Type 4 macular neovascularization in a consecutive age-related macular degeneration cohort: detected prevalence and multimodal imaging findings

Abstract

Background

To determine the detected prevalence of type 4 macular neovascularization (MNV) within a retrospectively screened age-related macular degeneration (AMD) cohort and to characterize its clinical, structural, and angiographic features.

Methods

This single-center retrospective cohort study included 4658 structural OCT examinations from 962 eyes of 500 patients with age-related macular degeneration (AMD). Eyes with structural OCT findings suggestive of type 4 MNV underwent multimodal evaluation via structural OCT and available OCT angiography (OCTA). Cases were categorized as definite, probable, or non–type 4 MNV according to predefined multimodal imaging criteria. Detected prevalence within the screened cohort was calculated at the patient and eye levels with exact 95% Clopper–Pearson confidence intervals (CIs).

Results

The detected prevalence of definite type 4 MNV within the entire consecutive AMD cohort was 2.00% (95% confidence interval [CI], 0.96–3.65%) at the patient level and 1.04% (95% CI, 0.50–1.90%) at the eye level. The hyperreflective oblique band (HOB), subretinal hyperreflective material (SHRM), and ellipsoid zone (EZ) loss were present in all the definite cases. Central posterior hyaloid fibrosis (CPHF) was observed in half of the definite cases, whereas the remaining eyes exhibited posterior hyaloid central hyperreflectivity (PHCH). By definition, the OCTA criterion of a continuous flow signal extending from the inner choroid to the preretinal space was fulfilled in all definite cases. During follow-up, HOB, CPHF, and PHCH remained stable, whereas exudative activity varied.

Conclusions

Within this consecutive AMD cohort, type 4 MNV was an uncommon phenotype characterized by recognizable structural and angiographic features. Structural OCT was used to identify eyes suspicious for type 4 MNV, whereas designation as definite type 4 MNV required demonstration on OCTA of the continuous flow signal specified in the predefined criteria. Our findings further characterize the clinical, structural, and angiographic features of type 4 MNV in this retrospective cohort.

Introduction

Age-related macular degeneration (AMD) is one of the leading causes of irreversible vision loss among older adults worldwide [1]. Approximately 85–90% of patients with AMD have the dry (atrophic) form, whereas 10–15% have the neovascular form [2]. Accurate anatomical classification of neovascular AMD (nAMD) is essential for predicting the disease course and guiding appropriate treatment strategies (3, 4). According to the current optical coherence tomography (OCT) consensus classification, macular neovascularization (MNV) is classified into type 1 (sub-RPE), type 2 (subretinal), and type 3 (intraretinal) lesions [3, 4].

However, with the widespread clinical use of optical coherence tomography angiography (OCTA) and high-resolution structural OCT, vitreomacular interface–related morphological features and complex neovascular architectures have become increasingly well characterized [5, 6]. Recently described by Chan et al., the type 4 MNV phenotype is characterized by a hyperreflective oblique band (HOB) extending from the subretinal lesion into the inner retina, central posterior hyaloid fibrosis (CPHF), and a continuous flow signal extending from the choroid to the preretinal space [7]. Furthermore, because of its preretinal neovascular extension, prominent fibroproliferative course, and overlapping features with type 3 MNV, it has been proposed as a distinct phenotype that may be difficult to differentiate within the current classification system [7, 8].

However, the available evidence on type 4 MNV remains limited and is based on only a small number of case series [7]. Notably, type 4 MNV has also recently been reported in nine eyes with advanced diabetic retinopathy without features of AMD, indicating that this morphological phenotype is not confined to AMD-associated disease [9]. The prevalence of this phenotype in a large, consecutive AMD cohort representative of real-world clinical practice remains unknown. In addition, it is unclear whether type 4 MNV represents an end-stage change secondary to chronic, inadequately treated neovascularization or a distinct MNV phenotype that can arise de novo in treatment-naïve eyes. Likewise, the relationship between the tractional or fibrotic bands observed on structural OCT and preretinal neovascular flow on OCTA remains incompletely characterized. Further characterization of these structural and flow features may improve the recognition of type 4 MNV and help distinguish potentially active exudative changes from advanced fibroproliferative alterations [10, 11].

The aim of this study was to determine the detected prevalence of the type 4 MNV phenotype recently described by Chan et al. [7] within a consecutive AMD cohort and to characterize its clinical, structural, and angiographic features.

Methods

Study design and patient selection

This single-center, observational, retrospective cohort study was conducted at the Retina Unit of the Department of Ophthalmology, Eskişehir Osmangazi University Faculty of Medicine, Eskişehir, Türkiye. The study protocol was approved by the Eskişehir Osmangazi University Non-Interventional Clinical Research Ethics Committee (approval no. 65; May 2, 2024).

