Seeing what visual acuity misses: Focal microperimetry in geographic atrophy

What if a functional end point could detect what lesion size and visual acuity miss in geographic atrophy (GA)? Tandon and colleagues report findings from a prospective, single-arm observational study, published in Ophthalmology Science, evaluating the test-retest repeatability of a novel focal photopic microperimetry approach using the Nidek MP-3 in patients with nonsubfoveal GA, and exploring how central cone-mediated sensitivity relates to lesion size and other visual function measures.¹

Why conventional measures leave a gap

GA, the non-neovascular advanced form of age-related macular degeneration (AMD), involves progressive degeneration of the retinal pigment epithelium and outer retinal layers. Although GA lesion growth is accepted as a structural end point in clinical trials, the authors note its limited correspondence with how patients actually see.¹ Lesions often spread outward from the fovea, sparing central vision for extended periods—a disconnect that may help explain why the FDA-approved therapies pegcetacoplan (Syfovre; Apellis Pharmaceuticals) and avacincaptad pegol (Izervay; Astellas Pharma) have slowed lesion growth without producing statistically significant gains in visual function.¹⁻³ With new treatments on the horizon, the authors argue that reliable, sensitive functional end points capable of detecting subtle changes in cone-mediated vision are increasingly important for trial design.¹

In AMD, photoreceptor loss does not affect all cell types equally or simultaneously. Evidence from histopathologic studies suggests that rods deteriorate before cones, and that this rod-dominant loss is concentrated in the parafoveal ring, while cone photoreceptors at the foveal center show relative resilience until later in the disease course.¹ This differential vulnerability provides the biological rationale for testing cone function under photopic conditions in the central foveal zone—the precise focus of the approach evaluated in this study. The Nidek MP-3 offers several features that make it particularly suited to this application: it supports the smaller Goldmann II stimulus size, which is not available on the Centervue MAIA; it spans a broad sensitivity range; it employs high-speed retinal tracking; and its grid patterns can be tailored to target specific retinal regions.¹

How the study was designed

The study was approved by the University of Texas Southwestern and WCG Institutional Review Board and conducted in accordance with the Declaration of Helsinki and the Health Insurance Portability and Accountability Act. Twelve participants (17 eyes) with non-subfoveal GA confirmed by spectral-domain optical coherence tomography and fundus autofluorescence underwent two focal photopic microperimetry tests on the MP-3 within a 2- to 4-week interval—short enough relative to expected GA progression to reflect test-retest variability rather than true disease change.¹

Each session used a 45-point focal grid centered on the fovea, with Goldmann II stimuli and 0.5° point spacing covering approximately 1.2 mm × 0.6 mm of the central retina. The MP-3's built-in follow-up alignment feature was used to re-register the stimulus grid between sessions. Primary repeatability outcomes were coefficient of repeatability (CoR) and intraclass correlation coefficient (ICC) for mean sensitivity (MS) and point-wise sensitivity (PWS), analyzed with Bland-Altman methods. Correlations between adjusted MS (excluding floor-effect loci at ≤3 dB) and GA lesion area, best-corrected visual acuity (BCVA), low-luminance visual acuity (LLVA), quick contrast sensitivity function (qCSF), and low-luminance qCSF (LL-qCSF) were examined using linear mixed-effects models with Bonferroni-Dunn correction.¹

What the data showed

The cohort had a median BCVA of 20/32 and a median GA lesion size of 3.76 mm², with a median of 2.6 weeks between test sessions. Mean MS across both visits was approximately 13.9 dB, with a clinically negligible between-session difference of −0.53 dB (P = .072).¹

MS demonstrated excellent repeatability, with a CoR of 2.42 dB and an ICC of 0.98—a threshold the authors note is meaningful because deviations beyond it are more likely to represent genuine functional shifts than instrument or testing variability. After excluding floor-effect loci, adjusted MS CoR improved further to 2.03 dB. PWS demonstrated moderate repeatability (CoR, 8.20 dB; ICC, 0.89), improving after floor-effect exclusion but with a corresponding drop in ICC to 0.79. Mean tracked test time was 4.33 minutes per eye, with fixation stability exceeding 89% across sessions. Eighty percent of test loci differed by ±4 dB or less between sessions.¹

Adjusted MS correlated significantly with LLVA (r=0.78; corrected P = .001) and LL-qCSF (r=0.76; corrected P = .01), but showed no significant association with BCVA. The correlation between GA lesion size and adjusted MS was moderate (r=−0.66; R²=0.44) but did not reach significance after multiple-comparison correction, which the authors attribute in part to limited statistical power rather than a true absence of relationship.¹

What the data cannot yet tell us

The 17-eye sample limits generalizability across GA phenotypes and left the study underpowered for structure-function analyses, with post hoc estimated statistical power of 0.56. No longitudinal data on sensitivity change over time were collected. The BCVA inclusion threshold of 20/80 or better may limit applicability to patients with more advanced central vision loss. qCSF and LL-qCSF data were unavailable for 2 eyes due to technical acquisition issues.¹

The case for a new functional end point in GA

Tandon and colleagues conclude that focal photopic microperimetry with the Nidek MP-3 and Goldmann II stimuli produced high test-retest repeatability for central cone-mediated sensitivity in nonsubfoveal GA, with short testing times and strong fixation stability across sessions.¹ The authors position MS as a dependable primary outcome measure for studies that track GA progression or evaluate treatment response over time, while noting that PWS—despite greater variability—may serve a complementary role in localized structure-function analyses.¹ The finding that foveal cone sensitivity was unrelated to BCVA and only weakly associated with lesion area points to a functional dimension of GA that neither structural imaging nor standard acuity testing is designed to capture, and which focal photopic microperimetry of the central 1 mm foveal region may be uniquely positioned to address.¹

References

1. Tandon NR, Vasquez E, Csaky KG. Test-retest repeatability of a novel photopic microperimetry approach in patients with geographic atrophy. Ophthalmol Sci. 2026. doi:10.1016/j.xops.2026.101267

2. Heier JS, Lad EM, Holz FG, et al. Pegcetacoplan for the treatment of geographic atrophy secondary to age-related macular degeneration (OAKS and DERBY): two multicentre, randomised, double-masked, sham-controlled, phase 3 trials. Lancet. 2023;402(10411):1434-1448. doi:10.1016/S0140-6736(23)01520-9

3. Khanani AM, Patel SS, Staurenghi G, et al. Efficacy and safety of avacincaptad pegol in patients with geographic atrophy (GATHER2): 12-month results from a randomised, double-masked, phase 3 trial. Lancet. 2023;402(10411):1449-1458. doi:10.1016/S0140-6736(23)01583-0