Structural-Functional Dissociation in Diabetic Retinopathy: New Longitudinal Insights on Perfusion and Sensitivity

---

1. Introduction: The Unsolved Mystery of Retinal Nonperfusion

Retinal capillary nonperfusion (CNP) remains a central pillar in the pathogenesis of sight-threatening diabetic retinopathy (DR), characterizing the transition toward proliferative DR (PDR) and driving the development of diabetic macular ischemia. However, the resulting clinical complications—from vitreous hemorrhage to tractional retinal detachment—manifest with a perplexing inconsistency among patients. Current therapeutic paradigms are notably constrained, lacking the capacity to reverse established ischemia or predictably prevent its onset. To address these clinical gaps, the Hamilton-Perais et al. study utilized ultra-widefield (UWF) technology to characterize the longitudinal, point-to-point relationship between retinal perfusion and functional sensitivity over a two-year follow-up period.

2. Study Architecture: Advanced Imaging and Functional Mapping

This prospective longitudinal cohort was conducted at the Belfast Health and Social Care Trust, Northern Ireland, involving 44 patients with a spectrum of disease ranging from moderate NPDR to less than high-risk PDR. The investigators employed a robust dual-modality approach:

• Structural Assessment: Ultra-widefield fluorescein angiography (UWF-FFA) using the Optos California system was performed to determine the Retinal Ischemic Index (RII) and precisely demarcate nonperfused regions.
• Functional Assessment: A MonCvONE perimeter was utilized to conduct 110° full-field projection perimetry, quantifying threshold sensitivity deficits (dB) across 57 distinct retinal points.

A hallmark of this investigation was the “Point-to-Point” topographical registration. As demonstrated in the following figure, perimetric sensitivity maps (represented by color-coded squares) were superimposed on graded angiograms (where nonperfused boundaries are demarcated in red). This rigorous alignment ensured that identical retinal coordinates were monitored for both structural and functional changes throughout the study.

3. Baseline Findings: The Perfusion-Sensitivity Association

Baseline data corroborated a significant association between structural perfusion and retinal function. Nonperfused retinal areas exhibited a mean sensitivity deficit of 11.8 dB, significantly higher than the 6.6 dB deficit observed in perfused regions (P < .001).

Perfusion Status	Normal Retinal Sensitivity (Deficit ≤5 dB)	Reduced Retinal Sensitivity (Deficit >5 dB)	Total Points <br>(% of Total Imaged)
Perfused	1,381 / 2,092 (66%)	711 / 2,092 (34%)	2,092 / 2,508 (83%)
Nonperfused	105 / 354 (30%)	249 / 354 (70%)	354 / 2,508 (14%)
Ungradable	—	—	62 / 2,508 (2.5%)
Total Studied (Gradable)	1,486 / 2,446 (61%)	960 / 2,446 (39%)	2,446 / 2,508 (97.5%)

Note: Across the 44 eyes evaluated at baseline, 62 of the 2,508 perimetric points (2.5%) were located in areas with eyelid, eyelash, or other transient artifacts and were thus ungradable perimetrically. Of the studied gradable points (2,446), a total of 1,486 (61%) demonstrated normal sensitivity, while 960 (39%) showed reduced sensitivity.

Multivariable analysis underscored that age was the only systemic variable positively correlated with functional deficit (estimate 0.2, P = .006). Notably, diabetes duration and HbA1C levels were not significantly associated with functional loss at baseline, suggesting that local retinal factors may predominate over systemic metrics in determining point-specific sensitivity.

4. The “Mismatch” Phenomenon: Function Beyond Perfusion

5. Longitudinal Evolution: Clinical Significant Functional Fluctuation

The two-year analysis revealed a paradoxical improvement in function despite the chronicity of the disease. While the structural perfusion status remained remarkably stable—with 97% of studied points showing no change in perfusion grade—functional sensitivity showed a progressive improvement (reduction in deficit) over time.

Rates of improvement were as follows:

• 1 Year: -0.20 to -0.28 dB/month.
• 2 Years: -0.16 to -0.34 dB/month.

This functional gain is largely attributed to a “learning effect” associated with repeated perimetric testing. Crucially, the most substantial improvements were observed in areas with the most severe baseline deficits. While biological compensatory mechanisms remain a possibility, these fluctuations demand caution in clinical trial interpretation. Despite these shifts, confirms a persistent positive correlation between the global Retinal Ischemic Index (RII) and the mean global sensitivity deficit.

6. Clinical Implications for the Retina Specialist

The data provide three critical takeaways for clinical practice and interventional trial design:

1. Endpoint Selection and the Learning Effect: Sensitivity deficits can improve by 1–3 dB over time due to perimetric experience. This must be meticulously accounted for in trials evaluating neuroprotective or revascularization therapies to avoid inflating treatment efficacy.
2. Predictive Complexity and Reliability: Perfusion status alone is an imperfect predictor of retinal function. Furthermore, researchers must note that in sensitivity analyses excluding unreliable tests (attention loss >33%), the association between RII and global mean sensitivity was no longer significant at two years. This highlights the need for rigorous “methodological scrutiny.”
3. Stability of CNP: The study corroborates the relative stability of capillary nonperfusion under observation; 97% of points remained unchanged over 24 months, even in a cohort with poor glycemic control (median HbA1C 9.1%).

7. Conclusion: Redefining the Ischemic Landscape

---

---