KENJI MORIYAMA, TAKASHI AKAMIZU, MITSUYOSr UMEMOTO, MASAxo MIURA, MISA SAIJO,

Transcription

1 Endocrine Journal 1999, 46 (5), A Case of Hashimoto's Thyroiditis with Markedly Elevated Serum Thyroglobulin and Evidence of its Influence on the Measurement of Anti-Thyroglobulin Antibody by Highly Sensitive Assays KENJI MORIYAMA, TAKASHI AKAMIZU, MITSUYOSr UMEMOTO, MASAxo MIURA, MISA SAIJO, KATSUMI TANIGUCHI* AND KAZUWA NAKAO Department of Medicine and Clinical Science, Kyoto University School of Medicine, 54 Shogoin Kawaharacho, Sakyo-ku, Kyoto , Japan *Institute of Immunochemistry, Eiken Chemical Co., Ltd., Shimoishigami, Dotawara, Tochigi 324, Japan Abstract. We present the case of a 66-year-old woman with Hashimoto's thyroiditis, who showed extremely high concentrations of serum thyroglobulin (Tg). Serum Tg levels were markedly elevated following a slight elevation of serum thyrotropin (TSH) (22,000 ng/ml and 11.1 pu/ml, respectively). Although elevated concentrations of serum Tg declined concomitant with decrease of serum TSH one month later, Tg concentrations remained high (>948 ng/ml) even at normal or suppressed TSH levels. There was no evidence of massive thyroid tissue damage or thyroid tumor. To our knowledge, there have been no case reports of such high concentrations of serum Tg (>2 x 104 ng/ml) in the clinical course of Hashimoto's thyroiditis. Furthermore, we showed evidence that extremely high Tg levels could possibly influence the measurement of anti-tg autoantibody using highly sensitive radioimmunoassays. Key words: Hyperthyroglobulinemia, Hashimoto's thyroiditis, (Endocrine Journal 46: ) THYROGLOBULIN (Tg) is a normal constituent of serum detectable by radioimmunoassay in all normal subjects [1]. The levels are altered in a number of physiological and pathological conditions. Abnormal concentrations of Tg are frequently observed in association with thyroid diseases; elevated serum Tg concentrations may result from excess thyroidal stimulation, physical thyroid damage or abnormalities in the thyroid mass. Patients with chronic autoimmune thyroiditis would be expected to have high serum Tg concentrations because of chronic thyroid inflammation and/or elevated thyrotropin (TSH) levels characteristic of this condition [1], but the levels of serum Tg rarely exceed 1000 ng/ml under normal TSH level [2, 3] At the same time, the exact mechanisms governing these phenomena have been poorly understood. We described here a case of markedly elevated serum Tg associated with Hashimoto's thyroiditis without any apparent causes, such as severe intrathyroidal destruction, association of Tg producible tumor or large elevation of serum TSH. Possible mechanisms of Tg elevation were discussed. Received: January 8, 1999 Accepted: June 25, 1999 Correspondence to: Takashi AKAMIZU, M.D., Department of Medicine and Clinical Science, Kyoto University School of Medicine, 54 Shogoin-Kawaharacho, Sakyo-ku, Kyoto , Japan. Case report A 66-year-old woman who has been suffering from primary hypothyroidism since 1977 when she was 46,

