Abstract
Purpose
To study associations of optical coherence tomography (OCT) features with presenting visual acuity (VA) in treatment naive neovascular age-related macular degeneration (nAMD).
Methods
Patients with nAMD initiated on aflibercept therapy were recruited from December 2019 to August 2021. Demographic and OCT (Spectralis, Heidelberg Engineering) features associated with good VA (VA ≥ 68 ETDRS letters, Snellen ≥ 6/12) and poor VA (VA < 54 letters, Snellen < 6/18) were analysed using Generalised Estimating Equations to account for inter-eye correlation.
Results
Of 2274 eyes of 2128 patients enrolled, 2039 eyes of 1901 patients with complete data were analysed. Mean age was 79.4 (SD 7.8) years, female:male 3:2 and mean VA 58.0 (SD 14.5) letters. On multivariable analysis VA < 54 letters was associated with increased central subfield thickness (CST) (OR 1.40 per 100 µm; P < 0.001), foveal intraretinal fluid (OR 2.14; P < 0.001), polypoidal vasculopathy (PCV) relative to Type 1 macular neovascularisation (MNV) (OR 1.66; P = 0.049), presence of foveal subretinal hyperreflective material (SHRM) (OR 1.73; P = 0.002), foveal fibrosis (OR 3.85; P < 0.001), foveal atrophy (OR 5.54; P < 0.001), loss of integrity of the foveal ellipsoid zone (EZ) or external limiting membrane (ELM) relative to their preservation (OR 3.83; P < 0.001) and absence of subretinal drusenoid deposits (SDD) (presence vs absence; OR 0.75; P = 0.04). These features were associated with reduced odds of VA ≥ 68 letters except MNV subtypes and SDD.
Conclusion
Presence of baseline fovea-involving atrophy, fibrosis, intraretinal fluid, SHRM, PCV EZ/ELM loss and increased CST determine poor presenting VA. This highlights the need for early detection and treatment prior to structural changes that worsen baseline VA.
Subject terms: Predictive markers, Business and industry
Introduction
Visual impairment due to neovascular age-related macular degeneration (nAMD) remains a major public health problem [1]. Eyes with macular neovascularisation (MNV) due to nAMD present with a wide range of visual acuity (VA) loss. Understanding the association of MNV characteristics with presenting VA may provide more information as to why initial VA is the strongest predictor of final visual outcome at 12 months [2–4]. Initial VA can be broadly classified into good, moderate and poor VA groups. Good VA group included those with presenting VA ≥68 Early Treatment Diabetic Retinopathy Study (ETDRS) letters (Snellen equivalent ≥6/12) [5]. Moderate VA group consisted of those who presented with 67–54 (Snellen <6/12 -6/24) [6, 7] and poor VA group were those initial VA < 54 letters (Snellen <6/24). Those eyes with severe visual loss (worse than 24 ETDRS letters or Snellen 6/96) are unlikely to benefit from treatment due to structural damage at the fovea and are usually not initiated on treatment [7, 8]. Identifying anatomical characteristics of those with good and poor VA groups may identify the anatomical surrogates of presenting VA [9].
MNVs usually develop in a retinal background of drusen, subretinal drusenoid deposits (SDD) and/or macular atrophy [10, 11]. The most striking characteristic of an active MNV on optical coherence tomography (OCT) is fluid exudation that may present as intraretinal fluid (IRF) and/or subretinal fluid (SRF). Some may have exudation of subretinal hyperreflective material (SHRM), mostly due to haemorrhage or fibrin. However, despite manifest exudation, some active MNV lesions have minimal influence on VA. Only a quarter of eyes with new-onset exudation in the fellow eyes of unilateral nAMD in the HARBOR study had a vision loss of 5 or more letters [12]. Other signs include visible contiguous or non-contiguous retinal haemorrhage and pigment epithelial detachment. With time, secondary structural changes due to the MNV may also influence VA such as atrophy, fibrosis and outer retinal tubulations. These are indicators of poor visual acuity [13, 14].
A combination of these morphological changes may determine presenting VA [12, 15]. Identifying optical coherence tomography (OCT) characteristics that best explain the presenting VA may in future also help to individualise treatment options as investigational agents are being evaluated for fibrosis and atrophy.
As VA is a foveal function, classification of each OCT feature based on its location within and outside 1-mm of the fovea also provides further information. The Comparison of Age-related macular degeneration Treatment Trials (CATT) study group concentrated on foveal-involving OCT features to study association and prediction of visual outcome while the VIEW study group showed that extrafoveal subretinal fluid (SRF) may indeed be a visual prognostic indicator [16, 17]. Therefore, both fovea and non-fovea-involving OCT parameters need to be considered to explain presenting VA [18]. Studies on artificial intelligence have shown that baseline VA is the strongest predictor of VA outcomes following treatment of nAMD [19–23]. Therefore, systematic approach to grading all OCT features of newly diagnosed MNV may provide further insight into the determinants of presenting VA. The aim of this study was to identify baseline OCT features that best explain presenting VA in nAMD
Study design and participants
The PRECISE study was a 10-centre predictive modelling study on treatment response after loading phase of aflibercept therapy for nAMD (ISRCTN 28276860). This study was conducted in accordance with the Declaration of Helsinki. Institutional review board approval was obtained from the National Research Service (REC number 19/LO/1385). All patients provided written informed consent at each centre before enrolment into the study. Patients were recruited for this study from 18/12/2019 to 04/08/2021. Inclusion criteria included patients aged 50 years or above presenting with treatment naive MNV due to nAMD initiated on aflibercept. The OCT scans had to be captured on Spectralis OCT (Heidelberg Engineering GmbH, Heidelberg, Germany) and MNV was confirmed by clinicians at the local sites. Both eyes of an individual were recruited, if eligible. Patients could be recruited retrospectively or prospectively if these criteria were met. Exclusion Criteria included co-existent ocular disease that, in the opinion of the investigator, could affect or alter VA during the study, poor image quality and missing baseline OCT scans.
