Incidence of the postpartum diagnosis of thyroid eye disease in relation to thyroid function in Graves’ disease

Abstract

It has been reported that Graves’ disease (GD) sometimes improves spontaneously during pregnancy, although exacerbation of GD during postpartum period or relapse of hyperthyroidism caused by GD might occur. This study aimed to investigate the incidence of postpartum diagnosis of thyroid eye disease (TED) in relation to thyroid dysfunction. This retrospective cross-sectional study enrolled 11,104 deliveries from the patients with GD between January 2004 and August 2022. Within the 12-month postpartum period, 72 patients (0.65%) were diagnosed with TED. The thyroid function of the 72 patients comprised 9 remission, 13 continued antithyroid medicine, and 50 thyroid dysfunction; 30 newly diagnosed GD, 1 hypothyroidism, and 19 relapse/recurrence of GD. In the 49 patients with thyroid dysfunction, no difference was observed in the median values of thyroid-stimulating hormone (TSH) receptor antibody (TRAb) and TSH receptor stimulating antibody between the TED diagnosis and the development of hyperthyroidism. However, when the patients were classified into the newly developed GD and relapse/recurrence of GD groups, the difference became significant and the TRAb level was high in the newly developed GD (16.1 vs. 5.0 IU/L, p < 0.0001, and 15.0 vs. 6.0 IU/L, p = 0.0003). Thyroid dysfunction preceded TED diagnosis in more than half of the patients and the median time for each event was 6.5 vs. 8.1 months. The active phase TED was observed in 8 of the 72 patients. Of the 72 patients newly diagnosed with TED in postpartum, two-thirds were accompanied by thyroid dysfunction and 8 of them were in active phase.

Introduction

Pregnancy can influence thyroid function. Approximately 1–3% of pregnant women experience gestational transient thyrotoxicosis in the first half of the gestation period [1], and this change is caused by high levels of serum human chronic gonadotropin and an increase in thyroxine-binding globulin [2]. Gestation also affects the clinical course of Graves’ disease (GD), but spontaneous improvement could be expected because of immunosuppression, particularly the decrease in TSH receptor antibody (TRAb) titers. Up to 20–30% of patients with GD can cease antithyroid drug (ATD) during pregnancy [3]. However, relapse or new development of hyperthyroidism may occur after delivery [3], with the highest incidence occurring in the first 7–9 months postpartum [4], followed by a rebound increase in TRAb levels in mothers [5]. Conversely, the clinical course of thyroid eye disease (TED) during the gestation and postpartum periods remains poorly investigated [6-11] and do not always follow a benign course. Indeed, a few cases of severe TED during pregnancy that required aggressive treatments such as intravenous glucocorticoid administration and decompression surgery were reported [9-12]. The eye symptoms of TED can vary depending on the underlying thyroid autoimmunities during pregnancy [8]. In addition, TRAb titers can fluctuate and do not always correlate with the severity of the orbital reaction in the perinatal period [7]. The clinical course of TED may differ from that of thyroid function during this period. Therefore, this study aimed to investigate the incidence of new postpartum diagnoses of TED in relation to thyroid dysfunction among women with GD.

Materials and methods

Participants

This was a cross-sectional analysis study of the retrospectively reviewed data of patients managed at the Ito Hospital, which specializes in thyroid diseases in Tokyo, Japan. Between January 2004 and August 2022, 39,762 patients with various thyroid diseases gave birth. A total of 12,322 children were born to 8,594 individuals with GD or those diagnosed with GD within the past 12 months. We excluded patients who gave birth more than 1 year before the initial visit, mothers of a child who did not show up in the perinatal period, and mothers of a child whose postpartum follow-up duration was less than 12 months. Ultimately, we analyzed the clinical course of thyroid function along with the eye symptoms of the 8,092 individuals of 11,104 deliveries (Fig. 1). Of the 8,092 patients, 111 were suspected of having developed TED and referred to a hospital specializing in TED. Of the 111 patients, 23 were excluded because of not visiting the hospital and 10 were excluded because of not being diagnosed with TED. In addition, three patients who did not undergo magnetic resonance imaging (MRI) test to exclude other causes of eye symptoms such as ocular tumor and inflammation were excluded. Finally, 72 patients (0.65%) were diagnosed with TED during the postpartum period. None of the patients had been diagnosed with TED based on the inspection of expert endocrinologists before and during pregnancy.

Fig. 1  Patient flowchart

This figure shows the number of potential participants, the inclusion and exclusion criteria, and the final analytical sample. In this study, 11,104 deliveries of 8,092 patients were the analytical sample, and 72 patients had thyroid eye disease (TED).

