Abstract
Purpose
To evaluate OCT vitreous signal intensity (OCT-VI) as a biomarker of vitreous haze (VH) in a longitudinal, prospective, and standardized manner in the First-Line Antimetabolites as Steroid-Sparing Treatment (FAST) Uveitis Trial.
Design
Secondary analysis of a block-randomized, observer-masked, multicenter clinical trial.
Subjects
Two hundred sixty-seven eyes from 147 patients with uveitis enrolled in the FAST Trial were evaluated in this analysis at baseline (BL), month 6 (M6), and month 12 (M12).
Methods
For each patient, Heidelberg Spectralis OCT volume scans (Heidelberg Engineering) were imported into the Doheny Reading Center 3D-OCTOR software and underwent a semi-automated segmentation. For each OCT volume, 5 B-scans were used to calculate the OCT-VI as a ratio of the vitreous space to the entire image.
Main Outcome Measures
OCT vitreous signal intensity at BL, M6, and M12 visits and the association with the National Eye Institute VH scale, anterior chamber (AC) cell grade, posterior synechiae (PS), lens status, visual acuity (VA), and central macular thickness (CMT).
Results
At BL, the median relative vitreous intensity was 0.43 (interquartile range, [0.33–0.54]) and reduced to 0.39 (0.27–0.53) at M6 and 0.39 (0.26–0.54) at M12. A mixed-effects linear regression model showed that OCT-VI and VH clinical grades were statistically significantly correlated (P < 0.001) when controlled for age, sex, time points, and eye. A statistically significant decrease of OCT-VI was observed over time with treatment (P < 0.001). A 2-step grade decrease in VH was associated with a –0.07 (standard deviation 0.19) change in OCT-VI. A mixed-effects linear regression model demonstrated a statistically significant association between OCT-VI and VA (b = 0.170, P = 0.020). A higher OCT-VI was associated with increasing grades of AC cell (P < 0.001) and presence of cataract (P < 0.001). There was no significant association (P > 0.05) of OCT-VI with PS or CMT.
Conclusions
In the FAST Uveitis Trial, OCT-VI was significantly correlated with the VH clinical grades and demonstrated significant longitudinal changes. OCT vitreous signal intensity may serve as a potential objective biomarker of VH.
Financial Disclosure(s)
Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
Keywords: Uveitis, OCT-based vitreous signal intensity, Standardization of uveitis nomenclature, Vitreous haze, FAST
Uveitis encompasses a broad spectrum of intraocular inflammatory conditions, which can be sight threatening, contributing substantially to vision impairment and ocular complications.1 Therefore, early diagnosis and timely intervention are critical to prevent blindness. However, accurately quantifying disease activity in patients with uveitis remains a clinical challenge.2
Vitreous haze (VH), an accumulation of inflammatory cells and proteinaceous material that obscures the fundus view, serves as a crucial indicator of uveitic activity. It has also been employed as an endpoint in numerous clinical trials. The Standardization of Uveitis Nomenclature (SUN) Working Group scale is the most frequently used scale for VH evaluation. This 6-point, stepwise grading system is based on a modified version of the National Eye Institute (NEI) VH scale.3 Despite routine use, the SUN scale's reliance on the examiner's subjective judgment represents a key limitation.4, 5, 6 Studies have demonstrated only moderate interobserver agreement among uveitis specialists and particular difficulty in distinguishing lower levels of inflammation (e.g., grades 0 and 0.5+).7,8 This poses a significant challenge in both clinical trial design and therapeutic decision-making. In response, several image-based surrogate measures have been proposed to provide more reliable and objective endpoints for VH assessment.9,10
Advancements in OCT, including eye tracking and precise image registration, have facilitated more objective approaches to the evaluation of vitreous (VIT) inflammation.11, 12, 13 Keane et al14 introduced a semiautomated OCT-based method for assessing VH by comparing VIT space signal intensity against a reference retinal pigment epithelium (RPE) intensity, producing a VIT-RPE relative intensity index. Subsequently, Sreekantam et al15 in a retrospective analysis, demonstrated that this VIT-RPE relative intensity index correlates strongly with treatment response in patients receiving sub-Tenon's triamcinolone acetonide for uveitic cystoid macular edema. More recently, Keane et al and Lee et al16,17 have developed fully automated models for VIT/RPE-relative intensity quantification, further streamlining the assessment process.
However, relying on RPE intensity as a reference can be confounded by conditions such as cystoid macular edema and choroidal neovascularization, where RPE integrity is compromised.18,19 To address this limitation, Liu et al20 introduced an OCT-based, normalized VIT intensity (OCT-VI) measure that considers the entire raw OCT image signal, rather than a single reference layer. This approach correlated well with clinical VH grading and showed high intereye repeatability in their cross-sectional study.
In this study, we employ a normalized OCT-VI measure that considers the entire raw OCT image signal, thereby accounting for various sources of image degradation. We apply this method to data from the First-Line Antimetabolites as Steroid-Sparing Treatment (FAST) Uveitis Trial,21, 22, 23 a multicenter randomized clinical trial encompassing a broad range of uveitis severities. By leveraging this robust dataset, our goal was to determine whether OCT-VI provides a reliable, objective, and reproducible biomarker of VH across different activity levels in a longitudinal setting.
