Current Issue

---

In this week’s issue

• Intraoperative floppy iris syndrome was shown to be associated with use of certain antidepressants as well as inhalers, with some sex-specific patterns. 
• Bisphosphonate use is linked to a higher cumulative incidence of age-related macular degeneration secondary to inflammatory signaling pathway activation. 
• Selective laser trabeculoplasty for patients with open-angle glaucoma or ocular hypertension is successful in reducing the need for IOP-lowering drops in stable patients. However, as a second line therapy for less stable patients, it may not prevent escalation to surgery. 
• Artificial intelligence shows promising performance in interpreting fluorescein angiography in the diagnosis of retinal vasculitis. 
• Patching later in childhood for unilateral congenital cataract does not necessarily yield better final visual acuity. 

Too busy to read the lens? Listen to our weekly summary on your daily https://creators.spotify.com/pod/show/lens-pod/episodes/The-Lens-Newsletter-March-5th--2025-e2vng7a

Ophthalmology
Patching for unilateral congenital cataract after age 4

Is patching still a visionary move for kids over age 4 with unilateral cataracts? The current standard of care for children with a unilateral congenital cataract (UCC) is early cataract extraction with refractive correction and patching. The reported amount of time spent patching has been associated with improved visual outcomes. Given the possible outcomes of unsuccessful treatment or poor vision, it may be helpful for caregivers to know if and when patching can be discontinued. This cohort study investigated which measures of early visual acuity (VA) may be predictive of final VA and how patching affects this relationship. The study used data from the Infant Aphakia Treatment Study (IATS), which was a multicenter clinical trial of treatment for UCC. Monocular VA was measured and assessed for any relationship with the average reported daily hours of patching. VA at age 4 years was predictive of VA at age 10.5 years on average (rSpearman = 0.83; P < 0.01). The amount of patching reported between ages 4 and 5 was unrelated to changes in VA between age 4 and 10.5 years (rSpearman = 0.002, P = 0.99). More hours of patching from 48-60 months of age was correlated with better VA at age 10.5 years overall (rPearson = –0.5, P = 0.01). These findings suggest that discontinuing patching or using a less-aggressive approach may be warranted in some patients.

American Journal of Ophthalmology

Risk factors for intraoperative floppy iris syndrome

Could antidepressants and inhalers rival alpha-blockers in causing the dreaded floppy iris during cataract surgery? Intraoperative floppy iris syndrome (IFIS), marked by iris billowing, poor dilation, and increased risk of surgical complications, can jeopardize cataract surgery outcomes, especially in patients with certain medication exposures. In this retrospective pharmacovigilance study, researchers analyzed over 12 million reports in the FDA Adverse Event Reporting System (FAERS) from 2003 to 2024, identifying 649 IFIS cases. Analysis revealed the highest associations with imipramine (OR = 251.66), tamsulosin (OR = 171.44), and chlorpromazine (OR = 91.30). Among women, brinzolamide (OR = 409.63) and salbutamol (OR = 67.12) were strongly linked to IFIS, while these associations were absent in men. These findings expand the list of high-risk medications beyond the well-known alpha-1 blockers and reveal novel, sex-specific patterns. For cataract surgeons, identifying these drugs preoperatively can inform tailored strategies such as intraoperative iris hooks or pharmacologic pupil dilation, to prevent complications. Given the frequency of cataract surgery and the widespread use of these medications, this study supports more vigilant screening and counseling to optimize outcomes for at-risk patients.

IOVS

Bisphosphonate use related to AMD risk

Another piece to the puzzle of AMD pathogenesis. The NLRP3 inflammasome has recently been shown to play a role in pathogenesis of age-related macular degeneration (AMD), a leading cause of blindness worldwide. Bisphosphonates are drugs used to treat skeletal disorders that are known to activate NLRP3 signaling pathways. This retrospective cohort study explored this relationship by investigating the effect of bisphosphonates on AMD incidence. The study included 291,921 patients who had undergone DEXA scans: 265,137 unexposed and 26,784 treated with a study bisphosphonate. After matching, bisphosphonate exposure was associated with a 4.7% increased hazard of AMD incidence (HR, 1.047; 95% CI, 1.02 - 1.074). Among bisphosphonate users, use of the direct NLRP3 inhibitors fluoxetine or fluvoxamine was associated with a lower hazard of incident AMD (HR, 0.814; 95% CI, 0.676 – 0.98). Overall, patients treated with bisphosphonates had a higher cumulative incidence of AMD. This supports the importance of the NLRP3 inflammasome in AMD pathogenesis, and provides evidence that it should be considered a target for future AMD drug development studies.