GD, Graves’ disease

GD diagnosis was based on thyrotoxicosis with TRAb or TSH receptor-stimulating antibody (TSAb) positivity or diffuse high uptake of radioactive iodine (¹²³I) by the thyroid in mothers who did not breastfeed. GD remission was defined as the maintenance of the euthyroid state for more than 1 year after the withdrawal of ATD. The development of hyperthyroidism after achieving remission indicated GD relapse, whereas that within a year after ATD withdrawal indicated GD recurrence.

The Ethics Committee of Ito Hospital approved this study (#IRB 390). Following an extensive dialog with the Ethics Committee of Ito Hospital, an opt-out consent method was approved.

Ophthalmological assessment and imaging

Patients suspected of having developed TED within 12 months after delivery were referred to the Olympia Eye Hospital, which specializes in TED in Tokyo, Japan. TED was diagnosed when the patients with autoimmune thyroid diseases have (1) distinctive eye symptoms or (2) MRI showing proptosis or enlargement or inflammation of orbital muscles, particularly extraocular muscles (EOM) or retrobulbar tissues. In addition, the imaging findings enabled us to exclude other causes that induce eye symptoms, such as ocular tumor and inflammation.

Ophthalmological examination categories are equivalent to the NOSPECS classification, comprising redness or swelling of the eyelid, eyelid fissure enlargement, proptosis, optic nerve dysfunction, corneal and conjunctival involvement, inflammation of the lacrimal caruncle and/or plica, and chemosis. With the wide accessibility of MRI in Japan, the Japanese guidelines for TED management edited by the Japan Thyroid Association and Japan Endocrine Society recommend the use of MRI for diagnosing and evaluating TED activity and severity (Supplementary Fig. 1) [13]. This is because the clinical activity score (CAS) is a subjective evaluation and is because of the racial differences in clinical manifestations [14, 15]. Thus, it may not adequately reflect the activity. Japanese have been reported to have this tendency, so even with CAS <3, active TED was recognized by MRI [14]. The MRI findings are objective and sensitive for evaluating TED more accurately than CAS; particularly, it excels in enhancing inflammation in orbital tissues [16]. Thus, TED was evaluated based on the CAS and MRI findings.

TED was evaluated based on the severity aspects using a modified NOSPECS classification and activity using CAS and MRI findings. In the modified NOSPECS classification, proptosis was diagnosed when patients have proptosis ≥15 mm. The other criteria for categorical items were the same as the original one. The severity of TED was classified as mild, moderate-to-severe, and severest. The activity was classified into active and inactive phases. The active phase is characterized by inflammatory changes [17, 18], and TED progresses and deteriorates in this phase. While, the inactive phase is characterized when the inflammation completely calms down. Active TED was defined as CAS >3, or MRI shows inflammation in EOM or retrobulbar tissues.

As an interpretation protocol for MRI [13], MRI T1-weighted images are used to measure proptosis, retro-orbital volume, enlargement of EOM, and positional relationship between EOM and optic nerve, whereas, MRI T2-weighted images and short TI inversion recovery (STIR) images are used to evaluate the TED activity, as well as to measure the signal intensity of enlarged EOM using STIR and to compare the signal intensity of EOM with that of cerebral white matter. If the ratio of EOM/cerebral white matter is >1, it means active.

Ophthalmological treatment was determined based on TED severity and activity (Supplementary Fig. 1) [13]. Combined therapy involving intravenous glucocorticoid, radiation, and decompression surgery is performed for the severest cases of TED. Intravenous glucocorticoid and/or radiation are treatment options for moderate-to-severe cases of active TED, whereas ophthalmologic functional recovery surgery can be performed for moderate-to-severe cases of inactive TED. Local injection of glucocorticoid may be indicated for mild cases of active TED, whereas observation and functional recovery surgery could be an option for those with inactive TED, if required.

Laboratory measurement

TSH, serum-free triiodothyronine (FT3), and serum-free thyroxine (FT4) were measured using electrochemiluminescence immunoassay kits during the whole observation period (ECLusys TSH (AB_2756377); reference range: 0.20–4.50 μU/mL, ECLusys FT3 (AB_2894892); reference range: 2.2–4.3 pg/mL, and ECLusys FT4 (AB_2893401); reference range: 0.8–1.6 ng/dL, respectively; Roche Diagnostics, Basel, Switzerland). TRAb was measured using two different testing kits depending on the time frame. A TRAb CT radioimmunoassay kit (AB_3095551) (Cosmic, Tokyo, Japan; cutoff value, <10%) was used between January 2004 and September 2008, and an ECLusys TRAb electrochemiluminescence immunoassay kit (AB_2801453) (Roche Diagnostics; cutoff value, <2.0 IU/L) was used from October 2008 onward. Given that the measurement method was different, the two values were compared using the following formula derived from the regression curve representing the relationship between TRAb CT and ECLusys TRAb: ECLusys TRAb = e^{(0.053 × TRAb CT – 0.533)} [19]. A TSAb radioimmunoassay and bioassay (RIA) kit (AB_3096097) (Yamasa, Choshi, Japan; cutoff value: <180%) was used to measure TSAb until June 2014, and a TSAb enzyme-immunoassay and bioassay (EIA) kit (AB_2784529) (Yamasa; cutoff value, <120%) was used from July 2014. Additionally, the TSAb value was measured using two different methods. Thus, the following formula derived from the regression analysis evaluating the relationship between TSAb RIA and TSAb EIA kit compared the two values: TSAb EIA = 0.988 × TSAb RIA + 142.7 [20].

