Characterization of peripapillary vascular density in traumatic optic neuropathy Authors: Blake M Snyder, Jordan Larsson, Alexander F Vu, Yin Allison Liu, Natalie Anne Homer Published: 2026 ABSTRACT AbstractPurposeOptical coherence tomography angiography is a novel imaging modality used to characterize pathologic vascular changes in the peripapillary region. Traumatic optic neuropathy is a common yet poorly defined entity, lacking evidence-based anatomic characterization. Optical coherence tomography angiography may be utilized in this pathology enhance patient evaluation and counseling.MethodsA case series of patients presenting to a tertiary care center after periorbital trauma with unilateral traumatic optic neuropathy were included. Patients underwent optical coherence tomography angiography, with the contralateral unaffected eye imaged as a control. Conventional automated visual field testing and optical coherence technology imaging of the retinal nerve fiber layer and macula were additionally obtained.ResultsTwelve eyes (six patients) were included. Eyes with traumatic optic neuropathy demonstrated significantly reduced vascular density in the inside disc and inferior regions compared with fellow control eyes on paired nonparametric analysis (both p = 0.0312), preceding detectable retinal nerve fiber layer thinning on structural coherence tomography or clinically apparent optic disc cupping. Greater inter-eye vessel density differences were observed in patients with stable or worsened visual acuity, with a trend toward association with final visual outcomes.ConclusionsOptical coherence tomography angiography may offer an early characterization of disease severity and inform prognosis in traumatic optic neuropathy. Inferior and internal optic nerve areas demonstrated early preferential vascular compromise, without consistent accompanying clinically apparent optic disc cupping and structural retinal nerve fiber layer thinning, and may be associated with final visual acuity. Findings suggest that vascular compromise may occur in the absence of detectable structural changes in the optic nerve that inform functional prognosis. Additional studies assessing vessel density longitudinally and correlating with management are warranted.Keywords: Imaging, Traumatic optic neuropathy, Optical coherence tomography angiography Introduction Traumatic optic neuropathy (TON) is an often-devastating eye condition caused by indirect or direct injury. Indirect TON results from blunt head or ocular trauma, with compressive and/or stretching forces transmitted to the optic nerve or vascular supply.1 Direct TON occurs from penetrating nerve injury from bone fragments or foreign body.2 Direct TON has been reported to carry a lower chance of visual recovery compared to indirect TON.2 A study by Gross et al. highlighted the mechanism by which traumatic forces may deform the orbital roof near the optic foramen, inducing damage to supporting blood vessels and exerting stress on the nerve.3 This typically occurs at the junction of intra-orbital and intracanalicular segments, compressing the optic nerve proper and disrupting pial vessels, thereby compromising the essential blood supply.3,4 Subsequent vascular ischemia-induced optic nerve swelling and compression within the narrow optic canal may ensue, further compromising blood supply to retinal ganglion cells and ensuing irreversible axon degeneration, often presenting with delayed optic disc pallor. This hypothesized compressive pathophysiology informs typical current-day treatment options, such as optic canal decompression and systemic steroid use. However, the early vascular changes in TON are not fully understood or adequately addressed with current management algorithms. Optical coherence tomography angiography (OCTA) provides precise and accurate measurements of vessel density in peripapillary and parafoveal regions and has proven valuable in quantifying vascular changes in various optic nerve and retinal conditions. TON, a frequently encountered yet poorly understood clinical condition, currently lacks well-defined anatomic vascular characterization. OCTA imaging may aid in characterizing pathological vessel changes in TON to guide treatment decisions and assist in patient counseling. This study aims to utilize OCTA imaging to evaluate angiographic peripapillary changes in traumatic optic neuropathy, with the goal of assessing the feasibility and efficacy of this testing modality for disease monitoring and management. Materials & methods A case series of patients with unilateral TON presenting to a single tertiary care center was performed. Institutional Review Board approval was obtained, and research was performed in accordance with the Declaration of Helsinki. TON diagnosis was made based on a history of acute vision loss following facial or head injury, accompanied by clinical findings of optic neuropathy, such as decreased visual acuity, a relative afferent pupillary defect, and optic disc edema or pallor, in absence of alternative acute ophthalmic pathology. For each case, the medical chart involved orbital