Methods
Study Participants
The FAST Uveitis Trial (ClinicalTrials.gov Identifier: NCT 01829295) was an NEI-funded, observer-masked, randomized clinical trial with 9 sites in the United States, India, Australia, Saudi Arabia, and Mexico between August 2013 and August 2017.21 Enrollment criteria included patients who were ≥16 years of age and had noninfectious intermediate, posterior uveitis, or panuveitis in ≥1 eye, requiring corticosteroid-sparing therapy at baseline (BL). In the study, patients all received a standardized oral corticosteroid course with taper at enrollment and were randomized to receive oral methotrexate 25 mg weekly or oral mycophenolate mofetil 1.5 g twice daily. Patients were followed over a period of 12 months. The FAST Trial was approved by the local institutional review board for each site, and all patients provided informed written consent. This trial was conducted in accordance with the tenets of the Declaration of Helsinki.
Imaging and Grading Procedures
In the FAST Trial, OCT scans were performed at every visit using Heidelberg Spectralis OCT (Heidelberg Engineering) and Cirrus OCT (Carl Zeiss Meditec). In this present study, only Spectralis OCT volume scans centered on the fovea (49 lines, 20° OCT B-scans, automatic real time ≥16) at BL, month 6 (M6), and month 12 (M12) visits were evaluated. OCT volume data were imported into a previously validated custom OCT grading software developed at the University of California, Los Angeles Doheny Eye Institute known as 3D-OCTOR software.24,25 All cases in the study cohort at BL, M6, and M12 were reviewed for quality and graded by trained, masked graders from the Doheny Imaging Reading & Research Lab (M.A., C.S., and D.O.). Only OCT scans with an image quality (Heidelberg Q score) of ≥20 dB were graded. OCT volumes with 49 scans were annotated, while raster or single-line scans were excluded. Additionally, inverted scans and those with unclear VIT or retinal boundaries were not included. Cases exhibiting vitreoretinal interface abnormalities or thickened posterior hyaloid were also excluded.
The OCTOR software was used to segment the VIT space from the rest of the image. The software has the capability to quantify and automatically compute the VIT signal intensity within a given segmented space defined by 2 boundaries.14
The graders defined the following boundaries: (1) VIT top, the uppermost border of the VIT space, (2) internal limiting membrane, (3) outermost margin of the OCT scan. These boundaries defined VIT space and “outside vitreous” space, which included whole retina, choroid, and space between the choroid and the end of the scan area (Fig 1).20
Figure 1.
An example of the semiautomated segmentation in an OCT B-scan using OCTOR software. Number 1 represents vitreous space; number 2 represents outside vitreous space (the whole retina, choroid, and the space between the choroid and the end of the scan area).
Similar to previous studies, to minimize potential artifacts, 5 OCT B-scans in each OCT volume were considered for annotations from each eye at each time: foveal B-scan (x), 2 OCT B-scans above the foveal B-scan (x+2, x+4), and 2 OCT B-scans below the foveal B-scan (x–2, x–4).14,15 For each assessed OCT scan, the foveal center location was identified individually. It is noteworthy that the foveal center often did not align with the central B-scan in the OCT volume.
Afterward, we calculated a normalized ratio by computing a ratio of the VIT intensity compared to the overall image signal intensity.20
Subjects’ data, such as age, sex, and visual acuity (VA), recorded in logarithm of the minimum angle of resolution (logMAR), in addition to uveitis characteristics, including anatomical site, SUN grade of anterior chamber (AC) cells, and clinical VH according to NEI scoring system were assessed by study ophthalmologists, masked to patients' assigned treatment groups. Posterior synechiae (PS) was graded from 0 to 360°. Lens status was noted. Central macular thickness (CMT) was obtained using manufacturer software.
Statistical Analysis
Descriptive statistics were calculated for demographic characteristics, including age and sex, and ophthalmic examination, including OCT-VI, NEI VH grade, AC cell grade, PS degree, lens status, logMAR VA, and CMT across the entire longitudinal data sample and separately for BL, M6, and M12 visits. A random-effects interclass correlation coefficient (absolute agreement) was used to evaluate intergrader reliability between 2 masked graders. To visualize relationships in the data, box plots were generated to examine OCT-VI distributions across VH grades and lens status categories, while scatter plots illustrated the association between OCT-VI and VA. To account for repeated measurements within subjects, we employed mixed-effects linear regression models with random intercepts for each subject. These models assessed temporal changes in OCT-VI and its associations with other clinical parameters (VH grade, AC cell grade, PS degree, lens status, logMAR VA, and CMT) while adjusting for age, sex, visit time, and eye. Additionally, a hybrid model was implemented to disentangle the longitudinal within-subject and between-subject relationship between OCT-VI and other clinical parameters. All statistical tests were carried out using R Statistical Software (v4.4.1; R Core Team 2024) with a 2-tailed statistical significance set at P < 0.05 and a 95% confidence interval.