Statistical analyses

Data are expressed as medians and ranges or interquartile ranges (IQRs) for continuous variables. We used the Wilcoxon rank-sum test to compare parameters between groups categorized by thyroid condition. Statistical data were analyzed using JMP version 14.0 software (SAS Institute, Cary, NC). A p-value of <0.05 was considered statistically significant.

Results

Incidence and characteristics of the patients with postpartum TED diagnosis

Of the 8,092 patients (11,104 deliveries) with GD, 7,750 (10,762 deliveries) were diagnosed with GD before getting pregnant and 342 (342 deliveries) were newly diagnosed with GD after giving birth. A total of 72 cases (0.65%) were diagnosed with TED after delivery. Of the 72 TED cases, 42 were among the 10,762 deliveries with GD (0.39%) and 30 were among the 342 postpartum newly developed patients with GD (8.8%) (Fig. 2). Among the patients who gave birth more than once, no cases of worsened TED in the perinatal period of another pregnancy were reported. Excluding the 342 patients newly developed GD after delivery, the median disease duration of GD until delivery was 14.1 (IQRs: 7.8–24.9) months in the patients postpartum diagnosed with TED. Compared with the disease duration in 7,708 patients without TED diagnosis (Fig. 2), the disease duration was significantly shorter in the patients with TED (p < 0.0001), suggested that TED tends to be diagnosed on early stage of GD development. The median ages at giving birth of all patients, patients with GD during pregnancy, and patients with newly developed GD were 32 (IQRs: 28–36), 33 (IQRs: 30–37), and 32 (IQRs: 29–33) years, respectively. In the 72 patients with TED, the median time from delivery to TED diagnosis was 7.3 (IQRs: 4.0–9.1) months, and the median values of TRAb and TSAb during TED diagnosis were 9.0 (IQRs: 4.6–32.7) IU/L and 1,410 (IQRs: 405.2–2,355.3) %.

Fig. 2  Characteristics of the patients divided by thyroid and eye condition

This figure shows the flowchart of the total participants divided by thyroid condition and eye condition. Herein, 10,762 pregnancies went through with the condition of GD and 342 patients developed GD after the deliveries. In total, 42 of 10,762 deliveries and 30 of 342 newly developed GD diagnosed TED in postpartum. The characteristics of the 72 patients diagnosed with TED were noted.

Fig. 3 shows the thyroid condition of the TED patients at 12 months postpartum and at TED assessment. Of the 72 patients, 9 showed no GD relapse and remained in a euthyroid state, 13 continued taking ATD or potassium iodide (KI), and 50 developed thyroid dysfunction in postpartum duration. Of the 50 patients, 30 patients newly developed GD in postpartum and the remaining 20 comprised 16 GD recurrences, 3 GD relapses, and 1 hypothyroidism. The time to diagnose TED was significantly longer in the thyroid dysfunction group than in the no-relapse group (8.1 vs. 4.0 months, p = 0.02) and in the ATD continued group (8.1 vs. 4.1 months, p = 0.0002). The number of patients with active and inactive TED in each group were shown in Fig. 3.

Fig. 3  Thyroid condition of the 72 patients at 12 months postpartum and TED assessment

This figure shows the thyroid condition of the 72 patients at 12 months postpartum on the left side and as the assessment of TED on the right side. In total, 9 remained in an euthyroid state, 13 continued medication, and 50 developed thyroid dysfunction. Of the 50 patients with thyroid dysfunction, 30 had newly developed GD, 16 experienced GD recurrence, 3 relapsed GD, and 1 developed hypothyroidism. At the assessment of TED, the patients were categorized based on their background condition and the treatment they required. In addition, they were categorized based on TED activity.

