Fluorescein Angiography Leakage Patterns in Pachychoroid Neovasculopathy: Neovascularization or Central Serous Chorioretinopathy? Authors: Soon Il Choi, Chui Ming Gemmy Cheung, Won Ki Lee Published: 2026 ABSTRACT IntroductionThe aim of this work is to characterize fluorescein angiography (FA) leakage patterns in eyes with pachychoroid neovasculopathy (PNV) and subretinal fluid (SRF), and to evaluate their associations with clinical features and treatment response.MethodsRetrospective observational study of consecutive patients with PNV presenting with SRF on optical coherence tomography and a clearly visible macular neovascular network on optical coherence tomography angiography at a single center. FA findings, multimodal imaging features, and clinical characteristics were analyzed. FA leakage was classified as neovascular (NV)-pattern leakage, characterized by diffuse leakage; central serous chorioretinopathy (CSC)-pattern leakage, characterized by pinpoint leakage; or an undetermined pattern. Main outcome measures were the distribution of FA leakage patterns and their associations with clinical features and treatment response.ResultsSeventy-one eyes from 68 patients were included. Baseline FA revealed NV pattern, CSC pattern, and mixed-pattern leakage in 43 (60.56%), 19 (26.76%), and three eyes (4.23%), respectively, while six eyes (8.45%) were indeterminate. Younger age, greater SRF height, and neurosensory retinal sagging were significantly associated with the CSC- pattern group. Among eyes that received anti–vascular endothelial growth factor (VEGF) therapy, complete SRF resolution after three loading injections occurred in 65.85% of the NV pattern group and 27.27% of the CSC pattern group (odds ratio, 5.14; 95% confidence interval, 1.18–22.49; P = 0.037). Of the seven eyes in the CSC pattern group that were observed without anti-VEGF therapy, five (71.43%) showed spontaneous resolution. Ninety-five percent of CSC leakage points (20/21) were located along the NV margin. Leakage pattern changes were observed in 19% of repeated FA examinations during follow-up.ConclusionsIn the pachychoroid context, the presence of neovascularization does not necessarily indicate that accumulated fluid originates from the neovascular complex. Combined CSC and nonexudative PNV appears relatively common, with CSC leakage points typically located along the NV margin. Given its frequency, the risk of neovascular activation, and the recurrent nature of CSC, FA remains valuable for identifying the origin of SRF when indicated. Key Summary Points Why carry out this study? In clinical practice, subretinal fluid (SRF) in pachychoroid neovasculopathy (PNV) with clearly identifiable neovascularization on optical coherence tomography angiography (OCTA) often leads to anti-vascular endothelial growth factor (VEGF) therapy.However, SRF may not always reflect neovascular activity, underscoring the value of fluorescein angiography (FA) leakage pattern analysis in determining the origin of SRF. What was learned from the study? A substantial proportion of eyes with PNV and SRF exhibited central serous chorioretinopathy (CSC)-pattern leakage on baseline FA.Eyes with CSC-pattern leakage had greater SRF height on OCT and a lower rate of complete fluid resolution after anti-VEGF loading, consistent with typical CSC.FA leakage patterns could change over time within the same eye, and CSC leakage points were predominantly observed along the margin of the neovascular complex, suggesting that neovascularization itself may give rise to CSC in PNV. Why carry out this study? In clinical practice, subretinal fluid (SRF) in pachychoroid neovasculopathy (PNV) with clearly identifiable neovascularization on optical coherence tomography angiography (OCTA) often leads to anti-vascular endothelial growth factor (VEGF) therapy.However, SRF may not always reflect neovascular activity, underscoring the value of fluorescein angiography (FA) leakage pattern analysis in determining the origin of SRF. What was learned from the study? A substantial proportion of eyes with PNV and SRF exhibited central serous chorioretinopathy (CSC)-pattern leakage on baseline FA.Eyes with CSC-pattern leakage had greater SRF height on OCT and a lower rate of complete fluid resolution after anti-VEGF loading, consistent with typical CSC.FA leakage patterns could change over time within the same eye, and CSC leakage points were predominantly observed along the margin of the neovascular complex, suggesting that neovascularization itself may give rise to CSC in PNV. Introduction Pachychoroid neovasculopathy (PNV) represents type 1 macular neovascularization (MNV) arising in eyes with a pachychoroid phenotype, characterized by choroidal thickening, dilated Haller’s layer vessels, and attenuation of the inner choroid, in the absence of typical age-related macular degeneration (AMD) signs [1]. On optical coherence tomography (OCT), a shallow irregular RPE elevation (SIRE), also referred to as the “double-layer sign,” is a characteristic finding [2]. The SIRE harboring type 1 MNV typically corresponds spatially to areas exhibiting pachychoroid features, particularly dilated Haller vessels and inner choroidal thinning. Since the advent of optical coherence tomography angiography (OCTA), there has been growing recognition of this distinct disease entity, which has been reported with notably high frequency in Asian populations [3]. The diagnosis of PNV can be confirmed when a tangled vascular network corresponding to a SIRE is identified on OCTA, where the neovascularization (NV) is often more clearly visualized than in typical neovascular AMD. As it develops gradually and insidiously, PNV is often identified as a relatively mature, planar vascular network at the time of detection. In addition, in the absence of OCT signal-confounding deposits such as drusen, OCTA provides high sensitivity and specificity for detecting NV. [4] Despite these advances, interpreting exudative changes in the pachychoroid context remains challenging. The presence of subretinal fluid (SRF) does not always indicate active neovascular exudation, as it may also result from choroidal hyperpermeability and dysfunction of retinal pigment epithelium (RPE) associated with central serous chorioretinopathy (CSC) in eyes harboring a non-exudative PNV [1]. Fluorescein angiography (FA), by depicting dynamic leakage patterns, helps determine the origin of SRF and thereby facilitates more tailored and effective management of pachychoroid-associated exudations. This study aimed to characterize FA leakage patterns in eyes with PNV presenting with SRF, and to determine whether SRF originated from neovascular activity or non-neovascular, CSC-related mechanisms. We also examined the correlations of FA leakage patterns with structural OCT and clinical features, as well as longitudinal changes in FA findings on serial examinations. Methods This retrospective study was conducted using the medical records of consecutive patients diagnosed with PNV who visited our institution between January 2019 and March 2022. Data collected included age, sex, history of CSC, type of anti-vascular endothelial growth factor (VEGF) agent used, and follow-up duration, as well as the number of dye-based angiographic examinations and additional treatments performed during follow-up. Clinical assessments included best-corrected visual acuity (BCVA), slit-lamp biomicroscopy, and dilated fundus examination. All patients underwent multimodal imaging at each visit, including color fundus photography (CR-2, Canon Inc., Tokyo, Japan), ultra-widefield imaging (California and Optos 200Tx, Optos PLC, Dunfermline, United Kingdom), OCT (Spectralis HRA-OCT, Heidelberg Engineering, Heidelberg, Germany), and OCTA (Plex Elite 9000, Carl Zeiss Meditec, Oberkochen, Germany). FA, indocyanine green angiography (ICGA), and fundus autofluorescence (Spectralis HRA-OCT, Heidelberg Engineering) were performed at baseline and as needed. Eyes were included based on multimodal imaging findings and included in this study if all of the following criteria were met: 1. SRF associated with a SIRE on OCT. 2. Dilated outer choroidal vessels (pachyvessels) accompanied by attenuation of the inner choroid and/or focal or diffuse choroidal thickening (no lower limit was defined for subfoveal choroidal thickness). 3. Absence or minimal presence of drusen or pseudodrusen, except for pachydrusen. 4. Presence of type 1 MNV identified on OCTA. Eyes were excluded if they had received prior anti-VEGF, photodynamic therapy (PDT), or focal laser treatment; had a follow-up of less than 1 year; or showed polypoidal lesions on baseline or follow-up imaging. OCTA analysis involved the examination of B-scans and a 6 × 6 mm macular cube to detect NV (Fig. 1A, D, G, and J). When automated segmentation did not reveal NV clearly, manual adjustment of the boundary slab between RPE and Bruch's membrane was performed. FA images were analyzed during an early phase (≤ 2 min after dye injection) and a late phase (≥ 6 min). Diffuse leakage or stippled early hyperfluorescence with progressive leakage was classified as the NV pattern (Fig. 1B, C). Focal leakage patterns, such as smoke-stack, ink-blot, pinpoint leakage, or small pigment epithelial detachment (PED) leakage were classified as the CSC pattern (Fig. 1E, F). Leakage patterns that could not be clearly assigned to either category were classified as undetermined (Fig. 1H, I). When NV and CSC patterns were observed concurrently, the eye was classified as having both patterns (Fig. 1K, L). If FA was repeated during follow-up, it was evaluated in the same manner and compared with the previous FA to assess any changes. All FA images were independently graded