To minimize selection bias, consecutive sampling was used. Consecutive patients followed at the retina unit with a diagnosis of AMD were screened retrospectively, beginning with the most recent follow-up visit (July 3, 2026) and proceeding backward in time until a cohort of 500 consecutive patients was assembled (February 2, 2024). A total of 38 eyes were excluded (Fig. 1). The remaining 4658 structural OCT images from 962 eyes were systematically screened for type 4 MNV, and eyes with suspicious findings were further evaluated. The sample size was not determined by a formal power calculation but was pragmatically defined to establish a sufficiently large cohort to enable the systematic evaluation of the detected prevalence and imaging characteristics of this newly described and uncommon phenotype. Because the cohort was assembled retrospectively on the basis of the most recent follow-up visit and all available historical imaging was subsequently reviewed, the resulting estimate represents the detected prevalence within the screened cohort rather than a cross-sectional point prevalence.

Fig. 1.

Study cohort selection and case-categorization workflow for type 4 MNV. AMD, age-related macular degeneration; OCT, optical coherence tomography; MNV, macular neovascularization

The cohort included all consecutive patients followed at the retina unit with a diagnosis of AMD and was not restricted to eyes with neovascular AMD (nAMD). As our institution is a tertiary referral retina center, a substantial proportion of patients are referred from other centers after a diagnosis of nAMD or initiation of anti-VEGF therapy, and previous diagnoses and treatment histories cannot always be reliably verified from institutional records alone. Consequently, some eyes with nAMD diagnosed at outside centers could be misclassified as having dry AMD if the classification relies solely on our institutional records. Therefore, the entire consecutive AMD cohort was used as the denominator for the prevalence calculations.

Although the cohort was assembled on the basis of the date of the most recent follow-up visit, all available clinical records, structural OCT imaging sessions, and OCTA images (when available) from each patient’s initial presentation to the final follow-up visit were reviewed. Consequently, the evaluated imaging sessions included examinations performed before the cohort assembly period. The follow-up duration was defined as the interval between the initial retina visit and the most recent follow-up visit, whereas a visit was defined as a structural OCT imaging session performed on a single date.

Inclusion and exclusion criteria

Inclusion criteria

• Diagnosis of AMD.

• Availability of at least one evaluable structural macular OCT imaging session

Exclusion criteria

Eyes meeting any of the following criteria were excluded:

• Type 2 macular telangiectasia (MacTel type 2), pathologic myopia, prosthetic eye, or pachychoroid spectrum disease.

• Structural OCT images that are unevaluable because of poor image quality (e.g., dense cataract, corneal opacity, dense vitreous opacity, vitreous hemorrhage, or imaging artifacts) or inadequate patient cooperation.

Data collection

Electronic medical records and available imaging data were reviewed retrospectively. For the entire study cohort, age, sex, date of the initial retina visit, date of the most recent follow-up visit, and follow-up duration were recorded. For all eyes categorized as definite, probable, or non–type 4 MNV, the affected eye, lens status, number of visits, AMD subtype in the fellow eye, history of anti-VEGF therapy, anti-VEGF agent, total number of intravitreal injections, best-corrected visual acuity, and history of vitreoretinal surgery (vitrectomy) were recorded. Age was calculated based on the date of the first suspicious structural OCT finding.

Best-corrected visual acuity data were obtained from the clinical records at the initial presentation and at the most recent follow-up visit using a Snellen chart. In eyes with low vision, visual acuity was recorded as counting fingers at 1–5 m. For statistical analysis and graphical presentation, all visual acuity measurements were converted to logMAR values via the standard logarithmic conversion for Snellen acuity and a commonly used clinical convention for counting fingers (counting fingers at X meters ≈ Snellen X/60; see Fig. 3).

Fig. 3.

Change in visual acuity (counting fingers and logMAR) from the initial presentation to the final follow-up visit in eyes with definite type 4 MNV (C1–C10) that had at least one follow-up visit (n = 8). The baseline value represents visual acuity at the patient’s initial presentation to the retina clinic rather than at the time of the type 4 MNV diagnosis. The duration displayed next to each case indicates the follow-up interval (months) between the initial and final examinations. The counting fingers values were converted to logMAR values according to the convention of X-meter counting fingers ≈ Snellen X/60 (1 m = 1.78, 2 m = 1.48, 3 m = 1.30, 4 m = 1.18, and 5 m = 1.08 logMAR)

Imaging protocol

All structural OCT images were acquired as part of routine clinical practice via a swept-source OCT device (DRI OCT Triton; Topcon Corporation, Tokyo, Japan). OCTA images were also obtained via the same device with a 6 × 6 mm scan protocol according to the routine clinical imaging protocol, with the built-in projection artifact removal (PAR) algorithm enabled. As part of the retrospective review, all available structural OCT imaging sessions for each patient were evaluated in chronological order. Volumetric OCT datasets and corresponding B-scan images from each session were systematically reviewed. Because of the retrospective study design, OCTA-based assessment relied on the available examinations, and no additional OCTA imaging could be performed for eyes without OCTA records.