Thyroid-related parameters and the time to diagnosis of TED and hyperthyroidism in 49 patients with postpartum hyperthyroidism

The median FT3, FT4, TRAb, and TSAb values and time to events, i.e., hyperthyroidism development and TED diagnosis, were compared between the onset of hyperthyroidism and TED diagnoses in 49 patients (Table 1). One case of hypothyroidism was excluded from the analysis. The median time to the onset of thyroid dysfunction and TED were 6.3 (IQRs: 3.4–8.9) and 8.0 (IQRs: 5.8–10.1) months, respectively, indicating that thyroid dysfunction tended to occur before the diagnosis of TED. In fact, 25 of 49 patients (51.0%) in the thyroid dysfunction group developed thyroid dysfunction before TED diagnosis, and the remaining 24 (49.0%) developed thyroid dysfunction at the same time of TED diagnosis. Considering that thyroid dysfunction developed before TED diagnosis in more than half of the patients, treatment for hyperthyroidism had already started when TED diagnosis was made. Therefore, FT3 and FT4 values were significantly higher at the development of hyperthyroidism than those at the diagnosis of TED (p = 0.005 and 0.002, respectively). However, no differences were observed in the levels of TRAb and TSAb between the onset of thyroid dysfunction and the diagnosis of TED (TRAb 9.2 IU/L vs. 9.1 IU/L, p = 0.82 and TSAb 1,080.3% vs. 1,297%, p = 0.79, respectively).

Table 1 Comparison of GD related parameters measured at hyperthyroidism development and TED diagnosis between patients newly developed GD in postpartum and GD relapse or recurrence. (n = 49)

Event at measurement	Number	FT3 (pg/mL)	FT4 (ng/dL)	TRAb (IU/L)	TSAb (%)	Required time (months)
		Median (IQRs)	Median (IQRs)	Median (IQRs)	Median (IQRs)	Median (IQRs)
Hyperthyroidism development
Total patients	49	16.4 (7.5–23.5)	4.7 (2.5–7.6)	9.2 (5.2–30.1)	1,080.3 (402.9–1,844.6)	6.3 (3.4–8.9)
New GD	30	20.7 (8.6–27.1)	5.8 (2.4–7.8)	16.1 (7.7–36.3)	1,237.8 (737.7–1,844.6)	7.4 (4.6–9.3)
Relapse/recurrence	19	12.7 (7.1–17.0)	3.9 (2.6–5.9)	5.0 (3.3–6.6)	391.2 (274.7–2,928.8)	4.5 (2.7–6.9)
p value		0.08	0.60	<0.0001	0.56	0.05
TED diagnosis
Total patients	49	9.7 (3.4–21.2)	2.2 (1.2–5.9)	9.1 (5.8–30.9)	1,297.0 (379.2–1,926.1)	8.0 (5.8–10.1)
New GD	30	15.0 (5.1–23.8)	3.4 (1.2–7.6)	15.0 (7.2–33.3)	1,297.0 (914–1,819.3)	8.6 (6.0–10.1)
Relapse/recurrence	19	4.3 (3.0–12.7)	1.7 (1.3–3.0)	6.0 (3.9–9.6)	483.0 (316.6–2,796.1)	7.5 (5.7–10.3)
p value		0.013	0.07	0.0003	0.13	0.61

Normal ranges: FT3 (2.2–4.3 pg/mL), FT4 (0.8–1.6 ng/dL)

Cut-off values: TRAb <2.0 IU/L, TSAb <120%

GD: Graves’ disease, IQRs: Interquartile ranges

Comparison of thyroid-related parameters between the newly developed GD and the relapse/recurrence of GD group

The median values of FT3, FT4, TRAb, and TSAb and the time of events were compared between the newly developed GD group (n = 30) and the GD relapse/recurrence group (n = 19) during postpartum (Table 1). At the onset of hyperthyroidism, the median level of TRAb was significantly higher in the newly developed GD group than that in the GD relapse/recurrence group (16.1 vs. 5.0 IU/L, p < 0.0001). However, the median values of FT3, FT4, and TSAb did not differ between the groups. Moreover, the median time to develop hyperthyroidism was longer in the newly developed GD group than in the GD relapse/recurrence group, but the difference was not significant (7.4 vs. 4.5 months) (p = 0.05).

During TED diagnosis, the median values of FT3 were significantly higher in the newly developed GD group than in the GD relapse/recurrence group (p = 0.013). The median value of FT4 was also higher in the newly developed GD group, but the difference was not significant (p = 0.07). Moreover, the median levels of TRAb was significantly higher in the newly developed GD group during TED diagnosis (15.0 vs. 6.0 IU/L, p = 0.0003), although no differences were observed in the values of TSAb between the two groups. In addition, the median time to TED diagnosis showed no difference between the groups (8.6 vs. 7.5 months) (p = 0.71).

Detailed findings and treatments for TED

Among the 72 patients, 38 were classified as mild TED, and 34 were classified as moderate-to-severe TED. No case was classified into the severest TED. Of the 72 patients, 8 (0.072%) were in the active phase and 64 (0.58%) were in the inactive phase (Table 2). Active TED comprised four mild and four moderate-to-severe cases, and inactive TED comprised 34 mild and 30 moderate-to-severe cases. Although the median CAS value in 72 patients was 1, it ranged from 0 to 3. When we analyzed CAS by TED activity, 62.5% of the patients with active TED and 42.9% with inactive TED were categorized in CAS of 1. In addition, the median CAS values in each group were 1 (IQRs: 1–2.5) in the active TED and 0 (IQRs: 0–1) in the inactive TED. For those who had inactive, moderate-to-severe TED (Table 2), the median CAS value was 1 and the most common findings were proptosis >18 mm (n = 19), followed by lid retraction >12 mm (n = 10) and soft tissue involvement in one case.