Results
Patient Characteristics
A total of 432 eyes from 216 patients were enrolled in the FAST Trial. For our OCT-VI analysis, at BL, 165 eyes from 95 patients were excluded, leaving 267 eyes from 147 patients for BL analysis. At M6, 197 eyes from 109 patients were included, and 70 eyes from 43 patients were excluded. At M12, 143 eyes from 79 patients were included, and 54 eyes from 36 patients were excluded. The reasons for patient exclusion including imaging on a non-Spectralis platform, poor quality imaging, and nonvolume scans, are shown in the flow chart in Figure 2. The included subjects' BL, M6, and M12 characteristics are illustrated in Table 1. Median Q score (interquartile range [IQR]) in decibels at BL was 24 (20–27), at M6 was 24 (22–28.3), and at M12 was 24 (20.8–28.3). Intergrader agreement for OCT-VI assessment was evaluated in a sub-cohort of 37 patients. The intergrader agreement was excellent (r = 0.983, P < 0.0001) with no significant mean difference between graders (P = 0.839). Bland–Altman analysis showed a very narrow 95% limits of agreement (±0.07) and a minimal bias (mean difference = 0.001), indicating excellent reproducibility of the OCT-VI measurement.
Figure 2.
Inclusion and exclusion flowchart of patients and images at each BL, M6, and M12. BL = baseline; M6 = month 6; M12 = month 12.
Table 1.
Patient and Eye Characteristics at BL, M6, and M12
| Cohort Characteristics | BL | M6 | M12 |
|---|---|---|---|
| Total eyes (n) | 267 | 197 | 143 |
| Age, yrs (median, IQR) | 39.18 [29.23–51.93] | 38.55 [28.99–51.95] | 39.79 [30.36–51.94] |
| Sex | |||
| Female (n, %) | 182 (68.2%) | 147 (74.6%) | 106 (74.1%) |
| Male (n, %) | 85 (31.8%) | 50 (25.4%) | 37 (25.9%) |
| Eye-level characteristics | |||
| Anterior chamber cells (n, %) | |||
| 0 | 122 (45.7%) | 151 (76.6%) | 116 (81.1%) |
| 1+ | 61 (22.8%) | 31 (15.7%) | 14 (9.8%) |
| 2+ | 58 (21.7%) | 8 (4.1%) | 6 (4.2%) |
| 3+ | 12 (4.5%) | 7 (3.6%) | 7 (4.9%) |
| 4+ | 12 (4.5%) | 0 (0%) | 0 (0%) |
| 5+ | 2 (0.7%) | 0 (0%) | 0 (0%) |
| Posterior synechiae degrees (n, %) | |||
| 0 | 222 (85.06%) | 158 (87.78%) | 102 (89.47%) |
| 1–90 | 28 (10.73%) | 9 (5%) | 4 (3.51%) |
| 91–180 | 1 (0.38%) | 3 (1.67%) | 4 (3.51%) |
| 181–270 | 5 (1.92%) | 5 (2.78%) | 4 (3.51%) |
| 271–360 | 5 (1.92%) | 5 (2.78%) | 0 (0%) |
| Vitreous haze (SUN VH) (n, %) | |||
| 0+ | 136 (50.9%) | 175 (88.8%) | 120 (83.9%) |
| 0.5+ | 44 (16.5%) | 12 (6.1%) | 11 (7.7%) |
| 1+ | 56 (21%) | 6 (3%) | 10 (7%) |
| 2+ | 28 (10.5%) | 4 (2%) | 2 (1.4%) |
| 3+ | 3 (1.1%) | 0 (0%) | 0 (0%) |
| Lens status (n, %) | |||
| No cataract | 186 (69.7%) | 115 (58.4%) | 77 (53.8%) |
| Cataract | 62 (23.2%) | 66 (33.5%) | 54 (37.8%) |
| Pseudophakic | 19 (7.1%) | 16 (8.1%) | 12 (8.4%) |
| Relative vitreous intensity (median, IQR) | 43 [0.33–0.54] | 0.39 [0.27–0.53] | 0.39 [0.26–0.54] |
| LogMAR visual acuity (median, IQR) | 0.2 [0–0.5] | 0 [–0.04 to 0.14] | 0.02 [0–0.2] |
| Central macular thickness, micron (μ), (median, IQR) | 297 [261–358] | 271 [239–300] | 268 [241.5–299.5] |
| Anatomical site of uveitis | |||
| Panuveitis | 84 (57.1%) | ||
| Posterior uveitis | 34 (23.1%) | ||
| Intermediate uveitis | 16 (10.9%) | ||
| Anterior and intermediate uveitis | 13 (8.8%) | ||
| Etiology | |||
| Undifferentiated | 24 (16.3%) | ||
| Vogt–Koyanagi–Harada disease | 65 (44.2%) | ||
| Pars planitis | 7 (4.8%) | ||
| Retinal vasculitis | 16 (10.9%) | ||
| Sarcoidosis | 13 (8.8%) | ||
| Sympathetic ophthalmia | 6 (4.1%) | ||
| Behcet disease | 3 (2.0%) | ||
| Birdshot chorioretinopathy | 4 (2.7%) | ||
| Other | 9 (6.1%) |
BL = baseline; IQR = interquartile range; logMAR = logarithm of minimum angle of resolution; M6 = month 6; M12 = month 12; SUN = Standardization of Uveitis Nomenclature; VH = vitreous haze.