Data collection
Data collected for the study in a web-based database (Playon Ltd, Bengaluru, India) included age and visual acuity at date of first aflibercept injection for treatment naive MNV in the first eye, gender and ethnicity (White, Black, South Asian, other Asian and other). The Spectralis raster scans centred at the fovea (range 20° × 25° to 30° × 20°) with line scans ranging 19 to 49 (19, 25, 31 or 49) accompanied with infra-red imaging were anonymised at each site and transferred by encrypted USB to Moorfields Eye Hospital for grading.
OCT grading protocol
A pre-defined OCT grading manual was used to grade the OCT scans. The images were graded by four Medical Retina Fellows after grading on a test set of 50 OCT images. Each scan was corrected for any segmentation errors and foveal centration before central subfield thickness was recorded. For gradable scans, MNV features, if present were graded as including fovea or not. These included the location of MNV, intraretinal (IRF) or subretinal fluid (SRF), pigment epithelial detachment (PED), contiguous blood, subretinal hyperreflective material (SHRM), hyperreflective foci, outer retinal tubulations, fibrosis or atrophy. SHRM was defined as fluffy material, ill-defined and isoreflective to inner retina located between retina and retinal pigment epithelium (RPE) [24]. Fibrosis was defined as subretinal hyperreflective thickening, well defined and dense as RPE, better observed by reducing the contrast [24]. Presence of atrophy was defined as per Classification of Atrophy Meeting (CAM) definition (Zone of hyper transmission of ≥250 µm; Zone of attenuation or disruption of RPE band of ≥250 µm; evidence of overlying photoreceptor degeneration characterised by features that include outer nuclear layer (ONL) thinning, external limiting membrane (ELM) loss, and ellipsoid zone (EZ) or interdigitating zone (IZ) loss) [25]. The MNV subtypes on OCT were defined based on the Consensus on Neovascular Age-Related Macular Degeneration Nomenclature Study Group (CONAN) classification [24]. Type 1 MNV was defined as presence of fibrovascular PED with fluid, in the absence of SHRM. Type 2 MNV was defined as hyperreflective areas in contact with or in front of the RPE and the pathology may be dome-shaped or appear as a thin formation (fusiform or nodular). Type 3 or retinal angiomatous proliferation (RAP) was defined as PED with RPE erosion with an overlying hyperreflective oval and IRF and polypoidal vasculopathy (PCV) was defined as presence of notched/tall peaked/ thumb like PED with a double layer sign or haemorrhage with or without presence of sub RPE hyperreflective oval structure of a polyp [26]. Other non-MNV related parameters graded included presence of epiretinal membrane and other vitreomacular interface abnormalities and the integrity of the outer retinal layers. Presence of SDD had to be confirmed on both OCT and infra-red reflectance [27, 28]. EZ and ELM were deemed ungradable if non-visible due to presence of SHRM, large PED, back shadowing due to intraretinal haemorrhage, large locules of IRF or large volumes of SRF, back shadowing due to fibrosis. Foveal location of a parameter was defined as within the central 1 mm area around the foveal depression on the ETDRS grid and manually marked. Non-foveal location was defined as scan area beyond the central 1 mm zone within the ETDRS grid. In case of disagreement, the case was reviewed by the team of graders and lead retina specialist and a final grade ascertained.
Outcomes
The baseline OCT features associated with presenting VA ≥68 ETDRS letters (Snellen equivalent ≥6/12) and those with VA <54 letters (Snellen <6/18). Comparisons between VA 54-67 (Snellen ≥6/18 & <6/12) vs VA ≥68 and VA 54–67 (Snellen ≥6/18 & <6/12) vs VA < 54 letters were also conducted.
Statistical analysis
Data were summarised with mean ± SD or median (interquartile range [IQR]) for normally and non-normally distributed continuous variables, respectively and n (%) for categorical variables. Univariate and multivariable associations between MNV features and the categorical outcomes for visual acuity (VA ≥ 68 ETDRS letters and VA < 54 ETDRS letters) were reported using Odds Ratio (95% CI) and P-value. Features associated with VA 54–67 ETDRS letters vs VA ≥ 68 letters and VA 54–67 vs VA < 54 letters were also reported. Generalised Estimating Equations (GEE) with an exchangeable working correlation structure were used to account for the within-participant correlation among those with data from both eyes [29, 30]. There were a small number of patients categorised as Non-White, therefore the non-Whites were not further categorised into Black, South Asians, other Asians and other in models. P-values < 0.05 were considered statistically significant. Statistical analysis was undertaken using the R version 4.1.2 statistical software package [31].
Results
A total of 2274 eyes of 2128 patients were enrolled in the study including 138 (7.3%) patients with both eyes included. A total of 2039 eyes of 1901 patients with treatment naive MNV were included in the sample for analysis. Notably, 33 eyes were not analysed due to absence of evident MNV or non-MNV diagnosis. The study flow diagram from recruitment to analysis sample is shown in Fig. 1. Table 1 shows the demographic and OCT characteristics in the whole population and in the three baseline vision categories: VA ≥ 68; VA 54–67 and those with VA < 54 ETDRS letters. The mean age of the study participants was 79.4 (SD 7.8) years. The female-to-male ratio was 3:2 and the study population were predominantly of white ethnic background (1808 [95%]). The mean presenting visual acuity of the study population was 58.0 (SD 14.5) ETDRS letters. The proportions of eyes with initial VA of >70, 61–70, 51–60, 41–50, 31–40, 30 ETDRS letters were 342 (16.8%), 622 (30.5%), 488 (23.9%), 301 (14.8%), 213 (10.4%), 73 (3.6%) respectively. When we consider the MNV types, Type 1 MNV was more frequent in the group presenting with VA ≥ 68 ETDRS letters (VA ≥68 letters—335 [51.5%]; VA 54–67 ETDRS letters—282 [38.7%]; VA < 54 ETDRS letters—185 [28.0%]). In contrast, polypoidal vasculopathy (PCV) was more common in those with VA < 54 ETDRS letters (VA ≥ 68 ETDRS letters—25 [3.8%]; VA 54–67 ETDRS letters—40 [5.5%]; VA < 54 ETDRS letters; 63 [9.5%]). Of note, there were more patients with SRF (1688 [82.8%]) than IRF (1,031 [50.6%]) in the study population and this pattern of distribution was similar in all the three vision categories. The median CST was 504.0 µm (IQR 395.0–632.0 µm) in those presenting with VA < 54 ETDRS letters, 413 µm (IQR 339.0–500.0 µm) in those with VA 54–67 ETDRS letters and 362.5 µm (IQR 312.0–430.0 µm) in those with VA ≥ 68 ETDRS letters. There were twice as many non-foveal atrophy compared to foveal atrophy (292 [14.3%] vs 138 [6.8%]) in the cohort. 212 of the 266 (79.7%) that presented with subfoveal fibrosis at baseline, presented with VA < 54 ETDRS letters. When considering loss of integrity of ellipsoid layer, only 1190 (58.4%) were gradable at presentation, of which 386 (32.4%) had subfoveal loss of ellipsoid layer and 303/386 (78.5%) presented with VA < 54 ETDRS letters. Similarly, out of the 1,448 (71.0%) eyes with gradable ELM, 330 (22.8%) had subfoveal loss of ELM and 273/330 (82.7%) presented with VA < 54 ETDRS letters.
