**3. Results**

In total, 36 patients were included in this study; 19 were patients with GO and 14 were normal participants. Table 1 shows the characteristics of the 19 GO patients, among whom 1 (2 eyes) had active GO, 18 had hyperthyroidism, and 1 had hypothyroidism. The mean duration of GO was 52.9 ± 21.5 months (Table 1). The mean age of all patients was 44.1 ± 7.9 years (range, 27–56 years). The mean age of GO patients was 44.0 ± 10.0 years (range, 27–56 years) and that of control participants was 44.6 ± 7.6 years (range, 27–56 years). There were no statistically significant differences between GO patients and normal patients (*p* = 0.4). The GO patients comprised of 17 women and 2 men whereas normal participants were all women (Table 2).


**Table 1.** Clinical characteristics and ophthalmologic parameter of the 19 patients with Graves' ophthalmopathy.

> GO: Graves' ophthalmopathy, CAS: clinical activity score.



TBUT = tear breakup time; SLK = superior limbic keratoconjunctivitis; DEQS = Dry Eye-related Quality-of-Life Score, GO: Graves' ophthalmopathy. All values are expressed as mean ± standard deviation (SD). *p*-value, vs. normal participants.

#### *3.1. Ocular Surface Examination*

In ocular surface investigations, the right eye was selected in GO patients, and their 19 eyes were compared to the 14 eyes of normal participants. Overall, 32 of 38 eyes (84.2%) of patients with GO were diagnosed with DED. Notably, all patients exhibited obstructive MGD. Objective ocular surface parameters are summarized in Table 2. TBUT was significantly shorter in patients with GO than that in normal participants (*p* = 0.000; Table 2). Vasculitis, meibum score, and DEQS were significantly higher in patients with GO than those in normal participants (*p* = 0.000, *p* = 0.000, *p* = 0.000 respectively; Table 2). Upper eyelid meiboscore and total meiboscore (upper and lower eyelid) were significantly higher in patients with GO than in normal participants (*p* = 0.002, *p* = 0.001 respectively; Table 2). Furthermore, abnormalities of the central part of the upper eyelid were significantly more common in patients with GO than in normal participants (*p* = 0.000; Table 2). The fluorescein staining score, Schirmer's test findings, and meiboscore of the lower lid were not significantly different between the two groups. There were no patients with SLK (Table 2; Figures 1 and 2).

#### *3.2. Examination for Graves' Ophthalmopathy*

Proptosis was observed in 38 of 38 eyes (100%); the mean proptosis value was 18.8 ± 1.8 mm. The mean palpebral fissure height was 9.0 ± 2.0 mm; of the 38 eyes, 33 (91.7%) exhibited palpebral fissure height > 7 mm. Twenty of 38 eyes (52.6%) exhibited thickening of the eyelids (eyelid swelling). Engorgement of the levator muscle on MR images was seen in 30 of 38 eyes (78.9%) (Figure 1; Figure 2). Engorgement of the extra-ocular muscles on MR images was seen in 17 of 38 eyes (44.7%). Lacrimal gland swelling was observed in 28 of 38 eyes (73.7%), and eyelid retraction was observed in 4 of 38 (10.5%) eyelids (Table 1). In terms of severity of GO, 10 patients were classified as having mild and 9 patients were classified as having moderate disease (Table 1). Only 1 patient was in the acute phase (Table 1).

Detailed data of the ocular surface parameters and GO parameters of the patients are presented in Table 3.

#### *3.3. Representative Cases*

#### 3.3.1. Case 1

A 53-year-old woman presented with GO, (disease duration of 60 months), in the inactive phase at presentation. The DEQS score was 25. The patient exhibited eyelid swelling, as well as exophthalmos (19 mm and 18 mm in the right and left eyes, respectively; Figure 1). There was no obvious eyelid retraction; the palpebral fissure height of the right and left eyes was relatively normal and mild (7 mm and 8 mm, respectively). The patient complained of dry eye symptoms, and the TBUT in both eyes was 6 s. In both eyes, the corneo-conjunctival staining score was 1, and the Schirmer's test value was 2 mm. The eyelid margins exhibited plugging (upper lid margin), displacement of the mucocutaneous junction, vascular engorgemen<sup>t</sup> (upper and lower lid margins), and inflammation (eyelid and ocular conjunctiva) (Figure 1d). The meibum grade was 1. On the basis of these findings, the patient was diagnosed with MGD. Meibography findings indicated meibomian gland dropout in the upper and lower eyelids of the right eye and the central region of the upper eyelid of the left eye (Figure 1b,c,e,f). Levator muscle swelling in the upper eyelids and lacrimal gland swelling were observed in MR images (Figure 1g–i). The CAS score was 0.