Association between VH Scales and OCT-VI
The median OCT-VI for patients with SUN VH grade 0 was 0.38 (IQR, 0.27–0.51); grade 0.5+ was 0.45 (0.34–0.55); grade 1+ was 0.46 (0.37–0.58); grade 2+ was 0.56 (0.39–0.70); and grade 3+ was 0.89 (0.65–0.94). The unadjusted mixed-effects linear regression model showed that OCT-VI was significantly correlated with NEI VH grades (b = 0.724, P < 0.001). When the model was fully adjusted by follow-up time, eye, age, and sex, OCT-VI scale remained statistically significantly correlated with NEI VH grades (b = 0.442, P < 0.001). Importantly, this association remained significant after adjustment for potential confounders, including lens status and AC cell grade (b = 0.247, P = 0.029; see Table S1, available at www.ophthalmologyscience.org).
Furthermore, a hybrid model results indicated that the association between OCT-VI and SUN VH was driven by both the between-subject effect (b = 1.24, P < 0.001) and the within-subject effect (b = 0.46, P = 0.003).
Change in OCT-VI over Follow-Up Time
Median OCT-VI at BL was 0.43 (IQR 0.33–0.54), 0.39 (0.27–0.53) at M6, and 0.39 (0.26–0.54) at M12. A statistically significant improvement of OCT-VI was observed over time with treatment—compared with BL, the OCT-VI was 0.052 unit lower both at M6 and M12 (P < 0.001) (Fig 3). We also evaluated a subset of eyes that demonstrated a 2-step improvement in NEI VH grades, a commonly accepted endpoint in clinical trials.3 From BL to the M6 primary endpoint, 38 eyes showed a 2-step decrease in VH. Of these eyes, the magnitude of OCT-VI change associated with a 2-step VH decrease was –0.07 (standard deviation 0.19). A hybrid model indicated that the association between OCT-VI and VH was significantly driven by the within-subject effect (b = 2.22, P = 0.001) but not significantly driven by the between-subject effect (b = 0.71, P = 0.242).
Figure 3.
Box and whisker plots demonstrating OCT-VI at each baseline, month 6, and month 12 compared to the NEI clinical grade of vitreous haze. NEI = National Eye Institute; OCT-VI = OCT vitreous intensity.
Effect of AC Cells on OCT-VI
Median OCT-VI for patients with SUN AC cell grade 0 was 0.39 (IQR 0.28–0.53), grade 0.5+ was 0.42 (0.29–0.54), grade 1+ was 0.45 (0.36–0.59), grade 2+ was 0.44 (0.39–0.55), grade 3+ was 0.55 (0.40–0.60), and grade 4+ was 0.46 (0.45–0.46). In the unadjusted mixed-effects linear regression model, AC cell grade was statistically significantly correlated with OCT-VI (b = 0.031, P < 0.001). When the model was fully adjusted for follow-up time, eye, age, and sex, the correlation with OCT-VI remained statistically significant (b = 0.026, P < 0.001).
Effect of PS on OCT-VI
Median OCT-VI for patients with PS (graded in degrees of iris involvement) were as follows: 0 degrees of PS was 0.40 (IQR 0.30–0.53), 1 to 90° was 0.44 (0.31–0.54), 91 to 180° was 0.56 (0.44–0.67), 181 to 270° was 0.56 (0.39–0.64), and 271 to 360° was 0.42 (0.37–0.53). A hybrid model indicated that eyes with higher PS degrees had trended toward higher OCT-VI, although it was not statistically significant (P = 0.054).
Effect of Phakic Status on OCT-VI
Mean OCT-VI for eyes with no cataracts was 0.38 (IQR 0.36–0.59), for eyes with cataracts was 0.49 (0.27–0.49), and for pseudophakic eyes was 0.47 (0.26–0.62). When the multivariable model was fully adjusted for follow-up time, eye, age, and sex, OCT-VI of eyes with cataracts was 0.079 higher than those without cataracts (P < 0.001). However, there was no reliable evidence showing that OCT-VI of pseudophakic eyes was higher than those without cataracts (P = 0.093) (Fig 4). Further stratified analyses by cataract subtype (nuclear sclerosis, n = 90; cortical cataract, n = 23; posterior subcapsular cataract, n = 138) demonstrated no statistically significant association between OCT-VI and nuclear sclerosis (P = 0.41) or cortical cataract (P = 0.72). In contrast, the presence of posterior subcapsular cataract was associated with higher OCT-VI values (P = 0.005).
Figure 4.
Box and whisker plots demonstrating OCT-VI in relationship to lens status. OCT-VI = OCT vitreous intensity.
Association between OCT-VI and Quantitative Treatment Outcomes (VA and CMT)
Visual acuity measured in logMAR was significantly associated with OCT-VI (b = 0.288, P < 0.001). When fully adjusted for follow-up time, eye, age, and sex, the correlation with OCT-VI remained statistically significant (b = 0.170, P = 0.020) (Fig 5).