injury without documented direct transection of the optic nerve and was clinically managed as traumatic optic neuropathy with compressive/apical features. As the mechanism of injury is often mixed or uncertain in TON, all clinically diagnosed TON presentations were included. The fellow contralateral eye served as an internal control for each patient, which is a standard approach in unilateral traumatic optic neuropathy and reduces inter-individual variability compared with external normative databases. Conditions were managed with observation, with or without systemic steroids, as deemed appropriate by the treating provider. Exclusion criteria included patients with preexisting ophthalmic pathology, including glaucoma, high myopia, amblyopia, concurrent/prior optic nerve disorders, previous retinal surgery, or concurrent vascular disorders. Pediatric patients were also excluded. All enrolled TON patients voluntarily participated in the study and were provided informed consent. Patients underwent comprehensive ophthalmic evaluation, including best-corrected visual acuity (BCVA), pupillary testing, color vision assessment, slit lamp and fundus examination, and ophthalmic imaging. The contralateral unaffected eye was also evaluated and imaged to serve as a control. Visual field perimetry was conducted using a Humphrey Field Analyzer (HFA; Carl Zeiss Meditec, Dublin, CA, USA) with the 24 − 2 Swedish Interactive Thresholding Algorithm (SITA)-standard protocol, with mean deviation used for analysis. Fundus photography and Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) scans were performed for all patients, with a 3.46-mm-diameter circle automatically centered on the optic disc. The Cirrus HD-OCT software calculated the average and sectoral macular ganglion cell-inner plexiform layer (GCIPL) thickness. Inclusion criteria required all scans to have a signal strength ≥6 without motion artifacts. The same OCTA imaging device was used for all patients (AngioVue; Optovue Inc.; RTVue software, Fremont, CA, USA) (Fig. 1, Fig. 2). A 4.5 × 4.5 mm2 cube angio scan centered at the optic nerve head (ONH) was performed for each patient's affected and control eye. Vessel density was defined as the percentage of perfused vascular area in the selected regions in en face views. OCTA output data from the 4.5 × 4.5 mm2 ONH scan included readings for vessel densities of the whole ONH image (onh-wiVD), inside-disc, and peripapillary sectors. OCTA also automated measurements of retinal thickness parameters, encompassing retinal nerve fiber layer (RNFL) at the ONH and a complexed layer (GCC) at the macular B-scans, comprising RNFL, retinal ganglion cell (RGC), and inner plexiform layers (IPL). All participants underwent both OCTA and Cirrus HD-OCT imaging on the same day for accurate and consistent data collection. Representative OCTA Peripapillary Vascular Density in Traumatic Optic Neuropathy Representative OCTA imaging of the peripapillary region of a study participant's affected and control eye, demonstrating relatively decreased vascular density in the affected eye at 40 days post-trauma. The affected eye and control eye demonstrated radial peripapillary capillary density (small vessels) of 40.1 and 50.0, respectively. Representative OCTA Peripapillary Vascular Density and Thickness in Traumatic Optic Neuropathy Representative OCTA imaging of the peripapillary region of a study participants affected and control eye at 40 days post-trauma, demonstrating relatively vessel dropout and decreased vascular density in the affected eye. Color scales inferior images represent relative thickness (left; microns), and vessel density (right; percent of radial peripapillary capillary density measurement). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Statistical analysis was performed at the eye level using paired comparisons between affected and fellow control eyes. Because of the small sample size and non-normal data distribution, continuous variables were summarized as medians and analyzed using exact two-sided Wilcoxon signed-rank tests. Effect size was reported as the median paired difference (affected minus control) with 95% confidence intervals estimated by bootstrap resampling (10,000 iterations). Analyses included complete pairs only, with p < 0.05 considered statistically significant, and no adjustment for multiple comparisons due to the exploratory case-series design. Additional exploratory analyses evaluated the relationship between time from injury to imaging and vascular density and visual outcomes using scatter plots and Spearman rank correlation. Patients were also stratified by imaging timing (≤ median vs > median days from injury) with nonparametric between-group comparisons. Results A total of 12 eyes from 6 patients were included, including 6 eyes with traumatic optic neuropathy, with a median patient age of 46 years (range 18–59) (Table 1). Patients underwent imaging at 29.5 days (range 7–126 days) after initial injury. Presenting optic nerve appearance was reported to be pale (n = 3), edematous (n = 1) and normal (n = 2) (see Table 2). Table 1Patient