Abstract
Background: Graves' ophthalmopathy (GO) is characterized by upper eyelid retraction (UER), the most prevalent clinical sign. We aimed to assess the clinical efficacy of a multimodal combination of steroids, 5-fluorouracil (5-FU), and botulinum neurotoxin A (BoNT-A) injections in managing UER with GO and analyze the clinical factors in relation to the injection response. Methods: A total of 37 eyes from 23 patients were enrolled for UER with GO. At the endocrinology clinic, the patients were referred to the ophthalmology clinic after taking antithyroid medication for an average of 5.76 months (13 patients), while 10 patients were initially diagnosed with GO and referred to the endocrinology clinic for management of the thyroid hormone function. They performed an orbital computed tomography (CT) scan and measured the cross-sectional area of the orbit, orbital fat, and each extra ocular muscle (EOM) except for the inferior oblique muscle 4 mm behind the eyeball. Each of the EOMs and orbital fat were calculated as a ratio to the total orbit area. A total of 0.1 cc of triamcinolone (40 mg/mL), dexamethasone (5 mg/mL), 5-FU, and BoNT-A (2.5 units) was injected transconjunctivally. Medical records were examined and photographs were utilized to assess MRD1, inferior palpebral fissure (IPF), and lid lag during down gaze before and after the injection. The patients were divided into two groups: responders (more than 1 mm decrease in MRD1 after injection) and non-responders. During the follow-up period (11.0 ± 11.6 months), any potential adverse effects were monitored. Results: CAS decreased from 3.0 ± 0.8 to 1.4 ± 0.5 after the injection, and MRD1 decreased from 5.0 ± 0.9 mm to 4.5 ± 1.3 mm. Sixty percent of the patients were responders. Before and after the injection, the difference between IPF and MRD1 in responders was 0.60 ± 1.10 mm and 0.90 ± 0.90 mm, respectively, whereas, in non-responders, it was -0.57 ± 0.88 mm and -0.15 ± 0.75 mm, respectively. In the responders, pre-injection IPF and FT4 were significantly higher (p < 0.05). Responders had a larger EOM cross-sectional area (153.5 ± 18.0 mm(2)), including a larger lateral rectus muscle cross-sectional area (37.6 ± 9.7 mm(2)) than non-responders (132.0 ± 27.9 mm(2); 29.1 ± 8.1 mm(2)). In responders, the treatment effect on IPF and MRD1 remained consistent at 1.2 ± 3.4 mm and 1.2 ± 1.6 mm, respectively, during the latest follow-up assessment. Conclusions: The combination injection of corticosteroids, 5-FU, and BoNT-A would be effective, especially, in patients with hyperthyroidism and an elongated IPF. Additionally, an increase in EOM cross-sectional area on CT, up to 150 mm(2), may serve as an additional positive indicator for the use of multimodal injections in UER with GO.