Figure 5.
Association between OCT-VI and logMAR visual acuity using pooled baseline and follow-up time points. logMAR = logarithm of minimum angle of resolution; OCT-VI = OCT vitreous intensity.
Median CMT at BL was 297 μm (IQR 261–358), 271 μm (239–300) at M6, and 268 μm (241.5–299.5) at M12. The unadjusted mixed-effects linear regression model showed that there was a statistically significant association between OCT-VI and CMT (b = 74.218, P = 0.002). However, this association becomes statistically insignificant (P = 0.072) when the model was fully adjusted for follow-up time, eye, age, and sex. A further hybrid model indicated that the association between OCT-VI and CMT was mainly driven by the within-subject effect (r = 73.97, P = 0.012), not the between-subject effect (r = 49.52, P = 0.291).
Discussion
This study provides a longitudinal evaluation of the relationship between OCT-based normalized relative OCT-VI and clinically graded VH. To our knowledge, this is the first study to evaluate the previously proposed OCT-VI measure, normalized to the total raw OCT image signal, within a clinical trial setting.20 Our results indicate that OCT-VI may serve as an objective, quantitative biomarker of VH. OCT vitreous intensity demonstrated strong statistically significant correlations with individual grades of the clinical NEI VH grades. Beyond these correlations, OCT-VI also showed potential as a predictor of functional outcomes, including changes in VA.
We adopted a normalized OCT-VI approach that incorporates the entire raw OCT signal to mitigate factors that affect image quality and subsequently, relative VIT intensity.26 Liu et al20 introduced this method to address the limitations of the VIT-RPE approach, which depends solely on RPE intensity.13, 14, 15 Retinal pigment epithelium intensity evaluation is problematic in conditions like chronic uveitis, cystoid macular edema, and choroidal neovascularization, where RPE integrity may be compromised.18,19 Although Liu et al20 found that OCT-VI effectively distinguished healthy eyes from those with uveitis and exhibited high repeatability, their study was cross-sectional and involved a small cohort with predominantly low-grade VH. In contrast, we examined a larger cohort of active uveitis patients representing a full range of disease severity and followed them longitudinally to assess OCT-VI changes over time. Consistent with previous findings,13,14,20 we observed a significant correlation between OCT-VI and the NEI clinical VH grade. This correlation persisted after adjusting for follow-up time, eye, age, and sex. Our model suggests that OCT-VI was significantly associated with VH grade, with each one-unit increase in OCT grade corresponding to a 0.442-unit increase in SUN score.
The significant correlation in OCT-VI at each clinical VH grade (P < 0.001) supports the potential utility of OCT-VI as an objective measure of inflammation and an indicator of treatment-induced changes in VH. This aligns with Sreekantam et al,15 who reported a substantial decrease in VIT-RPE intensity after sub-Tenon's triamcinolone acetonide therapy. Additionally, OCT-VI exhibited a modest but significant association with VA (logMAR), confirming that higher VIT intensity corresponds to worse visual function. Such findings are consistent with previous reports that OCT-based VIT metrics may better predict visual outcomes than the traditional clinical VH scale.20 In parallel, Barbosa et al27 found that the Visual Function Questionnaire-Utility Index scores exhibited a modest negative correlation with OCT-derived VIT-RPE intensity, reinforcing the broader functional relevance of objective VIT measurements.
A significant strength of our study was the ability to evaluate longitudinal changes in OCT-VI, finding that treatment improved OCT-VI over time. Moreover, when we evaluated the relationship between OCT-VI and the widely accepted endpoint of a 2-step decrease in VH from BL to M6, we found a significant decrease on a patient level. This suggests that we can potentially derive thresholds of change of OCT-VI as a treatment endpoint.
We did not find a statistically significant relationship between OCT-VI measurements and CMT when adjusted for follow-up time, age, sex, and eye, similar to Liu et al.20 Similar to other studies, this may be due to differing timelines of VH and macular edema emergence and resolution during uveitis course.
Posterior synechiae, a clinical sign of chronic anterior segment inflammation, may confound accurate measurement of OCT-VI due to light scatter, off-axis imaging, or signal attenuation. In our study, more extensive PS had a higher OCT-VI although this correlation did not reach statistical significance. Importantly, the overall prevalence of PS was low at BL, M6, and M12 indicating that the majority of OCT-VI measurements were unlikely to be affected by pupillary distortion.
Regarding phakic status, our results align with previous studies,28, 29, 30 indicating that media opacities, such as cataracts and intraocular lens, can affect OCT measurements due to light scattering.31 We observed a statistically significant 0.079-unit increase in OCT-VI in the presence of cataract when compared with eyes without cataracts. When stratified by cataract subtype, this increase was observed only in eyes with posterior subcapsular cataract, whereas nuclear sclerotic and cortical cataract did not significantly increase OCT-VI. An increase in AC cells was significantly associated with OCT-VI values, possibly indicating an increase in overall uveitic activity but also potentially increasing OCT signal scatter due to presence of cells. This association remained significant after adjustment for follow-up time, age, sex, and eye.
