Treatment of upper eyelid retraction related to thyroid-associated ophthalmopathy using subconjunctival triamcinolone injections

Sung Jun Lee, Tyler Hyung Taek Rim, Sun Young Jang, Chan Yun Kim, Dong Yeob Shin, Eun Jig Lee, Sang Yeul Lee, Jin Sook Yoon

Research output: Contribution to journalArticlepeer-review

45 Citations (Scopus)


Background: To evaluate the efficacy of subconjunctival triamcinolone injection for treating upper eyelid retraction caused by thyroid-associated ophthalmopathy (TAO). Methods: Prospective single blind randomized clinical trial. Patients diagnosed with TAO-associated eyelid retraction and/or swelling <6 months before study onset were randomly assigned to triamcinolone-injected (group I; 55 patients, 75 eyes) or observation-only (group II; 40 patients, 59 eyes) populations. Group I received 1-3 injections of 20 mg triamcinolone acetate into the subconjunctival eyelid, between the conjunctiva and Muller's muscle, at 3-week intervals. Group I was followed up at 3, 6, 9, and 24 weeks post-injection, and group II was evaluated at 9 and 24 weeks after initial visit. Eyelid swelling and retraction were each graded on a 0-3 scale. Treatment was stopped after 1-2 injections if both swelling and retraction resolved completely or if retraction and swelling scored 0/1 or 1/0 without functional or cosmetic patient concerns. Treatment success was defined at 9 and 24 weeks post-injection if eyelids were normal or when treatment was terminated because of early clinical resolution. We compared baseline clinical data between success and failure group evaluated at 9 and 24 weeks in group I, and investigated short-term and long-term success prognostic factor using multiple logistic regression analysis in each group I and II. Results: Swelling and retraction decreased significantly more in group I than in group II. Significantly more eyes had severe swelling (≥ grade 2) in group I (67 %) than in group II (34 %) upon initial evaluation (p < 0.01); this difference disappeared at 9 and 24 weeks. Fewer severely retracted eyes were observed in group I than in group II at 9 and 24 weeks (p < 0.01). Significantly more eyes in group I achieved success at both 9 and 24 weeks (59 %, 75 %) than in group II (39 %, 57 %) (p = 0.03, p = 0.04 respectively). Higher initial retraction grades (2-3) predicted a higher chance of post-injection failure versus the reference group (grade 0-1) at 9 and 24 weeks in group I, with adjusted odds ratios (aOR) 45.4 (95 % CI = 5.9-351.1, p < 0.01) and aOR 11.6 (95 % CI = 2.3-58.5, p < 0.01) respectively. Similarly in group II, initial retraction grade was associated with the failure at 9 and 24 weeks, with aOR 10.3 (95 % CI = 1.8-59.6, p < 0.01) and aOR 5.9 (95 % CI = 1.3-25.9, p < 0.05) respectively. Transient intraocular pressure elevation was observed in three eyes of two patients, although all ocular pressures were normalized within 1 month using anti-glaucoma medication. Conclusions: Subconjunctival triamcinolone injections were very effective in resolving eyelid swelling and retraction in recent-onset TAO. However, the symptom-reducing effect of triamcinolone was modest and less effective in patients initially presenting with severe retraction grades. As intraocular pressure may rise after steroid injection at upper eyelid, the treatment should be avoided in patients suspected to have glaucoma.

Original languageEnglish
Pages (from-to)261-270
Number of pages10
JournalGraefe's Archive for Clinical and Experimental Ophthalmology
Issue number1
Publication statusPublished - 2013 Jan

Bibliographical note

Funding Information:
Acknowledgment This study was supported by a grant of the Korea Health technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A101727).

All Science Journal Classification (ASJC) codes

  • Ophthalmology
  • Sensory Systems
  • Cellular and Molecular Neuroscience


Dive into the research topics of 'Treatment of upper eyelid retraction related to thyroid-associated ophthalmopathy using subconjunctival triamcinolone injections'. Together they form a unique fingerprint.

Cite this