**Figure 1.** Representative case of a 53-year-old woman with Graves' ophthalmopathy. The patient presented with vascular engorgement, lid-margin inflammation, and plugging of the lid margin (**a**). The meibomian glands at the central region of the eyelids exhibit abnormal findings, particularly in the upper eyelid (**b**,**<sup>c</sup>**). Magnetic resonance imaging findings show enlargement of the levator muscle and lacrimal gland (**d**–**f**).

#### 3.3.2. Case 2

A 56-year-old woman presented with GO (disease duration of 96 months), in the inactive phase at presentation. The DEQS score was 8.3. Initially, the patient exhibited swelling, chronic inflammation, and pigmentation of the eyelids. Exophthalmos was observed in both eyes (17 mm and 16 mm in the right and left eyes, respectively; Figure 2a). In the follow-up evaluation (current status), no lid retraction was observed, and the palpebral fissure height of both eyes was mild (8 mm, both eyes). The patient complained of dry eye symptoms at presentation, and the TBUT in the right and left eyes was 3 s and 5 s, respectively. The corneo-conjunctival score in both eyes was 0. Schirmer's test values in the right and left eyes were 4 mm and 13 mm, respectively. Evaluation of eyelid margins revealed plugging and vascular engorgemen<sup>t</sup> (upper and lower eyelid margins) as well as inflammation (palpebral and ocular conjunctiva) (Figure 2d). Meibum grades in the right and left eyes were 2 and 1, respectively. On the basis of these findings, the patient was diagnosed with DED and MGD. Meibography findings revealed meibomian gland dropout in the upper and lower eyelids of both eyes; both eyes exhibited a meiboscore of 3 (Figure 2b,c,e,f). Levator muscle swelling in the upper eyelid, as well as swelling of the superior, medial, and inferior rectus muscles and lacrimal glands, were observed on MR images (Figure 1g–i). The CAS score was 1.

**Figure 2.** Representative case of a 56-year-old woman with Graves' ophthalmopathy. The patient presented with lid-margin inflammation (**a**). The meibomian glands in the central region of the upper eyelid exhibit abnormal findings (**b**,**<sup>c</sup>**). Magnetic resonance imaging findings show (**d**,**<sup>e</sup>**) levator muscle and superior and inferior rectus muscle enlargement, as well as (**f**) medial rectus muscle enlargement and swelling of the lacrimal glands.