Several limitations should be acknowledged. First, our method may be less applicable in cases with severe inflammation or dense media opacities due to the effect of scatter on the OCT signal and subsequent poor image quality or inability to obtain OCT imaging. Additionally, we assessed 5 B-scans, which is under the assumption that VIT inflammation is homogeneously distributed, potentially overlooking VIT regions with early or patchy involvement. We also did not evaluate the anterior VIT, which may be differentially affected in certain uveitic conditions. Employing ultra-widefield OCT imaging may help overcome this limitation by providing broader coverage of VIT.
The generalizability of OCT-VI is critical for its broader adoption in clinical use. While OCT-VI shows promising correlation with clinical inflammation, its reproducibility across different OCT platforms and acquisition protocols is not fully established. Variability in imaging platform designs and signal scaling algorithms across OCT devices may affect the measurements, preventing direct comparisons. Cross-device validation studies are needed to integrate OCT-VI as a standardized and objective biomarker.
A strength of our study is its longitudinal clinical design, the first of its kind to assess OCT-VI in relation to VH severity over time. By demonstrating OCT-VI change over time, our findings highlight OCT-VI's potential as a surrogate marker. This measure could help standardize VH assessment and minimize subjective variability, even among experienced clinicians. Although limitations remain, this approach moves us closer to a comprehensive, noninvasive, OCT-based framework for monitoring uveitis activity, especially when combined with other objective assessments of CMT, VIT cells, and AC cells.13,32
In conclusion, our study builds upon earlier research by showing a strong, longitudinal correlation between OCT-VI and NEI VH grades. Over a 12-month follow-up, OCT-VI reliably tracked changes in disease activity and treatment response. These results support the utility of OCT-VI as an objective, quantitative biomarker capable of capturing subtler changes than the clinical scale. By refining our ability to measure and monitor uveitis severity, OCT-VI may ultimately enhance clinical decision-making and inform the design of future clinical trials.
Data Availability
The data used to support these findings of this study are available from the corresponding author upon request.
Manuscript no. XOPS-D-25-00539.
Footnotes
Supplemental material available atwww.ophthalmologyscience.org.
This work was presented in part at the UCLA/American Uveitis Society second International Workshop on Objective Measures for Use in Clinical Trials on September 27-28, 2025, in Los Angeles, CA.
Disclosure(s):
All authors have completed and submitted the ICMJE disclosures form.
The authors made the following disclosures:
L.L.L.: Consultant — Roche, Bayer, and Novotech; Research funding — Bayer and Roche.
E.B.S.: Consultant — Roche/Genentech, Priovant, Gilead, Acelyrin, Kriya, Alumis, Merck and Regeneron; Research support — Roche/Genentech, Priovant, Gilead, Acelyrin.
N.R.A.: Consultant — Roche; Nonmonetary support — AbbVie, Inc (in the form of a drug donation).
S.R.S.: Consultant — 4DMT, AbbVie/Allergan Inc., Alexion, Alnylam Pharmaceuticals, Amgen Inc., Apellis Pharmaceuticals, Inc., Astellas, Bayer Healthcare Pharmaceuticals, Biogen MA Inc., Boehringer Ingelheim, Carl Zeiss Meditec, ONL Therapeutics, Catalyst Pharmaceuticals Inc., CharacterBio, iCare/Centervue Inc., GENENTECH, Ocular Therapeutics, Eyepoint, Heidelberg Engineering, Hoffman La Roche, Ltd., Iveric Bio, Janssen Pharmaceuticals Inc., Nanoscope, Notal Vision Inc., Novartis Pharma AG, Optos Inc., Oxurion/Thrombogenics, Oyster Point Pharma, Regeneron Pharmaceuticals Inc., Samsung Bioepis, Topcon Medical Systems Inc.; Grant support — Carl Zeiss Meditec, Heidelberg Engineering, Nidek Incorporated, Novartis Pharma AG, Topcon Medical Systems Inc.; Roche. He has received financial support from Carl Zeiss Meditec, Heidelberg Engineering, Optos Inc., Nidek, Topcon, iCare/Centervue, Intalight.
E.T.: Consultant — AbbVie, Kodiak Sciences, ANI Pharmaceuticals, Kowa, Eyepoint, Oculis.
This study was supported by the National Eye Institute (NEI) of the National Institutes of Health under award number (R21EY034543) to Dr Tsui. The FAST Trial was supported by the NEI under grant U10 EY021125 to Dr Acharya. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Funding for publishing this article open access was provided by the University of California Libraries under a transformative open access agreement with the publisher.
HUMAN SUBJECTS: Human subjects were included in this study. The FAST Trial was approved by the local Institutional Review Board (IRB) for each site, and all patients provided informed written consent. This trial was conducted in accordance with the tenets of the Declaration of Helsinki.
No animal subjects were used in this study.
Srinivas R. Sadda, MD, an associate editor of this journal, was recused from the peer-review process of this article and had no access to information regarding its peer review.
