Patients with systemic lupus erythematosus (SLE) had about 3 times the incidence of neovascular age-related macular degeneration (AMD) compared with matched controls.
Investigators conducted a retrospective cohort study using South Korea's National Health Insurance Service claims database, identifying 9,496 patients aged 45 years and older with incident SLE between 2011 and 2020. The patients were matched 1:3 with 28,488 controls without SLE according to sex, birth year, index year, and Charlson Comorbidity Index. Those with AMD diagnosed within 12 months prior to or following the index date were excluded. Outcomes were assessed through December 2023, with a mean follow-up of 6.7 years among patients with SLE and 7 years among controls.
The investigators evaluated incident AMD, including dry and neovascular subtypes, and examined whether incidence differed according to treatment-defined SLE severity and hydroxychloroquine exposure duration. Severe SLE was defined by sustained treatment with intravenous cyclophosphamide or mycophenolatemofetil for at least 6 months, serving as a claims-based marker of organ-threatening disease requiring prolonged immunosuppression. Neovascular AMD identification required a diagnosis code and a copayment registration code indicating ophthalmologist-confirmed choroidal neovascularization.
During 63,696 person-years of follow-up, 313 AMD events occurred among patients with SLE, including 246 dry AMD and 67 neovascular AMD cases. Overall AMD incidence was 49.1 per 10,000 person-years among patients with SLE vs. 26.7 among controls.
The largest relative difference involved neovascular AMD, with incidence rates of 10.5 vs. 3.3 per 10,000 person-years among patients with SLE and controls, respectively. Dry AMD incidence was higher among patients with SLE at 38.6 vs. 23.4 per 10,000 person-years.
Despite the elevated AMD incidence associated with SLE, greater treatment-defined disease severity was not associated with a corresponding increase. Patients classified as having severe SLE did not have statistically significantly higher incidence rates of overall, dry, or neovascular AMD compared with those with nonsevere disease. Adjusted analyses similarly found no association between severe SLE and higher AMD risks, indicating that the excess incidence observed in the SLE cohort was not concentrated among patients receiving prolonged major immunosuppression.
The association between SLE and AMD persisted across hydroxychloroquine exposure-duration groups. Compared with controls, patients with SLE who had received hydroxychloroquine for less than 5 years had about 2.7 times the likelihood of neovascular AMD, whereas those with at least 5 years of exposure had about 4 times the likelihood. However, analyses restricted to patients with SLE did not identify a statistically significant association between longer hydroxychloroquine exposure and higher neovascular AMD hazards. Treatment duration was an imperfect marker of ophthalmic surveillance and that the findings should not be interpreted as evidence of a causal medication effect.
The study had several limitations. Its retrospective claims-based design prevented causal inference, while the absence of information on smoking, body mass index, family history, and genetic susceptibility introduced potential residual confounding. Patients with SLE may have undergone more frequent ophthalmic examinations, particularly during hydroxychloroquine treatment, increasing the likelihood of AMD detection. The database lacked retinal imaging findings, visual acuity measurements, and detailed information on AMD severity or progression. Additionally, the treatment-based severity definition did not directly measure disease activity, and the study population was limited to Korean adults aged 45 years and older.
"AMD, particularly neovascular AMD, may represent an additional component of ocular morbidity in SLE," wrote lead study author Hong Ki Min, of the Division of Rheumatology at Konkuk University Medical Center in the Republic of Korea, and colleagues.
The study authors declared no conflicts of interest.
Source: Scientific Reports