Image evaluation

All structural OCT imaging sessions were systematically screened for features suggestive of type 4 MNV by two ophthalmology residents and two retina specialists. During this initial screening, only the imaging data were evaluated, and the clinical records were not reviewed. Images considered suspicious for type 4 MNV were subsequently reassessed independently by two retina specialists.

For each patient, structural OCT imaging sessions were reviewed in chronological order to identify the first visit showing structural findings suspicious for type 4 MNV. From this time point onward, the corresponding clinical records and available OCTA images were reviewed, and the eyes were categorized according to the predefined case-definition criteria described below (see Case definitions). Following the independent assessments, all suspicious cases were reviewed jointly by the two retina specialists to reach a consensus on the final case categorization, after which the study database was finalized.

Assessment of type 4 MNVs

The assessment of suspected type 4 MNV was based on the multimodal imaging criteria proposed by Chan et al. [7]. Structural OCT examinations were systematically evaluated for morphological features associated with type 4 MNV. Central posterior hyaloid fibrosis (CPHF) was assessed according to the structural OCT characteristics described by Khan et al. and Venkatesh et al. [5, 6].

Structural OCT examinations were systematically screened for the following morphological features suggestive of type 4 MNV:

• Type 1, type 2, or mixed type 1–type 2 MNV morphology.

• Subretinal hyperreflective material (SHRM).

• Hyperreflective oblique band (HOB).

• Structural changes suggestive of inner retinal or preretinal proliferation.

• Central posterior hyaloid fibrosis (CPHF).

In eyes suspected of having type 4 MNV, additional descriptive structural OCT findings, including vitreomacular adhesion/traction (VMA/VMT), ellipsoid zone (EZ) loss, intraretinal fluid (IRF), subretinal fluid (SRF), prechoroidal cleft, and the presence of hemorrhagic MNV at the initial presentation, were also recorded.

Case definitions

Eyes were categorized as definite type 4 MNV when all of the following criteria were fulfilled:

• MNV secondary to age-related macular degeneration (AMD);

• Structural OCT findings consistent with type 1, type 2, or mixed type 1–type 2 MNV morphology together with structural changes suggestive of inner retinal or preretinal proliferation;

• On OCTA, a continuous flow signal extending from the inner choroid to the preretinal space was observed.

In addition to these OCTA findings, supportive OCTA features, including type 1, type 2, or mixed MNV flow patterns; flow signals within the HOB; and flow signals within the CPHF, were also evaluated.

Eyes with structural OCT findings suggestive of type 4 MNV but without corresponding OCTA imaging at the same visit were categorized as probable type 4 MNV. Eyes with suspicious structural OCT findings in which a continuous flow signal extending from the inner choroid to the preretinal space could not be demonstrated on OCTA were categorized as non–type 4 MNVs.

Statistical analysis

The detected prevalence of definite type 4 MNV within the screened cohort and the corresponding expanded estimate including probable cases were calculated separately at the patient and eye levels. Exact 95% confidence intervals (CIs) were calculated via the Clopper–Pearson method in Python with the SciPy statistical library (version 1.15.3). Because no patient had bilateral definite or probable type 4 MNV, the assumption of independence was preserved for eye-level analyses. The demographic, clinical, and imaging characteristics of definite, probable, and non–type 4 MNV patients were summarized descriptively. Categorical variables are presented as numbers (percentages), and continuous variables are presented as medians (ranges). Owing to the limited number of type 4 MNV cases, no inferential statistical analyses were performed.

Results

A total of 4658 structural OCT images from 962 eyes of 500 consecutive patients were systematically reviewed. Based on this screening, 16 eyes with morphological findings suggestive of type 4 MNV were selected for further evaluation (Fig. 1). Of these, 10 (62.5%) were categorized as definite type 4 MNVs, 3 (18.8%) as probable type 4 MNVs, and 3 (18.8%) as non–type 4 MNVs. Each eye belonged to a different patient, and no bilateral involvement was identified. Cases were assigned consecutive identifiers (C, case): C1–C10 for the definite group, C11–C13 for the probable group, and C14–C16 for the non–type 4 MNV group. These identifiers are used throughout the manuscript.

Characteristics of the definite type 4 MNV group

The detected prevalence of definite type 4 MNV within the entire consecutive AMD cohort—not restricted to neovascular AMD—was 2.00% at the patient level (10/500; 95% Clopper–Pearson CI, 0.96%–3.65%) and 1.04% at the eye level (10/962; 95% CI, 0.50%–1.90%). When probable type 4 MNV cases (n = 3) were included, the expanded detected prevalence within the screened cohort was 2.60% at the patient level (13/500; 95% CI, 1.39%–4.41%) and 1.35% at the eye level (13/962; 95% CI, 0.72%–2.30%) (Fig. 2).

Fig. 2.