Author Contributions:
Conception and design: Alhelaly, Soylu, Oncel, Corradetti, Sadda, Acharya, Tsui
Data collection: Alhelaly, Soylu, Oncel, Rathinam, Gonzales, Thundikandy, Kanakath, Murugan, Vedhanayaki, Lim, Suhler, Al-Dhibi, Doan, Sadda, Madow, Coyne, Acharya, Tsui
Analysis and interpretation: Alhelaly, Soylu, Chen, Jackson, Sadda, Madow, Acharya, Tsui
Obtained funding: Sadda, Acharya, Tsui
Overall responsibility: Alhelaly, Soylu, Oncel, Corradetti, Chen, Jackson, Rathinam, Gonzales, Thundikandy, Murugan, Vedhanayaki, Lim, Suhler, Al-Dhibi, Doan, Sadda, Madow, Coyne, Acharya, Tsui
Supplementary Data
References
- 1.Durrani O.M., Meads C.A., Murray P.I. Uveitis: a potentially blinding disease. Ophthalmologica. 2004;218:223–236. doi: 10.1159/000078612. [DOI] [PubMed] [Google Scholar]
- 2.Rosenbaum J.T., Bodaghi B., Couto C., et al. New observations and emerging ideas in diagnosis and management of non-infectious uveitis: a review. Semin Arthritis Rheum. 2019;49:438–445. doi: 10.1016/j.semarthrit.2019.06.004. [DOI] [PubMed] [Google Scholar]
- 3.Jabs D.A., Nussenblatt R.B., Rosenbaum J.T., Standardization of Uveitis Nomenclature SUN Working Group Standardization of uveitis nomenclature for reporting clinical data. Results of the first international workshop. Am J Ophthalmol. 2005;140:509–516. doi: 10.1016/j.ajo.2005.03.057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Khairallah M. Are the Standardization of the Uveitis Nomenclature (SUN) working group criteria for codifying the site of inflammation appropriate for all uveitis problems? Limitations of the SUN working group classification. Ocul Immunol Inflamm. 2010;18:2–4. doi: 10.3109/09273940903348835. [DOI] [PubMed] [Google Scholar]
- 5.Denniston A.K., Holland G.N., Kidess A., et al. Heterogeneity of primary outcome measures used in clinical trials of treatments for intermediate, posterior, and panuveitis. Orphanet J Rare Dis. 2015;10:97. doi: 10.1186/s13023-015-0318-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Williamson P.R., Altman D.G., Blazeby J.M., et al. Developing core outcome sets for clinical trials: issues to consider. Trials. 2012;13:132. doi: 10.1186/1745-6215-13-132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hornbeak D.M., Payal A., Pistilli M., et al. Interobserver agreement in clinical grading of vitreous haze using alternative grading scales. Ophthalmology. 2014;121:1643–1648. doi: 10.1016/j.ophtha.2014.02.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kempen J.H., Ganesh S.K., Sangwan V.S., Rathinam S.R. Interobserver agreement in grading activity and site of inflammation in eyes of patients with uveitis. Am J Ophthalmol. 2008;146:813–818.e1. doi: 10.1016/j.ajo.2008.06.004. [DOI] [PubMed] [Google Scholar]
- 9.Denniston A.K., Keane P.A., Srivastava S.K. Biomarkers and surrogate endpoints in uveitis: the impact of quantitative imaging. Invest Ophthalmol Vis Sci. 2017;58:BIO131–BIO140. doi: 10.1167/iovs.17-21788. [DOI] [PubMed] [Google Scholar]
- 10.Liu X., Hui B.T.K., Way C., et al. Noninvasive instrument-based tests for detecting and measuring vitreous inflammation in uveitis: a systematic review. Ocul Immunol Inflamm. 2022;30:137–148. doi: 10.1080/09273948.2020.1799038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pichi F., Carreño E. The Vitreous in uveitis: characterizing the invisible with optical coherence tomography. Ocul Immunol Inflamm. 2022;30:690–696. doi: 10.1080/09273948.2022.2075399. [DOI] [PubMed] [Google Scholar]
- 12.Jacquot R., Sève P., Jackson T.L., et al. Diagnosis, classification, and assessment of the underlying etiology of Uveitis by artificial intelligence: a systematic review. J Clin Med. 2023;12:3746. doi: 10.3390/jcm12113746. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zicarelli F., Ometto G., Montesano G., et al. Objective quantification of posterior segment inflammation: measuring vitreous cells and haze using optical coherence tomography. Am J Ophthalmol. 2023;245:134–144. doi: 10.1016/j.ajo.2022.08.025. [DOI] [PubMed] [Google Scholar]
- 14.Keane P.A., Karampelas M., Sim D.A., et al. Objective measurement of vitreous inflammation using optical coherence tomography. Ophthalmology. 2014;121:1706–1714. doi: 10.1016/j.ophtha.2014.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sreekantam S., Macdonald T., Keane P.A., et al. Quantitative analysis of vitreous inflammation using optical coherence tomography in patients receiving sub-tenon’s triamcinolone acetonide for uveitic cystoid macular oedema. Br J Ophthalmol. 2017;101:175–179. doi: 10.1136/bjophthalmol-2015-308008. [DOI] [PubMed] [Google Scholar]