by a single retina specialist (Dr. W.K.L.), who was masked to the clinical outcomes during image assessment. From the B-scan OCT images, the maximum height of SRF was measured. The presence of neurosensory retinal sagging (Fig. 2A), hyperreflective bands in the subretinal space (Fig. 2B), and RPE defects with choroidal hypertransmission (Fig. 2C) was assessed at baseline [5].Fig. 1Representative baseline FA leakage patterns. A–C A vascular network on OCTA (A) shows spatially correlated diffuse leakage on FA (B, C), consistent with a NV leakage pattern. D–F OCTA (D) demonstrates NV, and FA (E, F) shows focal leakage pattern. A window defect is also noted in the corresponding area of the inferior portion of the neovascularization. Leakage originating from such a focal point is classified as a CSC leakage pattern. G–I Leakage patterns resembling focal CSC-type leakage but lacking definitive features or clear distinction are classified as undetermined. J–L Concurrent CSC and NV leakage patterns are classified as both patternsFig. 2Baseline OCT findings. A Sagging of the neurosensory retina (arrowheads). B A hyperreflective band within the subretinal space (arrowheads). C A focal retinal pigment epithelium (RPE) defect with associated choroidal hypertransmission (arrowheads) Representative baseline FA leakage patterns. A–C A vascular network on OCTA (A) shows spatially correlated diffuse leakage on FA (B, C), consistent with a NV leakage pattern. D–F OCTA (D) demonstrates NV, and FA (E, F) shows focal leakage pattern. A window defect is also noted in the corresponding area of the inferior portion of the neovascularization. Leakage originating from such a focal point is classified as a CSC leakage pattern. G–I Leakage patterns resembling focal CSC-type leakage but lacking definitive features or clear distinction are classified as undetermined. J–L Concurrent CSC and NV leakage patterns are classified as both patterns Baseline OCT findings. A Sagging of the neurosensory retina (arrowheads). B A hyperreflective band within the subretinal space (arrowheads). C A focal retinal pigment epithelium (RPE) defect with associated choroidal hypertransmission (arrowheads) When intravitreal anti-VEGF treatment was initiated at baseline, a loading series of three-monthly injections was administered, and treatment response was evaluated one month after the final injection. For eyes initially managed with observation, SRF changes were assessed at 3 months. This retrospective observational study adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Nune Eye Hospital (IRB approval number: N-2309-001-999). The requirement for informed consent was waived due to the retrospective nature of the study. Statistical analysis was conducted using SPSS software (version 26.0; IBM Corp., Armonk, NY, USA), and a P value of < 0.05 was considered statistically significant. The Mann–Whitney U test was used for continuous variables. For categorical variables, Fisher’s exact test or chi-square test was used, depending on the expected frequency in each cell. Results A total of 71 eyes from 68 patients (41 males) were included, with a mean age of 59.19 ± 10.09 years (range 39–83 years). Eighteen eyes (25.35%) had a documented history of CSC. The mean duration of follow-up was 38.23 ± 17.21 months (12–71 months). All patients diagnosed with PNV at the baseline visit had SRF on OCT, neovascular network on OCTA and FA images were obtained at that time. Among the 71 eyes, 43 eyes (60.56%) demonstrated an NV-pattern leakage. Nineteen eyes (26.76%) exhibited a CSC-pattern leakage. In three eyes (4.23%), both patterns coexisted, with CSC leaking points at different locations, whereas the leakage pattern could not be determined precisely in the remaining six eyes (8.45%). Representative examples of each FA leakage pattern are shown in Fig. 1. According to the baseline FA leakage pattern, eyes were divided into two groups: the NV-pattern group (43 eyes) and the CSC-pattern group (19 eyes). Baseline characteristics are summarized in Table 1. Patients in the NV-pattern group were significantly older (P < 0.001). Baseline OCT findings (Fig. 2, Table 2) showed that the maximum SRF height was significantly greater in the CSC-pattern group (P < 0.001), whereas subfoveal choroidal thickness (SFCT) did not differ significantly. Neurosensory retinal sagging was more frequent in the CSC-pattern group (P = 0.024), while hyperreflective bands and focal RPE defects showed a nonsignificant trend toward higher prevalence in this group.Table 1Baseline characteristics according to FA leakage patternCharacteristicsNV pattern group (n = 43 eyes)CSC pattern group (n = 19 eyes)P valueNumber of patients4119Sex distribution (male/female)30/1111/8Age, years (mean ± SD)62.84 ± 9.3451.32 ± 6.63 < 0.001History of CSC, n (%)10 (24.39%)8 (42.10%)0.144logMAR BCVA (mean ± SD)0.16 ± 0.180.10 ± 0.130.165FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, BCVA best-corrected visual acuity, SD standard deviationTable 2Baseline OCT findings according to FA leakage patternOCT findingsNV pattern group (n = 43 eyes)CSC pattern group (n = 19 eyes)P valueMaximum SRF height, µm (mean ± SD)169.72 ± 110.03298.21 ± 143.05 < 0.001SFCT, µm: mean ± SD (range)401.00 ± 142.20 (215–629)396.47 ± 102.96 (226–619)0.906Sagging of neurosensory retina, n (%)2/43 (4.65%)5/19 (26.32%)0.024Hyperreflective band in subretinal space, n (%)5/43 (11.63%)6/19 (31.58%)0.077Focal RPE defect with hypertransmission, n (%)6/43 (13.95%)7/19 (36.84%)0.087OCT optical coherence tomography, FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, SRF subretinal fluid, SFCT subfoveal choroidal thickness, RPE retinal pigment epithelium, SD standard deviation Table 1Baseline characteristics according to FA leakage patternCharacteristicsNV pattern group (n = 43 eyes)CSC pattern group (n = 19 eyes)P valueNumber of patients4119Sex distribution (male/female)30/1111/8Age, years (mean ± SD)62.84 ± 9.3451.32 ± 6.63 < 0.001History of CSC, n (%)10 (24.39%)8 (42.10%)0.144logMAR BCVA (mean ± SD)0.16 ± 0.180.10 ± 0.130.165FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, BCVA best-corrected visual acuity, SD standard deviation Baseline characteristics according to FA leakage pattern FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, BCVA best-corrected visual acuity, SD standard deviation Table 2Baseline OCT findings according to FA leakage patternOCT findingsNV pattern group (n = 43 eyes)CSC pattern group (n = 19 eyes)P valueMaximum SRF height, µm (mean ± SD)169.72 ± 110.03298.21 ± 143.05 < 0.001SFCT, µm: mean ± SD (range)401.00 ± 142.20 (215–629)396.47 ± 102.96 (226–619)0.906Sagging of neurosensory retina, n (%)2/43 (4.65%)5/19 (26.32%)0.024Hyperreflective band in subretinal space, n (%)5/43 (11.63%)6/19 (31.58%)0.077Focal RPE defect with hypertransmission, n (%)6/43 (13.95%)7/19 (36.84%)0.087OCT optical coherence tomography, FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, SRF subretinal fluid, SFCT subfoveal choroidal thickness, RPE retinal pigment epithelium, SD standard deviation Baseline OCT findings according to FA leakage pattern OCT optical coherence tomography, FA fluorescein angiography, NV neovascularization, CSC central serous chorioretinopathy, SRF subretinal fluid, SFCT subfoveal choroidal thickness, RPE retinal pigment epithelium, SD standard deviation Among the 71 included eyes, 54 received three consecutive monthly anti-VEGF injections (including 43 in the NV-pattern group and 11 in the CSC-pattern group): aflibercept in 29 eyes and bevacizumab in 12 eyes. In the CSC-pattern group, 11 eyes were treated with anti-VEGF injections: aflibercept in six eyes, bevacizumab in four eyes, and brolucizumab in one eye, following the same treatment regimen as the NV-pattern group. At month 3, complete resolution of SRF was significantly more common in the NV-pattern group (65.85% vs. 27.27%, P = 0.037). In the CSC-pattern group, among the seven eyes observed without treatment at baseline, spontaneous resolution of SRF occurred in five eyes (71.43%) within 3 months. Thermal laser treatment was applied to one eye, and PDT was later performed in four eyes, all achieving complete fluid resolution. During follow-up, FA was performed two or more times in 47 eyes (66.20%), primarily because of recurrent fluid or suboptimal treatment response. Twenty-six eyes underwent two examinations, 17 underwent three, and four underwent four, yielding a total of 73 follow-up FAs. Each examination was evaluated for the leakage pattern and its changes from the previous examination. Among the 73 examinations, persistent NV-pattern leakage and persistent CSC-pattern leakage were observed in 34 and 13 examinations, respectively. A change in leakage pattern was identified in 14 examinations (19.18%), including six shifts from NV to CSC and eight from CSC to NV. When analyzed per eye, eight of the 47 eyes (17.02%) with serial FA exhibited a change in leakage pattern during follow-up (Fig. 3).Fig. 3CSC leakage at baseline and NV leakage on follow-up FA. A 63-year-old woman presents with decreased vision in the right eye. A–C At baseline, OCT shows shallow subretinal fluid (SRF), and OCT angiography demonstrates a neovascular lesion corresponding to diffuse leakage on FA. D–F At 11 months, recurrent SRF is associated with focal smoke-stack-pattern leakage at the temporal margin of the NV lesion. The SRF resolves spontaneously within 3 months without treatment. G–I At 20 months, recurrent SRF shows diffuse leakage on FA, and the fluid resolves after three aflibercept injections CSC leakage at baseline and NV leakage on follow-up FA. A 63-year-old woman presents with decreased vision in the right eye. A–C At baseline, OCT shows shallow