Evolving Treatment Patterns and Outcomes of Neovascular Age-Related Macular Degeneration Over a Decade

Maintenance of Vision Needed to Drive after Intravitreal Anti-VEGF Therapy in Patients with Neovascular Age-related Macular Degeneration and Diabetic Macular Edema

OBJECTIVE

To evaluate the association between intravitreal anti-VEGF therapy and visual acuity (VA)/driving vision maintenance over 4 years in patients with neovascular age-related macular degeneration (nAMD) or diabetic macular edema (DME).

DESIGN

Retrospective, observational, clinical practice cohort study using data from the Vestrum Health database.

PARTICIPANTS

Initial diagnosis (January 1, 2014 to June 30, 2019) of nAMD or DME and ≥ 1 year of treatment/follow-up history. The VA analysis required 4 years of treatment/follow-up history. For the driving vision maintenance analysis, patients required Snellen VA of 20/40 or better at baseline and for ≥ 6 months during year 1 after index intravitreal anti-VEGF treatment in the better-seeing eye.

METHODS

A loss-of-driving event was the first clinic visit with VA worse than 20/40 sustained for ≥ 6 consecutive months. Kaplan-Meier analyses estimated the probability of maintaining driving vision over 4 years stratified by year-1 injection number. Cox proportional hazard models examined associations between baseline clinical characteristics and year-1 injection frequency and the risk of losing driving vision.

MAIN OUTCOME MEASURES

Mean change in VA over time and by baseline VA, driving vision maintenance probability over time and stratified by anti-VEGF injection frequency, and baseline factors predictive of driving vision maintenance.

RESULTS

In year 1, the nAMD and DME cohorts gained 8.5 and 9.5 ETDRS letters, respectively. Between years 1 and 4, patients with nAMD and DME lost 6.6 and 2.7 ETDRS letters, respectively. The probability of maintaining driving vision over 4 years was 56% (nAMD) and 72% (DME); among patients who received 1 to 5, 6 to 7, and ≥ 8 anti-VEGF injections in year 1, corresponding probabilities were 50%, 56%, and 65% (nAMD; P < 0.001) and 63%, 72%, and 77% (DME; P < 0.001). Baseline factors associated with driving vision loss included older age, worse index VA, geographic atrophy (nAMD), and worsening baseline diabetic retinopathy (DME).

CONCLUSIONS

Older age and worse index VA were risk factors for driving vision loss, whereas a greater year-1 injection number was associated with driving vision maintenance through year 4, supporting early initiation and frequent anti-VEGF injections for maintaining driving vision in nAMD or DME.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Aflibercept monotherapy or bevacizumab first for diabetic macular edema

PURPOSE

Clinical outcome and switch patterns with bevacizumab first treatment strategy for patients with newly diagnosed neovascular age-related macular degeneration (nAMD).

METHODS

Retrospective observational study of the number of intravitreal injections of bevacizumab and treatment switch in patients who started intravitreal bevacizumab injections between January 1, 2016 and December 30, 2018.

RESULTS

From January 1 2016 to December 31 2018, 608 eyes of 565 patients started intravitreal injections of bevacizumab for a new diagnosis of nAMD. Average visual acuity (VA) at presentation was 0.60 logarithm of the minimum angle of resolution (logMAR), which improved to 0.47 after six injections (P < 0.001) and decreased to 0.63 at the last follow-up (P = 0.543). Switch of treatment was recommended for 190 eyes (31.3%), and of them, 91 patients (15%) were switched during the first 6 months and defined as primary failure of bevacizumab. The switch of treatment resulted in a statistically significant improvement in VA in the first 6 months after the switch. The gain in VA was not sustainable over time.

CONCLUSION

Bevacizumab first treatment strategy produced results that were comparable to previous real-world outcomes publications of ranibizumab and aflibercept treatment with low rates of failure of bevacizumab treatment. Treatment switch to second-line treatment yielded a significant VA improvement, mainly in patients with primary bevacizumab failure.

Age-Related Macular Degeneration: A Review

Importance

Age-related macular degeneration (AMD) affects approximately 20 million people in the US and 196 million people worldwide. AMD is a leading cause of severe vision impairment in older people and is expected to affect approximately 288 million people worldwide by 2040.

