Changes of Circulating Thyroxine, Triiodothyronine and Reverse Triiodothyronine After Radiographic Contrast Agents HANS BURGI, CLAUS WIMPFHEIMER, ALBERT BURGER, WOLFGANG ZAUNBAUER, HELMUT ROSLER, AND THfiRfiSE LEMARCHAND-BfiRAUD Departments of Medicine and Diagnostic Radiology, University of Berne, Department of Medicine, University of Geneva, and Department of Medicine, University of Lausanne, Switzerland ABSTRACT. Thyroid function was studied for 42 days in 58 patients, 28 of whom had euthyroid goiter, after urography (diatrizoic acid), cholangiography (ioglycamic acid), and cholecystography (Naiopanoate). After urography and cholangiography shortlived increases of the serum thyroxine occurred in a few patients, but the mean thyroxine and triiodothyronine concentration did not change. By contrast, 7 days after oral cholecystography serum thyroxine had risen consistently by 22% with a concomittant rise of the free thyroxine, while triiodothyronine declined by 15%. The thyroxine metabolite 3,3',5'-triiodo-l-thyronine (reverse T3) rose by 50% and serum thyrotropin concentration doubled. After
B
Y VIRTUE OF their high iodine content, radiographic contrast agents depress thyroidal radioiodine uptake (1,2) and elevate plasma protein-bound iodine (3), but it is generally assumed that thyroid hormone secretion remains constant despite wide fluctuations in iodine supply (4,5). In view, however, of some recent reports of thyrotoxicosis after radiographic contrast media (612), we decided to analyze thyroid function prospectively after 3 common radiographic examinations involving iodinated substances. Patients with euthyroid simple goiter were included in the study because they seem to be particularly at risk for iodineinduced thyrotoxicosis (13-15) and because Switzerland, a previously endemic area, still has a high goiter prevalence (16). The study showed that oral cholecystogReceived January 20, 1976. Supported by grants 3.7550.72 and 3.799.72 of the Schweizerische Nationalfonds and by a contribution of Schering AC, Zurich. Correspondence to: Dr. H. Biirgi, Medizinische Klinik, Burgerspital, CH-4500 Solothurn, Switzerland.
42 days thryoxine and triiodothyronine had returned to baseline, and none of the 58 patients developed clinical hyperthyroidism. In patients with severe myxoedema kept on a constant replacement dose with 1-thyroxine Na-iopanoate produced similar changes with the exception of the rise of the serum thyroxine. The primary event after Na-iopanoate seems to be a fall of the serum triiodothyronine, which in turn augments thyrotropin and indirectly thyroxine secretion. The marked and sometimes sustained rise of serum thyroxine after cholecystography may lead to the erroneous diagnosis of hyperthyroidism. (J Clin Endocrinol Metab 43: 1203, 1976)
raphy with Na-iopanoate consistently leads to changes in circulating thyroid hormones due to alterations in peripheral hormone metabolism. Intravenous cholangiography and urography had no such effect. No case of hyperthyroidism occurred in 58 patients followed for 6 weeks. Subjects and Methods Study A Patients referred to the department of diagnostic radiology for oral cholecystography, iv cholangiography and iv urography were included in the study provided that a) they agreed to participate, b) appropriate blood samples could be obtained, c) there had been no known exposure to excess iodine in the past 6 months, and d) there was no evidence of actual or past thyroid disease except euthyroid goiter. 11 subjects were volunteers who took Na-iopanoate, but who were not examined radiologically. All subjects were examined by one of us (C.W.) and goiter was considered present when the thyroid was well palpable and unequivocally enlarged. Blood was drawn prior to and for intervals up to 42 days after the contrast agent. In cases of 1203
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1204
JCE & M • 1976 Vol 43 • No 6
BURGI ET AL.
TABLE 1. Non-hormonal iodine in serum, calculated as the difference between total and thyroxine iodine after radiographic examinations; before the radiographic examination the non-hormonal iodine was below the detection limit of the method (below 1 yug/100 ml) Non-hormonal iodine of serum Mean ± SEM (jag/100 ml) Day 3 Oral cholecystogram with 2 g Na-iopanoate (N = 10) Intravenous cholangiogram with 4 g Na-ioglycamate + 6 g methyl-glucamine ioglycamate (N = 10) Intravenous urogram with 8.0 g Na-diatrizoate and 52.8 methyl-glucamine diatrizoate (N = 9)
striking changes of serum hormones the patients
or their physicians were contacted to obtain additional clinical information. Thyroid scintiscans were done in a few selected patients 3 months or more after the cholecystograms. In cases where the serum thyroxine had not returned to normal after 42 days, additional blood samples were taken after 2 to 3 months. For oral cholecystography 2 g Na-iopanoate (Bilijodon) containing 1.3 g iodine were given per os in 4 tablets. For iv cholangiography 4 g Na-ioglycamate and 6 g methylglucamine ioglycamate (Bilivistan) were injected in a 20 ml solution. The iodine content was 5.6 g. For iv urography 8.0 g Na-diatrizoate and 52.8 g methylglucamine diatrizoate (Urografin) were injected in a volume of 80 ml. The iodine in this mixture amounted to 29.6 g. Study B Four women (ages 58 to 68 years) with well established severe idiopathic myxoedema (serum thyroxine originally below 1.5 /Ag/100 ml) who were taking a constant dose of L-thyroxine (0.1 to 0.2 mg per day) for more than 6 months and who appeared clinically euthyroid agreed to participate in the study. Blood was drawn prior to and 4 days after 2 g Na-iopanoate p.o.
