You are on page 1of 4

Endocrine Journal 2015, 62 (6), 557-560

Note

Serum human chorionic gonadotropin levels and thyroid
hormone levels in gestational transient thyrotoxicosis: Is the
serum hCG level useful for differentiating between active
Graves’ disease and GTT?
Ai Yoshihara, Jaeduk Yoshimura Noh, Koji Mukasa, Miho Suzuki, Hidemi Ohye, Masako Matsumoto,
Yo Kunii, Natsuko Watanabe, Nami Suzuki, Toshiaki Kameda, Kiminori Sugino and Koichi Ito
Ito Hospital, Tokyo 150-8308, Japan

Abstract. Gestational transient thyrotoxicosis (GTT) is defined as transient thyrotoxicosis caused by the stimulating effect
of human chorionic gonadotropin (hCG) during pregnancy. We attempted to identify the serum hCG level that causes GTT,
and we compared the serum hCG levels and thyroid hormone levels of GTT patients according to whether they had a
background of thyroid disease. We also evaluated serum hCG as a parameter for differentiating between active Graves’
disease (GD) and GTT. We reviewed the 135 cases of pregnant women who came to our hospital to be evaluated for
thyrotoxicosis during their 7th to 14th week of pregnancy, and their serum hCG level was measured at that time. Among
the 135 pregnant women with thyrotoxicosis; 103 of the women had GTT, and the other 32 women had active GD. There
were no correlations between their serum hCG levels and free thyroid hormone levels. There were no significant differences
in thyroid hormone levels or hCG levels among the GTT groups with different thyroid disease backgrounds; i.e., the GTT
group without thyroid disease, GTT group with chronic thyroiditis, GTT group with non-functioning thyroid nodules, and
GTT group with GD in remission. The serum hCG level of the GTT group was significantly higher than in the active GD
group, but it was not a good parameter for differentiating between the two groups. The FT3/FT4 ratio of the active GD was
significantly higher than in GTT group, and was a better parameter for differentiation.
Key words: Human chorionic gonadotropin, Gestational transient thyrotoxicosis, Graves’ disease

Gestational transient thyrotoxicosis (GTT) is
defined as transient thyrotoxicosis caused by the stimulating effect of human chorionic gonadotropin (hCG)
on the TSH receptor in the first trimester of gestation.
The prevalence of GTT has been reported to be 5.5%
in the Japanese population [1] and 11% in Singapore
[2]. GTT is characterized by an elevated serum thyroid hormone level and low or undetectable serum TSH
level in the absence of TSH receptor antibody. It is
very important to differentiate between GTT and active
Graves’ disease (GD) during pregnancy in GD patients
in remission, and having additional parameters besides
the serum TRAb titer as a means of making the differential diagnosis would be very helpful in clinical practice.
In the present study we investigated the hCG levels of
Submitted Dec. 9, 2014; Accepted Mar. 10, 2015 as EJ14-0596
Released online in J-STAGE as advance publication Mar. 29, 2015

Correspondence to: Ai Yoshihara, M.D., Ito Hospital, 4-3-6
Jingumae, Shibuya-ku, Tokyo 150-8308, Japan.
E-mail: a-yoshihara@ito-hospital.jp
©The Japan Endocrine Society

GTT patients and whether the serum hCG levels of the
GTT patients correlated with their thyroid hormone levels. We also compared the characteristics of GTT according to whether the patients had a background of thyroid
disease, and more specifically whether they had a background of inactive GD, chronic thyroiditis, or non-functioning thyroid nodules. We compared the serum hCG
levels and FT3/FT4 ratios of patients with active GD and
GTT patients to determine which was a better parameter
for differentiating between active GD and GTT.
Patients
We reviewed the cases of pregnant women who
came to our hospital during their 7th to 14th weeks of
pregnancy to be evaluated for thyrotoxicosis between
November 1, 2005 and December 31, 2012, and their
serum hCG level was measured at that time. Patients
who had been treated with drugs that affect thyroid
function, including antithyroid drugs and levothyroxine, at the time of the blood tests were excluded as