Fig. 1. Participant flow.
AMD age-related macular degeneration, Angioid streak CNV Angioid streak related choroidal neovascularization, AOFVD adult onset foveomacular vitelliform dystrophy, CSCR CNV central serous chorioretinopathy related choroidal neovascularization, CNV choroidal neovascularization, dB decibel, Mactel macular telangiectasia, Myopic CNV myopic choroidal neovascularization, PAMM paracentral acute middle maculopathy, PEVAC perifoveal exudative vascular anomalous complex, RAM retinal artery macroaneurysm, RVO retinal vein occlusion, V1 Visit 1, V4 Visit 4, µ microns.
Table 1.
Demographic, ocular and OCT characteristics of study participants overall and by visual acuity categories.
| Visual acuity, ETDRS letter score | ||||
|---|---|---|---|---|
| Variable | Overall, N = 2039 eyes of 1901 patients | <54 letters, N = 660 eyes of 644 patientsa | 54-67 letters, N = 729 eyes of 710 patientsa | ≥68 letters, N = 650 eyes of 629 patientsa |
| Patient level (N = 1901) | ||||
| Age, mean(SD) | 79.4 (7.8) | 80.3 (7.8) | 79.6 (7.6) | 78.3 (7.8) |
| Age, years | ||||
| <60 | 27 (1.4%) | 9 (1.4%) | 9 (1.3%) | 9 (1.4%) |
| 60–69 | 175 (9.2%) | 43 (6.7%) | 65 (9.2%) | 74 (11.8%) |
| 70–79 | 672 (35.3%) | 213 (33.1%) | 246 (34.6%) | 244 (38.8%) |
| ≥80 | 1027 (54.0%) | 379 (58.9%) | 390 (54.9%) | 302 (48.0%) |
| Gender | ||||
| Female | 1155 (60.8%) | 394 (61.2%) | 436 (61.4%) | 374 (59.5%) |
| Male | 746 (39.2%) | 250 (38.8%) | 274 (38.6%) | 255 (40.5%) |
| Ethnicity | ||||
| Black | 12 (0.6%) | 5 (0.8%) | 5 (0.7%) | 2 (0.3%) |
| Other | 56 (2.9%) | 22 (3.4%) | 18 (2.5%) | 17 (2.7%) |
| Other Asian | 9 (0.5%) | 2 (0.3%) | 5 (0.7%) | 2 (0.3%) |
| South Asian | 16 (0.8%) | 6 (0.9%) | 6 (0.8%) | 4 (0.6%) |
| White | 1808 (95.1%) | 609 (94.6%) | 676 (95.2%) | 604 (96.0%) |
| Number of patients with two eyes | 138 (7.3%) | 16 (2.5%) | 19 (2.7%) | 21 (3.3%) |
| Eye level (N = 2039) | ||||
| Location of MNV | ||||
| Subfoveal (central 1 mm region) | 1951 (95.7%) | 652 (98.8%) | 703 (96.4%) | 596 (91.7%) |
| Non-foveal location | 88 (4.3%) | 8 (1.2%) | 26 (3.6%) | 54 (8.3%) |
| MNV Type (CONAN OCT Classification) | ||||
| Type 1 | 802 (39.3%) | 185 (28.0%) | 282 (38.7%) | 335 (51.5%) |
| Type 2 and mixed | 667 (32.7%) | 262 (39.7%) | 230 (31.6%) | 175 (26.9%) |
| RAP | 442 (21.7%) | 150 (22.7%) | 177 (24.3%) | 115 (17.7%) |
| PCV | 128 (6.3%) | 63 (9.5%) | 40 (5.5%) | 25 (3.8%) |
| Subfoveal presence of any component of MNV complex | 1983 (97.3%) | 655 (99.2%) | 713 (97.8%) | 615 (94.6%) |
| Central subfield thickness in microns, Median (IQR) | 414.0 (339.5, 525.0) | 504.0 (395.0, 632.0) | 413.0 (339.0, 500.0) | 362.5 (312.0, 430.0) |
| Presence of Intraretinal fluid | ||||
| No | 1008 (49.4%) | 241 (36.5%) | 366 (50.2%) | 401 (61.7%) |
| Yes, foveal involving | 771 (37.8%) | 350 (53.0%) | 262 (35.9%) | 159 (24.5%) |
| Yes, non-foveal involving | 260 (12.8%) | 69 (10.5%) | 101 (13.9%) | 90 (13.8%) |
| Presence of subretinal fluid | ||||
| No | 351 (17.2%) | 118 (17.9%) | 136 (18.7%) | 97 (14.9%) |
| Yes, foveal involving | 973 (47.7%) | 262 (39.7%) | 357 (49.0%) | 354 (54.5%) |
| Yes, non-foveal involving | 715 (35.1%) | 280 (42.4%) | 236 (32.4%) | 199 (30.6%) |
| Presence of pigment epithelial detachment | ||||
| No | 103 (5.1%) | 46 (7.0%) | 37 (5.1%) | 20 (3.1%) |
| Yes, foveal involving | 1593 (78.1%) | 535 (81.1%) | 573 (78.6%) | 485 (74.6%) |
| Yes, non-foveal involving | 343 (16.8%) | 79 (12.0%) | 119 (16.3%) | 145 (22.3%) |
| Presence of atrophy | ||||
| No | 1609 (78.9%) | 485 (73.5%) | 586 (80.4%) | 538 (82.8%) |
| Yes, foveal involving | 138 (6.8%) | 109 (16.5%) | 22 (3.0%) | 7 (1.1%) |
| Yes, non-foveal involving | 292 (14.3%) | 66 (10.0%) | 121 (16.6%) | 105 (16.2%) |
| Presence of fibrosis | ||||
| No | 1740 (85.3%) | 438 (66.4%) | 667 (91.5%) | 635 (97.7%) |
| Yes, foveal involving | 266 (13.0%) | 212 (32.1%) | 47 (6.4%) | 7 (1.1%) |
