [-) dit ook langer blijven dan bij patiënten die minder erg ziek zijn. Bij i Hi patiënten die overleden, daalde het T3 voor de dood en steeg het rT3 nog f | verder. Een kwantitatief verband tussen de ernst van een ziekte en de H veranderingen in de conversie van T4 konden wij duidelijk aantonen bij Î f patiënten met een acuut hartinfarct. De ernst van een hartinfarct is ¿f i gerelateerd aan de grootte van het infarct. Wij vonden een rechtlijnig k f verband tussen de grootte van het infarct en de mate van T3 daling en rT3 f J§/-' stijging. '| ¥ Patienten met hyperthyreoïdie worden vaak behandeld met proprano- ;

•&. V 203 ' {• ••iv. i? '•% lol, een middel dat de effecten van heel andere hormonen dan het . .t >:> schildklierhormoon blokkeert. Het was dan ook niet geheel duidelijk, % fk waarom dit medicijn bij overmatige schildklierhormoonproductie vaak zo 'f f-;;: goed helpt. Wij hebben kunnen aantonen dat propranoiol de conversie van .. f \t T4 beinvloedt, zodanig dat minder van het sterk werkzame hormoon T3 en ; ,. f I' meer van het inactieve rT3 ontstaat. Dit verklaart althans ten dele het $ I therapeutisch effect ervan bij verhoogde schildklierwerking. De remmende : |i I> werking van propranoiol op de vorming van T3 uit T4 hebben wij ook |'i \.J aangetoond in een kunstmatig systeem van rattelevercellen. Voegt men •-,> I. aan levercellen T4 toe, dan maken de levercellen T3 uit het toegevoegde T4; ;' dit is een fraai model voor bestudering van de T4 conversie in een • reageerbuisje. Bestudering van de conversie van T4 heeft inmiddels al een nieuw geneesmiddel voor de behandeling van hyperthyreoidie opgeleverd: ' \ een van de door Chopra en medewerkers onderzochte stoffen remt de \ ;' conversie van T4 in T3 zo sterk, dat de verschijnselen van teveel { X schildklierhormoon erdoor verminderen. • I

Veranderingen in de perifere conversie van T4 treden op onder tal van omstandigheden. Men heeft het iaag T3, hoog rT3' beeld gevonden bij het ongeboren kind, bij een laag lichaamsgewicht, tijdens vasten, bij ziek zijn in het algemeen, na operaties, en bij gebruik van diverse medicamenten. De betekenis van deze veranderingen in de perifere conversie van T4 is onduidelijk. Waarschijnlijk is het een aanpassingsmechanisme, recht- streeks betrokken bij de energiehuishouding van het lichaam. Men kan zich tenminste goed indenken dat bij ziekte, als de omstandigheden voor de opbouwende processen in het lichaam ongunstig zijn, een overmatige afbraak voorkomen kan worden door verlaging van de T3 productie, waardoor de stofwisseling en het energieverbruik van de weefsels wordt verminderd. Recente experimenten bij vrijwilligers die vastten,versterken ; dit vermoeden. Vasten leidt tot een dating in het T3-gehalte en tot eiwitafbraak (~ spierverlies). Voorkomt men de T3-daling door toediening \ van T3, dan is de eiwitafbraak groter dan zonder T3 toediening. Dit betekent dus, dat de T3-daling bij vasten een eiwitsparend effect heeft. : Het vermogen van veel organen om zelf T3 uit T4 te maken, en dit mogelijk ook zelf te regelen, houdt in dat het hormonaal effect van schildklierhormoon op deze organen goeddeels ter plaatse wordt bepaald, i tot op zekere hoogte onafhankelijk van de schildklier. Dit inzicht heeft het \ klassieke endocrinologische denken in 'centrale' regelmechanismen ingrijpend gewijzigd. • e

5>' ir'- ABBREVIATIONS tí' - symbol for radioactivity; the labeled hormone i Ab - antibody B.I Ag - antigen AMI - acute myocardial infarction a-MPT -oi-methyl-p-tyrosine AS - antiserum b - bovine B« - initial binding (the percentage of tracer bound to the antibody in the first zero-standard point of the standard curve) - percentage of tracer specifically bound to antibody BSA - bovine serum albumin CG - choriogonadotropin; chorionic gonadotropin CHO - carbohydrates c.p.m. - counts per minute DA - double antibody DASP - double-antibody solid phase DPH - diphenylhydantoin DTT - dithiothreitol FD - final dilution FFA - free fatty acids FSH - follitropin; follicle-stimulating hormone FT, - free triiodothyronine

FT4 - free thyroxine

FT4 índex - free thyroxine index (T3U xT4) GH - growth hormone h - human Ht - hematocrit LH - lutropin; luteinizing hormone MCR - metabolic clearance rate MRC • Medical Research Council PBS - phosphate buffered saline PEG - polyethylene glycol PR - production rate PTÜ - propylthiouracil RIA - radioimmunoassay rT, - reverse triiodothyronine; 3,3',5'-triiodo-L-thyronine SGOT - serum glutamic oxalacetic transaminase SH - non-protein sulfhydryl groups - triiodothyronine; 3,5,3'-triiodo-L-thyronine T3 - triiodothyronine uptake ratio T3U 205 - thyroxine; 3,5,3',5'-tetraiodo-L-thyronine

T4(D) • T4-assay by displacement, the competitive protein binding method of Murphy-Pattee TBG - thyroxine binding globulin •_ -'._. _z _._-._• TBP - thyroid hormone binding proteins TBPA - thyroxine binding pre-albumin -- . ~. TETRAC - tetraiodothyroacetic acid TRH - thyroliberin; thyrotropin-releasing hormone; protirelin TRIAC - triiodothyroacetic acid TSH - thyrotropin; thyrotropic hormone