Forest plot showing the detected prevalence of type 4 MNV within the screened cohort at the patient and eye levels. Detected prevalence estimates and exact 95% Clopper–Pearson confidence intervals are presented for the definite and expanded (including probable cases) definitions. Filled circles represent definite estimates, whereas open squares represent expanded estimates

The demographic and clinical characteristics of the 10 definite type 4 MNV patients are summarized in Table 1. No surgical intervention was performed in any patient during follow-up.

Table 1.

Demographic and clinical characteristics of eyes with definite type 4 MNV (n = 10)

Variable	Value
Age at diagnosis, median (range), years	75.5 (55–89)
Sex, male/female, n	8/2
Affected eye, right/left, n	4/6
Lens status, phakic/pseudophakic, n	5/5
Number of visits, median (range)	3 (1–9)
Follow-up duration, median (range), months	14.1 (0–26.5)
Patients receiving anti-VEGF therapy, n (%)	5/10 (50%)
Anti-VEGF agent	Bevacizumab (5/5)
Number of injections, median (range)	3 (3–4)
Patients receiving anti-VEGF injections before the first suspicious OCT finding, n (%)	0/10 (0%)
Fellow-eye AMD subtype, dry/neovascular, n (%)	6/4 (60%/40%)

Visual acuity was evaluated in the eight patients with at least two visits by comparing the measurements obtained at the initial presentation and the final follow-up visit, irrespective of the date of the type 4 MNV diagnosis. Patients C3 (counting fingers at 1 m) and C8 (counting fingers at 2 m), who had only a single visit, were excluded from this analysis. The results are presented in Fig. 3. Visual acuity remained stable in most cases, with the greatest change observed in Patient C9 (from counting fingers at 2 m to counting fingers at 5 m).

Among the 10 eyes with definite type 4 MNVs, central posterior hyaloid fibrosis (CPHF) was identified in five eyes, whereas the remaining five exhibited a similar hyperreflective pattern with separation of the posterior hyaloid from the fovea. This latter finding was defined as a distinct structural OCT feature and termed posterior hyaloid central hyperreflectivity (PHCH). The structural OCT and OCTA findings at diagnosis are summarized in Table 2, and representative structural OCT images of CPHF and PHCH patients are shown in Fig. 4.

Table 2.

OCT and OCTA findings at diagnosis (n = 10)

Optical coherence tomography (OCT) findings	n (%)
Hyperreflective oblique band	10/10 (100%)
Subretinal hyperreflective material	10/10 (100%)
Ellipsoid zone loss	10/10 (100%)
Central posterior hyaloid fibrosis	5/10 (50%)
Posterior hyaloid central hyperreflectivity	5/10 (50%)
Vitreomacular adhesion/traction	5/10 (50%)
Intraretinal fluid	9/10 (90%)
Subretinal fluid	6/10 (60%)
Hemorrhagic macular neovascularization at diagnosis	2/10 (20%)
Prechoroidal cleft	2/10 (20%)
Optical coherence tomography angiography (OCTA) findings
Mixed type 1–type 2 MNV	10/10 (100%)
Flow signal within the hyperreflective oblique band	10/10 (100%)
Flow signal within the central posterior hyaloid fibrosis	5/5 (100%)
Preretinal neovascular extension	10/10 (100%)
Continuous flow signal extending from the inner choroid to the preretinal space	10/10 (100%)

Fig. 4.

Fundus photographs and corresponding structural OCT images of all eyes with definite type 4 MNVs (n = 10). Left panel (red asterisks): eyes with central posterior hyaloid fibrosis (CPHF; C1, C2, C3, C7, and C9). Right panel (yellow asterisks): eyes with posterior hyaloid central hyperreflectivity (PHCH; C4, C5, C6, C8, and C10). The white arrows indicate the hyperreflective oblique band (HOB). A disciform scar was present in the fundus photographs of all the eyes

On OCTA, a mixed type 1–type 2 MNV pattern, a flow signal within the hyperreflective oblique band (HOB), and preretinal neovascular extension were identified in all eyes (Table 2). Flow signals within central posterior hyaloid fibrosis (CPHF) were detected in all five eyes with CPHF (5/5, 100%). A continuous flow signal extending from the inner choroid to the preretinal space, which formed part of the predefined case definition for definite type 4 MNV, was observed in all definite cases (10/10, 100%) (Fig. 5).

Fig. 5.

Multimodal OCTA images from Patient C6. The upper and lower panels show images acquired during the same visit from two different scan lines through the lesion. (a–d) En face OCTA images at different segmentation levels: (a) superficial capillary plexus, (b) deep capillary plexus, (c) outer retina, and (d) choriocapillaris. (e) Cross-sectional OCTA, (f) en face structural OCT, and (g) color fundus photograph. The cross-sectional OCTA images revealed a continuous flow signal extending from the choriocapillaris through the outer and inner retina to the preretinal space

The follow-up findings of the eight eyes with at least one postdiagnosis visit (C1, C2, C4, C5, C6, C7, C9, and C10) are summarized in Table 3. The hyperreflective oblique band (HOB), the overall type 4 MNV phenotype, and the findings of central posterior hyaloid fibrosis (CPHF) and posterior hyaloid central hyperreflectivity (PHCH) were preserved in all eyes in which they were initially identified, whereas subretinal fluid (SRF) and intraretinal fluid (IRF) showed variable changes over time.