.0001). subjects of this study. Roche Diagnostics GmbH. ≤40 IU/mL). version 11. Cardiff. All participants gave informed consent. Japan).01-4. manufacturer’s reference limits: 0. UK. non-functioning thyroid nodules. but since October 1. chronic thyroiditis.000 mIU/mL) and 51.900–150. 1). There were no significant differences between the serum hCG levels among the GTT group according to thyroid disease background (no thyroid disease. and TSH at 4-12 weeks of gestation were 2. The TSH level was measured by an ECLIA (ECLusys TSH. TRAb-CT.500 mIU/mL (range 8. Roche Diagnostics GmbH. The 103 GTT patients consisted of 30 women with no underlying thyroid disease.0009). HCG was determined with an enzyme immunoassay kit (Fujirebio Inc. and the other 21 patients had a recurrence of GD.558 Yoshihara et al.0001).0. Germany. We diagnosed GTT when the results of the TRAb test remained negative throughout the hyperthyroid state and the hyperthyroid state spontaneously improved in the absence of treatment. There were no correlations between the serum hCG levels and free thyroid hormone levels (either FT3 or FT4) in either the GTT group or active GD group (Fig. Germany. Cary. Mannheim. GTT without thyroid disease was diagnosed when serum TRAb.2-4.05 was considered significant. Receiver operating characteristic (ROC) curve analysis was performed to assess the accuracy of serum hCG levels and FT3/FT4 ratios of patients for differentiating between active GD and GTT. and 0. Tokyo.35 mIU/L. GD in remission). Mannheim.0 IU/L). Results The 135 hyperthyroid women consisted of 103 women with GTT and 32 women with active GD. and 30 women with GD in remission. Roche Diagnostics GmbH. (SAS Institute Inc. TgAb. TgAb. the serum FT3 level and FT4 levels of the active . Based on the results of our previous study of a large population. reference values <10%). thyroglobulin antibody [TgAb]. manufacturer’s reference limits: 2. This study received approval by the local ethics committee. and TPOAb tests were all negative and thyroid ultrasonography showed normal findings. 2008.5 mIU/L).77-1. The serum hCG level of the active GD group was significantly lower than in the GTT group (p = 0. Thyroid ultrasonography and serum tests for the presence of thyroid-related antibodies (anti-TSH receptor antibody [TRAb]. Germany. Mannheim. reference values <2. reference values were: TPOAb. they have been determined with an electrochemiluminescence immunoassay kit (ECLusys TRAb. The data were analyzed by the Wilcoxon/ Kruskal-Wallis test for differences between two groups. 0. Until September 30. respectively. A diagnosis of non-functioning thyroid nodule was made when ultrasonography showed one or more thyroid nodules and the hyperthyroid state spontaneously improved to a normal thyroid state in the absence of treatment.0. Also.2-4. Tukey-Kramer’s test and ANOVA were used to compare the hCG levels and thyroid hormone levels among different groups. Correlations between two parameters were evaluated by calculating Spearman’s rank correlation coefficient. and the FT3/FT4 ratios were also significantly higher in the active GD group than in the GTT group (p < 0.000 mIU/mL (range 20.91 ng/dL.3 pg/mL. 0.000 mIU/mL) in the active GD group. Germany. None of the women had a hydatidiform mole or choriocarcinoma. The characteristics of the GTT group and active GD group are shown in Table 1. 2008. 11 patients had been newly diagnosed with GD.. Statistical analysis The statistical analysis was performed with JMP software. Mannheim. and underlying thyroid disorders were diagnosed. Chronic thyroiditis was diagnosed when the serum TgAb test and/or TPOAb test was positive. and anti-thyroid peroxidase antibody [TPOAb]) were performed.01-3. 19 women with chronic thyroiditis.000–220. TPOAb and TgAb were determined by an ECLIA performed with Roche ECLusys Anti-Tg and Anti-TPO (Roche Diagnostics GmbH. The median serum hCG of the GTT group was 71. The serum FT3 level and FT4 levels of the active GD group were significantly higher than in the GTT group (p < 0.8-1.6 ng/ dL respectively). Laboratory methods Free triiodothyronine (FT3) and free thyroxine (FT4) levels were measured by an electrochemiluminescence immunoassay (ECLIA) (ECLusys FT4. A p-value <0. and FT4. the reference intervals for FT3. Active GD was diagnosed when the serum TRAb test was positive and antithyroid drugs were required to suppress the thyroid hormone level.9 pg/mL. TRAb levels were determined with a solid-phase immunoradioreceptor assay kit (RSR. 24 women with nonfunctioning thyroid nodule. NC). Of the 32 active GD patients. ≤28 IU/mL.