Table 2 Number of patients and CAS classified by activity and severity

	Activity
	Active (n = 8)	Inactive (n = 64)
Severity
Mild (n = 38)	4 CAS 1 (IQR: 0.25–1)	34 CAS 0 (IQR: 0–1)
Moderate to severe (n = 34)	4 CAS 2 (IQR: 1–3)	30 CAS 1 (IQR: 0–1)

IQR: interquartile range, CAS: clinical activity score.

Patients in the active phase required treatment for TED; their details are summarized in Table 3. Of the eight patients, two patients had CAS ≥3. Although the remaining 6 patients had CAS <3, inflammatory findings were seen with MRI on EOM, levator palpebrae superioris muscle, and caruncle, and, they were classified as an active phase based on these findings. Treatments provided for active TED were five orbital radiation therapy for EOM and three subcutaneous triamcinolone acetonide injections into the levator palpebrae superioris muscle. In addition, from the aspect of thyroid condition, four of the eight patients were among those with newly developed GD, one developed hypothyroidism in the postpartum period, one was in the GD recurrence group, and the remaining two required ATD at delivery. No patients with active TED remained in the remission during the postpartum period.

Table 3 Detailed parameters and findings of the eight patients with active TED at diagnosis

No.	Age	Thyroid condition		TRAb (IU/L)	TSAb (%)	Time to develop TED (M)	Treatment for TED	Examination findings (R/L, mm)	Inflammatory findings on MRI	CAS	Severity
		During pregnancy	At postpartum 12M
1	27	Remission	Hypothyroidism in PP 3M	1.6	—	9.4	Radiation	LA 8/8 Proptosis 21/20 ELS	EOM	1	moderate to severe
2	24	Withdrew MMI at 5 weeks pregnant	Recurrence	40.0	2,400	7.6	Local GC injection	LA 9/9 Proptosis 14/15 Corneal infiltration, ELS	LPS and caruncle	1	mild
3	32	—	Newly developed GD	30.1	—	7.6	Radiation	LA 9/8 Proptosis 18/18 Diplopia, ELS	EOM and LPS	1	mild
4	35	—	Newly developed GD	21.9	3,110	8.0	Radiation	LA 9/9 Proptosis 23/23 Diplopia, ELS	EOM and LPS	1	moderate to severe
5	32	Continued PTU at delivery	Continued PTU	40.0	2,017	1.2	Radiation	LA 10/8, Proptosis 22/20 Diplopia, ELS	EOM and LPS	3	moderate to severe
6	32	Continued KI at delivery	Switched from KI to MMI after birth	5.9	2,865	2.7	Radiation	LA 10/10, Proptosis 17/17 ELS	EOM	3	moderate to severe
7	31	—	Newly developed GD	7.2	1,600	8.8	Local GC injection	LA 9/9 Proptosis 14/14	LPS	0	mild
8	35	—	Newly developed GD	4.4	1,441	9.0	Local GC injection	LA 9/9 Proptosis 15/15 ELS	LPS	1	mild

LA: Lid aperture, ELS: Eyelid swelling, LPS: levator palpebrae superioris muscle, EOM: extraocular muscles, GD: Graves’ disease, GC: glucocorticoid, PP: postpartum, MMI: Thiamazole, PTU: propylthiouracil, KI: potassium iodide.

Normal ranges: FT3 (2.2–4.3 pg/mL), FT4 (0.8–1.6 ng/dL), TRAb (<2.0 IU/L), Normal range of TSAb was <180% until June 2014, and <120% from July 2014.

The Graphical Abstract summarized the study. The incidence, activity phase, treatments for TED, and the thyroid function of TED in postpartum were depicted.

Graphical Abstract 

Discussion

This study showed that in the postpartum period, 0.65% of patients with GD were diagnosed with TED. The incidence among those who already had GD before pregnancy was 0.39%, and among those who newly developed GD after giving birth was 8.8%. Regarding TED activity, the incidence of active phase TED was 0.072% in total patients, half of these patients with active TED patients belonged to the newly developed GD group. Of the 72 patients with postpartum TED, 70% (50/72) also developed thyroid dysfunction and 11% (8/72) had active TED. The incidence of newly diagnosed TED in the postpartum period was rare because the previous report using similar evaluation methods showed that the incidence of TED in patients with newly diagnosed untreated GD was 63% [21]; however, active phase TED could occur in the postpartum period, particularly for patients with newly developed hyperthyroidism.