Observations

Older age, genetic factors, and environmental factors, such as cigarette smoking, are associated with development of AMD. AMD occurs when extracellular deposits accumulate in the outer retina, ultimately leading to photoreceptor degeneration and loss of central vision. The late stages of AMD are characterized by outer retinal atrophy, termed geographic atrophy, or neovascularization associated with subretinal and/or intraretinal exudation, termed exudative neovascular AMD. The annual incidence of AMD ranges from 0.3 per 1000 in people who are aged 55 to 59 years to 36.7 per 1000 in people aged 90 years or older. The estimated heritability of late-stage AMD is approximately 71% (95% CI, 18%-88%). Long-term prospective cohort studies show a significantly higher AMD incidence in people who smoke more than 20 cigarettes per day compared with people who never smoked. AMD is diagnosed primarily with clinical examination that includes a special lens that focuses light of the slit lamp through the pupil. Exudative neovascular AMD is best identified using angiography and by optical coherence tomography. Individuals with AMD who take nutritional supplements consisting of high-dose vitamin C, vitamin E, carotenoids, and zinc have a 20% probability to progress to late-stage AMD at 5 years vs a 28% probability for those taking a placebo. In exudative neovascular AMD, 94.6% of patients receiving monthly intravitreal anti-vascular endothelial growth factor (anti-VEGF) injections experience less than a 15-letter visual acuity loss after 12 months compared with 62.2% receiving sham treatment.

Conclusions and Relevance

The prevalence of AMD is anticipated to increase worldwide to 288 million individuals by 2040. Intravitreally administered anti-VEGF treatment is first-line therapy for exudative neovascular AMD.

Clinically important change for the FACIT-Fatigue scale in paroxysmal nocturnal hemoglobinuria: a derivation from international PNH registry patient data

BACKGROUND

Fatigue is the most common symptom associated with paroxysmal nocturnal hemoglobinuria (PNH). The objective of this analysis was to estimate values that would suggest a clinically important change (CIC) for the functional assessment of chronic illness therapy-fatigue scale (FACIT-Fatigue) in patients with PNH.

METHODS

Adults with PNH who initiated eculizumab within 28 days of enrollment in the International PNH Registry as of January 2021 with baseline FACIT-Fatigue scores were included in the analysis. Distribution-based estimates of likely difference were calculated using 0.5 × SD and SEM. Anchor-based estimates of CIC considered the European Organization for Research and Treatment of Cancer (EORTC) global health status/quality of life summary score and the EORTC Fatigue Scale score. Changes in anchors and high disease activity (HDA) shift from start of eculizumab treatment to each follow-up visit were then assessed by FACIT-Fatigue score change (≤ 1 CIC, no change, or ≥ 1 CIC).

RESULTS

At baseline, 93% of 423 patients had fatigue documented in their medical history. The distribution-based estimates for FACIT-Fatigue were 6.5 using 0.5 × SD and 4.6 using SEM; internal consistency was high (α = 0.87). For anchor-based estimates, the FACIT-Fatigue CIC ranged from 2.5 to 15.5, and generally supported 5 points as a reasonable lower end of the value for meaningful individual change. The percentage of patients who changed from having HDA at baseline to no HDA at eculizumab-treated follow-up visits increased over time.

CONCLUSION

These results support the use of 5 points as the CIC for FACIT-Fatigue in patients with PNH, which is within range of the CICs reported in other diseases (3-5 points).

Volumetric subfield analysis of cynomolgus monkey’s choroid derived from hybrid machine learning optical coherence tomography segmentation

This study aimed to provide volumetric choroidal readings regarding sex, origin, and eye side from healthy cynomolgus monkey eyes as a reference database using optical coherence tomography (OCT) imaging. A machine learning (ML) algorithm was used to extract the choroid from the volumetric OCT data. Classical computer vision methods were then applied to automatically identify the deepest location in the foveolar depression. The choroidal thickness was determined from this reference point. A total of 374 eyes of 203 cynomolgus macaques from Asian and Mauritius origin were included in the analysis. The overall subfoveolar mean choroidal volume in zone 1, in the region of the central bouquet, was 0.156 mm³ (range, 0.131–0.193 mm³). For the central choroid volume, the coefficient of variation (CV) was found of 6.3%, indicating relatively little variation. Our results show, based on analyses of variance, that monkey origin (Asian or Mauritius) does not influence choroid volumes. Sex had a significant influence on choroidal volumes in the superior-inferior axis (p ≤ 0.01), but not in the fovea centralis. A homogeneous foveolar choroidal architecture was also observed.

Longer-Term Anti-VEGF Therapy Outcomes in Neovascular Age-Related Macular Degeneration, Diabetic Macular Edema, and Vein Occlusion-Related Macular Edema: Clinical Outcomes in 130 247 Eyes

PURPOSE

The clinical practice visual acuity (VA) outcomes of anti-VEGF therapy for up to 5 years were assessed in patients with neovascular age-related macular degeneration (nAMD), diabetic macular edema (DME), branch retinal vein occlusion-related macular edema (BRVO-ME), and central retinal vein occlusion-related macular edema (CRVO-ME).