2 688 MORIYAMA et al. was diagnosed as Hashimoto's thyroiditis based on the presence of diffuse goiter, positive anti-tg and anti-microsome antibody tests measured by the hemagglutination assay, and the finding of iodine (131I)-scintigraphy, and mild hypothyroidism (T3i 1.38 ng/ 100 ml; T4, 5.7 ug/ 100 ml; and TSH, 7.6 uu/ml). She had been treated with 50 pg of levothyroxine sodium (T4). She was also regularly given a single daily dose of 15 mg prednisolone orally to control interstitial nephritis possibly due to Sjogren's syndrome. Her compliance with these drugs was good. In addition, she had hypertension, hyperlipidemia and hyperuricemia. On May 14, 1977, she was admitted to Kyoto University Hospital for evaluation of cardiac function and control of the systemic edema. On admission, she had no complaints. Her blood pressure was 154/86 mmhg. A moderate-sized diffuse goiter with an elastic consistency was palpable without any tenderness or lymphadenopathy. Her face was mildly edematous. There were no abnormal findings in the chest, abdomen or neurological examinations. Laboratory data The laboratory findings of the patient on admission demonstrated a mild leucocytosis, hyperuricemia (8.8 mg/100 ml). Total cholesterol was increased (total cholesterol 367 mg/100 ml). Indices of renal function revealed renal insufficiency (creatinine 2.9 mg/ 100 ml, blood urea nitrogen 56 mg/ 100 ml), and anti-hepatitis C virus antibody (HCV) was positive though of unclear etiology. Although serum C- reactive protein (CRP) was not noticeably increased, erythrocyte sedimentation rate (ESR) was slightly accelerated (34 mm/hr). Data obtained from immunological examinations on admission were compatible with Sjogren's syndrome. Interstitial nephritis was diagnosed based on increased urine 3- microglobulin (5264pg/day) and al- microglobulin (81.9 fig/day) levels, and impaired urine concentration function. Renal biopsy was not performed because only the left kidney remained after right nephrectomy at age 25 due to renal tuberculosis. Thyroidal function tests on admission (Table 1) showed a slightly decreased T3 level (44 ng/ 100 ml), though both serum free thyroxine (FT4) and triiodine thyronine (FT3) were within normal range under replacement therapy with 50 ig of levothyroxine sodium. Serum Tg level on admission was already extremely high at 2851 ng/100 ml (Ab Bead Thyroglobulin `Eiken', Eiken Co. Ltd., Tokyo, Japan; see also Fig. 1). Anti-thyroid peroxidase Table 1. Thyroid function data on admission

3 HIGH THYROGLOBULIN LEVEL IN A HASHIMOTO PATIENT 689 antibody and anti-tg antibody levels (Thyroid Peroxidase Antibody Direct Assay Kit; Thyroglobulin Antibody Direct Assay Kit, RSR Ltd., Cardiff, U.K.) were slightly increased (0.6 U/ml and 0.45 U/ml, respectively), although both of the antibodies determined by hemagglutination assay were negative at admission. TSH binding inhibitor immunoglobulins (TBII) (RSR Ltd., Cardiff, U.K.) were not detected. Ultrasonography of the thyroid gland showed only a moderately-sized diffuse goiter with entirely homogeneous low echogenicity (diameter, 66 mm) (data not shown). A fine needle-biopsy specimen of the thyroid gland showed various degenerative eosinophils compatible with Hashimoto's thyroiditis. Pathological diagnosis based on Papanicolau stain was class II (data not shown). An open biopsy has not been performed for ethical reasons. Fig. 1. Clinical course of the patient. Top panel shows the alterations of serum Tg level ( ). The broken line indicates the upper limit of normal range. Lower panel shows changes of TSH ( ), free T4 (A), and free T3 ( ). The middle panel shows doses of drugs administered.

4 690 MORIYAMA et al. Serum Tg levels and other hormonal status during clinical course Changes in serum Tg and endocrinological parameters of the thyroid gland, such as TSH, FT4, and FT3 after admission, are illustrated in Figure 1. Serum Tg rose to very high levels following the elevation of TSH, probably due to the pause in medication for examinations of gastrointestinal fiberscopy for several days after admission. Elevated Tg levels fell following the decline of TSH level after resuming administration of T4, but still remained extremely high even at normal or suppressed TSH levels (Fig. 1). To examine the abnormalities in iodine transport, radioactive iodine (123I) uptake was performed and counted at 3 and 24 hours after administration. Results were apparently low at 0.5% (3 h) and 1.4% (24 h), probably due to suppressed TSH secretion (< 0.16,uU/ml) after replacement therapy with levothyroxine sodium. Perchlorate discharge test could not be performed because of the low iodine uptake. A TRH-test was also performed to examine the acute effects of TSH on Tg release. No effects of TSH on Tg level were observed within 3 hrs after TRH administration, although TSH responses to TRH were slightly weak and delayed (Table 1). Table 2. Tg determinations by other kits using the serum of May 27, 1997 Table 3. Dilution tests of patient serum thyroglobulin

5 HIGH THYROGLOBULIN LEVEL IN A HASHIMOTO PATIENT 691 Results To eliminate the possibility of artifacts in Tg measurements, we at first tested the reproducibility of Tg values and the existence of human anti-murine antibodies (HAMA). Repeated Tg determinations using the same kit confirmed the reproducibility, and the determinations by other kits using the serum of May 27, 1997 gave similar values (Table 2). Dilution test gave good recovery rates (Table 3). No HAMA was detected (data not shown). Next, we examined whether the Tg of the patient was of a native form or not, since heterogeneity is a key feature of Tg and Fig. 2. Elution profile of serum Tg with HPLC Sephacryl 5300 HR column (Pharmacia Biotech, Co., Ltd, Tokyo, Japan) was employed for gel filtration. Bed volume of the column and each fraction volume were 80 ml and 2 ml respectively. Five hundred microliters of the patient's serum of the May 30, 1997 was applied. Solid ( ), open ( 0) circles, and open triangle (o) show the elution pattern of Tg of the patient's serum, 123I-control Tg and 1231-human IgG, respectively. The sample of each fraction was also assayed with the Ab Bead Thyroglobulin `Eiken' kit (Eiken Chemical Co., Ltd., Tokyo, Japan). Amounts of 123I-control Tg and 1231-IgG were measured by counting the radioactivities. Table 4. Effects of Tg on the measurement of titration of anti-tg antibodies