Table 3.

Follow-up findings in eyes with at least one postdiagnosis visit (n = 8)

Finding	Value
The hyperreflective oblique band was preserved	8/8
* The central posterior hyaloid fibrosis was preserved	4/4
* The posterior hyaloid central hyperreflectivity was preserved	4/4
Overall type 4 phenotype preserved	8/8
Persistent SRF (eyes with SRF at baseline)	2/5
Resolution of SRF (eyes with SRF at baseline)	3/5
New-onset SRF (eyes without SRF at baseline)	1/3
Persistent IRF (eyes with IRF at baseline)	6/8
Resolution of IRF (eyes with IRF at baseline)	2/8

*Only eyes with corresponding findings at baseline were included in the analysis

Abbreviations: IRF, intraretinal fluid; SRF, subretinal fluid

Comparison of probable type 4 MNV and non–type 4 MNV cases

The demographic, clinical, and structural OCT characteristics of the probable type 4 MNV (n = 3; C11–C13) and non–type 4 MNV (n = 3; C14–C16) groups are compared in Table 4. Unlike in the definite type 4 MNV group, anti-VEGF therapy was initiated before the first suspicious OCT finding in all eyes with probable type 4 MNV. Patient C13 received both bevacizumab and ranibizumab, and all three patients had dry AMD in the fellow eye. In contrast, none of the eyes in the non–type 4 MNV group, nor their fellow eyes, had a history of anti-VEGF treatment. The most notable structural OCT difference between the two groups was the absence of the hyperreflective oblique band (HOB) and central posterior hyaloid fibrosis (CPHF) in all eyes in the non–type 4 MNV group, whereas these findings were frequently observed in the probable type 4 MNV group (Table 4). Representative structural OCT images demonstrating HOB, CPHF, and posterior hyaloid central hyperreflectivity (PHCH) in the probable type 4 MNV group are shown in Fig. 6.

Table 4.

Demographic, clinical, and structural OCT characteristics of eyes categorized as probable type 4 MN’V (n = 3) and non–type 4 MNV (n = 3)

Variable	Probable type 4 MNV (n = 3)	Non–type 4 MNV (n = 3)
Age at the first suspicious OCT finding, median (range)	64 (61–74)	74 (63–84)
Sex, male/female	3/0	1/2
Affected eye, right/left	3/0	1/2
Lens status, phakic/pseudophakic	2/1	3/0
Number of visits, median (range)	10 (9–12)	1 (1–3)
Follow-up duration, median (range), months	27.9 (21–31.2)	0 (0–3.3)
Visual acuity at initial presentation	CF 2 m (1.48), CF 2 m (1.48), CF 1 m (1.78)	0.3** (0.52), CF 1 m (1.78), CF 4 m (1.18)
Visual acuity at final visit	CF 2 m (1.48), 0.1* (1.00), CF 1 m (1.78)	0.3** (0.52), CF 1 m (1.78), CF 4 m (1.18)
Patients receiving anti-VEGF therapy, n (%)	3/3	0/3
Number of injections, median (range)	6 (5–8)	–
Hyperreflective oblique band	3/3	0/3
Subretinal hyperreflective material	3/3	2/3
Ellipsoid zone loss	3/3	3/3
Central posterior hyaloid fibrosis	2/3	0/3
Posterior hyaloid central hyperreflectivity	1/3	3/3
Intraretinal fluid	2/3	3/3
Subretinal fluid	0/3	3/3
Vitreomacular adhesion/traction	2/3	2/3

*In Patient C12, visual acuity improved beyond the counting fingers threshold at the final visit and was therefore measured via a Snellen chart (0.1)

**In Patient C14, visual acuity was above the counting fingers threshold at baseline and was therefore measured via a Snellen chart (0.3)

CF: counting fingers

Fig. 6.

Color fundus photographs and corresponding structural OCT images of eyes with probable type 4 MNVs (C11–C13, n = 3). The white arrows indicate posterior hyaloid central hyperreflectivity (PHCH) in Patient C11 and central posterior hyaloid fibrosis (CPHF) in Cases C12 and C13. The magnified structural OCT images in the right panel demonstrate the hyperreflective oblique band (HOB) in all three eyes

The OCTA findings of the three eyes categorized as non–type 4 MNV were heterogeneous. Patients C14 and C15 presented flow signals consistent with type 1 MNVs only, whereas Patient C16 presented type 1, type 2, and mixed type 1–type 2 MNV patterns. However, none of the three eyes demonstrated preretinal neovascular extension or a continuous flow signal extending from the inner choroid to the preretinal space. Therefore, regardless of the OCTA flow pattern, none of the three eyes fulfilled the predefined criteria for definite type 4 MNV, and all were categorized as non–type 4 MNV.