**Table 3.** Ocular surface parameters and Graves' ophthalmopathy (GO) severity.

#### *J. Clin. Med.* **2020**, *9*, 2814

## **4. Discussion**

This study investigated the relationship between ocular surface parameters, including MGD and DED, and GO in patients with dry eye symptoms. A high percentage (84.2%) of the patients were diagnosed with DED. More surprisingly, all patients in this study exhibited obstructive MGD; this proportion was higher than the proportion of patients with DED. Our findings revealed that patients presented with clinical features of vascular engorgemen<sup>t</sup> and inflammation of eyelid margins, levator muscle thickening, and in some patients, meibomian gland changes in the central region of the eyelids, based on meibography. TBUT, vasculitis, DEQS, and meiboscores (upper lid and total score) were significantly worse in GO patients than those in the normal controls. Recently, Kim et al. reported that patients with GO exhibited morphological changes in the meibomian glands, which correlated with proptosis and palpebral fissure height [24]. The authors considered that lack of blinking by proptosis and palpebral fissure height in patients with GO was caused by decreased excretion of meibum and gave rise to obstructive MGD. The patients in this study had high meibum grades, which were similar to those previously reported in 60-year-old patients with MGD [25,26]. Recent studies reported that oxidative stress is associated with GO [27,28]. According to another study, oxidative stress resulted in changes in the meibomian glands and meibum composition [29]. In this study, most patients exhibited proptosis and increased palpebral fissure height. Based on previous studies and the results of the present study, we speculated that MGD and deterioration of meibum were caused by morphological changes related to GO and oxidative stress. The meiboscores of patients with GO were significantly higher than that of normal participants in this study, and those previously reported [24]. Furthermore, we found that patients with GO exhibited meibomian gland changes in the central region of the eyelids, particularly of the upper lid, which has not been reported previously. In a previous study of contact-lens users [30], meibography findings revealed negative changes. Hence, the authors concluded that this condition was correlated to the duration of use and chronic irritation caused by the contact-lenses. Most patients (nearly 80%) in the study exhibited levator muscle enlargement. This position might correspond to the superior levator muscles and also SLK. SLK is often observed in patients with GO and is caused by abnormal friction and ocular surface inflammation [31,32]. In a previous study, SLK had a stronger correlation with lid retraction and extra-ocular muscle enlargement in GO patients than in individuals without GO [31,33]. Although, none of our patients exhibited SLK and eyelid retraction, because most patients were in the inactive phase of GO. These results sugges<sup>t</sup> that levator muscle condition may a ffect the upper central meibomian gland morphology. However, considering the results of our own and previous reports, patients with meibography changes in the center of the eyelids might have inflammation and friction. Furthermore, they may have SLK and lid retraction during the active phase of GO. Hence, we speculated that friction and inflammation may be caused by GO; prolonged duration of GO might lead to changes in the center of the eyelids that are noted by meibography. We also hypothesized that these findings would be a characteristic meibography sign in some GO patients with dry eye symptoms. As in previous studies, TBUT in this study was lower than normal participants. [3,7,8,24]. Proptosis and increased palpebral fissure height have been reported as causes of ocular surface damage in patients with GO [2,4]. In particular, increased palpebral fissure height has been reported to be the most influential factor for tear evaporation [10,34,35]. In our study, most patients exhibited proptosis and increased palpebral fissure height; therefore, we speculated that these anatomical factors were among the reasons for the relatively short TBUT. Furthermore, excessive tear evaporation has been proposed as a cause of hyperosmolarity [35,36].

Tear hyperosmolarity stimulates the production of proinflammatory cytokines at the ocular surface [35,36]. These inflammatory responses are known to be related to T-lymphocytes. It has been reported that GO is caused by T-lymphocyte-mediated stimulation of orbital cells [2,4]. Expression of TSH receptors in GO patients has also been demonstrated, and the lacrimal gland is known as the target organ for TSH [2]. Previously, computed tomography findings have revealed swollen lacrimal glands in patients with GO [37]. In this study, MRI findings showed lacrimal gland swelling, and our patients' Schirmer's test values were lower than those reported in previous studies [7,33]. The rate of

vascular engorgemen<sup>t</sup> of the lid margin in this study was 97.3%, although, a previous study showed that the rate of this phenomenon in normal controls in women of the same age range was only 14% [20]. We also recognized inflammatory signs around the eyes. Since GO is caused by inflammation and immune-related mechanisms, GO is likely to result in MGD and DED.

This study had a few limitations. First, this was a pilot study and, therefore, the sample size was small; however, we will be conducting a similar study with a larger sample size in the future and include additional comparisons such as those of the active and inactive phases of GO. Second, some selection bias might have occurred, although we included consecutive patients with GO. Inclusion of patients with GO at di fferent stages of activity (active/chronic) and di fferent disease durations might have helped to define the details of the relationship between GO and MGD more specifically, thereby helping to elucidate the point of occurrence of MGD in GO eyes. In addition, in further studies, quantitative markers, such as those reflecting ocular surface oxidative stress and inflammation, should be studied in patients with GO. Furthermore, we would like to investigate the pathophysiological mechanisms and inflammatory pathways (e.g., T-lymphocyte-related pathways and oxidative stress-related pathways) related to MGD in GO patients, using experimental models.