6 692 MORIYAMA et al. involves its three-dimensional structure and immunoreactivity [4, 5]. When the Tg size of the sample dated May 30, 1997, was evaluated by high pressure liquid chromatography (HPLC), the eluted size of Tg in the main fractions was not deviated from those of 123I-Tg control (Fig. 2). Thus, abnormal-sized molecules, at least fragmented ones, were not detected in the patient Tg. It has been known that anti-tg autoantibodies in sera influence Tg measurements by the RIA of double antibody method [6]. However, this effect is much less in the kit of Ab Bead Thyroglobulin `Eiken', since the monoclonal anti-tg antibody used in this kit does not cross-react markedly with the autoantibodies. Kasagi et al. reported that the autoantibodies, whose titer is less than a thousand fold in the hemagglutination test or 70 U/mL in RIA, hardly influence Tg measurements by this kit [7]. Since the autoantibody of this patient was at a negative or low level in RIA, it may be assumed that this kit gave accurate Tg values in this patient. On the contrary, to detect the influence of high levels of Tg on the assay of anti-tg antibody, we tested if the addition of various amounts of Tg affected the titers of anti-tg antibodies. Various amounts of Tg were added to reference sera with a low, mild or high titer of anti-tg (Table 4). In all of the reference sera, the higher the amount of Tg added to the serum, the lower the titer of Tg antibodies in the assay (Tg antibody kit `Eiken', Eiken Chemical Co., Ltd.). Even the lower concentration of Tg (12,500 ng/ml) than the peak level of this patient (22,000 ng/ml) gave a false-negative result to a serum with a low titer of anti-tg. These results suggested that the titrations of patient anti-tg antibodies determined by this assay or by another radioimmunoassay (Thyroglobulin Antibody Direct Assay Kit, RSR Ltd.) were likely to be underestimated. Discussion This patient has been diagnosed as Hashimoto's thyroiditis based on the presence of antibodies against Tg and thyroid microsomal antigen, diffuse hypoechogenicity of the thyroid on ultrasonography, low uptake of 99mTc-scintigraphy and the presence of various degenerative eosinophils in the histological findings. No space occupying lesion was detected in the thyroid gland by scintigraphy or ultrasonography. Although the precise mechanisms of Tg elevation in Hashimoto's thyroiditis have not been clarified, there are several possibilities: 1) thyroid tissue destruction, 2) iodide organification defect, 3) excess thyroidal stimulation. Destructive changes in the thyroid such as painless thyroiditis were, though not completely neglected, considered unlikely based on the following findings: serum TSH elevated even at high levels of serum Tg, which elevated following the increase of TSH; 1231 intake was present, although low probably due to suppressed TSH secretion after replacement therapy with levothyroxine sodium (performed on June 22, 1997; data not shown); no systemic inflammatory reactions including fever, increased CRP, or high ESR. The possibility of iodide organification defect was not excluded by a perchlorate discharge test; however, Tg elevation in this defect is usually not so marked [8]. Excess thyroidal stimulation is especially true among patients with high levels of serum TSH (>40 mu/l) [2]. Patients who have some degree of hypothyroidism have elevated serum TSH concentrations and show high concentrations of Tg. In this patient, the elevation of serum TSH was rather slight compared to that of the serum Tg levels. This may be explained by the TSH stimulatory actions to the production of Tg which are thought to be different among cases. Supporting this speculation, there is a report that in some patients the Tg levels paralleled those of TSH behavior, while in others the TSH-Tg correlation was not so obvious [9]. Since Tg concentrations of this patient remained high even in the absence of increased TSH level, factors other than TSH or thyroid damage might be considered to influence the Tg level. The humoral factors controlling Tg gene expression include thyrotropin (TSH), insulin and insulin-like growth factor-1 (IGF-1), which synergistically stimulate transcription of Tg mrna [10]. In addition, cytokines and transcription factors, such as TTF-1 and Pax-8, may affect Tg gene transcription [11, 12]. Abnormal processes involved in the release of Tg into the circulation and clearances should also be considered. In this patient, Tg clearance could be slower possibly due to renal insufficiency, although extrathyroidal Tg clearance is supposed to be mainly performed by the metabolism in the liver [1].