Discussion

Systematic screening of 962 eyes from 500 consecutive patients with AMD yielded a detected prevalence of definite type 4 MNV of 2.00% at the patient level and 1.04% at the eye level. These findings indicate that type 4 MNV is an uncommon phenotype within the AMD spectrum (Figs. 1 and 2). To date, however, no previous study has specifically investigated the prevalence of type 4 MNV. In a cohort of 266 eyes with newly diagnosed neovascular AMD, Jung et al. classified lesions via fluorescein angiography and structural OCT as type 1 (39.9%), type 2 (9.0%), type 3 (34.2%), or mixed type 1–type 2 (16.9%) MNVs [12]. In our study, the detected prevalence of type 4 MNV was calculated using the entire consecutive AMD cohort as the denominator, without restricting the analysis to patients with neovascular AMD. Therefore, the resulting detected prevalence estimates are not directly comparable with studies reporting the distribution of MNV subtypes exclusively among eyes with neovascular AMD. Nevertheless, the rarity of type 4 MNVs in the literature and their limited systematic characterization highlight the value of our findings in expanding the current epidemiological understanding of this phenotype. Beyond estimating the detected prevalence, postdiagnosis follow-up was available for eight of the ten definite cases (Table 3), allowing a preliminary assessment of the temporal behavior of the recorded structural and exudative features.

The multimodal imaging characteristics of type 4 MNV, originally proposed by Chan et al. (7) as a novel subtype within the OCT-based MNV classification, were examined in our retrospective cohort. We applied the proposed criteria to categorize suspicious eyes, which showed structural and angiographic features broadly consistent with the original description. However, the current definition of type 4 MNV is based on a limited number of descriptive case series, and its diagnostic performance has not yet been evaluated in large prospective cohorts [7]. Therefore, we adopted a conservative case-definition approach and categorized only eyes fulfilling all predefined structural OCT and OCTA criteria as definite type 4 MNV. To the best of our knowledge, this is the first study to report the detected prevalence of type 4 MNV based on systematic screening of a consecutive AMD cohort.

Structural OCT was used to identify eyes with findings suspicious for type 4 MNV, whereas designation as definite type 4 MNV required demonstration on OCTA of continuous neovascular extension from the inner choroid to the preretinal space. This distinction reflects the predefined case definition: three eyes had suspicious structural OCT findings but were not designated as definite type 4 MNV because the required continuous neovascular extension was not demonstrated on the available OCTA images (Table 4). Because the OCTA finding was part of the predefined case definition for definite type 4 MNV, the present study was not designed to assess the diagnostic performance of OCTA.

In the original descriptive series, the hyperreflective oblique band (HOB) was present in all eyes, and its persistence even in eyes without central posterior hyaloid fibrosis (CPHF) suggested that it may represent one of the core morphological features of type 4 MNVs (7). Consistent with the original report, HOB was identified in all eyes (10/10) in our series. Moreover, the equal distributions of central posterior hyaloid fibrosis (CPHF) and posterior hyaloid central hyperreflectivity (PHCH) further support HOB as one of the most consistent structural OCT features of type 4 MNV.

In a previous case series describing CPHF, this finding was associated primarily with chronic, highly fibrotic type 2 MNV and was interpreted as a manifestation of the profibrotic influence of the neovascular complex on the posterior hyaloid [5]. Venkatesh et al. reported CPHF in 4% of eyes in an overall neovascular AMD cohort and in 6% of eyes within the predominantly scarred MNV subgroup [5]. In our series, CPHF was present in 50% of definite type 4 MNV cases (Table 2). This proportion is not directly comparable with the frequencies reported by Venkatesh et al. because their estimates were calculated in broader neovascular AMD populations, whereas our estimate was restricted to eyes with definite type 4 MNV [5]. Nevertheless, the frequent presence of CPHF in our series is compatible with the advanced fibroproliferative characteristics of this phenotype [5, 7]. However, the absence of CPHF in the remaining half of our definite cases indicates that this finding, although characteristic, was not included among the required criteria for definite type 4 MNV. Notably, posterior hyaloid central hyperreflectivity (PHCH) was identified in all definite type 4 MNV eyes without CPHF. According to the proposed mechanism, disruption of the retinal pigment epithelium barrier allows proinflammatory and profibrotic mediators to enter the subretinal space, thereby strengthening posterior hyaloid adhesion, whereas progressive subretinal fibrosis may further contribute to this process through mechanical traction [5, 6]. Although vitreomacular adhesion has been implicated in the development of neovascular AMD [13], our findings suggest that CPHF and PHCH may not simply represent sequential stages of the same pathological process but may instead reflect different underlying clinical contexts.