| Yes, non-foveal involving | 33 (1.6%) | 10 (1.5%) | 15 (2.1%) | 8 (1.2%) |
| Presence of subretinal hyperreflective material | ||||
| No | 849 (41.6%) | 191 (28.9%) | 306 (42.0%) | 352 (54.2%) |
| Yes, foveal involving | 966 (47.4%) | 417 (63.2%) | 341 (46.8%) | 208 (32.0%) |
| Yes, non-foveal involving | 224 (11.0%) | 52 (7.9%) | 82 (11.2%) | 90 (13.8%) |
| Presence of outer retinal tubulation | ||||
| No | 1987 (97.4%) | 636 (96.4%) | 714 (97.9%) | 637 (98.0%) |
| Yes, foveal involving | 23 (1.1%) | 13 (2.0%) | 5 (0.7%) | 5 (0.8%) |
| Yes, non-foveal involving | 29 (1.4%) | 11 (1.7%) | 10 (1.4%) | 8 (1.2%) |
| Presence of subretinal drusenoid deposit | 612 (30.0%) | 166 (25.2%) | 235 (32.2%) | 211 (32.5%) |
| Presence of hyperreflective foci | 1431 (70.2%) | 471 (71.4%) | 517 (70.9%) | 443 (68.2%) |
| Loss of ellipsoid zone | ||||
| No | 591 (29.0%) | 79 (12.0%) | 220 (30.2%) | 292 (44.9%) |
| Yes, foveal involving | 386 (18.9%) | 303 (45.9%) | 70 (9.6%) | 13 (2.0%) |
| Yes, non-foveal involving | 213 (10.4%) | 29 (4.4%) | 92 (12.6%) | 92 (14.2%) |
| Ungradable | 849 (41.6%) | 249 (37.7%) | 347 (47.6%) | 253 (38.9%) |
| Loss of external limiting membrane | ||||
| No | 900 (44.1%) | 167 (25.3%) | 332 (45.5%) | 401 (61.7%) |
| Yes, foveal involving | 330 (16.2%) | 273 (41.4%) | 51 (7.0%) | 6 (0.9%) |
| Yes, non-foveal involving | 218 (10.7%) | 38 (5.8%) | 93 (12.8%) | 87 (13.4%) |
| Ungradable | 591 (29.0%) | 182 (27.6%) | 253 (34.7%) | 156 (24.0%) |
| EZ loss/ELM loss | ||||
| Neither foveal involving | 1073 (52.6%) | 175 (26.5%) | 414 (56.8%) | 484 (74.5%) |
| Either or both ELM and EZ loss foveal involving | 387 (19.0%) | 304 (46.1%) | 70 (9.6%) | 13 (2.0%) |
| Both ungradable | 579 (28.4%) | 181 (27.4%) | 245 (33.6%) | 153 (23.5%) |
| Presence of epiretinal membrane | ||||
| No | 1821 (89.3%) | 584 (88.5%) | 644 (88.3%) | 593 (91.2%) |
| Yes, foveal involving | 70 (3.4%) | 28 (4.2%) | 23 (3.2%) | 19 (2.9%) |
| Yes, non-foveal involving | 148 (7.3%) | 48 (7.3%) | 62 (8.5%) | 38 (5.8%) |
VA visual acuity, OR Odds Ratio, MNV macular neovascularisation, SD standard deviation, IQR interquartile range, PCV polypoidal vasculopathy, RAP retinal angiomatous proliferation, CONAN Consensus on neovascular AMD nomenclature.
aEyes from the same patient may be present in different visual acuity categories, hence the number of patients may not add up to the total of 1901 patients included in this analysis.
Associations of presenting visual acuity
Univariate and multivariable analysis of demographic and OCT characteristics were done to study the associations of VA ≥ 68 ETDRS letters and VA < 54 ETDRS letters at presentation (Fig. 2 and Table S1).
Fig. 2. Plot of Odds Ratio with 95% CIs showing ocular and OCT characteristics associated with VA ≥ 68 ETDRS letter score and VA < 54 ETDRS letter score—univariate and multivariable analysis using Generalised Estimating Equations (GEE).
OCT Optical coherence tomography, ETDRS Early Treatment Diabetic Retinopathy Study, GEE Generalised Estimating Equation, VA visual acuity, OR Odds Ratio, CI confidence interval, MNV macular neovascularisation, PCV polypoidal vasculopathy, RAP retinal angiomatous proliferation. The ratio axis is displayed on the logarithmic scale to provide a visual description of the uncertainty associated with each estimate. Reference categories were: a<60 years. bFemale. cNon-White (Black/South Asian/other Asian/other). dCNV fovea-involving. eType 1. fAbsent. gNeither fovea-involving.
None of the demographic factors were associated with presenting visual acuity.
Although non-foveal location was associated with both VA ≥ 68 and VA < 54 ETDRS letters on univariate analysis, multivariable analysis showed that MNV location was not a determinant of presenting visual acuity. PCV relative to Type 1 MNV were found to be present in eyes with VA < 54 ETDRS letters (multivariable OR 1.66 (1.00–2.76); P = 0.049).