subretinal fluid (SRF), and OCT angiography demonstrates a neovascular lesion corresponding to diffuse leakage on FA. D–F At 11 months, recurrent SRF is associated with focal smoke-stack-pattern leakage at the temporal margin of the NV lesion. The SRF resolves spontaneously within 3 months without treatment. G–I At 20 months, recurrent SRF shows diffuse leakage on FA, and the fluid resolves after three aflibercept injections To investigate the spatial relationship between CSC leakage points and NV lesions, the locations of focal leaks on FA were compared with the corresponding NV complexes delineated on OCTA. At baseline, two eyes had two leakage sites, and 20 of 21 focal leaks (95.24%) were located along the margin of the neovascular complex. Including those identified during follow-up, a total of 39 CSC leakage points were observed, 37 of which (94.87%) were within the NV boundaries, predominantly along the lesion margin (Fig. 4).Fig. 4Changes in the location of CSC leakage points over time. At baseline, OCTA (A) demonstrates neovascularization involving the fovea, and FA (B) shows active focal leakage at the temporal margin of the NV. At 14 months, recurrent subretinal fluid (SRF) is associated with focal leakage at a location different from the previous site (C). At 30 months, focal leakage develops at a new location (D). Note that all leaking points are located at the margins of the neovascular network Changes in the location of CSC leakage points over time. At baseline, OCTA (A) demonstrates neovascularization involving the fovea, and FA (B) shows active focal leakage at the temporal margin of the NV. At 14 months, recurrent subretinal fluid (SRF) is associated with focal leakage at a location different from the previous site (C). At 30 months, focal leakage develops at a new location (D). Note that all leaking points are located at the margins of the neovascular network Discussion In this study, approximately 30% of eyes with OCTA-confirmed PNV and SRF exhibited a CSC-pattern leakage on FA. Previous studies have reported that the prevalence of NV in the setting of CSC ranges widely from 15 to 40% [6–11]. In contrast, our study investigated the reverse scenario, the prevalence of CSC in a consecutive PNV cohort. To date, few studies have addressed the role of FA in PNV. One case report identified a CSC- leakage pattern in quiescent PNV, and another study examined 1-year outcomes according to baseline FA leakage patterns [12, 13]. The present findings have meaningful clinical implications. In current practice, FA is often avoided because of its invasive nature, and anti-VEGF therapy is typically initiated in eyes presenting with SRF once NV is detected on OCTA. However, a considerable proportion of these cases may actually represent CSC with nonexudative NV, for which anti-VEGF monotherapy is not the most appropriate initial approach. PNV has been reported to respond better to anti-VEGF therapy than typical neovascular AMD, often requiring fewer injections [14–16]; yet a subset of eyes shows inconsistent or refractory responses, for which PDT has been effective as rescue or adjunctive treatment [17–19]. These earlier observations may, in fact, reflect the underlying heterogeneity revealed in our study. In addition to these cross-sectional findings, long-term follow-up provided further insights. Patients were followed for more than 3 years, during which repeated FA examinations were performed when fluid recurred or when the treatment response or clinical course was not as expected. Among the eyes that underwent repeated FA, a change in FA leakage pattern was observed in about 20%, suggesting a shift in the underlying mechanism of fluid accumulation, from neovascular activity to a CSC-related process or vice versa. These longitudinal findings carry further clinical implications. The FA leakage pattern observed at baseline may not reliably predict a patient’s subsequent clinical course or prognosis. Given that the nature of accumulated fluid can evolve over time, albeit only in a subset of eyes, OCT plays an essential yet complementary role in ongoing monitoring and treatment decision-making. We also attempted to identify OCT and clinical features that may help differentiate between the two mechanisms of fluid accumulation in the absence of FA. Although none are specific, findings such as a relatively greater SRF height at presentation, a sudden increase in SRF during the disease course, neurosensory retinal sagging, and a relatively younger age may be indicative of a CSC-related process. In addition, a suboptimal response to anti-VEGF therapy may also imply that the underlying mechanism is CSC-related rather than neovascular. Among eyes with CSC-pattern leakage, anti-VEGF therapy was administered in 11 of 19 eyes, representing a relatively large proportion. Although we recognized that typical