Eye and Eyelid Abnormalities Are Common in Patients with Hashimoto's Thyroiditis

Introduction

T hyroid-associated ophthalmopathy is an autoimmune disorder of the extraocular muscles and surrounding orbital connective tissue and fat and is generally associated with Graves' disease. Although case studies of ophthalmopathy in association with Hashimoto's thyroiditis have been reported (1 –3), the occurrence of overt ophthalmopathy in patients with Hashimoto's thyroiditis is much less common than in Graves' disease (4,5). However, the spectrum of eye signs and their prevalences in Hashimoto's thyroiditis are not well documented, mainly because in the absence of stare and proptosis, the eyes and eyelids are not usually examined in detail.

The pathogenesis of ophthalmopathy and the mechanism for its link to thyroid autoimmunity are poorly understood (6 –9). Several antigens have been identified as possible autoantibody targets, including the thyroid stimulating hormone (TSH)-receptor (TSH-r) (10 –12), the skeletal muscle calcium binding protein calsequestrin (13 –15), and the fibroblast cell membrane protein collagen XIII (16 –18). A popular working hypothesis for this association is that the TSH-r, a shared antigen in the thyroid and orbit, is targeted in those patients who develop ophthalmopathy (6 –8). However, the TSH-r hypothesis cannot apply to Hashimoto's thyroiditis, as antibodies against the TSH-r are generally not found in these patients. The aim of the present study was to determine the true prevalence of eye signs in patients with Hashimoto's thyroiditis. We prove that about one-third of patients have eye signs, which are usually mild and often manifest as dominant upper eyelid retraction (UER).

Materials and Methods

Results

We have examined 91 recently diagnosed patients with Hashimoto's thyroiditis for eye and upper eyelid abnormalities. We measured serum orbital antibodies in serum from 20 patients. The overall prevalences of eye changes were 34%. In more detail, 29 patients were classified as Nunery type I (inflammatory disease without restrictive myopathy), and two patients were classified as Nunery type II (with restrictive myopathy). The disease was inactive (CAS score 0) in 13 patients; mildly active (CAS 1–3) in 8 patients; and active (CAS > 4) in 10 patients. According to the NOSPECS classification, 15 patients were class 1; 11 were class 2; 3 were class 3; and 2 were class 4 (eye muscle involvement, i.e., Nunery type II), whereas no patient was class 5 or 6. Only two patients had eye muscle dysfunction manifest as intermittent double vision and reduced eye movements in up gaze (Nunery type II, CAS 7). Both of these patients also had UER. Thirteen patients had significant UER, taken as a margin-reflex distance of >5 mm, which was the only abnormality except for mild soft tissue swelling (NOSPECS class 2) in five patients, associated with eye muscle damage in two patients (NOSPECS class 4, Nunery type II), and with more severe soft tissue inflammation in six patients (NOSPECS classes 2 or 3), which was often active (CAS 4–6) (Table 1).

We measured calsequestrin and collagen XIII antibodies in serum from 20 patients. Mean ± standard deviation values were not significantly different between patients with and without any eye signs, for either antibody (Mann–Whitney test, p = not significant (NS), p = NS respectively) (Fig. 1). Moreover, tests were also positive in about 40% of patients with no evident eye signs at the time of the study.

OPEN IN VIEWER

FIG. 1.

Levels of serum collagen XIII (COL XIII) and calsequestrin (CSQ1) antibodies in 20 patients with Hashimoto`s thyroiditis with (n = 13) and without (n = 7) eye signs. Results were expressed as mean optical density (OD) at 405 nm × 1000. A positive test was taken as an optical density greater than the upper limit of normal for 30 healthy men below 30 years and with no known medical history and no family history of thyroid-autoimmune, which was 194 for calsequestrin and 193 for collagen XIII antibodies.

Finally, there was no significant relationship between cigarette smoking and either eye abnormalities or calsequestrin or collagen XIII antibodies in these patients (results not shown).

Discussion

Ophthalmopathy is diagnosed clinically from the well-known signs of proptosis, soft tissue swelling, decreased eye muscle function, and chronic UER and lagophthalmos (24,25). UER and lagophthalmos may also occur in the absence of other signs of ophthalmopathy except for mild proptosis (26). Hashimoto's thyroiditis is not usually associated with overt ophthalmopathy, with an estimated prevalence as low as 2–5%. In contrast, Graves' disease is strongly associated with ophthalmopathy, with an occurrence rate as high as 50%, depending on how the eye disorder is defined. Here, we determined the prevalences of the full spectrum of ophthalmopathy and eyelid changes in patients with Hashimoto's thyroiditis seen at a single thyroid clinic in Sydney, Australia. To summarize the main findings, the overall prevalence of any eye signs was 34%. The eye signs were mostly mild and often manifest as dominant UER. However, two of the patients had eye muscle dysfunction and about a third had periorbital swelling, chemosis, conjunctival injection, and/or proptosis, that is, more severe inflammatory changes. The eye signs correlated only modestly with calsequestrin and collagen XIII antibodies. In contrast to patients with Graves' disease, where cigarette smoking is shown to be a major risk factor for ophthalmopathy (27 –29), there was no relationship between smoking status and eye abnormalities in patients with Hashimoto's thyroiditis, as studied here.

Eye changes, in particular UER and lag, due to levator muscle damage, are thus common in patients with Hashimoto's thyroiditis and are generally associated with antibodies against calsequestrin and collagen XIII. This novel observation raises questions about the pathogenesis of ophthalmopathy in patients with thyroid autoimmunity. Since TSH-r antibodies are expected to be negative in the majority of patients with Hashimoto's thyroiditis, autoimmune-mediated damage of the levator palpebrae superioris muscle cannot be due to TSH-r antibodies. An alternative explanation for ophthalmopathy in patients with Hashimoto's thyroiditis is specific production of antibody against eye muscle antigens such as calsequestrin and flavoprotein, shown to be good markers of eye muscle damage in patients with Graves' disease (15,17). Here, we were not able to show a close relationship between eye signs and collagen XIII and calsequestrin antibodies. This may be due to one of the following reasons: other antibodies, such as those targeting flavoprotein or G2s (30,31), play a more important role; the main cell membrane autoantigen has not yet been identified; or cytotoxic T lymphocytes targeting this, or another, antigen are the principal pathogenetic mechanisms of the UER in patients with Hashimoto`s thyroiditis.

In conclusion, severe ophthalmopathy is uncommon in patients with Hashimoto's thyroiditis. However, mild eye signs, in particular UER, are found in about a third of patients when looked for carefully. While we did not show a close relationship between eye signs and calsequestrin and collagen XIII antibodies, this may reflect the natural history of the thyroid disorder, which may not be diagnosed until several years after its onset, at which time the antibodies may have become negative. Although eyelid abnormalities may be a minor problem for most patients, for some there are major cosmetic implications requiring surgical management. Future studies should address the natural history of eye changes in patients with Hashimoto's thyroiditis before the onset of any ophthalmopathy and following treatment with thyroxin.