demographics.Table 1n-value (%)Gender Male3 (50%) Female3 (50%)Laterality Right Eye1 (17%) Left Eye5 (83%)Pathology Direct TON1 (17%) Indirect TON5 (83%)Past Ophthalmologic/Medical History Prior Intraocular Surgery2 (33%) Hypertension0 (0%) Diabetes0 (0%) Hyperlipidemia1 (17%) Coronary Artery Disease0 (0%)Management Systemic Corticosteroids4 (67%) Optic Canal Decompression0 (0%) Observation2 (33%) Patient demographics. Table 2OCTA vascular density.Table 2Median ChangeInterquartile RangeP-valueTotal Vessel Density Whole Image−4.534.910.0312∗ Inside Disc−5.504.320.0312∗ Peripapillary−6.868.480.0312∗ Superior Disc−6.1110.460.0625 Inferior Disc−8.646.270.0312∗Small Vessel Density Whole Image−4.794.120.0312∗ Inside Disc−10.206.330.0312∗ Peripapillary−8.159.180.0938 Superior Disc−5.7011.270.1380 Inferior Disc−10.907.170.0625 OCTA vascular density. Eighty-three percent of patients (n = 5) had a diagnosis of indirect TON. The single patient with direct TON sustained BB-gun injury to the orbit. Four patients were treated with high-dose systemic corticosteroids: two received oral and two intravenous formulations. No patients underwent surgical optic canal decompression. Median presenting visual acuity was logMAR 2.00 (range 0.88–4.70) (Table 3). All patients demonstrated an ipsilateral afferent pupillary defect and relative dyschromatopsia. Table 3Visual acuity and vessel density.Table 3VA PresentationVA Follow-upVA ChangeSmall VD InsideTotal VD InsidePatient 11.300.181.124.85.3Patient 20.881.30−0.4211.010.7Patient 30.880.100.781.32.6Patient 42.703.70−1.0012.75.7Patient 54.704.700.0014.510.6Patient 63.003.000.009.45.0 Visual acuity and vessel density. On OCTA analysis using paired Wilcoxon signed-rank testing, affected eyes demonstrated significantly lower total vessel density compared with fellow control eyes across multiple regions, including whole image total density (median paired difference −4.53, 95% CI −9.10 to −2.00; p = 0.0313), inside disc total density (−5.50, 95% CI −10.65 to −3.79; p = 0.0313), peripapillary total density (−6.86, 95% CI −11.85 to −1.90; p = 0.0313), and inferior total density (−8.65, 95% CI −11.65 to −2.65; p = 0.0313) (Table 2, Table 4, Fig. 3). Superior total density showed a similar trend but did not reach statistical significance (−6.11, 95% CI −12.00 to −0.35; p = 0.0625). Affected eyes also showed reduced ganglion cell complex thickness compared with controls (median paired difference −14.5 μm, 95% CI −27.5 to −3.5; p = 0.0313). Differences in retinal nerve fiber layer thickness and RNFL average did not reach statistical significance (RNFL thickness −20.0 μm, 95% CI −66.5 to 19.0; p = 0.3125; RNFL average −5.0 μm, 95% CI −37.5 to 13.5; p = 0.345). Table 4Per-patient vessel density values.Table 4PatientDaysInitial VAFinal VASteroidsOutcomeTypeAff WholeAff InsideAff PeripapCtrl WholeCtrl InsideCtrl Peripap1151.300.18YesImprovedTotal54.3057.7058.4058.9063.0061.30Small47.8049.1052.2052.0053.9054.3021260.881.30YesWorsenedTotal49.1055.4046.7057.9066.1058.70Small43.0044.2041.1051.9055.2052.803440.880.10NoImprovedTotal51.6056.8055.2053.3059.4056.10Small46.5050.7052.5046.8052.0049.404122.703.70NoWorsenedTotal43.8051.0043.3046.1056.7046.40Small37.3040.8036.4041.1053.5041.005544.704.70YesStableTotal45.8051.2046.3055.2061.8058.00Small40.1041.6040.6050.0056.1052.70673.003.00YesStableTotal51.9654.3149.4056.4259.2860.01Small45.0140.6342.7650.3850.0254.73 Per-patient vessel density values. Vessel density in affected versus control eyes in traumatic optic neuropathy. Box-and-whisker plots comparing vessel density (%) across five OCTA-derived metrics: (A) whole image small vessel density, (B) inside disc small vessel density, (C) inside disc total vessel density, (D) inferior peripapillary small vessel density, and (E) inferior peripapillary total vessel density. Affected eyes (black boxes) are compared with control eyes (white boxes). Individual patient values are overlaid, with paired eyes connected by lines. Boxes represent the interquartile range with median indicated by the horizontal line; whiskers denote the range. Paired comparisons were performed using the Wilcoxon signed-rank test, with p-values shown above each comparison; asterisks indicate statistical significance. Final follow-up occurred at median 195 days (range 12–366 days). Visual acuity improved in 2 patients, worsened in 2 patients and remained stable in 2 patients (Table 3). Patients with worse final visual acuity tended to demonstrate larger mean inter-eye vessel density differences when summarized across OCTA-derived metrics (Fig. 5). This association did not reach statistical significance (R2 = 0.53, p = 0.102). No statistically significant associations were detected between vascular density measures and clinical variables such as age, gender, or injury mechanism on nonparametric testing (Fig. 4, Fig. 5). Relationship between mean inter-eye vessel density difference and final visual acuity in traumatic optic neuropathy. Scatter plot illustrating the relationship between final visual acuity (VA logMAR, x-axis) and the mean inter-eye difference in vessel density (control minus affected; percentage points, y-axis), calculated for each patient as the average difference across five OCTA-derived metrics: whole image small vessel density, inside disc small vessel density, inferior peripapillary small vessel density, inside disc total vessel density, and inferior peripapillary total vessel density. Each point represents an individual patient and is labeled accordingly. A linear regression line is shown with corresponding R2 and p-value. The horizontal dashed line indicates no inter-eye difference. Relationship between inter-eye vessel density difference and change in visual acuity in traumatic optic neuropathy. Scatter plot illustrating the relationship between change in visual acuity (VA logMAR change, x-axis) and inter-eye difference in vessel density (control minus affected; percentage points, y-axis) across five OCTA-derived metrics. Each point represents a vessel density measurement for an individual patient. A pooled linear regression line is shown with corresponding R2 and p-value. The horizontal dashed line indicates no inter-eye difference. Exploratory analysis addressing timing variability showed no statistically significant association between time from injury to imaging and inside disc total vessel density difference (Spearman rho = −0.60, p = 0.21) or visual acuity change (rho = 0.03, p = 0.96). Stratified analysis using a median timing split (29.5 days) demonstrated similar median visual acuity change in earlier versus later imaging groups, although later imaging showed a larger median vessel density reduction (−10.6 vs −5.3). Scatter plots are provided in Fig. 6, Fig. 7. Given the small sample size, associations between vessel density differences and visual outcomes should be interpreted as exploratory. Relationship between time from injury to imaging and vessel density difference Scatter plot showing days from injury to OCTA imaging versus paired inside disc total vessel density difference between affected and fellow control eyes (affected minus control) for each patient (n = 6). No statistically significant monotonic association was observed on Spearman rank correlation (rho = −0.60, p = 0.21). Relationship between time from injury to imaging and visual acuity change Scatter plot showing days from injury to imaging versus change in visual acuity (logMAR follow-up minus initial) for each patient (n = 6). No statistically significant association was observed on Spearman rank correlation (rho = 0.03, p = 0.96). Discussion Traumatic optic neuropathy, resulting from indirect or direct trauma to the optic nerve, poses challenges in early diagnosis and prognostication due to the current lack of evidence-based anatomic characterization. This study evaluated early angiographic peripapillary changes in TON using OCTA and found a statistically significant decrease in whole image small vessel density, inside disc small vessel density, inferior peripapillary region small vessel density, inside disc total vessel density, and inferior peripapillary region total vessel density when comparing the initial presentations of the affected eye to the contralateral eye. Patients with smaller inter-eye vessel density differences appeared to demonstrate greater visual improvement, although this relationship was exploratory and not statistically significant. Detailed clinical characterization of TON is poor, limiting the practitioner's ability to predict visual recovery potential, accurately counsel patients, and guide management decisions. Visual improvement in traumatic optic neuropathy irrespective of treatment has been reported to be seen in 27-82% of cases.5,6 In the largest review of traumatic eye injuries from the operations of Iraqi and Enduring Freedom, 103 of 523 periorbital injuries were traumatic optic neuropathy, accounting for nearly 20% of the ocular combat injuries, with 48% from indirect trauma.7 Poor clinical prognostic factors were found to be initial no-light-perception vision, concomitant orbital fractures, and presenting flash visual evoked potentials less than 50% compared to the contralateral eye.8,9 Imaging predictors of visual prognosis have been lacking. OCTA, with its ability to provide precise measurements of vessel density, may offer an early characterization of disease severity and an early objective indication of prognosis. The greater involvement of the inferior and inside disc vascular regions may reflect regional susceptibility of papillomacular and inferior nerve fiber bundles to traumatic shearing and microvascular compromise at the level of the lamina cribrosa. These regions contain densely packed axons and capillary networks and may therefore demonstrate regional perfusion differences on OCTA that may be detectable even when global structural thinning is not yet apparent. From a clinical perspective, region-specific OCTA metrics may provide adjunctive value in early assessment and longitudinal monitoring of traumatic optic neuropathy, particularly when conventional structural measures are equivocal. Prior studies have demonstrated the utility of OCTA in various challenging optic