Eyes with increased CST (OR 1.40 per 100 μm increase; 95% CI [1.30–1.51]; P < 0.001), foveal presence of IRF (OR 2.14 [95% CI 1.55–2.94]; P < 0.001), presence of fovea-involving SHRM (OR 1.73 [95% CI 1.21–2.46]; P = 0.002), foveal fibrosis (OR 3.85 [95% CI 2.34–6.34]; P < 0.001), foveal atrophy (OR 5.54 [95% CI 2.94–10.5]; P < 0.001), loss of integrity of the foveal ellipsoid or external limiting membrane (OR 3.83 [95% CI 2.34–6.27]; P < 0.001) and absence of subretinal drusenoid deposits (SDD present relative to absent; OR 0.75 [95% CI 0.57–0.99]; P = 0.04) were associated with presenting VA < 54 ETDRS letters. It is interesting to note that although presence of foveal SRF and PED were associated with reduced odds of VA < 54 ETDRS letters, they were not significant in multivariable analysis.
Increasing CST (OR 0.71 per 100 μm increase [95% CI 0.65–0.78]; P < 0.001) was associated with reduced odds of VA ≥ 68 ETDRS letter. Similarly, other features associated with reduced odds of VA ≥ 68 ETDRS letters were foveal IRF (OR 0.60 relative to IRF absent [95% CI 0.43–0.84]; P = 0.003), fovea-involving SHRM (OR 0.63 relative to SHRM absent [95% CI 0.46–0.88]; P = 0.007), fovea-involving fibrosis (OR 0.18 relative to fibrosis absent [95% CI 0.07–0.48]; P < 0.001), foveal atrophy (OR 0.27 relative to atrophy absent [95% CI 0.09–0.85]; P = 0.03), fovea-involving loss of ELM or EZ (OR 0.22 relative to preservation of fovea-involving ELM or EZ [95% CI 0.09–0.57]; P = 0.002) and ungradable OCT scan (OR 0.70 relative to preservation of fovea-involving ELM or EZ [95% CI 0.52–0.96]; P = 0.03).
The MNV features were compared in eyes with VA 54–67 ETDRS letters relative to VA ≥ 68 and VA < 54 ETDRS letters to assess the associations with VA 54–67 ETDRS letters (Table S2). In multivariable analysis, increasing CST, fovea-involving SHRM, fibrosis, EZ/ELM and non-fovea-involving epiretinal membrane were associated with VA 54–67 relative to VA ≥ 68 ETDRS letters. While, increasing CST, eyes with fovea-involving IRF, SHRM, atrophy, fibrosis and EZ/ELM were at reduced odds of VA 54-67 relative to VA < 54 ETDRS letters. RAP compared to Type 1 eyes on the other hand exhibited increased odds of presenting VA 54–67 relative to VA < 54 ETDRS letters (Table S2).
Discussion
This study on 2039 eyes with new-onset MNV due to nAMD in the UK highlight the improvement in mean presenting VA of eyes initiated on anti-VEGF therapy (58 ETDRS letters or approximate Snellen equivalent 6/24), with a third of patients presenting with 68 letters (approximate Snellen equivalent of 6/12 or better). This finding is a likely reflection of the improved fast-track referral and treatment services for this condition in the UK. The earlier studies done in 2010–2015 reported average baseline VA of <50 ETDRS letters. In 2016, the UK Aflibercept Users Group reported outcomes of the first cohort of patients with treatment naive initiated on aflibercept therapy and the mean baseline visual acuity was 53.7 (SD 0.4) ETDRS letters [32, 33].
The key finding of our study is that presenting VA can be partly explained by certain OCT features. Increased central subfield thickness, foveal presence of intraretinal fluid, fovea-involving SHRM, foveal atrophy or fibrosis and the loss of integrity of the ellipsoid layer or external limiting membrane at the fovea were associated with poor VA at presentation i.e., less than 54 letters (Snellen equivalent <6/18). Indeed, foveal fibrosis followed by foveal atrophy showed the strongest associations.
However, fovea-involving subretinal fluid or pigment epithelial detachment were not associated with poor presenting VA. This finding is in keeping with most previous reports that have shown that baseline presence of foveal SRF or PED do not influence final visual outcome at 12 months [34–36]. Therefore, baseline OCT features of foveal intraretinal fluid, fibrosis or atrophy in a treatment naive MNV provide significant information not only explaining poor presenting VA but also their poor visual outcome at 12 months. In fact, the recent analysis of the TENAYA and LUCERNE pooled data showed that those who required 8 weekly injections post-loading had poorer baseline VA compared to those who met the criteria for 12 or 16-weekly re-treatment (Lai et al. at the 22nd EURETINA Congress, Hamburg, Germany).
Surprisingly, we found that eyes with SDD were unlikely to present with poor VA. SDD is a known poor prognostic indicator for both geographic atrophy and nAMD. One explanation may be that eyes with SDD usually present with RAP and these MNVs develop outside the fovea [37]. There were 21.7% of eyes with RAP in this study cohort. As these lesions are bilateral, it is likely that these eyes also present earlier [38]. If they develop intraretinal fluid, these are usually located outside the fovea at presentation.
An eye with increased CST is likely to present with VA < 54 ETDRS letters and had reduced odds to present with VA ≥ 68 ETDRS letters, while foveal intraretinal fluid or fibrosis with fovea-involving external limiting layer or ellipsoid zone has increased likelihood of VA < 54 ETDRS letters. These observations validate the relation of foveal structural integrity with presenting visual acuity.
Other prognostic indicators of visual outcome at 12 months such as older age, MNV subtypes such as PCV and type 2 MNV, and outer retinal tubulations were not associated with presenting visual acuity in this study. Although PCV and type 2 MNV, foveal SHRM, fovea-involving MNV and ORT were more prevalent in people with poor presenting visual acuity and showed significant associations with poor vision on univariate analysis, these did not reach statistical significance on multivariable analysis. There were very few 8 (1.2%) non-fovea-involving MNV in eyes with VA < 54 ETDRS letters. Similarly, although presence of foveal SRF and PED were associated with reduced odds of VA < 54 ETDRS letters, they were not significant in multivariable analysis.
Our study has several strengths. Firstly, our study sample represents one of the largest cohorts of treatment naive nAMD patients. This multicentre study was done across 10 UK centres and so the study is generalisable to the UK as the care pathway and referral system are guided by National Institute of Health and Care Excellence (NICE) guidelines. Late presentations due to delayed diagnosis or referral have reduced over the last decade [33]. The MNV subtypes were defined by the CONAN classification and SDD was graded as present only if seen both on infra-red reflectance imaging and OCT. All images were graded by trained MR fellows and minimum intergrader agreement of >0.7. Statistical analysis was done to account for potential within-participant correlation among those who had both eyes included in the analysis.