19 (1.0) 2. the cut-off FT3/FT4 ratio for differentiating between active GD and GTT was 2.0001.91~7.7.000 IU/L lack overt hyperthyroid symptoms [5]. FT4.31~2.18) *p < 0.5~17.94~3. The area under the ROC curve was 0. FT4 0.21~4. In addition.90) 2.2) 2. and specificity was 88%.0) 2.28 (1. in comparison with the GTT group **p < 0. in comparison with the GTT with GD in remission group ††p < 0.21 (1. 1 Serum hCG levels and free thyroid hormone levels (FT3 or FT4) in the GTT group and active GD group. .96~7.7. a hydatidiform mole or choriocarcinoma.3~32. The ROC curve analysis revealed that the cut-off hCG level for differentiating between active GD and GTT was 70.26 (1. reported that hCG concentrations higher than 400.559 HCG level in gestational thyrotoxicosis Table 1 Characteristics of the patients in the GTT group and active GD group Total GTT 103 Without thyroid disease 30 With chronic thyroiditis 19 With non-functioning thyroid nodule 24 With GD in remission 30 Active GD 32 hCG (mIU/mL) FT3 (pg/mL) FT4 (ng/dL) FT3 / FT4 median (range) median (range) median (range) median (range) 71000 (20000~220000) 5. p = 0. and we compared the serum hCG levels and thyroid hormone levels of GTT in patients according to whether they had a background of thyroid disease.001.36 (1.76) 62000 (26000~220000) 5. and sensitivity was 84%.32 (1.77-1.9 pg/mL.91~7.000 IU/L suppresses TSH level and most patients with hCG higher than 200.42 (1.95~7.80 (3.0~13.5) 2.56 (1. Furthermore.0001).15 (3. 2] Lockwood et al.96) 68000 (20000~160000) 5.80) 2. The prevalence of GTT among pregnant woman in Europe has been estimated to be 2-3% [4].03**†† (2.34~3. There were no significant differences in thyroid hormone levels among the GTT groups (Table 1).31 (1.10) 80500 (33000~180000) 6.0001.65~2.01-4.00**†† (4.83**† (1.76) 73000 (35000~190000) 5.31~3.68) 2.1) 2.72) 51500*† (8900~150000) 9.0004).68) 2.30 (2.5~ 9.0001.9~11.30 (2.35 (3.28 (1.9~17. sensitivity was 77%.91 ng/dL Fig. p < 0.93~3. Discussion We attempted to identify the serum hCG level that causes GTT. and the FT3/FT4 ratios were also significantly higher in the active GD group than in the GTT with GD in remission group (p < 0. and a higher prevalence is estimated in Asia [1. in comparison with the GTT group †p < 0.77~3.77) 3. GD group were significantly higher than in the GTT with GD in remission group (FT3. The area under the ROC curve for differentiating between active GD and GTT was 0. in comparison with the GTT with GD in remission group Reference intervals at 4-12 weeks of gestation: FT3 2.2) 2. and specificity 51%.01) 2.87.000 mIU/mL. HCG shares a common α-subunit with TSH and possesses thyroid stimulating activity [3].01.