7 HIGH THYROGLOBULIN LEVEL IN A HASHIMOTO PATIENT 693 However, the indices of renal function remained unchanged throughout this period, suggesting that renal insufficiency was unlikely to be one of the major factors raising her serum Tg. In Hashimoto's thyroiditis, when the presence of anti-tg antibodies, the concentration of serum Tg is expected to decrease [6]. It is, however, unlikely in the Tg measurement of this patient using the Tg assay kit (Ab Bead Thyroglobulin `Eiken', Eiken Chemical Co., Ltd.) in which monoclonal anti-tg antibodies do not cross-react markedly with anti-tg autoantibodies [7]. Conversely, very high levels of Tg may affect the assay of anti-tg antibodies [13, 14]. Ohtaki and Endo previously showed that the effects of Tg on the assay of anti-tg antibodies could be observed if the ratio of Tg to anti-human Tg antibodies were higher than 1:10 [13]. In fact, as shown in Table 4, these observations appear still to be true in antibody titers measured by the highly sensitive Tg antibody kit `Eiken' and also in those of this patient measured by Thyroglobulin Antibody Direct Assay Kit (RSR Ltd.) In summary, this is, as far as we know, the first report of a patient with Hashimoto's thyroiditis who showed extremely elevated serum Tg levels during the clinical course. Because the causes and mechanisms of elevation of serum Tg with Hashimoto's thyroiditis remain to be clarified, this unusual case should be studied further. Acknowledgment We thank Ms. M. Kouchi and Ms. H. Hiratani for their secretarial assistance. References Spencer CA (1996) Thyroglobulin. In: Braverman LE, Utiger RD (eds.) The Thyroid. Lippincott-Raven Publishers, Philadelphia, Seventh edition: Sato K, Okamura K, Ikenoue H, Shiroozu A, Yoshinari M, Fujishima M (1988) TSH dependent elevation of serum thyroglobulin in reversible primary hypothyroidism. Clin Endocrinol 29: Yamaguchi N, Kumamoto Y, Aoyama A, Natori Y, Kusakabe K, Idemura R (1995) Fundamental and clinical evaluation of Ab Bead Thyroglibulin `Eiken'. Japanese J Med Pharmaceut Sci 33: Dunn JT, Ray SC (1978) Variation in the structure of thyroglobulin from normal and goitrous human thyroids. J Clin Endocr Metab, 47: Edelhoch H (1965) The structure of thyroglobulin and its role in iodination. Recent Prog Horm Res 21: 1. Schneider AB, Pervos R (1978) Radioimmunoassay of human thyroglobulin: effect of antithyroglobulin autoantibodies. J Clin Endocr Metab 47: Kasagi K, Miyamoto S, Misaki T, Konishi J (1996) Evaluation of an IRMA kit for thyroglobulin. Japanese J Med Pharmaceut Sci 35: Medeiros-Neto G, Targovinik HM, Vassart G (1993) Defective thyroglobulin synthesis and secretion causing goiter and hypothyroidism. Endocr Rev 14: Sack J, Kaiserman I (1995) The importance of thyroglobulin levels in monitoring the treatment of congenital hypothyroidism. Horm Res 44: Kohn LD, Shimura H, Shimura Y, Hidaka A, Giuliani C, Napolitano G, Ohmori M, Laglia G, Saji M (1995) The thyrotropin receptor. Vitam Horm 50: Graves P, Neufeld DS, Davies TF (1989) Differential cytokine regulation of MHC class II and thyroglobulin mrnas in rat thyroid cells. Mol Endocrinol 3: Kambe F, Seo H (1996) Mediation of the hormoneand serum-dependent regulation of thyroglobulin gene expression by thyroid-transcription factors in rat thyroid FRTL-5 cells. J Endocrinol 150: Ohtaki S, Endo Y (1979) Study on enzyme immunoassays for the measurement of thyroglobulin and anti-thyroglobulin autoantibody in human serum. Folia Endocrinol Japon 55: Pinchera A, Mariotti S, Vitti P, Tosi M, Grasso L, Baschieri L, Pacini F, Buti R (1977) Interference of serum thyroglobulin in the radioassay for serum antithyroglobulin antibodies. J Clin Endocr Metab 45: