ANNALS O F CLINICAL AND LABORATORY SCIENCE, Vol. 28, No. 2 Copyright © 1998, Institute for Clinical Science, Inc.

Abnormal Proinsulin Levels in Thyroid Dysfunction Measured by a Sensitive Proinsulin Immunochemiluminoassay*

TA-JEN WU, M.D.,t SHIH-MING HUANG, M .D.,i

ROBERT L. TAYLOR, B.S.,§ and PAI C. KAO, Ph.D.§

fSection o f Endocrinology and Metabolism , D epartm ent o f Internal Medicine,

National Cheng R ung University Hospital, Tainan, Taiwan ROC

and$D epartm ent o f Surgery,

Tzu Chi Buddhist General Hospital, Hualien, Taiwan ROC

and§Section o f Clinical Biochemistry,

Departm ent o f Laboratory Medicine and Pathology, Mayo Clinic and Mayo Foundation,

Rochester, M N 55905

ABSTRACT

Abnormalities of carbohydrate metabolism in hyperthyroid patients have been long noted. Elevated proinsulin level is considered as an early marker of B-cell impairment. Proinsulin levels in hyperthyroid patients decreased after antithyroid drug therapy. However, proinsulin in hypothyroid patients was only rarely reported, and the difference was only demonstrated after glucose stimulation—there was a greater response of proinsulin secretion after thyroxine therapy—and the basal fast­ing proinsulin level was not different after therapy. One of the reasons might be that the assay was not sensitive enough to detect the change of basal proinsulin levels in patients with hypothyroidism after therapy. A newly developed immunochemilumi- nometric assay of proinsulin was used to demonstrate that the suppressed proinsulin level increased after thyroxine therapy in hypothyroid patients (4.2 ± 2.4 vs. 10.0 ±5.6 pmol/L, p < 0.05; n = 7). On the other hand, our study also confirmed that the proinsulin levels decreased in hyperthyroid patients after antithyroid therapy by methimazole (27.8 ± 26.0 vs. 15.8 ± 15.7 pmol/L, p < 0.05; n = 12). In conclusion, proinsulin increased in hypothyroid patients after thyroxine therapy and decreased in hyperthyroid patients after methimazole therapy. The results demonstrated there is a high correlation between thyroid function and B-cell function in hypothyroid as well as hyperthyroid patients.

* Address reprint requests to: Pai C. Kao, Ph.D ., Mayo Clinic, 200 First Street, SW, Rochester, MN 55902.

820091-7370/98/0300-0082 $01.50 © Institute for Clinical Science, Inc.

PROINSULIN IN THYROID DYSFUNCTION 8 3

Introduction

Impaired glucose tolerance in hyperthyroid patients has been recognized and studied.1,2,3,4 H y p erin su lin em ia has b een freq u en tly observed in hyperthyroid patients.3,4 Hyperin­sulinemia was regarded as a manifestation of insulin resistance. Recently, it was suggested that proinsulin makes up a great proportion of increased im m unoreactive insulin (IRI) results. This was found by using specific pro­insulin assays to measure a patient’s speci­mens.5 The increased proinsulin has been pro­posed as an early marker of B-cell defect.6,7,8,9 Although insulin metabolism in hyperthyroid­ism was widely studied, the study of proinsulin levels in hyperthyroidism was still limited.

The proinsulin levels in hypothyroidism were even less studied. The reports of proin­sulin level in hypothyroidism are rare. In addi­tion, results from animal and human studies were rather conflicting.10,11,12’13 Searching Medline from 1966, only two articles were found to have been published. One of the studies applied proinsulin radioimmunoassay14 and the other study applied immunoradiomet- ric assay.15 Both articles indicated that the pro­insulin levels of hypothyroid patients were not different from proinsulin levels of euthyroid controls and did not increase after thyroxine therapy. However, proinsulin in response to glucose stimulation was increased after thyrox­ine therapy.15 The response of proinsulin increase in thyroxine treated hypothyroid patients was greater than that of untreated patients. These data suggested that the early assay might not be sensitive enough to detect the change of basal proinsulin levels in patients with hypothyroidism after therapy, but it might be sensitive enough to detect the increase after glucose stimulation. For this rea­son, a newly developed immunochemilumino- metric assay (ICMA) of proinsulin16 was used to determine the changes of basal proinsulin levels in hypothyroid patients after thyroxine treatment. The decrease of proinsulin levels of hyperthyroid patients was determined after antithyroid treatment.

Patients and Methods

P a t ie n t s

Twelve hyperthyroid patients (9 female and 3 male) and 7 hypothyroid patients (2 male and 5 female) who w ere trea ted at the National Cheng Kung University Hospital were enrolled in this study after informed con­sent. None of them had a family history of diabetes mellitus. Eleven of 12 thyrotoxic patients were diagnosed as Graves’ disease; the other one was diagnosed as multinodular toxic goiter. The clinical diagnoses of 7 patients with hypothyroidism were atrophic thyroid in 3 and Hashimoto’s thyroiditis in 4 patients. The clini­cal features of the patients are listed in table I. After confirming thyroid status by clinical manifestations and thyroid function tests, fast­ing plasma glucose, creatinine and proinsulin were measured. After methimazole treatment for hyperthyroidism or thyroxine replacement therapy for hypothyroidism, the thyroid func­tion tests (including serum T3, T4, and TSH), fasting plasma glucose, and proinsulin levels were studied again 3 months later.

B io c h e m is t r y

Plasma glucose was determined by glucose oxidase reaction (Hitachi 747). Serum creati­nine was measured with Hitachi 747.

Thyroid Function Tests: Commercial kits were used to measure T3 (Coat-A-Count Total T3, Abbott; Solid-phase RIA; normal range 86-187 ng/dL; intraassay CV 5.5 percent, interassay CV 7.6 percent), T4 (Coat-A-Count Total T4, Abbott; Solid-phase RIA, normal range 4.5-12.5 |xg/dL; intraassay CV 3.1 per­cent, interassay 8.2 percent), and TSH levels (SPAC-S TSH kit, IRMA method, Daiichi; normal range: 0.5-5.6 (xU/mL; intraassay CV 2.1 percent, interassay CV 2.5 percent).

Immunochemiluminometric Assay (ICMA): Proinsulin levels w ere m easured w ith a po lyclonal im m u n o ch em ilu m in o m etric assay.16,17,18 In brief, one goat was immunized with synthetic C-peptide conjugated with bovine serum albumin (BSA). The antisera

TABLE I8 4 WU, HUANG, TAYLOR, AND KAO

Clinical Features of the Studied Patients

Hvperthvroidism (n = 12) Hypothyroidism (n = 7)Mean± S.D. Min Max Mean ± SD Min Max

Age 35 ± 11 19 48 38 ±13 19 60BMI 22.0 ± 3.3 18.5 28.3 24.1 ± 3.8 19.5 31.2Cre 0.6 ± 0.1 0.4 0.8 1.0 ± 0.6 0.7 2.3Glu 92 ± 8 82 103 84 ± 8 76 99Glu* 90 ± 8 80 105 85 ± 6 78 96T4 18.0 ± 2.7 15.0 22.1 1.6± 1.4 0 3.9T4* 8.0 ± 3.2“ 1.4 11.3 7.8 ± 1.8* 5.5 10.5T3 399 ±116 221 600 64 ±31 37 128T3* 141 ± 52* 59 220 108 ± 23b 79 142TSH 0.02 ± 0.02 0.01 0.05 143 ±80 47 229TSH* 8.8 ± 25.9 0.05 90.8 32.7 ± 62.0° 0.27 168.3

Min = Minimum.Max = Maximum.Age = Years.BMI = Body mass index [= body weight (kg)/body height (m)2].Cre = Creatinine (mg/dL).Glu = Fasting plasma glucose (mg/dL).T4 = ng/dL.T3 = ng/dL.TSH = nll/mL‘ Denotes value after therapy.ap < 0.0001, bp < 0.05, and °p < 0.01, comparison before and after therapy.

were immunopurified on a C-peptide affinity column. The purified antibodies were labeled with acridinium ester and used as a signal anti­body. A second goat was immunized with pork insulin conjugated to BSA; the antiserum was immunopurified by human insulin affinity col­umn. The purified insulin antibodies were immobilized on a plastic bead and used as a capture antibody, and paired with acridinium ester-labeled C-peptide signal antibody to form a “sandwich” ICMA.

After washing to remove “free” signal anti­bodies, chemiluminescence of the signal anti­body linked on the bead vs. proinsulin was detected by luminometer. Sensitivity (deter­mined by two standard deviations above back­

ground) was 0.2 pmol/L for proinsulin. Intra­assay and interassay coefficients of variations were 5.0 percent and 9.0 percent, respectively, for proinsulin. The assays showed no cross­reactivity with insulin or C-peptide. The ref­erence range for proinsulin ICMA, which was determined with 120 healthy volunteers, was 3 to 20 pmol/L.