Day 7
Day 14
Day 28
Day 42
2,171 ± 1,325
81 ± 28
15.9 ±5
5.4 ± 1.8
2.0 ±0.6
2,627 ± 718
192 ± 69
42 ±7
12 ±2
12 ± 3
1.5 ± 1.5
1.2 ± 1.2
34.3 ± 18.4
2.8 ± 1.6
1.8 ± .18
Analytical methods The following measurements in serum were made: Thyroxine (T4D) by a commercial competitive protein binding set (Tetrasorb Abbott) and in selected cases also by radioimmunoassay (T4RIA) (17); free thyroxine (FT4) by a modified equilibrium dialysis method (18); 3, 5, 3'-triiodo1-thyronine (T3 RIA) (19), 3,3',5'-triiodo-l-thyronine ("reverse T3") (20) and thyrotropin (TSH) (21) by specific radioimmunoassays; total iodine by a eerie acid color reaction (22). Non-hormonal iodine was calculated by subtraction of thyroxine iodine from total serum iodine. All sera of each patient were analyzed in the same assay. The ranges of these tests for normal persons are given in Table 6. Na-iopanoate added in vitro in a concentration up to 100 /ug/100 ml did not interfere with the assays of T4, T3 and reverse T3; moreover, in 3 untreated hypothyroid patients Na-iopanoate caused only an insignificant rise of T4 by 0.2 /Ltg/100 ml and reverse T4 remained immeasurable. T3 decreased from 61 to 43 /ng/100 ml. This renders an interference of Na-iopanoate with the assays quite unlikely.
Results
Study A Study C 4 healthy men (ages 23 to 40 years) were given 0.2 mg L-thyroxine p.o. daily for 21 days to suppress endogenous thyroid hormone secretion. They took 2 g Na-iopanoate p.o. on day 17. Blood drawn at days 15, 17, and 21 was analyzed.
Table 1 gives the time course of the serum concentrations of the three contrast media, expressed as non-hormonal iodine. After urography and after oral cholecystography, non-hormonal iodine has returned to below 2 jLig/100 ml within 2 and 6 weeks respec-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1205
T4, T3 AND REVERSE T3 AFTER X-RAY DYES tively. After iv cholangiography, nonhormonal iodine remains elevated to 12 /xg/ 100 ml after 6 weeks. After oral cholecystography, serum T4 rose consistently within 3 to 7 days in all 10 patients with and in the 12 patients without goiter (Fig. 1). At both times the rise was statistically highly significant. The rise was comparable in the 2 patient groups (1.8 and 2.6 /Ltg/100 ml respectively). In 5 patients the serum T4 was also measured by a radioimmunoassay, which showed good agreement with the routine competitive binding assay (Table 2). FT 4 was measured in 9 patients and rose markedly in 8 of them (Table 2). After 42 days, the average T4 had returned to baseline in both the goitrous and non-goitrous subjects. In 2 subjects, the T4 was still above 12 /xg/100 ml, but it returned to normal within the following 2 months (Fig. 1). The serum T3 showed a behavior opposite to that of the T4 (Fig. 2). It dropped in the majority of patients within 3 to 7 days by an average of 29 ng/100 ml in the nongoitrous and 22 ng/100 ml in the goitrous persons. After 2 weeks, the T3 rose again and after 6 weeks its concentration was slightly but significantly higher than baseline. 3,3',5'-triiodo-l-thyronine (reverse T3) was measured in 7 patients who had shown clear-cut decreases of T3 after oral cholecystography (Table 3). It increased from a basal value of 72 ng to 158 ng after 3 days and then slightly decreased again to 129 ng/100 ml. The reverse T3 thus behaved inversely to T3. When the magnitude of the changes of T3 and of reverse T3 between day 0 and day 7 were analyzed for all 7 patients, a highly significant linear inverse relationship between the two variables emerged (correlation coefficient r = -0.946; P < 0.001). In all 9 patients in whom it was measured, TSH rose after cholecystography, and 3 days after the contrast agent the mean serum concentration had doubled and was well above the normal range (Table 3).
Tt(O) pgHOOml 6
CHOLECYSTOGRAM
WITHOUT GOITER
WITH GOITER
42 DAYS
2-4 MONTHS
FIG. 1. Serum thyroxine in 10 patients with euthyroid goiter and in 12 patients without thyroid abnormality after oral cholecystography with Na-iopanoate. The thin lines connect the values of individual patients. The thick bars indicate the mean values with the standard errors. The abscissa gives the days after the cholecystography. The results of Student's t test for paired comparisons to the initial measurement are indicated on top: +, + + , and + + + mean P < 0.05, 0.01 and 0.001, respectively.
Thyroid scans with 131I were obtained in 3 goitrous patients who had shown a marked rise of the serum T4 after cholecystography. In all patients, the thyroid was moderately enlarged, but no autonomous nodules were seen, although this was not further confirmed by scans after T3 suppression. After iv cholangiography, there was on the average a slight increase of the T4 and a decrease of T 3 (Table 4); however the response was highly variable from one patient to the other and the changes were not statistically
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1976 Vol 43 • No 6
BURGI ET AL.
1206
TABLE 2. Serum thyroxine by radioimmunoassay (T4 RIA) or by displacement (T4 D) and free thyroxine (FT4) after radiographic contrast agents T4 (RIA) Ozg/100 ml) Subject C 4 C 9 Cll C16 C 19 C 2 C 8 C 10 C22
Goiter
+ + + +
initial
peak
10.2
11.3
12.6 10.6 11.2 11.2
16.5 16.5 15.8 13.7
Mean ± SEM of all 9 patients Mean ± SEM of 5 patients in whom both T4D and T4 RIA were done
11.16 ±0.40
14.76 ± 1.00
FT4 (ng/100 ml)
T 4 (D) (/ug/100 ml) initial
peak
initial
peak
8.0 7.2 8.5 10.4 9.9 11.9 9.8 11.9 11.4
10.4 10.4 11.7 13.0 11.6 17.9 14.4 15.0 13.5
1.76 1.43 1.59 2.69 2.28 2.93 1.27 2.21 2.37
2.16 2.35 2.16 2.86 2.32 5.60 1.99 2.79 3.47
9.89 ±0.57
13.10 ±0.81
11.08 ± 0.42
14.76 ±0.85
2.06 ±0.19
2.85* ±0.37
Patients with clear-cut increases of the T4D were selected for measurement of FT4 and in 5 cases also of T4 RIA. The initial and the peak values of each individual patient (day 3 or 7) are given. * t test for paired samples: P < 0.025.
significant. 6 of the 16 patients showed transient rises of the T4 above 12 />ig/100 ml. These elevations took place after different intervals and therefore have no effect on the group average. Table 5 shows that T4 and reverse T3 remained on the average remarkably stable after iv urography. T3 showed a rise after 42 days, but the change did not quite reach statistical significance. None of the 58 patients of study A developed clinical evidence of hyperthyroidism. Study B The 4 hypothyroid patients on constant replacement therapy were all clinically euthyroid. Nonetheless, TSH in serum was moderately elevated in 3 of them, indicating that the replacement dose was slightly too low. In these 3 patients, Na-iopanoate produced a further marked rise of TSH which, due to wide individual scatter of values, was not statistically significant (Table 3). The fourth patient had an unmeasurable TSH, both before and after Naiopanoate. Thus, her replacement dose was
probably slightly too high. In all four patients, Na-iopanoate produced a highly significant fall of T3RIA and a rise of reverse T3, while the T4D and the FT4 remained unchanged (Table 6). Study C In normal individuals given 0.2 mg Lthyroxine to suppress endogenous thyroid secretion, the thyroid hormones and reverse T3 had reached a steady state in serum after 15 days of administration of 0.2 mg L-thyroxine, since there was no change of these indices between day 15 and day 17 (table 6). Na-iopanoate again produced a fall of T3 and a rise of reverse T3 in serum. T4 and FT4 rose slightly, but the changes were not significant. Discussion Our studies establish that cholecystography consistently produces changes in circulating thyroid hormones. Since iopanoic acid (Telepaque) or its sodium salt (Bilijodon) are among the most frequently
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1207
T4, T3 AND REVERSE T3 AFTER X-RAY DYES used radiographic contrast agents in many countries, the findings are of general clinical interest. An elevated serum T4 might easily mislead one to diagnose hyperthyroidism, unless other indices such as the T3 or TSH are also measured. A change in thyroxine-binding globulin or a technical artefact does not account for the rise in serum T4, since the rise was confirmed by a second independent method of measurement and since FT4 also rose (Table 1). The serum T4 rose in all persons studied, irrespective of whether they had simple goiter or a normal thyroid gland. Thus the observed change was not due to an uptake of excess iodine by autonomous nodules with consequent hypersecretion of T3(WA) ng/100 ml 300 ,.
CHOLECYSTOGRAM
200
100
WITHOUT GOITER
300
200
100
WITH GOITER 0
3
7
28
42 DAYS
FiG. 2. Serume triiodothyronine after oral cholecystography in the same patients as shown in Fig. 1. The meaning of the symbols is the same as in Fig. 1.
TABLE 3. Mean and standard error for T3) reverse T3
and TSH in serum after cholecystography with Na-iopanoate
Day 0 Day 3 Day 7
T3 (RIA) ng/100 ml (N = 7)
Reverse T3 (RIA) ng/100 ml (N = 7)
TSH /xU/ml (N = 9)
194.0 ± 12.8 142.6 ± 9.1*** 156.6 ± 17.9*
57.0 ± 7.8 121.0 ± 15.0*** 97.7 ± 12.5*
5.5 + 0.8 12.2 ± 2.1**
Asterisks give the results of t tests for paired samples between day zero and day 3 and 7: * = P < 0.02, ** = P < 0.01, *** = p < 0.005. Due to limited availability of serum, tests could not be done in all 22 study subjects.
hormone (23). Moreover such nodules were not found in scintigrams performed in patients several months after Na-iopanoate. Our results establish that the lowering of the T3, the rise of the reverse T3 and the rise of TSH all take place in persons in whom the supply of thyroxine is kept constant (studies B and C). On the other hand, the rise of serum T4 only occurs when there is functioning thyroid tissue (study A). Admittedly the study does not establish the exact temporal relationship of events, but it seems likely that the primary event after Na-iopanoate is the fall of the serum T3. The resulting state of T3 deficiency of the peripheral tissues causes a rise in TSH, which in turn stimulates T4 secretion by the thyroid gland. Our studies do not allow us to decide whether the lowering of the serum T3 is due to decreased production or enhanced degradation of this hormone. A large part of the circulating T3 (24) and virtually all of the circulating reverse T3 (25) arise in peripheral tissues by monodeiodination of T4. The fact that Na-iopanoate affects both T3 and reverse T3 and that the changes in addition are strictly reciprocal with a high degree of correlation, suggests that the contrast agent interferes somehow with the monodeiodination of T4 and thereby with the production of T3 and reverse T3. However, more studies are needed to strictly exclude an effect on T3 and reverse T3 degradation. Similar reciprocal changes of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1208
BURGIETAL.
^ S l
TABLE 4. Thyroid hormone parameters in serum after intravenous cholangiography with ioglycamic acid Day 0 T, (D) (/i.g/100 ml) N = 15 T3 (RIA) (ng/100 ml) N=15 reverse T3 (RIA) (ng/100 ml) N= 3
Day 3
Day 7
Day 14
Day 28
Day 42
9.2 ± 0.5
9.6 ± 0.6
10.1 + 0.5
9.1 ±: 0.6
8.5 ± 0.6
8.0 ± 0.5
166.5 ± 12.9
147.1 ± 17.5
158.6 ± 15.7
172.7 ±: 13.1
185.1 ± 16.6
171.8 ± 14.6
—
—
47
± 3
57
± 2
54
± 10
The data for the 7 patients with euthyroid goiter and the 8 patients with a normal thyroid gland were pooled since there was no difference between the two groups. Reverse T3 was only measured in 3 patients. The data give the mean with the standard error. None of the changes from baseline was statistically significant.
T3 and reverse T3 occur in the newborn (26,27) and during fasting (28). Chemically, iopanoic acid is a benzene ring substituted with 3 iodine atoms, one amino group and an isovaleric acid side chain (Fig. 3). Superficially at least the molecule resembles thyroxine and it is conceivable that it interferes with the enzymatic deiodination of thyroxine as shown for other iodinated benzene derivatives (29) and thyroid hormone analogs (30,31). In these latter studies T3, reverse T3 and TSH were not measured so that a direct comparison is not possible. Iopanoic acid is mostly excreted as the glucuronide (32). Deiodination has not been demonstrated and, if it occurs, it accounts for only a very small fraction of its metabolism (33). Nonetheless, even with a very low affinity for deiodinating enzymes iopanoic acid could interfere with T4 metabolism, since its initial molar serum concentration exceeds that of T4 by a factor of 250 (Table 1). The observation that iopanoic acid depresses radioio-
dine uptake for several weeks provides some indirect evidence that part of it is deiodinated in the body (1). Iodide excess per se could partly explain the initial decrease of T3 (4), but neither iodide excess alone nor urography with diatrizoic acid nor cholangiography with ioglycamic acid (tables 4 and 5) causes similar changes of T3 and reverse T3. Unlike iopanoic acid, diatrizoic and ioglycamic acid have a benzoic acid group on the iodine-substituted benzene ring (Fig. 3). The strong negative charge of this group could radically change the affinity of these agents to a deiodinating enzyme system as compared to iopanoic acid. Severe iodine deficiency may lower the circulating T4 and raise the T3 (34), and Na-iopanoate might have acted by simply improving the iodine supply in previously iodine deficient patients. However, the fact that other iodinated contrast agents did not cause the same changes is strong evidence against such an interpretation. Moreover,
TABLE 5. Thyroid hormone parameters in serum after intravenous urography with diatrizoic acid
T4 (D) ()Ltg/100 ml) N = 19 T3 (RIA) (ng/100 ml) N=19 reverse T3 (RIA) (ng/100 ml) N = 3
Day 0
Day 3
Day 7
Day 14
Day 28
Day 42
9.1i :0.3
9.2 ±:0.4
9.6 ±:0.3
9.0:t 0.4
9.1:t 0.3
9.5 ± 0.5
185.5 ±:9.6
178.6 ±:7.6
186.7 ±:9.8
214.4 ± 8.8*
64 ±: 9
65 ±: 9
—
57
*: 4
The data for the 9 patients with euthyroid goiter and the 10 patients with a nomial thyroid gland were pooled, since there was no difference between the two groups. Reverse T3 was only measured in three patients. The data give the mean with the standard error. * t test compared to day zero F < 0.10.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
T4, T3 AND REVERSE T3 AFTER X-RAY DYES
1209
TABLE 6. Mean and standard error for T4D, FT4, T3RIA, reverse T3 and TSH in serum
Normal range
T4D (/xg/100 ml)
FT4 (ng/100 ml)
T3RIA (ng/100 ml)
Reverse T3 (ng/100 ml)
TSH OiU/ml)
5.5-12.0
1.33-2.41
100-220
20-80
B
Y VIRTUE OF their high iodine content, radiographic contrast agents depress thyroidal radioiodine uptake (1,2) and elevate plasma protein-bound iodine (3), but it is generally assumed that thyroid hormone secretion remains constant despite wide fluctuations in iodine supply (4,5). In view, however, of some recent reports of thyrotoxicosis after radiographic contrast media (612), we decided to analyze thyroid function prospectively after 3 common radiographic examinations involving iodinated substances. Patients with euthyroid simple goiter were included in the study because they seem to be particularly at risk for iodineinduced thyrotoxicosis (13-15) and because Switzerland, a previously endemic area, still has a high goiter prevalence (16). The study showed that oral cholecystogReceived January 20, 1976. Supported by grants 3.7550.72 and 3.799.72 of the Schweizerische Nationalfonds and by a contribution of Schering AC, Zurich. Correspondence to: Dr. H. Biirgi, Medizinische Klinik, Burgerspital, CH-4500 Solothurn, Switzerland.
42 days thryoxine and triiodothyronine had returned to baseline, and none of the 58 patients developed clinical hyperthyroidism. In patients with severe myxoedema kept on a constant replacement dose with 1-thyroxine Na-iopanoate produced similar changes with the exception of the rise of the serum thyroxine. The primary event after Na-iopanoate seems to be a fall of the serum triiodothyronine, which in turn augments thyrotropin and indirectly thyroxine secretion. The marked and sometimes sustained rise of serum thyroxine after cholecystography may lead to the erroneous diagnosis of hyperthyroidism. (J Clin Endocrinol Metab 43: 1203, 1976)
raphy with Na-iopanoate consistently leads to changes in circulating thyroid hormones due to alterations in peripheral hormone metabolism. Intravenous cholangiography and urography had no such effect. No case of hyperthyroidism occurred in 58 patients followed for 6 weeks. Subjects and Methods Study A Patients referred to the department of diagnostic radiology for oral cholecystography, iv cholangiography and iv urography were included in the study provided that a) they agreed to participate, b) appropriate blood samples could be obtained, c) there had been no known exposure to excess iodine in the past 6 months, and d) there was no evidence of actual or past thyroid disease except euthyroid goiter. 11 subjects were volunteers who took Na-iopanoate, but who were not examined radiologically. All subjects were examined by one of us (C.W.) and goiter was considered present when the thyroid was well palpable and unequivocally enlarged. Blood was drawn prior to and for intervals up to 42 days after the contrast agent. In cases of 1203
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1204
JCE & M • 1976 Vol 43 • No 6
BURGI ET AL.
TABLE 1. Non-hormonal iodine in serum, calculated as the difference between total and thyroxine iodine after radiographic examinations; before the radiographic examination the non-hormonal iodine was below the detection limit of the method (below 1 yug/100 ml) Non-hormonal iodine of serum Mean ± SEM (jag/100 ml) Day 3 Oral cholecystogram with 2 g Na-iopanoate (N = 10) Intravenous cholangiogram with 4 g Na-ioglycamate + 6 g methyl-glucamine ioglycamate (N = 10) Intravenous urogram with 8.0 g Na-diatrizoate and 52.8 methyl-glucamine diatrizoate (N = 9)
striking changes of serum hormones the patients
or their physicians were contacted to obtain additional clinical information. Thyroid scintiscans were done in a few selected patients 3 months or more after the cholecystograms. In cases where the serum thyroxine had not returned to normal after 42 days, additional blood samples were taken after 2 to 3 months. For oral cholecystography 2 g Na-iopanoate (Bilijodon) containing 1.3 g iodine were given per os in 4 tablets. For iv cholangiography 4 g Na-ioglycamate and 6 g methylglucamine ioglycamate (Bilivistan) were injected in a 20 ml solution. The iodine content was 5.6 g. For iv urography 8.0 g Na-diatrizoate and 52.8 g methylglucamine diatrizoate (Urografin) were injected in a volume of 80 ml. The iodine in this mixture amounted to 29.6 g. Study B Four women (ages 58 to 68 years) with well established severe idiopathic myxoedema (serum thyroxine originally below 1.5 /Ag/100 ml) who were taking a constant dose of L-thyroxine (0.1 to 0.2 mg per day) for more than 6 months and who appeared clinically euthyroid agreed to participate in the study. Blood was drawn prior to and 4 days after 2 g Na-iopanoate p.o.
Day 7
Day 14
Day 28
Day 42
2,171 ± 1,325
81 ± 28
15.9 ±5
5.4 ± 1.8
2.0 ±0.6
2,627 ± 718
192 ± 69
42 ±7
12 ±2
12 ± 3
1.5 ± 1.5
1.2 ± 1.2
34.3 ± 18.4
2.8 ± 1.6
1.8 ± .18
Analytical methods The following measurements in serum were made: Thyroxine (T4D) by a commercial competitive protein binding set (Tetrasorb Abbott) and in selected cases also by radioimmunoassay (T4RIA) (17); free thyroxine (FT4) by a modified equilibrium dialysis method (18); 3, 5, 3'-triiodo1-thyronine (T3 RIA) (19), 3,3',5'-triiodo-l-thyronine ("reverse T3") (20) and thyrotropin (TSH) (21) by specific radioimmunoassays; total iodine by a eerie acid color reaction (22). Non-hormonal iodine was calculated by subtraction of thyroxine iodine from total serum iodine. All sera of each patient were analyzed in the same assay. The ranges of these tests for normal persons are given in Table 6. Na-iopanoate added in vitro in a concentration up to 100 /ug/100 ml did not interfere with the assays of T4, T3 and reverse T3; moreover, in 3 untreated hypothyroid patients Na-iopanoate caused only an insignificant rise of T4 by 0.2 /Ltg/100 ml and reverse T4 remained immeasurable. T3 decreased from 61 to 43 /ng/100 ml. This renders an interference of Na-iopanoate with the assays quite unlikely.
Results
Study A Study C 4 healthy men (ages 23 to 40 years) were given 0.2 mg L-thyroxine p.o. daily for 21 days to suppress endogenous thyroid hormone secretion. They took 2 g Na-iopanoate p.o. on day 17. Blood drawn at days 15, 17, and 21 was analyzed.
Table 1 gives the time course of the serum concentrations of the three contrast media, expressed as non-hormonal iodine. After urography and after oral cholecystography, non-hormonal iodine has returned to below 2 jLig/100 ml within 2 and 6 weeks respec-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1205
T4, T3 AND REVERSE T3 AFTER X-RAY DYES tively. After iv cholangiography, nonhormonal iodine remains elevated to 12 /xg/ 100 ml after 6 weeks. After oral cholecystography, serum T4 rose consistently within 3 to 7 days in all 10 patients with and in the 12 patients without goiter (Fig. 1). At both times the rise was statistically highly significant. The rise was comparable in the 2 patient groups (1.8 and 2.6 /Ltg/100 ml respectively). In 5 patients the serum T4 was also measured by a radioimmunoassay, which showed good agreement with the routine competitive binding assay (Table 2). FT 4 was measured in 9 patients and rose markedly in 8 of them (Table 2). After 42 days, the average T4 had returned to baseline in both the goitrous and non-goitrous subjects. In 2 subjects, the T4 was still above 12 /xg/100 ml, but it returned to normal within the following 2 months (Fig. 1). The serum T3 showed a behavior opposite to that of the T4 (Fig. 2). It dropped in the majority of patients within 3 to 7 days by an average of 29 ng/100 ml in the nongoitrous and 22 ng/100 ml in the goitrous persons. After 2 weeks, the T3 rose again and after 6 weeks its concentration was slightly but significantly higher than baseline. 3,3',5'-triiodo-l-thyronine (reverse T3) was measured in 7 patients who had shown clear-cut decreases of T3 after oral cholecystography (Table 3). It increased from a basal value of 72 ng to 158 ng after 3 days and then slightly decreased again to 129 ng/100 ml. The reverse T3 thus behaved inversely to T3. When the magnitude of the changes of T3 and of reverse T3 between day 0 and day 7 were analyzed for all 7 patients, a highly significant linear inverse relationship between the two variables emerged (correlation coefficient r = -0.946; P < 0.001). In all 9 patients in whom it was measured, TSH rose after cholecystography, and 3 days after the contrast agent the mean serum concentration had doubled and was well above the normal range (Table 3).
Tt(O) pgHOOml 6
CHOLECYSTOGRAM
WITHOUT GOITER
WITH GOITER
42 DAYS
2-4 MONTHS
FIG. 1. Serum thyroxine in 10 patients with euthyroid goiter and in 12 patients without thyroid abnormality after oral cholecystography with Na-iopanoate. The thin lines connect the values of individual patients. The thick bars indicate the mean values with the standard errors. The abscissa gives the days after the cholecystography. The results of Student's t test for paired comparisons to the initial measurement are indicated on top: +, + + , and + + + mean P < 0.05, 0.01 and 0.001, respectively.
Thyroid scans with 131I were obtained in 3 goitrous patients who had shown a marked rise of the serum T4 after cholecystography. In all patients, the thyroid was moderately enlarged, but no autonomous nodules were seen, although this was not further confirmed by scans after T3 suppression. After iv cholangiography, there was on the average a slight increase of the T4 and a decrease of T 3 (Table 4); however the response was highly variable from one patient to the other and the changes were not statistically
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1976 Vol 43 • No 6
BURGI ET AL.
1206
TABLE 2. Serum thyroxine by radioimmunoassay (T4 RIA) or by displacement (T4 D) and free thyroxine (FT4) after radiographic contrast agents T4 (RIA) Ozg/100 ml) Subject C 4 C 9 Cll C16 C 19 C 2 C 8 C 10 C22
Goiter
+ + + +
initial
peak
10.2
11.3
12.6 10.6 11.2 11.2
16.5 16.5 15.8 13.7
Mean ± SEM of all 9 patients Mean ± SEM of 5 patients in whom both T4D and T4 RIA were done
11.16 ±0.40
14.76 ± 1.00
FT4 (ng/100 ml)
T 4 (D) (/ug/100 ml) initial
peak
initial
peak
8.0 7.2 8.5 10.4 9.9 11.9 9.8 11.9 11.4
10.4 10.4 11.7 13.0 11.6 17.9 14.4 15.0 13.5
1.76 1.43 1.59 2.69 2.28 2.93 1.27 2.21 2.37
2.16 2.35 2.16 2.86 2.32 5.60 1.99 2.79 3.47
9.89 ±0.57
13.10 ±0.81
11.08 ± 0.42
14.76 ±0.85
2.06 ±0.19
2.85* ±0.37
Patients with clear-cut increases of the T4D were selected for measurement of FT4 and in 5 cases also of T4 RIA. The initial and the peak values of each individual patient (day 3 or 7) are given. * t test for paired samples: P < 0.025.
significant. 6 of the 16 patients showed transient rises of the T4 above 12 />ig/100 ml. These elevations took place after different intervals and therefore have no effect on the group average. Table 5 shows that T4 and reverse T3 remained on the average remarkably stable after iv urography. T3 showed a rise after 42 days, but the change did not quite reach statistical significance. None of the 58 patients of study A developed clinical evidence of hyperthyroidism. Study B The 4 hypothyroid patients on constant replacement therapy were all clinically euthyroid. Nonetheless, TSH in serum was moderately elevated in 3 of them, indicating that the replacement dose was slightly too low. In these 3 patients, Na-iopanoate produced a further marked rise of TSH which, due to wide individual scatter of values, was not statistically significant (Table 3). The fourth patient had an unmeasurable TSH, both before and after Naiopanoate. Thus, her replacement dose was
probably slightly too high. In all four patients, Na-iopanoate produced a highly significant fall of T3RIA and a rise of reverse T3, while the T4D and the FT4 remained unchanged (Table 6). Study C In normal individuals given 0.2 mg Lthyroxine to suppress endogenous thyroid secretion, the thyroid hormones and reverse T3 had reached a steady state in serum after 15 days of administration of 0.2 mg L-thyroxine, since there was no change of these indices between day 15 and day 17 (table 6). Na-iopanoate again produced a fall of T3 and a rise of reverse T3 in serum. T4 and FT4 rose slightly, but the changes were not significant. Discussion Our studies establish that cholecystography consistently produces changes in circulating thyroid hormones. Since iopanoic acid (Telepaque) or its sodium salt (Bilijodon) are among the most frequently
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1207
T4, T3 AND REVERSE T3 AFTER X-RAY DYES used radiographic contrast agents in many countries, the findings are of general clinical interest. An elevated serum T4 might easily mislead one to diagnose hyperthyroidism, unless other indices such as the T3 or TSH are also measured. A change in thyroxine-binding globulin or a technical artefact does not account for the rise in serum T4, since the rise was confirmed by a second independent method of measurement and since FT4 also rose (Table 1). The serum T4 rose in all persons studied, irrespective of whether they had simple goiter or a normal thyroid gland. Thus the observed change was not due to an uptake of excess iodine by autonomous nodules with consequent hypersecretion of T3(WA) ng/100 ml 300 ,.
CHOLECYSTOGRAM
200
100
WITHOUT GOITER
300
200
100
WITH GOITER 0
3
7
28
42 DAYS
FiG. 2. Serume triiodothyronine after oral cholecystography in the same patients as shown in Fig. 1. The meaning of the symbols is the same as in Fig. 1.
TABLE 3. Mean and standard error for T3) reverse T3
and TSH in serum after cholecystography with Na-iopanoate
Day 0 Day 3 Day 7
T3 (RIA) ng/100 ml (N = 7)
Reverse T3 (RIA) ng/100 ml (N = 7)
TSH /xU/ml (N = 9)
194.0 ± 12.8 142.6 ± 9.1*** 156.6 ± 17.9*
57.0 ± 7.8 121.0 ± 15.0*** 97.7 ± 12.5*
5.5 + 0.8 12.2 ± 2.1**
Asterisks give the results of t tests for paired samples between day zero and day 3 and 7: * = P < 0.02, ** = P < 0.01, *** = p < 0.005. Due to limited availability of serum, tests could not be done in all 22 study subjects.
hormone (23). Moreover such nodules were not found in scintigrams performed in patients several months after Na-iopanoate. Our results establish that the lowering of the T3, the rise of the reverse T3 and the rise of TSH all take place in persons in whom the supply of thyroxine is kept constant (studies B and C). On the other hand, the rise of serum T4 only occurs when there is functioning thyroid tissue (study A). Admittedly the study does not establish the exact temporal relationship of events, but it seems likely that the primary event after Na-iopanoate is the fall of the serum T3. The resulting state of T3 deficiency of the peripheral tissues causes a rise in TSH, which in turn stimulates T4 secretion by the thyroid gland. Our studies do not allow us to decide whether the lowering of the serum T3 is due to decreased production or enhanced degradation of this hormone. A large part of the circulating T3 (24) and virtually all of the circulating reverse T3 (25) arise in peripheral tissues by monodeiodination of T4. The fact that Na-iopanoate affects both T3 and reverse T3 and that the changes in addition are strictly reciprocal with a high degree of correlation, suggests that the contrast agent interferes somehow with the monodeiodination of T4 and thereby with the production of T3 and reverse T3. However, more studies are needed to strictly exclude an effect on T3 and reverse T3 degradation. Similar reciprocal changes of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
1208
BURGIETAL.
^ S l
TABLE 4. Thyroid hormone parameters in serum after intravenous cholangiography with ioglycamic acid Day 0 T, (D) (/i.g/100 ml) N = 15 T3 (RIA) (ng/100 ml) N=15 reverse T3 (RIA) (ng/100 ml) N= 3
Day 3
Day 7
Day 14
Day 28
Day 42
9.2 ± 0.5
9.6 ± 0.6
10.1 + 0.5
9.1 ±: 0.6
8.5 ± 0.6
8.0 ± 0.5
166.5 ± 12.9
147.1 ± 17.5
158.6 ± 15.7
172.7 ±: 13.1
185.1 ± 16.6
171.8 ± 14.6
—
—
47
± 3
57
± 2
54
± 10
The data for the 7 patients with euthyroid goiter and the 8 patients with a normal thyroid gland were pooled since there was no difference between the two groups. Reverse T3 was only measured in 3 patients. The data give the mean with the standard error. None of the changes from baseline was statistically significant.
T3 and reverse T3 occur in the newborn (26,27) and during fasting (28). Chemically, iopanoic acid is a benzene ring substituted with 3 iodine atoms, one amino group and an isovaleric acid side chain (Fig. 3). Superficially at least the molecule resembles thyroxine and it is conceivable that it interferes with the enzymatic deiodination of thyroxine as shown for other iodinated benzene derivatives (29) and thyroid hormone analogs (30,31). In these latter studies T3, reverse T3 and TSH were not measured so that a direct comparison is not possible. Iopanoic acid is mostly excreted as the glucuronide (32). Deiodination has not been demonstrated and, if it occurs, it accounts for only a very small fraction of its metabolism (33). Nonetheless, even with a very low affinity for deiodinating enzymes iopanoic acid could interfere with T4 metabolism, since its initial molar serum concentration exceeds that of T4 by a factor of 250 (Table 1). The observation that iopanoic acid depresses radioio-
dine uptake for several weeks provides some indirect evidence that part of it is deiodinated in the body (1). Iodide excess per se could partly explain the initial decrease of T3 (4), but neither iodide excess alone nor urography with diatrizoic acid nor cholangiography with ioglycamic acid (tables 4 and 5) causes similar changes of T3 and reverse T3. Unlike iopanoic acid, diatrizoic and ioglycamic acid have a benzoic acid group on the iodine-substituted benzene ring (Fig. 3). The strong negative charge of this group could radically change the affinity of these agents to a deiodinating enzyme system as compared to iopanoic acid. Severe iodine deficiency may lower the circulating T4 and raise the T3 (34), and Na-iopanoate might have acted by simply improving the iodine supply in previously iodine deficient patients. However, the fact that other iodinated contrast agents did not cause the same changes is strong evidence against such an interpretation. Moreover,
TABLE 5. Thyroid hormone parameters in serum after intravenous urography with diatrizoic acid
T4 (D) ()Ltg/100 ml) N = 19 T3 (RIA) (ng/100 ml) N=19 reverse T3 (RIA) (ng/100 ml) N = 3
Day 0
Day 3
Day 7
Day 14
Day 28
Day 42
9.1i :0.3
9.2 ±:0.4
9.6 ±:0.3
9.0:t 0.4
9.1:t 0.3
9.5 ± 0.5
185.5 ±:9.6
178.6 ±:7.6
186.7 ±:9.8
214.4 ± 8.8*
64 ±: 9
65 ±: 9
—
57
*: 4
The data for the 9 patients with euthyroid goiter and the 10 patients with a nomial thyroid gland were pooled, since there was no difference between the two groups. Reverse T3 was only measured in three patients. The data give the mean with the standard error. * t test compared to day zero F < 0.10.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 April 2015. at 00:05 For personal use only. No other uses without permission. . All rights reserved.
T4, T3 AND REVERSE T3 AFTER X-RAY DYES
1209
TABLE 6. Mean and standard error for T4D, FT4, T3RIA, reverse T3 and TSH in serum
Normal range
T4D (/xg/100 ml)
FT4 (ng/100 ml)
T3RIA (ng/100 ml)
Reverse T3 (ng/100 ml)
TSH OiU/ml)
5.5-12.0
1.33-2.41
100-220
20-80