During pregnancy, an immunosuppressive microenvironment exists to protect the fetus. This microenvironment comprised several cellular immunities, such as natural killer cells, macrophages, effector T cells, regulatory T cells, and specific human leukocyte antigens, interacting with leukocyte Ig-like receptor subfamily B. Of these factors, the regulatory T cells play an important role together with homeostatic cells to represent tightly controlled microenvironment [22]. In addition to a decrease in the TRAb value caused by immune change, significant hormonal changes are the underlying mechanism of the spontaneous improvement of GD in pregnancy, particularly an increase in the levels of estrogen, progesterone, and glucocorticoid. By contrast, a rebound immune activation starts after the termination of pregnancy, comprising two phases [23]. The activation of cellular immunity comes first, from 1 to 4 months postpartum, owing to an increase in cytotoxic T and natural killer cells, which trigger tissue injury. Next, humoral immunity is enhanced, from 7 to 10 months postpartum because of an increase in the number of cells producing autoantibodies. These immunological changes seem an activation of pre-existing autoimmunity involved in GD and TED. Cellular and humoral immunity enhancement induces postpartum thyroid dysfunction and TRAb level increase. Consequently, the postpartum exacerbation of hyperthyroidism caused by GD or a new onset of GD postpartum occurs.

Some of our study results are consistent with the course of humoral immunity change because the median time to TED diagnosis and thyroid dysfunction development in 49 patients with hyperthyroidism was 8.0 and 6.3 months, respectively, and thyroid dysfunction preceded in more than half of them. However, when we compared the time to TED diagnosis among the group with thyroid dysfunction, ATD continued, no relapse occurred, and the latter two groups had considerably shorter time than thyroid dysfunction group. This might be because of the individual difference in the reactivity to immune change or direct orbital tissue injury owing to cellular immunity.

As for TED, Eckstein et al. observed that the TRAb level was considerably higher in severe TED cases than in mild one and was an independent risk factor for triggering and maintaining the TED process [24]. Similarly, TRAb was an independent risk factor for patients with active phase TED who required anti-inflammatory and immunosuppressive treatments [25]. In addition, a review report showed that the TRAb level correlated with CAS [26]. Herein, the level of TRAb was significantly higher in patients who newly developed GD than in those who had relapse/recurrence of GD at the time of hyperthyroidism onset and TED diagnosis. This presents that patients who already had GD were susceptible to stimulation by TRAb and that hyperthyroidism and TED were induced with lower titer of TRAb in these patients.

However, TRAb is not the only risk factor that related to TED. According to the American Thyroid Association and European Thyroid Association guidelines for the Management of TED, the risk factors for the development and severity of TED are older age and male sex as nonmodifiable factors, and smoking (active and second-hand), thyroid dysfunction, and RAI for GD as modifiable factors [17, 27]. Regarding the relationship between TED and thyroid dysfunction, Prummel et al. reported that patients with more severe TED had a considerably higher rate of thyroid dysfunction than those with milder TED and confirmed the relationship between abnormal thyroid dysfunction and more severe eye disease [28]. In addition, the American Thyroid Association guidelines for the diagnosis and management of hyperthyroidism stated that euthyroidism should be expeditiously achieved and maintained in patients with TED or risk factors for TED development. Thyroid dysfunction referred to not only hyperthyroidism but also hypothyroidism because thyroid hormone replacement therapy is recommended for patients with hypothyroidism to prevent TED deterioration [2]. Considering that 50 of the 72 patients with TED also had thyroid dysfunction, i.e., 1 case of hypothyroidism and 49 cases of hyperthyroidism, and 6 of 8 patients with active TED had thyroid dysfunction, thyroid dysfunction is an important factor related to TED development. Still, TRAb elevation and oxidative stress are also regarded as potential risk factors [17, 27]. However, during pregnancy, TRAb level does not always correlate with TED severity, and TED development in patients with thyroid autoimmunity may depend on other factors [8]. Moreover, the TRAb and TSAb values at TED onset widely ranged, as shown in Fig. 2, and some patients showed negative results for these antibodies. Considering that 70% of the 72 patients with TED also had thyroid dysfunction and the results for TRAb or TSAb were negative at TED diagnosis in this study, thyroid dysfunction, and TRAb might be the factors responsible for TED development in the postpartum period.

This study has a few limitations. First, it is a retrospective study; hence, some data are lacking. Second, KI was used as a treatment for hyperthyroidism in 6.7% of the participants. KI is not commonly administered worldwide, and it is recommended for limited situations such as thyroid storm and preoperative preparation [2]. However, treatment with KI for patients with GD is widely accepted in Japan [29, 30], and combined therapy with ATD is recommended by GD treatment guidelines published by the Japan Thyroid Association. Our results may not be applicable worldwide because of the difference in iodine nutrition. Finally, the eye condition of our study participants was not evaluated before and during pregnancy by the ophthalmologists specializing in TED. However, visual examination was performed daily by expert thyroidologists in our clinical practice and none of the participants showed any signs or symptoms suggesting TED during the term. It suggests that the onset of TED may be at least after delivery for the patients with active TED and the diagnoses of TED were newly made for the first time after delivery. Considering that TED could develop in any point of the life time in patients with GD even during pregnancy, TED has to be evaluated and monitored before and during pregnancy and also after delivery. Further analysis is required to investigate the clinical course of TED in the perinatal period.