- 16.Lee H., Kim S., Chung H., Kim C. Automated quantification of vitreous hyperreflective foci and vitreous haze using optical coherence tomography in patients with uveitis. Retina. 2021;41:2342–2350. doi: 10.1097/IAE.0000000000003190. [DOI] [PubMed] [Google Scholar]
- 17.Keane P.A., Balaskas K., Sim D.A., et al. Automated analysis of vitreous inflammation using spectral-domain optical coherence tomography. Transl Vis Sci Technol. 2015;4:4. doi: 10.1167/tvst.4.5.4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sevgi D.D., Davoudi S., Comander J., Sobrin L. Retinal pigmentary changes in chronic uveitis mimicking retinitis pigmentosa. Graefe's Archive Clin Exp Ophthalmol. 2017;255:1801–1810. doi: 10.1007/s00417-017-3689-7. [DOI] [PubMed] [Google Scholar]
- 19.Nassisi M., Fan W., Shi Y., et al. Quantity of intraretinal hyperreflective foci in patients with intermediate age-related macular degeneration correlates with 1-year progression. Invest Ophthalmol Vis Sci. 2018;59:3431–3439. doi: 10.1167/iovs.18-24143. [DOI] [PubMed] [Google Scholar]
- 20.Liu X., Kale A.U., Ometto G., et al. OCT assisted quantification of vitreous inflammation in uveitis. Transl Vis Sci Technol. 2022;11:3. doi: 10.1167/tvst.11.1.3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Rathinam S.R., Gonzales J.A., Thundikandy R., et al. Effect of corticosteroid-sparing treatment with Mycophenolate Mofetil vs methotrexate on inflammation in patients with uveitis: a randomized clinical trial. JAMA. 2019;322:936–945. doi: 10.1001/jama.2019.12618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tsui E., Rathinam S.R., Gonzales J.A., et al. Outcomes of uveitic macular edema in the first-line antimetabolites as steroid-sparing treatment uveitis trial. Ophthalmology. 2022;129:661–667. doi: 10.1016/j.ophtha.2022.02.002. Elsevier Inc. [DOI] [PubMed] [Google Scholar]
- 23.Sura A.A., Sun Y., Reddy A.K., et al. Reduced dose methotrexate and mycophenolate Mofetil in noninfectious uveitis: a sub-analysis from the first-line antimetabolites as steroid sparing therapy (FAST) trial. Ocul Immunol Inflamm. 2024;32:955–960. doi: 10.1080/09273948.2023.2165949. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sadda S.R., Keane P.A., Ouyang Y., et al. Impact of scanning density on measurements from spectral domain optical coherence tomography. Invest Ophthalmol Vis Sci. 2010;51:1071–1078. doi: 10.1167/iovs.09-4325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sadda S.R., Joeres S., Wu Z., et al. Error correction and quantitative subanalysis of optical coherence tomography data using computer-assisted grading. Invest Ophthalmol Vis Sci. 2007;48:839–848. doi: 10.1167/iovs.06-0554. [DOI] [PubMed] [Google Scholar]
- 26.Montesano G., Way C.M., Ometto G., et al. Optimizing OCT acquisition parameters for assessments of vitreous haze for application in uveitis. Sci Rep. 2018;8:1648. doi: 10.1038/s41598-018-20092-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Barbosa J., Syeda S., Rodriguez-Torres Y., et al. Quantifying vitreous inflammation in uveitis: an optical coherence tomography prospective study. Can J Ophthalmol. 2020;55:352–358. doi: 10.1016/j.jcjo.2020.04.003. [DOI] [PubMed] [Google Scholar]
- 28.Zarranz-Ventura J., Keane P.A., Sim D.A., et al. Evaluation of objective vitritis grading method using optical coherence tomography: influence of phakic status and previous vitrectomy. Am J Ophthalmol. 2016;161:172–180.e4. doi: 10.1016/j.ajo.2015.10.009. [DOI] [PubMed] [Google Scholar]
- 29.Van Velthoven M.E.J., Van Der Linden M.H., De Smet M.D., et al. Influence of cataract on optical coherence tomography image quality and retinal thickness. Br J Ophthalmol. 2006;90:1259–1262. doi: 10.1136/bjo.2004.097022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.García-Bella J., Talavero-González P., Carballo-Álvarez J., et al. Changes in retinal nerve fiber layer thickness measurements in response to a trifocal intraocular lens implantation. Eye (Basingstoke) 2018;32:1574–1578. doi: 10.1038/s41433-018-0141-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Azhdam A.M., Goldberg R.A., Ugradar S. In vivo measurement of the human vitreous chamber volume using computed tomography imaging of 100 eyes. Transl Vis Sci Technol. 2020;9:2. doi: 10.1167/tvst.9.1.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Agarwal A., Ashokkumar D., Jacob S., et al. High-speed optical coherence tomography for imaging anterior chamber inflammatory reaction in uveitis: clinical correlation and grading. Am J Ophthalmol. 2009;147:413–416.e3. doi: 10.1016/j.ajo.2008.09.024. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data used to support these findings of this study are available from the corresponding author upon request.