DESIGN

A retrospective analysis was performed using the Vestrum Health Retina Database.

PARTICIPANTS

Treatment-naive patients with nAMD, DME, BRVO-ME, or CRVO-ME who received anti-VEGF injections between 2014 and 2019 and had follow-up data for ≥12 months.

METHODS

Data on age, sex, the number of anti-VEGF treatments, and VA were analyzed.

MAIN OUTCOME MEASURES

Mean VA change up to 3 years (BRVO-ME and CRVO-ME) and 5 years (nAMD and DME).

RESULTS

At 1, 3, and 5 years, in 67 666, 21 305, and 5208 eyes with nAMD, after a mean of 7.6, 19.5, and 32 injections, there was a mean change of +3.1, -0.2, and -2.2 letters, respectively. At 1, 3, and 5 years, in 40 832, 7728, and 1192 eyes with DME, after a mean of 6.2, 15.4, and 26.0 injections, there was a mean change of +4.7, +3.3, and +3.1 letters, respectively. At 1 and 3 years, in 12 451 and 3027 eyes with BRVO-ME, after a mean of 7.1 and 18.2 injections, there was a mean change of +9.5 and +7.7 letters, respectively. At 1 and 3 years, in 9298 and 2264 eyes with CRVO-ME, after a mean of 7.3 and 18.8 injections, there was a mean change of +8.3 and +6.0 letters, respectively (P < 0.01 for all VA changes of > 1 letter). In all 4 conditions, the mean VA increased with the mean number of anti-VEGF injections, eyes with a baseline VA of 20/40 or better tended to lose VA, and eyes with progressively worse baseline VA experienced a progressively greater VA gain at 3 years.

CONCLUSIONS

In practice, patients with nAMD, DME, BRVO-ME, and CRVO-ME showed limited visual outcomes, with patients with nAMD tending to lose VA at 3 and 5 years. Across all 4 disorders, the mean change in VA correlated with treatment intensity at 1, 3, and 5 years. Patients with better baseline VA are more vulnerable to vision loss.

Artificial intelligence-based predictions in neovascular age-related macular degeneration

PURPOSE OF REVIEW

Predicting treatment response and optimizing treatment regimen in patients with neovascular age-related macular degeneration (nAMD) remains challenging. Artificial intelligence-based tools have the potential to increase confidence in clinical development of new therapeutics, facilitate individual prognostic predictions, and ultimately inform treatment decisions in clinical practice.

RECENT FINDINGS

To date, most advances in applying artificial intelligence to nAMD have focused on facilitating image analysis, particularly for automated segmentation, extraction, and quantification of imaging-based features from optical coherence tomography (OCT) images. No studies in our literature search evaluated whether artificial intelligence could predict the treatment regimen required for an optimal visual response for an individual patient. Challenges identified for developing artificial intelligence-based models for nAMD include the limited number of large datasets with high-quality OCT data, limiting the patient populations included in model development; lack of counterfactual data to inform how individual patients may have fared with an alternative treatment strategy; and absence of OCT data standards, impairing the development of models usable across devices.

SUMMARY

Artificial intelligence has the potential to enable powerful prognostic tools for a complex nAMD treatment landscape; however, additional work remains before these tools are applicable to informing treatment decisions for nAMD in clinical practice.

Labelling of data on fundus color pictures used to train a deep learning model enhances its macular pathology recognition capabilities

Retinal diseases remain one of the leading causes of visual impairments in the world. The development of automated diagnostic methods can improve the efficiency and availability of the macular pathology mass screening programs. The objective of this work was to develop and validate deep learning algorithms detecting macular pathology (age-related macular degeneration, AMD) based on the analysis of color fundus photographs with and without data labeling. We used 1200 color fundus photographs from local databases, including 575 retinal images of AMD patients and 625 pictures of the retina of healthy people. The deep learning algorithm was deployed in the Faster RCNN neural network with ResNet50 for convolution. The process employed the transfer learning method. As a result, in the absence of labeling, the accuracy of the model was unsatisfactory (79%) because the neural network selected the areas of attention incorrectly. Data labeling improved the efficacy of the developed method: with the test dataset, the model determined the areas with informative features adequately, and the classification accuracy reached 96.6%. Thus, image data labeling significantly improves the accuracy of retinal color images recognition by a neural network and enables development and training of effective models with limited datasets.