However, our study has some limitations. Presenting visual acuity is not refracted visual acuity as this is not routine practice in the UK [33, 39]. The VA categories are pragmatically defined as the ICD-10 or WHO criteria of visual impairment do not apply to this study population because patients with presenting VA less than 24 ETDRS letters are usually not initiated on anti-VEGF treatment in the UK.
In conclusion, our study reports the baseline OCT features that determine different ranges of baseline VA in exudative nAMD. Poor presenting VA is associated with high CST, atrophy, fibrosis, intraretinal fluid, and ellipsoid zone/external limiting membrane (ELM) loss within the central 1-mm region. As these features form the surrogate for poor VA that eventually determine poor VA outcomes, we recommend early detection of nAMD through screening of second eyes of patients with unilateral nAMD.
Summary
What was known before
Visual impairment due to neovascular age-related macular degeneration (nAMD) remains a major public health problem and eyes with macular neovascularisation (MNV) due to nAMD present with a wide range of visual acuity (VA) loss. Visual acuity in these eyes is a foveal function and both foveal and non-foveal-involving OCT parameters need to be considered to explain presenting VA. Multiple studies have investigated biomarkers that predict final visual acuity in these eyes and also that baseline VA is the most important determinant of final VA. However what OCT features determine baseline VA remains elusive.
What this study adds
This study shows systematic approach to grading all OCT features of newly diagnosed MNV provides further insight into the determinants of presenting VA. This study identified presence of foveal-involving atrophy, fibrosis, intraretinal fluid, subretinal hyperreflctive material, polypoidal choroidal vasculopathy, ellipsoid zone/external limiting membrane loss and increased central subfield thickness as determinants of poor baseline VA. This also highlights the need for early detection and treatment prior to structural changes that worsen baseline VA.
Supplementary information
Table S1. Ocular and OCT characteristics associated with VA>=68 ETDRS letters and VA<54 ETDRS letters – Odds Ratios (95% CI) and P-values from univariate and multivariable analysis using Generalised E
Table S2. Ocular and OCT characteristics associated with moderate VA comparing moderate (VA 54-67) vs good VA (VA>=68) and moderate (VA 54-67) vs poor VA (VA<54) – univariate and multivariable analysi
Acknowledgements
The research was funded by Boehringer Ingelheim and supported by the NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology and the NIHR Moorfields Clinical Research Facility. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.
Author contributions
Conceptualization: ShC and SS; Data curation: ShC, ST, RMP, SC, and AM; Formal analysis: SG and SS; Funding acquisition: AGi, VC and SS; Investigation: ShC, SG and SS; Methodology: ShC, SG, and SS; Project administration: AGi. and SS; Resources: ShC, GM, BJB, IP, MM, ST, SC, RPM, AM, AK, JT, AGr, FG, RG, BP and SS; Supervision: AGi, VC and SS; Visualisation: SG and ShC; Writing – original draft: ShC, SG and SS; Writing - review & editing: AGi, VC and SS Review and approval of final manuscript: ShC, SG., AGi, VC, SS, GM, BJB, IP, MM, ST, RPM, SC, AM, AK, JT, AGr, CC, FG, RG, and BP.
Data availability
The anonymised PRECISE clinical database analysed during the current study is available from author SS on approval of a data sharing agreement. Sharing of retinal images requires patient consent and sponsor approval.
Competing interests
SS received consultancy fees from Bayer, Allergan, Novartis Pharma AG, Roche, Boehringer Ingelheim, Optos, Apellis, Oxurion, Oculis and Heidelberg Engineering. VC is an employee of Janssen R&D and previously of Boehringer Ingelheim. AGi is an employee of Boehringer Ingelheim. TCNY is an employee of Boehringer Ingelheim. BB is in the advisory board and received international conference attendance sponsored by Novartis and Bayer; GM has conducted consultancy-advisory boards for Novartis, Bayer and Allergan, received educational travel grants from Novartis, Bayer, Allergan; IP has received lecture fees from Allergan, Bayer, Heidelberg and Novartis, consultancy fees from Allergan, Alimera, Bayer and Novartis and travel fees from Allergan, Bayer and Novartis. FG has received honorarium for consultancy-advisory boards from Alimera, Allergan, Bayer, Novartis, Oxford BioElectronics, Roche; educational travel grants from Allergan, Bayer, Novartis. MM has received lecture and advisory board honoraria from Bayer and Novartis and an educational travel grant from Bayer. RG has conducted consultancy-advisory boards for Novartis, Bayer and Allergan, Alimera, Santen, received educational travel grants from Novartis, Bayer, Allergan, Heidelberg Engineering. JT is a consultant for Bayer and Novartis, received grant support from Bayer, Novartis and Heidelberg Engineering, and is involved in research for Allergan, Roche, Bayer, Novartis and Boehringer Ingelheim. AK received travel support from Novartis, Bayer, and Allergan, and speaker fees from Allergan and Bayer. Bishwanath Pal received travel support and received advisory boards honoraria from Novartis and Bayer. ShC, SG, ST, RMP, SC and AM have no financial disclosures. SS, FG, and ShC are members of the Eye editorial board.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41433-023-02769-5.