Conclusion

The incidence of TED postpartum diagnosis was 0.65% (72/11,104), and that of active phase TED was 0.072% (8/11,104). Of the 72 newly diagnosed patients with TED in the postpartum period, two-thirds were accompanied by thyroid dysfunction and 8 of them were in active phase.

Author disclosure statement

All authors declared no conflicts of interest in association with this manuscript.

Kiminori Sugino and Natsuko Watanabe are members of Endocrine Journal’s Editorial Board.

Funding statement

No funding was received for this study.

Acknowledgement

We would like to thank all physicians and ophthalmologists who contributed to this study.

Authors’ contribution

Conceptualization: N.S.

Data curation: N.S., A.K., T.I., M.F., M.M., S.H., H.I., M.I., M.K., A.S., and R.H.

Formal analysis: N.S.

Project administration: J.Y.N., A.Y., and N.W.

Supervision: A.Y., K.S., and K.I.

Writing-original draft: N.S.

Supplementary Fig. 1  Flowchart for TED management

This flowchart shows the recommendations for TED management in Japan. Regardless of the severity of TED, MRI is recommended for all patients with TED for evaluating activity.

References

• 1   Alexander  EK,  Pearce  EN,  Brent  GA,  Brown  RS,  Chen  H, et al. (2017) 2017 Guidelines of the American Thyroid Association for the Diagnosis and Management of Thyroid Disease During Pregnancy and the Postpartum. Thyroid 27: 315–389.
• 2   Ross  DS,  Burch  HB,  Cooper  DS,  Greenlee  MC,  Laurberg  P, et al. (2016) 2016 American Thyroid Association Guidelines for Diagnosis and Management of Hyperthyroidism and Other Causes of Thyrotoxicosis. Thyroid 26: 1343–1421.
• 3   Peng  CC,  Pearce  EN (2022) An update on thyroid disorders in the postpartum period. J Endocrinol Invest 45: 1497–1506.
• 4   Kahaly  GJ,  Bartalena  L,  Hegedus  L,  Leenhardt  L,  Poppe  K, et al. (2018) 2018 European Thyroid Association Guideline for the Management of Graves’ Hyperthyroidism. Eur Thyroid J 7: 167–186.
• 5   Bucci  I,  Giuliani  C,  Napolitano  G (2017) Thyroid-stimulating hormone receptor antibodies in pregnancy: clinical relevance. Front Endocrinol (Lausanne) 8: 137.
• 6   Ford  HF,  Spies  A (2001) Postpartum proptosis with ophthalmopathy. Optom Vis Sci 78: 75–78.
• 7   Gilbert  AL,  Prasad  S,  Mallery  RM (2019) Neuro-ophthalmic disorders in pregnancy. Neurol Clin 37: 85–102.
• 8   Wall  JR,  Lahooti  H,  Hibbert  EJ,  Champion  B (2015) Relationship between clinical and immunological features of thyroid autoimmunity and ophthalmopathy during pregnancy. J Thyroid Res 2015: 698470.
• 9   Stafford  IP,  Dildy  GA 3rd,  Miller  JM Jr (2005) Severe Graves’ ophthalmopathy in pregnancy. Obstet Gynecol 105: 1221–1223.
• 10   Aranyosi  JK,  Deli  T,  Erdei  A,  Toth  G,  Jakab  A, et al. (2020) Unusual onset of thyroid associated orbitopathy during pregnancy: case report and review of literature. BMC Endocr Disord 20: 183.
• 11   Sakurai  K,  Inoue  T,  Niitsuma  S,  Sato  R,  Takahashi  K, et al. (2020) Sight-threatening Graves’ ophthalmopathy during the third trimester of pregnancy: a case report. Tohoku J Exp Med 252: 321–327.
• 12   Koshibu  M,  Watanabe  N,  Suzuki  N,  Takahashi  Y,  Onda  H, et al. (2024) Exacerbation of thyroid eye disease and dysthyroid optic neuropathy during pregnancy: a case report. Intern Med doi: 10.2169/internalmedicine.3872-24. Online ahead of print.
• 13  The Japan Thyroid Association, The Japan Endocrine Society (2023) Management of thyroid eye disease; a statement by the Japan Thyroid Association and the Japan Endocrine Society 2023 (the 3rd edition):19. https://www.j-endo.jp/uploads/files/edu/koujyousengansyo_3.pdf accessed on September 16, 2024 (in Japanese).
