Real-World Data Reveal Impact of Macular Fluid Volume on nAMD Treatment

Ursula Schmidt-Erfurth, MD

Credit: Medical University of Vienna

A new analysis of ophthalmology registry data revealed greater macular fluid volume after initial anti-VEGF treatment correlates with poorer vision outcomes and requires more intravitreal injections in eyes with neovascular age-related macular degeneration (nAMD).¹

Using real-world data obtained from the Fight Retinal Blindness! Registry, a team of investigators observed consistent outcomes across eyes with intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (RPE), leading to a greater burden on both patients and health institutions.

“Identification of patients with high and low fluid volumes and the use of their information to personalize treatment regiments may be keys to improving management of nAMD in the real world,” wrote the investigative team, led by Ursula Schmidt-Erfurth, MD, of the department of ophthalmology and optometry at the Medical University of Vienna.

Intravitreal anti-VEGF therapy often lacks effectiveness in visual outcomes in the real-world, compared with clinical trials, with nonadherence and diminished treatment frequencies as potential factors.² Artificial intelligence algorithms can identify subtle patterns and individual treatment responses in retinal diseases, but the need to train and validate these algorithms for further testing has proven challenging.

Fluid dynamics are important in treatment requirements and prognosis in nAMD and could assist in the personalization nAMD treatment regimens.³ Advances in automated deep learning algorithms have allowed for reliable determination of fluid quantity and location. Fluid measures, including IRF, SRF, and RPE, are important biomarkers in the prediction of visual outcomes, treatment needs, and late-stage outcomes in clinical trials and daily practice.

For this analysis, Schmidt-Erfurth and colleagues investigated the effect of post-initial treatment volumes on short- and long-term treatment frequency, visual acuity, and structural changes for each fluid type.¹ A total 209 treatment-naive eyes with nAMD from the Fight Retinal Blindness! Registry treated with a treat-and-extend regimen were included. Medical records were reviewed for demographic data, BCVA, number of anti-VEGF treatments, and the development of macular atrophy or fibrosis over 12 and 48 months.

Macular fluid on optical coherence tomography (OCT) was automatically quantified using an approved AI algorithm. The high-fluid-volume subgroup was defined as those with the highest 25% quartile of mean residual fluid volumes after 2 initial treatments, while the remaining 75% were classified as the low-fluid-volume subgroup.

After initial treatment, patient with high IRF volumes differed by 2.6 letters (P = .021) and 7.4 letters (P = .007) at months 12 and 48, respectively, compared with low IRF volumes. Those eyes with high IRF additionally received significantly more treatments, with mean differences of +1.6 (P <.001) and +5.3 (P = .002) injections at months 12 and 48, respectively.

Meanwhile, patients with high SRF volumes after initial treatment had no statistically significant differences in BCVA outcomes, compared with the low-SRF subgroup. There were, however, significantly more injections in the high-SRF subgroup, showing mean differences of +2.4 injections (P <.001) and +11.4 injections (P <.001) at months 12 and 48, respectively.

Analysis of PED outcomes showed similar results as SRF, with a lack of statistically significant differences in BCVA outcomes between high- and low-volume subgroups, but an increased treatment need in the high-PED-volume subgroup following initial treatment. The mean differences in injection frequency between high- and low-PED volumes were +1.2 (P = .001) and +7.8 (P <.001) at months 12 and 48, respectively.

Further analysis of new-onset atrophy and fibrosis revealed macular atrophy occurred in 71 of 187 eyes (40%) included in the data set. Among the higher 25% IRF volume subgroup, patients had a 1.81-times higher risk of developing atrophy (P = .016) compared with the lower 75% fluid volume subgroup. Fibrosis was present in 43 of the 187 eyes (23%), but there were no significant differences between the high- and low-volume subgroups after initial treatment in all fluid compartments in both the univariate and multivariate models (all P >.05).

Schmidt-Erfurth and colleagues noted that segmentation errors and generalizability are still limiting factors inherent to studies using AI-based algorithms. Given the fluid monitor’s use in real-world data sets an in clinical practice, however, they noted the tool can provide valuable insights and assist physicians in making more informed decisions.

“With the help of automated fluid quantifications using AI, we can tailor individual treatment regimens, predict the risk of disease progression, and use this information to increase patient adherence during their lifelong disease management,” investigators wrote.

References

1. _{Reiter GS, Mares V, Leingang O, et al. Long-term effect of fluid volumes during the maintenance phase in neovascular age-related macular degeneration in the real world: results from Fight Retinal Blindness!. Can J Ophthalmol. Published online November 18, 2023. doi:10.1016/j.jcjo.2023.10.017}
2. _{Ciulla TA, Hussain RM, Pollack JS, Williams DF. Visual Acuity Outcomes and Anti-Vascular Endothelial Growth Factor Therapy Intensity in Neovascular Age-Related Macular Degeneration Patients: A Real-World Analysis of 49 485 Eyes. Ophthalmol Retina. 2020;4(1):19-30. doi:10.1016/j.oret.2019.05.017}
3. _{Bogunović H, Mares V, Reiter GS, Schmidt-Erfurth U. Predicting treat-and-extend outcomes and treatment intervals in neovascular age-related macular degeneration from retinal optical coherence tomography using artificial intelligence. Front Med (Lausanne). 2022;9:958469. Published 2022 Aug 9. doi:10.3389/fmed.2022.958469}