This large, retrospective review finds that prophylactic topical nonsteroidal anti-inflammatory drugs (NSAIDs) have limited utility for preventing macular edema after cataract surgery, especially in patients with diabetic retinopathy.
The finding that NSAIDs are not associated with a decrease in macular edema in patients with diabetic retinopathy is especially noteworthy since many physicians prescribe perioperative NSAIDs for their diabetic patients because they are a high-risk population for macular edema.
A total of 89,731 patients undergoing cataract surgery from 2007 to 2014 were included in the analysis. At 90 days postop, the incidence of developing macular edema was significantly lower among those prescribed NSAIDs, falling from 1.7% to 1.3%; however, because the incidence of edema is already low, the decrease amounts to only a small reduction in the number of patients with macular edema.
Based on their analysis, the authors conclude that to prevent a single case of macular edema one would need to treat 320 patients with perioperative NSAIDs.
Additionally, the authors found that most cases of macular edema were mild and resolved within 3 to 4 months without treatment. Only a minority developed chronic edema (1% to 3%). Therefore, the long-term visual consequences of developing macular edema in the absence of perioperative NSAIDs would be nominal for the majority of patients.
A subgroup analysis showed NSAID use conferred a lower incidence of macular edema in patients without diabetes [relative risk (RR) 0.68] and in those with diabetes but without retinopathy (RR 0.51); however, there was no association between NSAID use and macular edema risk among patients with diabetic retinopathy (RR 1.06).
Why the difference in patients with diabetic retinopathy? The authors suggest that postoperative macular edema in diabetic patients is the result of different mechanisms than in nondiabetic patients, such as an exacerbation of native diabetic eye disease, the result of pre-existing microvascular damage or activation of different inflammatory pathways.
The lower incidence of macular edema in diabetic subjects without retinopathy using NSAIDs may indicate that differences in NSAID response may not be rooted in intrinsic aspects of being diabetic (such as blood glucose levels) but rather are the result of diabetic damage to the retina.
The retrospective nature of the study and its reliance on a prescription database limit this study’s findings. Physicians may be more likely to prescribe NSAIDs in high-risk patients. Also, not all patients had an OCT, so macular edema may be under-reported in patients with good vision. A prospective, controlled trial should be conducted to confirm these results.