• 14   Tachibana  S,  Murakami  T,  Noguchi  H,  Noguchi  Y,  Nakashima  A, et al. (2010) Orbital magnetic resonance imaging combined with clinical activity score can improve the sensitivity of detection of disease activity and prediction of response to immunosuppressive therapy for Graves’ ophthalmopathy. Endocr J 57: 853–861.
• 15   Chng  CL,  Seah  LL,  Khoo  DH (2012) Ethnic differences in the clinical presentation of Graves’ ophthalmopathy. Best Pract Res Clin Endocrinol Metab 26: 249–258.
• 16   Zhang  H,  Fan  J,  Qu  J,  Han  Q,  Zhou  H, et al. (2023) Predictive markers for anti-inflammatory treatment response in thyroid eye disease. Front Endocrinol (Lausanne) 14: 1292519.
• 17   Burch  HB,  Perros  P,  Bednarczuk  T,  Cooper  DS,  Dolman  PJ, et al. (2022) Management of thyroid eye disease: a consensus statement by the american thyroid association and the european thyroid association. Thyroid 32: 1439–1470.
• 18   Rundle  FF (1957) Management of exophthalmos and related ocular changes in Graves’ disease. Metabolism 6: 36–48.
• 19   Suzuki  N,  Noh  JY,  Yoshimura  R,  Mikura  K,  Kinoshita  A, et al. (2021) Does age or sex relate to severity or treatment prognosis in Graves’ disease? Thyroid 31: 1409–1415.
• 20   Kamijo  K,  Maruyama  H,  Togashi  K (2014) Fundamental and clinical studies on modified bioassay for TSAb using porcine thyroid cells. Japanese Journal of Medicine and Pharmaceutical Science 71: 903–911 (in Japanese).
• 21   Noh  JY,  Hamada  N,  Inoue  Y,  Abe  Y,  Ito  K, et al. (2000) Thyroid-stimulating antibody is related to Graves’ ophthalmopathy, but thyrotropin-binding inhibitor immunoglobulin is related to hyperthyroidism in patients with Graves’ disease. Thyroid 10: 809–813.
• 22   Miller  D,  Gershater  M,  Slutsky  R,  Romero  R,  Gomez-Lopez  N (2020) Maternal and fetal T cells in term pregnancy and preterm labor. Cell Mol Immunol 17: 693–704.
• 23   Amino  N,  Arata  N (2020) Thyroid dysfunction following pregnancy and implications for breastfeeding. Best Pract Res Clin Endocrinol Metab 34: 101438.
• 24   Eckstein  AK,  Plicht  M,  Lax  H,  Neuhauser  M,  Mann  K, et al. (2006) Thyrotropin receptor autoantibodies are independent risk factors for Graves’ ophthalmopathy and help to predict severity and outcome of the disease. J Clin Endocrinol Metab 91: 3464–3470.
• 25   Watanabe  N,  Yoshimura Noh  J,  Kozaki  A,  Yoshimura  R,  Yoshihara  A, et al. (2023) Incidence and risk factors for Graves’ orbitopathy in patients who underwent anti-inflammatory and immunosuppressive treatment during medical treatment for Graves’ disease: investigation of 1,553 cases with newly diagnosed Graves’ disease and proposal of a predictive score. Endocr J 70: 1087–1096.
• 26   Bartalena  L,  Gallo  D,  Tanda  ML,  Kahaly  GJ (2023) Thyroid eye disease: epidemiology, natural history, and risk factors. Ophthalmic Plast Reconstr Surg 39: S2–S8.
• 27   Burch  HB,  Perros  P,  Bednarczuk  T,  Cooper  DS,  Dolman  PJ, et al. (2022) Management of thyroid eye disease: a consensus statement by the American Thyroid Association and the European Thyroid Association. Eur Thyroid J 11: e220189.
• 28   Prummel  MF,  Wiersinga  WM,  Mourits  MP,  Koornneef  L,  Berghout  A, et al. (1990) Effect of abnormal thyroid function on the severity of Graves’ ophthalmopathy. Arch Intern Med 150: 1098–1101.
• 29   Yoshihara  A,  Noh  JY,  Watanabe  N,  Kunii  Y,  Suzuki  M, et al. (2020) Characteristics of patients with Graves’ disease whose thyroid hormone levels increase after substituting potassium iodide for methimazole in the first trimester of pregnancy. Thyroid 30: 451–456.
• 30   Sato  S,  Noh  JY,  Sato  S,  Suzuki  M,  Yasuda  S, et al. (2015) Comparison of efficacy and adverse effects between methimazole 15 mg + inorganic iodine 38 mg/day and methimazole 30 mg/day as initial therapy for Graves’ disease patients with moderate to severe hyperthyroidism. Thyroid 25: 43–50.