nerve disorders, including non-arteritic anterior ischemic optic neuropathy (NAION), demyelinating optic neuritis, and compressive optic neuropathy 10, 11, 12, 13, 14. These early OCTA studies demonstrated promising wider application of this technology. Two recent studies analyzed use of OCTA in the assessment of vascular changes over time in indirect TON and vascular recovery after surgical correction.15,16 These studies observed time-dependent thinning of the RNFL layer that correlated with diffuse peripapillary vascular density decrease in their patients. The study herein did not find RNFL thinning to be consistently observed in our patients at the time of OCTA imaging. These authors suggest that early vascular changes of TON were captured may be associated with more diffuse vascular dropout and accompanying conventional RNFL thinning observed over time. Further longitudinal imaging in the patients presented herein may have revealed delayed RNFL changes. The aforementioned study also found decreased radial peripapillary capillary (RPC) vessel density in all regions and quadrants of the affected retina.15 Conversely, the current study showed statistically significant differences in vessel density specifically inside of the disc and in the inferior peripapillary region that may reflect localized vascular alterations associated with broader patterns of compromise. Understanding the early vascular compromise in TON via OCTA imaging may serve to enhance our knowledge of the particular areas of optic nerve vascular vulnerability. The management of traumatic optic neuropathy remains controversial. High-dose intravenous systemic corticosteroids used for management in acute spinal cord injuries have been applied to TON patients. However, the 2004 Corticosteroid Randomization After Significant Head Injury (CRASH) study revealed increased mortality with delayed treatment in the setting of traumatic brain injury, raising concerns about the safety of high-dose steroid therapy.17 The study herein did not demonstrate less vascular density loss with use of steroids, though with only a small sample size included. The lack of clear management guidelines, balancing potential benefits and risks, presents a challenge for practitioners. Additionally, management strategies to specifically target early ischemia from vascular changes is lacking. Exploring the cellular-level progression of retinal and optic nerve damage post-trauma and identifying reliable clinical biomarkers for TON detection may aid in determining best-practice neuroprotective strategies. This study is limited by small patient numbers and variable follow-up duration. Patient participation and uniform timing of clinical evaluation were limited by concomitant non-ophthalmic traumatic injuries, distance from the tertiary care center, and challenging psychosocial variables often present in this patient population. The sensitivity was impacted by the heterogeneity and complexity of presentation of TON. These authors believe this study provides promising proof of the feasibility and efficacy of this emerging imaging technology in enhancing the pathophysiologic understanding and prognostic evaluation of this potentially vision-devastating pathology. The study also novelly suggests particular vascular changes that may be detectable even when clinical and conventional imaging findings are not yet apparent and may correlate with long-term visual prognosis. Confirmation of these findings via a larger prospective study with increased follow-up duration may further enhance our understanding of the underlying vascular changes in TON, be a tool for management surveillance, and better inform prognosis. This study suggests the potential utility of emerging OCTA technology in characterizing the peripapillary vascular characterization in the early stages of traumatic optic neuropathy that may provide complementary information alongside clinical and standard imaging findings. There appears to be a significant decrease in internal optic nerve and inferior peripapillary vascular density in patients with TON, which may be associated with decreased visual recovery. This detailed vascular characterization using OCTA may enhance early patient counseling and guide management. Given the cross-sectional design, temporal relationships between vascular and structural changes cannot be determined. Further larger long-term studies are indicated to confirm these findings and evaluate the effect of various management strategies. CRediT authorship contribution statement Blake M. Snyder: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Jordan Larsson: Writing – review & editing. Alexander F. Vu: Conceptualization. Yin Allison Liu: Writing – review & editing, Conceptualization. Natalie Anne Homer: Writing – review & editing, Supervision, Project administration, Conceptualization. Propriety interest statement The authors have no commercial relationships to disclose. Financial Support This project was funded in part by the Hagen Schroeter Incentive Research Fund. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments None. References