References
- 1.Rahman F, Zekite A, Bunce C, Jayaram H, Flanagan D. Recent trends in vision impairment certifications in England and Wales. Eye. 2020;34:1271–8. doi: 10.1038/s41433-020-0864-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ying GS, Maguire MG, Daniel E, Ferris FL, Jaffe GJ, Grunwald JE, et al. Association of baseline characteristics and early vision response with 2-year vision outcomes in the comparison of AMD treatments trials (CATT) Ophthalmology. 2015;122:2523–31.e1. doi: 10.1016/j.ophtha.2015.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ying GS, Huang J, Maguire MG, Jaffe GJ, Grunwald JE, Toth C, et al. Baseline predictors for one-year visual outcomes with ranibizumab or bevacizumab for neovascular age-related macular degeneration. Ophthalmology. 2013;120:122–9. doi: 10.1016/j.ophtha.2012.07.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ho AC, Albini TA, Brown DM, Boyer DS, Regillo CD, Heier JS. The potential importance of detection of neovascular age-related macular degeneration when visual acuity is relatively good. JAMA Ophthalmol. 2017;135:268–73. doi: 10.1001/jamaophthalmol.2016.5314. [DOI] [PubMed] [Google Scholar]
- 5.Diabetic Retinopathy Clinical Research, N. Wells JA, Glassman AR, Ayala AR, Jampol LM, Aiello LP, et al. Aflibercept, bevacizumab, or ranibizumab for diabetic macular edema. N Engl J Med. 2015;372:1193–203. doi: 10.1056/NEJMoa1414264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ho AC, Kleinman DM, Lum FC, Heier JS, Lindstrom RL, Orr SC, et al. Baseline visual acuity at wet AMD diagnosis predicts long-term vision outcomes: an analysis of the IRIS registry. Ophthalmic Surg Lasers Imaging Retina. 2020;51:633–9. doi: 10.3928/23258160-20201104-05. [DOI] [PubMed] [Google Scholar]
- 7.Chae B, Jung JJ, Mrejen S, Gallego-Pinazo R, Yannuzzi NA, Patel SN, et al. Baseline predictors for good versus poor visual outcomes in the treatment of neovascular age-related macular degeneration with intravitreal Anti-VEGF therapy. Invest Ophthalmol Vis Sci. 2015;56:5040–7. doi: 10.1167/iovs.15-16494. [DOI] [PubMed] [Google Scholar]
- 8.Rao P, Lum F, Wood K, Salman C, Burugapalli B, Hall R, et al. Real-world vision in age-related macular degeneration patients treated with single anti-VEGF Drug Type for 1 Year in the IRIS Registry. Ophthalmology. 2018;125:522–8. doi: 10.1016/j.ophtha.2017.10.010. [DOI] [PubMed] [Google Scholar]
- 9.Fragiotta S, Rossi T, Cutini A, Grenga PL, Vingolo EM. Predictive factors for development of neovascular age-related macular degeneration: a spectral-domain optical coherence tomography study. Retina. 2018;38:245–52. doi: 10.1097/IAE.0000000000001540. [DOI] [PubMed] [Google Scholar]
- 10.Naysan J, Jung JJ, Dansingani KK, Balaratnasingam C, Freund KB. Type 2 (Subretinal) neovascularization in age-related macular degeneration associated with pure reticular pseudodrusen phenotype. Retina. 2016;36:449–57. doi: 10.1097/IAE.0000000000000758. [DOI] [PubMed] [Google Scholar]
- 11.Schlanitz FG, Baumann B, Kundi M, Sacu S, Baratsits M, Scheschy U, et al. Drusen volume development over time and its relevance to the course of age-related macular degeneration. Br J Ophthalmol. 2017;101:198–203. doi: 10.1136/bjophthalmol-2016-308422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hu X, Waldstein SM, Klimscha S, Sadeghipour A, Bogunovic H, Gerendas BS, et al. Morphological and functional characteristics at the onset of exudative conversion in age-related macular degeneration. Retina. 2020;40:1070–8. doi: 10.1097/IAE.0000000000002531. [DOI] [PubMed] [Google Scholar]
- 13.Roberts PK, Schranz M, Motschi A, Desissaire S, Hacker V, Pircher M, et al. Baseline predictors for subretinal fibrosis in neovascular age-related macular degeneration. Sci Rep. 2022;12:88. doi: 10.1038/s41598-021-03716-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Foss A, Rotsos T, Empeslidis T, Chong V. Development of macular atrophy in patients with wet age-related macular degeneration receiving anti-VEGF Treatment. Ophthalmologica. 2022;245:204–17. doi: 10.1159/000520171. [DOI] [PubMed] [Google Scholar]
- 15.Schmidt-Erfurth U, Bogunovic H, Sadeghipour A, Schlegl T, Langs G, Gerendas BS, et al. Machine learning to analyze the prognostic value of current imaging biomarkers in neovascular age-related macular degeneration. Ophthalmol Retina. 2018;2:24–30. doi: 10.1016/j.oret.2017.03.015. [DOI] [PubMed] [Google Scholar]
- 16.Jaffe GJ, Ying GS, Toth CA, Daniel E, Grunwald JE, Martin DF, et al. Macular morphology and visual acuity in year five of the comparison of age-related macular degeneration treatments trials. Ophthalmology. 2019;126:252–60. doi: 10.1016/j.ophtha.2018.08.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Waldstein SM, Simader C, Staurenghi G, Chong NV, Mitchell P, Jaffe GJ, et al. Morphology and visual acuity in aflibercept and ranibizumab therapy for neovascular age-related macular degeneration in the VIEW Trials. Ophthalmology. 2016;123:1521–9. doi: 10.1016/j.ophtha.2016.03.037. [DOI] [PubMed] [Google Scholar]
- 18.Rim TH, Lee AY, Ting DS, Teo K, Betzler BK, Teo ZL, et al. Detection of features associated with neovascular age-related macular degeneration in ethnically distinct data sets by an optical coherence tomography: trained deep learning algorithm. Br J Ophthalmol. 2021;105:1133–9. doi: 10.1136/bjophthalmol-2020-316984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.von der Emde L, Pfau M, Dysli C, Thiele S, Möller PT, Lindner M, et al. Artificial intelligence for morphology-based function prediction in neovascular age-related macular degeneration. Sci Rep. 2019;9:11132. doi: 10.1038/s41598-019-47565-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Moraes G, Fu DJ, Wilson M, Khalid H, Wagner SK, Korot E, et al. Quantitative analysis of OCT for neovascular age-related macular degeneration using deep learning. Ophthalmology. 2021;128:693–705. doi: 10.1016/j.ophtha.2020.09.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Golbaz I, Ahlers C, Stock G, Schütze C, Schriefl S, Schlanitz F, et al. Quantification of the therapeutic response of intraretinal, subretinal, and subpigment epithelial compartments in exudative AMD during anti-VEGF therapy. Invest Ophthalmol Vis Sci. 2011;52:1599–605. doi: 10.1167/iovs.09-5018. [DOI] [PubMed] [Google Scholar]