A male predominance was observed among the definite type 4 MNV cases in our series (8/10). The original description by Chan et al. [7] also included more men than women (7/11; 63.6% male). However, the small number of cases precludes any inference regarding a sex-related predisposition, and the present retrospective data cannot determine whether this distribution reflects a biological difference, variations in referral or presentation patterns, or chance.

The age distribution of the CPHF patients in our series was generally consistent with that reported by Venkatesh et al. (median age, 69 years [range, 54–89 years] vs. mean age, 72.76 ± 9.09 years, respectively) [5]. In contrast, the PHCH cases were clustered within a narrower and older age range (median age, 76 years; range, 75–79 years). Because the prevalence of posterior vitreous detachment (PVD) increases with advancing age, this age distribution may partly explain the predominance of PHCH in older patients [14]. The observed difference between CPHF and PHCH may therefore be more closely related to vitreous status at presentation than to progression along a single fibrotic pathway. These findings suggest that CPHF may develop in eyes in which the posterior hyaloid remains attached to the fovea, whereas PHCH may be observed during the more advanced stages of posterior vitreous detachment (PVD; Stages 3–4). In the PHCH group, Stage 3 PVD could be identified on structural OCT in two eyes. However, in complete PVD (stage 4), the detached posterior hyaloid may move away from the retina and beyond the imaging depth of structural OCT, limiting the reliable assessment of this stage via macula-centered structural OCT (14). This interpretation is further supported by the absence of reported posterior vitreous detachment in the CPHF cases described by Venkatesh et al. [5].

Clinically, visual acuity was already at the level of counting fingers in most eyes at initial presentation to the retina service, and ellipsoid zone loss was observed in all eyes (Table 2; Fig. 3), indicating that advanced structural and functional macular damage had already been established by the time type 4 MNV was diagnosed. None of the definite cases had documented anti-VEGF treatment before the first suspicious OCT finding in the records available at our institution (Table 1). Thus, whether anti-VEGF treatment had been administered before recognition of the phenotype could not be confirmed from the available institutional documentation. However, because most eyes already showed advanced structural and functional macular damage at presentation, undocumented treatment at another center, an unrecognized chronic neovascular process, and delayed clinical detection cannot be excluded. The retrospective records do not allow us to determine the precise onset of neovascularization or to establish the circumstances surrounding the absence of documented earlier treatment.

A notable observation was the apparent dissociation between the stability of the recorded structural features and the variability of exudative activity during follow-up. Among the eight definite cases with postdiagnosis follow-up, the HOB and overall type 4 MNV phenotype were preserved in all eight eyes, while CPHF and PHCH were each preserved in all four eyes in which they were initially identified. In contrast, SRF resolved in 3 of 5 eyes in which it was initially present and developed in 1 of 3 eyes without baseline SRF; IRF resolved in 2 of 8 eyes and persisted in 6 of 8 eyes (Table 3). Although these descriptive findings are based on a small cohort with variable follow-up durations and variable anti-VEGF treatment histories, they suggest that the recorded structural phenotype may remain relatively stable once established, whereas fluid status represents a more dynamic component of the clinical course. Visual acuity remained stable in most eyes throughout follow-up, with the only notable exception being Case C9, in which vision improved from counting fingers at 2 m to counting fingers at 5 m (Fig. 3). Despite this functional improvement, the HOB, CPHF, and overall type 4 MNV phenotypes remained structurally unchanged throughout the follow-up period (Table 3). This dissociation suggests that the observed visual improvement cannot be attributed to structural regression of the underlying neovascular complex and was more likely related to factors independent of the structural phenotype, such as changes in exudative activity. All 16 eyes identified during screening were from different patients, and no bilateral involvement was observed, indicating that the phenotype was unilateral in this cohort. However, given the limited sample size, no conclusion can be drawn regarding unilateral involvement as a characteristic feature of type 4 MNV.

Several methodological limitations should be considered when interpreting the findings of this study. First, the retrospective, single-center design may limit the generalizability of the results to other clinical populations. In addition, because the cohort was assembled retrospectively on the basis of the most recent follow-up visit and included variable follow-up durations, the incidence of type 4 MNV could not be assessed; therefore, only its detected prevalence within the screened cohort could be estimated. Furthermore, because the cohort comprised all patients followed with a diagnosis of AMD rather than being restricted to neovascular AMD (nAMD), the true prevalence of type 4 MNV within the nAMD subgroup cannot be determined directly from the present study. In addition, the retrospective records did not provide standardized information regarding symptom duration, referral delay, previous counseling, treatment received at other centers, or the reasons for the absence of documented earlier treatment; therefore, the pathway by which some eyes reached an advanced stage before recognition of the phenotype could not be established.