there was no correlation between the serum hCG levels and thyroid hormone levels of the GTT patients. there were no significant differences between the thyroid hormone levels or hCG levels of the GTT group without thyroid disease. et al. 2. Kashiwai T. we did not follow the hCG levels throughout the pregnancy of any of the subjects. However. and it is important to differentiate GTT from GD in order to prevent patients from undergoing unnecessary ATD treatment. Miyabe Y. GTT group with non-functioning thyroid nodules. We attempted to identify hCG and FT3/FT4 ratio cut-off levels for differentiating between active GD and GTT by performing an ROC curve anal- ysis. Amino N. Usui T. Eng PH.000 mIU/mL. Disclosure Statement The authors declare that they have no competing financial interests. Tan HK. We also evaluated serum hCG levels as a parameter for differentiating between active GD and GTT. The thyroid function of most untreated GTT patients becomes normal by the second trimester. Yeo CP. Kimura T. Furuhashi M. and the area under the ROC curve was 0. (2001) Prevalence of gestational thyrotoxicosis in Asian women evaluated in the 8th to 14th weeks of pregnancy: correlations with total and free beta human chorionic gonadotrophin. we concluded that the FT3/FT4 ratio is a better parameter than the serum hCG level for differentiating between active GD and GTT. The FT3/FT4 ratio was a better parameter than the hCG level for differentiating between GTT and active GD. and there were no correlations between the serum hCG levels and thyroid hormone levels. 8. including GTT with GD in remission. 3. Glinoer D (1997) The regulation of thyroid function in pregnancy: pathways of endocrine adaptation from physiology to pathology. emesis. et al. Hagiwara H. and fetal and child development. Burrow GN (1993) Thyroid function and hyperfunction during gestation. There is one report of a change in the thyroid stimulating activity of hCG as a result of a modification of the carbohydrate chain [8]. The exact molecular variants of hCG responsible for thyrotrophic effects are still unknown. since hCG was measured only once in every subject.7. Oda H. De Nayer P. Grenache DG. Our study had the limitations of being retrospective and the numbers of patients in each group being small. (2007) The incidence of gestational hyperthyroidism and postpartum thyroiditis in treated patients with Graves’ disease. Clin Endocrinol (Oxf) 46: 719-725.000 to 220. the diagnosis of GTT in each subject was confirmed by reviewing the subjects’ clinical course. Tagami et al. Thyroid 17: 767-772. 4. Tamaki H. Shimogaki K. Tada H. Gronowski AM (2009) Serum human chorionic gonadotropin concentrations greater than 400. Kubota S. and there were no significant differences between the thyroid hormone levels or hCG levels of the GTT groups. 7].87. Thyroid 19: 863-868. 7. 5. . J Clin Endocrinol Metab 94: 1683-1688. Endocr Rev 18: 404-433. Our results in the present study were consistent with the findings in that report. Orito Y. References 1. Yo SL. Endocr J 45: 585-590.7.The hCG levels in the GTT group in this study ranged from 20. and GTT group with GD in remission. Endocr Rev 14: 194-202. Katsumata Y. This result suggests that the characteristics of GTT do not vary with thyroid disease background. Clin Endocrinol (Oxf) 55: 391-398. Imai N (1998) Gonadotropin-releasing hormone-induced elevation of serum hCG in choriocarcinoma: a case report. (1995) Pathogenic role of asialo human chorionic gonadotropin in gestational thyrotoxicosis.560 Yoshihara et al. Tagami T. Oku H. et al. Also. Khoo DH. ated with suppressed serum thyrotropin concentrations. In conclusion. Lockwood CM. Shimatsu A. et al. Asahi K. Grun JP. (2009) Thyroid function in early pregnancy in Japanese healthy women: relation to urinary iodine excretion. evaluated the serum hCG values and FT3/FT4 ratios of pregnant GD patients and found that the hCG values were higher and FT3/FT4 ratios were lower in the GTT group than in the non-GTT group [9]. 9. and the weeks of gestation when the measurements were made differed from subject to subject. and twin pregnancies are also known to cause GTT [6. Meuris S. J Clin Endocrinol Metab 80: 350-355. GTT group with chronic thyroiditis. Tsuruta E. In addition. Since the FT3/FT4 ratio cut-off level for differentiating between active GD and GTT was 2. The hCG cut-off level for differentiating between active GD and GTT was 70000 mIU/mL and the area under ROC curve for differentiating active GD between GTT was 0. Glinoer D (1997) The thyrotrophic role of human chorionic gonadotrophin (hCG) in the early stages of twin (versus single) pregnancies.000 IU/L are invariably associ- 6.