- 22.Abbas A, O'Byrne C, Fu DJ, Moraes G, Balaskas K, Struyven R, et al. Evaluating an automated machine learning model that predicts visual acuity outcomes in patients with neovascular age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol. 2022;260:2461–73. doi: 10.1007/s00417-021-05544-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Yeh TC, Luo AC, Deng YS, Lee YH, Chen SJ, Chang PH, et al. Prediction of treatment outcome in neovascular age-related macular degeneration using a novel convolutional neural network. Sci Rep. 2022;12:5871. doi: 10.1038/s41598-022-09642-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Spaide RF, Jaffe GJ, Sarraf D, Freund KB, Sadda SR, Staurenghi G, et al. Consensus nomenclature for reporting neovascular age-related macular degeneration data: consensus on neovascular age-related macular degeneration nomenclature study group. Ophthalmology. 2020;127:616–36. doi: 10.1016/j.ophtha.2019.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sadda SR, Guymer R, Holz FG, Schmitz-Valckenberg S, Curcio CA, Bird AC, et al. Consensus definition for atrophy associated with age-related macular degeneration on OCT: classification of atrophy report 3. Ophthalmology. 2018;125:537–48. doi: 10.1016/j.ophtha.2017.09.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cheung CMG, Lai T, Teo K, Ruamviboonsuk P, Chen SJ, Kim JE, et al. Polypoidal choroidal vasculopathy: consensus nomenclature and non-indocyanine green angiograph diagnostic criteria from the Asia-Pacific Ocular Imaging Society PCV Workgroup. Ophthalmology. 2021;128:443–52. doi: 10.1016/j.ophtha.2020.08.006. [DOI] [PubMed] [Google Scholar]
- 27.Spaide RF, Ooto S, Curcio CA. Subretinal drusenoid deposits AKA pseudodrusen. Surv Ophthalmol. 2018;63:782–815. doi: 10.1016/j.survophthal.2018.05.005. [DOI] [PubMed] [Google Scholar]
- 28.Wu Z, Fletcher EL, Kumar H, Greferath U, Guymer RH. Reticular pseudodrusen: a critical phenotype in age-related macular degeneration. Prog Retin Eye Res. 2022;88:101017. doi: 10.1016/j.preteyeres.2021.101017. [DOI] [PubMed] [Google Scholar]
- 29.Hardin, JW, Hilbe JM, Generalized estimating equations. 1st ed. New York: Chapman and Hall/CRC; 2002.
- 30.Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22. doi: 10.1093/biomet/73.1.13. [DOI] [Google Scholar]
- 31.Team, RC, R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2013. http://www.R-project.org/.
- 32.Talks JS, Lotery AJ, Ghanchi F, Sivaprasad S, Johnston RL, Patel N, et al. First-year visual acuity outcomes of providing aflibercept according to the VIEW study protocol for age-related macular degeneration. Ophthalmology. 2016;123:337–43. doi: 10.1016/j.ophtha.2015.09.039. [DOI] [PubMed] [Google Scholar]
- 33.Mehta H, Kim LN, Mathis T, Zalmay P, Ghanchi F, Amoaku WM, et al. Trends in real-world neovascular AMD treatment outcomes in the UK. Clin Ophthalmol. 2020;14:3331–42. doi: 10.2147/OPTH.S275977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ohji M, Okada AA, Sasaki K, Moon SC, Machewitz T, Takahashi K, et al. Relationship between retinal fluid and visual acuity in patients with exudative age-related macular degeneration treated with intravitreal aflibercept using a treat-and-extend regimen: subgroup and post-hoc analyses from the ALTAIR study. Graefes Arch Clin Exp Ophthalmol. 2021;259:3637–47. doi: 10.1007/s00417-021-05293-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Guymer RH, Markey CM, McAllister IL, Gillies MC, Hunyor AP, Arnold JJ, et al. Tolerating subretinal fluid in neovascular age-related macular degeneration treated with ranibizumab using a treat-and-extend regimen: FLUID Study 24-month results. Ophthalmology. 2019;126:723–34. doi: 10.1016/j.ophtha.2018.11.025. [DOI] [PubMed] [Google Scholar]
- 36.Kim KT, Lee H, Kim JY, Lee S, Chae JB, Kim DY. Long-term visual/anatomic outcome in patients with fovea-involving fibrovascular pigment epithelium detachment presenting choroidal neovascularization on optical coherence tomography angiography. J Clin Med. 2020;9:1863. doi: 10.3390/jcm9061863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Haj Najeeb B, Deak G, Schmidt-Erfurth U, Gerendas BS. THE RAP STUDY, REPORT TWO: The regional distribution of macular neovascularization type 3, a novel insight into its etiology. Retina. 2020;40:2255–62. doi: 10.1097/IAE.0000000000002774. [DOI] [PubMed] [Google Scholar]
- 38.Haj Najeeb B, Schmidt-Erfurth U. Do patients with unilateral macular neovascularization type 3 need AREDS supplements to slow the progression to advanced age-related macular degeneration? Eye (Lond). 2023;37:1751–3. [DOI] [PMC free article] [PubMed]
- 39.Fasler K, Moraes G, Wagner S, Kortuem KU, Chopra R, Faes L, et al. One- and two-year visual outcomes from the Moorfields age-related macular degeneration database: a retrospective cohort study and an open science resource. BMJ Open. 2019;9:e027441. doi: 10.1136/bmjopen-2018-027441. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Ocular and OCT characteristics associated with VA>=68 ETDRS letters and VA<54 ETDRS letters – Odds Ratios (95% CI) and P-values from univariate and multivariable analysis using Generalised E
Table S2. Ocular and OCT characteristics associated with moderate VA comparing moderate (VA 54-67) vs good VA (VA>=68) and moderate (VA 54-67) vs poor VA (VA<54) – univariate and multivariable analysi
Data Availability Statement
The anonymised PRECISE clinical database analysed during the current study is available from author SS on approval of a data sharing agreement. Sharing of retinal images requires patient consent and sponsor approval.