Statistics: P aired t- te s t was used toanalyze the results. A p < 0.05 was consid­ered significant.

Results

Eleven out of 12 hyperthyroid patients had their proinsulin levels decreased after antithy-

PROINSULIN IN THYROID DYSFUNCTION 8 5

roid drug—methimazole therapy (average reduction of mean ± S.D. from 27.8 ± 26.0 to 15.8 ± 15.7 pmol/L, p < 0.05). Five of the 12 hyperthyroid patients had their proinsulin lev­els greater than the upper limit of reference range, 20 pmol/L. After a 3-month methima­zole treatm ent, all of the 5 patients had decreased proinsulin levels; however, 2

patients still had proinsulin levels greater than the upper limit of normal range. In total, 11 of the 12 patients had decreased proinsulin levels but one patient had an increased proinsulin level after therapy (figure 1). In hypothyroid patients, three of the 7 had lower proinsulin levels than the lower limit of reference range (3 pmol/L). All 7 patients had significantly

P r o i n s u l i n(pmol/L)

90 -|- p<0.05 - j

80

70 -

60 -

50 -

40 -

# : HyperthyroidismO : Hypothyroidism

|- p<0.05 ~|

F i g u r e 1. Changes o f proin- sulin levels before and after 3 months o f therapy. Hyperthy­roidism by the antithyroid drug, methimazole; hypothyroidism by thyroxine. Shaded areas indicate reference range o f proinsulin.

30

20

10

Before Tx After Tx Before Tx After Tx

86 WU, HUANG, TAYLOR, AND KAO

increased proinsulin levels after thyroxine replacement therapy (4.2 ± 2.4 vs 10.0 ± 5.6 pmol/L, p < 0.05) (figure 1).

Discussion

Using our im m unochem ilum inom etric assay, it was demonstrated that the results of increased proinsulin levels in hyperthyroidism were consistent with the results measured by radioimmunoassay14 or by immunoradiometric assay.15 The results suggested impaired B-cell function in hyperthyroidism. A study of carbo­hydrate metabolism in hypothyroidism was more limited. Since 1966, there were only two studies published; one by Sestoft et al14 and the other by Beer et al.15 They both indicated that there was no change in proinsulin levels after thyroxine treatm ent in hypothyroid patients. Beer et al reported that the basal (fasting) proinsulin levels were not changed; however, there was a significant increase in proinsulin levels after glucose stimulation per­formed by glucose tolerance test in the group of thyroxine-treated hypothyroid patients. These data suggested that the no difference in basal proinsulin level of hypothyroid patients was due to their assays which were not sensi­tive enough to measure the minute change in the proinsulin level after thyroxine therapy.

In contrast to both studies, the decrease in basal proinsulin levels in our hypothyroid patients was statistically significantly lower than the control. In comparison with previous studies, our assay had an improved sensitivity (0.2 pmol/L vs. 10 pmol/L1419 and 0.5 pmol/ L 15), which may contribute the difference. Since insulin sensitivity reported no decrease in hypothyroidism,20 our result suggested that decreased B-cell activity may have occurred in hypothyroidism.21'

Two of the hyperthyroid patients showed marked elevated proinsulin levels, which decreased after methimazole therapy but were still greater than the upper limit of reference range (figure 1). One of these two patients was overweight. Obesity, which was currently con­sidered to be related to insulin resistance, might be a contributing factor of the elevated

p ro in su lin level. H ow ever, p ro in su lin decreased after 3 months of methimazole therapy, when body weight was increased slightly (from 66 kg to 67.5 kg) after therapy. Oral glucose tolerance test after therapy showed normal (not shown) indicating insulin resistance was unlikely. A pparently the decrease in proinsulin level after therapy in this patient was due to effect of improved thy­roid status.

The other patient with marked elevated pro­insulin, who was lean (body mass index 19.3), had a history of hyperthyroidism for more than2 years and did not receive any treatment until the study. After 3 months of methimazole therapy, the thyroid function test still revealed mild T3 toxicosis. It was suspected that the long-term hyperthyroid disease had a pro­found effect on B-cell function, which led to a much increased proinsulin level. The mecha­nisms of proinsulin level changes in thyroid dysfunction should be studied further.