CSC generally responds poorly to anti-VEGF therapy, treatment was nevertheless administered at the time, as the concurrent NV was regarded as a potential contributor to fluid accumulation. However, complete fluid absorption was achieved in only 27% of these eyes after three loading injections, which was significantly lower than that in the NV-pattern group (66%) and generally consistent with the limited efficacy observed in typical CSC. In addition, the high rate of spontaneous improvement after observation and the favorable response to thermal laser or subsequent PDT further suggest that CSC with concomitant non-exudative NV behaves similarly to typical CSC in terms of treatment response and clinical course. A noteworthy and novel finding in this study was that most focal leakage sites were localized along the margin of the neovascular complex. These spatial concordances suggest that the RPE barrier is intrinsically compromised at the proliferating edge of the NV. We postulate that active proliferation at the NV border imposes mechanical and metabolic stress on the marginal RPE, thereby accelerating RPE dysfunction at the interface. Considering the vertical anatomy, fluid originating from the choroid may gain access to the subretinal space through RPE defects at these marginal sites, rather than being blocked by the type 1 neovascular membrane. These observations support the notion that, in PNV, the presence of NV itself can give rise to CSC. This framework also accounts for our findings of a relatively high rate of CSC coexisting with non-exudative NV and for the time-dependent transition between NV-pattern and CSC-pattern leakage in the same eye. This perspective contrasts with the conventional view that PNV develops secondary to longstanding CSC and highlights the need for further investigation. In this study, we categorized FA leakage patterns into focal and diffuse types based on the assumption that leakage in PNV generally originates from the entire neovascular network, whereas leakage in CSC typically arises from the choroid through a focal RPE defect. These fundamental differences underscore the importance of distinguishing neovascular from CSC-related mechanisms when interpreting FA. However, FA may also have limitations in precisely identifying the underlying true source of fluid. As seen in polypoidal choroidal vasculopathy, another neovascular pachychoroid entity, exudation may originate from a specific portion of the vascular complex, often at marginal proliferative areas, producing a focal-appearing leak rather than a diffuse neovascular pattern [20]. In addition, CSC-related focal leakage becomes more apparent when the neovascular network is nonexudative, whereas in eyes with active neovascular leakage, identifying a CSC-related focal leak can be challenging. Choroidal leakage from a CSC may also secondarily stain the neovascular complex, further complicating interpretation. Finally, in chronic CSC or in cases with multiple recurrences, leakage may appear as diffuse ooze, and ambiguous hyperfluorescence from RPE window defects may also be present, both of which can limit accurate differentiation [21]. In our study, cases with an FA leakage pattern that could not be clearly classified were designated as undetermined and excluded from analysis. Outcomes were compared only between the NV- and CSC-pattern groups to maintain objectivity in image interpretation, which may limit the generalizability of our findings. This study is also limited by its relatively small sample size and retrospective design. As this study was conducted at a single center within a clinic-based cohort, the generalizability of the findings may be limited. Larger multicenter investigations will be necessary to confirm these results and strengthen their broader applicability. In addition, variability in the choice of anti-VEGF agents and treatment protocols among PNV cases necessitates caution in interpretation. To address the inherent constraints of the retrospective design, outcome assessment was limited to short-term (3-month) clinical results. Future prospective studies incorporating the PNV features delineated in this study are anticipated to provide further insight. Conclusions In eyes with PNV and SRF, the presence of neovascularization does not necessarily indicate that accumulated fluid originates solely from the neovascular complex. CSC-pattern leakage was frequently observed on FA and was associated with greater SRF height and a lower rate of complete fluid resolution after anti-VEGF loading. FA leakage patterns may change over time within the same eye, and CSC leakage points were typically located along the NV margin. Given the frequency of CSC-pattern leakage, the potential risk of neovascular activation, and the recurrent nature of CSC, FA remains valuable for identifying the origin of SRF when indicated.