JAMA Ophthalmology

Is selective laser trabeculoplasty an effective secondary treatment?

Even the steadiest hands may tremble after a night of drinking. Routine alcohol testing in physicians is a debated topic due to the lack of standard guidelines and concern for individual privacy. This study aimed to assess how different levels of alcohol consumption impact simulated surgical dexterity and tremor in experienced vitreoretinal surgeons. 11 senior vitreoretinal surgeons completed Eyesi surgical simulations after drinking alcohol to achieve a low-dose (0.06%-0.10%) or high-dose (0.11%-0.15%) BAC. They also completed surgical simulations after drinking a high-dose BAC and after a night of sleep to simulate a hangover. Performance was assessed the next morning after a night of high-dose alcohol consumption and sleep. Performance scores (0-700) (worst-best) and hand tremors (0-100) (best-worst) were recorded. Surgical performance significantly decreased after high-dose vs. low-dose alcohol use (−8.60±10.77 vs. −1.21±7.71). Hangover performance in the morning after high-dose alcohol use was similar to low-dose alcohol use (−1.76±14.47 vs. −1.21±7.71). A greater tremor was experienced during the hangover and after alcohol use, but a hangover caused a greater hand tremor score than after high-dose alcohol (7.33±21.65 vs. −4.12±17.17). Alcohol-related declines in surgical dexterity were dose-dependent, with notable impairment at higher doses. While dexterity improved after sleep, tremors worsened, raising concerns about next-day surgical performance during hangovers. Further studies are needed to assess real-world implications for patient safety, but this study suggests surgeons should not drink the same day of surgery and avoid high-dose alcohol the day before surgery as well.

Ocular Immunology and Inflammation

Using deep learning to segment retinal vascular leakage & occlusion in retinal vasculitis

Retinal vasculitis, meet artificial intelligence. Retinal vasculitis (RV) is a complicated disease, sometimes subtle, often vision-threatening, and difficult to quantify on fluorescein angiography (FA) alone. While FA is a mainstay for detecting leakage and occlusion, clinical interpretation is inherently subjective, and lacks a standard grading system for RV severity. To increase objectivity and consistency, this study developed deep learning models to automatically segment retinal vascular leakage and occlusion in RV using widefield FA images. Researchers collected 463 FA images from 82 patients across two centers and trained 12 segmentation models (variations of UNet, UNet++, and DeepLabV3+ with different backbones and preprocessing) to identify leakage and occlusion. Manual segmentations were used as ground truth. The best-performing model for leakage achieved a Dice score of 0.628, while the occlusion model scored 0.699, showing promising reliability across both Optos and Heidelberg images. Notably, occlusion was easier to segment than leakage, likely due to more consistent anatomical patterns. This study supports the feasibility of deep learning–based segmentation in RV and highlights its potential to support clinicians in more consistent disease monitoring, especially in multicenter trials or clinical settings with limited uveitis expertise.

VIEW1 & VIEW2 (Aflibercept)

Two heads may be better than one, but are two injections really needed each month? In 2012, the VIEW1 & VIEW2 trials aimed to compare the efficacy and safety of monthly and every-two-month aflibercept to monthly ranibizumab for the treatment of wet age-related macular degeneration (AMD). These large studies (total n=2419) were conducted in the wake of the prior studies, which revealed the safety shortcomings of both monthly bevacizumab and ‘as needed’ ranibizumab. The VIEW studies demonstrated the statistical non-inferiority (margin of 10%) and clinical equivalence (margin of 5%) of three treatment regimens of aflibercept to monthly ranibizumab measured by preservation of vision and reduction in central retinal thickness.

Key Points:

• All three monthly aflibercept treatment regimens were non-inferior and clinically equivalent to monthly ranibizumab in preventing moderate visual acuity loss at 52 weeks.
• The sustained durability of intravitreal aflibercept demonstrated by the every-two-month regimen is consistent with prior evidence of higher binding affinity to VEGF than ranibizumab.

Overall, VIEW1 & VIEW2 are landmark studies by providing evidence of the clinical efficacy for less frequent anti-VEGF injections with aflibercept. These studies came at a time when prior studies highlighted the need for monthly rather than as-needed injections, raising concerns of the large treatment burden of monthly visits. However, the VIEW studies revived the hope for extended treatment regimens, which have become part of today’s clinical practice.