Conclusion

Increased proinsulin levels in patients with hyperthyroidism and decreased proinsulin lev­els in hypothyroid patients were demonstrated in this study. After therapy, the patients showed decreased and increased proinsulin, respectively. The results imply there is a high correlation between B-cell function and thy­roid function.

References

1. Maxon HR, Kreines KW, Goldsmith RE, et al. Long­term observations o f glucose tolerance in thyrotoxic patients. Arch Intern Med 1975;135:1477-80.

2. Ikejiri K, Yamada T, Ogura H, et al. Age-related glu­cose intolerance in hyperthyroid patients. Diabetes 1978;27:543-9.

3. Osei K, Falko JM, O’Dorisio TM, et al. Decreased serum C-peptide/insulin molar ratios after oral glu­cose ingestion in hyperthyroid patients. Diabetes Care 1984;7:471-5.

4. Ikeda T, Fujiyama K, Hoshino T, et al. Oral and intra­venous glucose-induced insulin secretion in hyperthy­roid patients. Metabolism 1990;39:633-7.

5. Temple RC, Clark PMS, Nagi DK, et al. Radioimmu­noassay may overestim ate insulin in non-insulin- dependent diabetics. Clin Endocrinol 1990;32:689-93.

PROINSULIN IN THYROID DYSFUNCTION 8 7

6. Spinas GA, Snorgaard O, Hartling SG, et al. Elevated proinsulin levels related to islet cell antibodies in first- degree relatives o f ID D M patients. Diabetes Care 1992;15:632-7.

7. Deacon CE, Schleser-Mohr S, Ballmann M, et al. Preferential release o f proinsulin relative to insulin in non-insulin-dependent diabetes mellitus. Acta Endo- crinologica 1988;119:549-54.

8. Persson B, Hanson U, Hartling SG, et al. Follow-up of wom en with previous GDM. Insulin, C-peptide, and proinsulin responses to oral glucose load. Diabetes 1991;40(Suppl. 2):136-41.

9. Davies MJ, Rayman G, Gray IP, et al. Insulin defi­ciency and increased plasma concentration of intact and 32/33 split proinsulin in subjects with impaired glucose tolerance. Diabetic Med 1993;10:313-20.

10. Achmadi J, Terashima Y. The effect of propylthioura­cil-induced low thyroid function on secretion response and action o f insulin in sheep. D om estic Animal Endocrinol 1995;12:157-66.

11. Jolin T, Morreale de Escobar G, Escobar del Rey F. Differential effects in the rat o f thyroidectomy, pro­pylthiouracil and other goitrogens on plasma insulin and thyroid weight. Endocrinology 1970;87:99-110.

12. Shah JH, Cerchio GM. Hypoinsulinemia of hypothy­roidism. Arch Intern Med 1973;132:657-61.

13. Hays JH, Silverman E, Potter BB, et al. Normal gas­tric inhibitory polypeptide response to oral glucose in hypothyroidism. J Endocrinol 1994;140:309-12.

14. Sestoft L, Christensen NJ, Saltin B. Responses o f glu­cose and glucoregulatory hormones to exercise in thy­rotoxic and myxoedematous patients before and after3 months o f treatment. Clin Sci 1991;81:91-9.

15. Beer SF, Parr JH, Temple RC, et al. The effect of thyroid disease on proinsulin and C-peptide levels. Clin Endocrinol 1989;30:379-83.

16. Kao PC, Taylor RT, Heser DW. C-peptide immuno- chemiluminometric assay developed from two seem ­ingly identical polyclonal antisera. Ann Clin Lab Sci 1992;22:307-16.

17. Kao PC, Taylor RL, Service FJ. Proinsulin by immu- nochemiluminometric assay for the diagnosis o f insu­linoma. J Clin Endocrinol Metab 1994;78:1048-51.

18. Wu T-J, Lin C-L, Taylor RL, et al. Proinsulin level in diabetes mellitus measured by a new immunochemi- luminometric assay. Ann Clin Lab Sci 1995;25:467—74.

19. Heding LG. Specific and direct radioimmunoassay for human proinsulin in serum. Diabetologia 1977;13: 467-74.

20. Shah JH, Motto GS, Papagiannes E, et al. Insulin metabolism in hypothyroidism. Diabetes 1975;24: 922-5.

21. Kung AW, Lam JS, Pun KK, et al. Circulating somato­statin after oral glucose in hypothyroidism. J Endocri­nol Invest 1990;13:403-6.