Because OCTA evaluates flow signals from volumetric scans via cross-sectional sampling, vascular segments with low-flow signals, particularly within densely fibrotic tissue, may not be visualized on every B-scan. Consequently, the absence of the characteristic OCTA flow pattern does not necessarily exclude the presence of a retinochoroidal anastomotic network. Therefore, the possibility of misclassification in some eyes because of the technical limitations of OCTA cannot be excluded [15]. In addition, because visual acuity was at the counting fingers level in most eyes, conversion to logMAR values was based on a widely used clinical convention (counting fingers at X meters ≈ Snellen X/60) rather than an empirically validated conversion method. Consequently, quantitative comparisons with other studies should be interpreted with caution. During image evaluation, all suspicious cases were independently reviewed by two retina specialists, followed by a consensus meeting to establish the final case categorization. Because the aim of the study was not to assess interobserver agreement, a formal agreement analysis (e.g., Cohen’s kappa coefficient) was not performed. Furthermore, as image evaluation was not masked, the possibility of observer bias cannot be completely excluded.

The present findings provide systematic estimates of the detected prevalence of type 4 MNV within a large consecutive AMD cohort and further characterize the structural and angiographic features of this phenotype. Our findings further suggest that CPHF and PHCH may represent distinct morphological manifestations influenced by vitreous status. Further multicenter, prospective studies incorporating standardized imaging protocols are needed to better define the location and clinical significance of type 4 MNVs within the spectrum of neovascular AMD. Such studies should estimate the frequency of this phenotype across different populations, further characterize its multimodal imaging features, clarify its natural history, and identify features that may help distinguish it from type 3 MNVs. In addition, further investigation is needed to determine which anatomical or exudative features are associated with a favorable response to anti-VEGF therapy and at what stage treatment may provide clinical benefit. Addressing these questions may help identify patients who are more likely to benefit from treatment and clarify the potential limitations of anti-VEGF therapy in eyes with advanced fibroproliferative changes.

Conclusions

In this large consecutive AMD cohort, the detected prevalence of definite type 4 MNV was 2.00% at the patient level and 1.04% at the eye level, increasing to 2.60% and 1.35%, respectively, when probable cases were included. Definite type 4 MNVs exhibit a recognizable morphological pattern characterized by the hyperreflective oblique band (HOB), subretinal hyperreflective material (SHRM), ellipsoid zone loss, and either central posterior hyaloid fibrosis (CPHF) or posterior hyaloid central hyperreflectivity (PHCH) on structural OCT. In this study, structural OCT was used to identify suspicious eyes, and a case was designated as definite type 4 MNV only when the predefined OCTA criterion of continuous neovascular extension from the choroid to the preretinal space was demonstrated.

The evaluation of suspicious cases further illustrates how the predefined multimodal imaging criteria were applied to categorize eyes as definite, probable, or non–type 4 MNV. During follow-up, the structural features remained largely unchanged, whereas exudative activity showed variable changes, and visual acuity remained stable in most eyes. Overall, our findings suggest that type 4 MNV is an uncommon phenotype within the spectrum of AMD, with recognizable structural and angiographic features. Further prospective studies incorporating serial multimodal imaging are warranted to clarify its natural history, pathogenesis, and clinical significance.

Abbreviations

AMD

Age-related macular degeneration

CF

Counting fingers

CPHF

Central posterior hyaloid fibrosis

EZ

Ellipsoid zone

HOB

Hyperreflective oblique band

IRF

Intraretinal fluid

MNV

Macular neovascularization

nAMD

Neovascular age-related macular degeneration

OCT

Optical coherence tomography

OCTA

Optical coherence tomography angiography

PHCH

Posterior hyaloid central hyperreflectivity

PVD

Posterior vitreous detachment

RPE

Retinal pigment epithelium

SHRM

Subretinal hyperreflective material

SRF

Subretinal fluid

VEGF

Vascular endothelial growth factor

VMA

Vitreomacular adhesion

VMT

Vitreomacular Traction

Author contributions

MDB designed the study. MDB, MKÖ, ET, and NE collected the data. MDB and NE analyzed the data. MKÖ and ET drafted the manuscript. MDB and NE revised the manuscript. All the authors read and approved the final manuscript.

Funding

This study did not receive any specific funding.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available because they contain potentially identifiable clinical information but are available from the corresponding author upon reasonable request.

Decleration

Ethics approval and consent to participate

This study was conducted in accordance with the tenets of the Declaration of Helsinki. The study protocol was approved by the Non-Interventional Clinical Research Ethics Committee of Eskişehir Osmangazi University (approval no. 65; May 2, 2024). Owing to the retrospective design of the study and the use of anonymized patient data, the requirement for informed consent was waived by the Ethics Committee.

Consent for publication

Not applicable.

Declaration of generative AI in the writing process

During the preparation of this work, the authors used ChatGPT (OpenAI) and Claude (Anthropic) to improve the language, readability, and overall clarity of the manuscript. After using these tools, the authors critically reviewed and edited the content as needed and took full responsibility for the content of the publication.

Competing interests

The authors declare no competing interests.