Purpose
Ocular tuberculosis (OTB) affecting the posterior segment commonly presents with choroiditis lesions. However, tubercular retinitis (TBR) is a rare and poorly characterized phenotype. This study aims to define the clinical spectrum, multimodal imaging characteristics, natural course of TBR lesions, and differentiate it from similar appearing chorioretinal lesions in OTB.
Design
Single-center, retrospective case series.
Methods
The study included patients presenting with lesions clinically suggestive of retinitis. Clinical data, fundus findings, and multimodal imaging (ultrawidefield fundus photography, fluorescein and indocyanine green angiography, and optical coherence tomography [OCT]) were analyzed. The diagnosis of OTB was established based on the Collaborative Ocular Tuberculosis Study criteria, which includes supportive immunological and radiological evidence (positive Mantoux test or interferon-gamma release assay, and chest imaging). All patients received standard four-drug antitubercular therapy (ATT) with adjunctive corticosteroids and were followed for anatomical and functional outcomes.
Results
Eight patients (six males; age range: 21-54 years) were included. A total of 37 lesions were identified in 12 eyes (median: 2 lesions per eye). Twenty-five lesions (67.6%) were purely retinal with no choroidal involvement and were classified as TBR. Twelve lesions (32.4%) showed involvement of the choroid, retinal pigment epithelium (RPE), and outer retina, and were classified as tubercular retinochoroiditis (TBRC). All the eyes demonstrated occlusive retinal vasculitis. Fluorescein angiography of TBR lesions demonstrated early central hypofluorescence with circumferential perilesional hyperfluorescence (“moat sign”). On OCT, TBR lesions were intraretinal and characterized by full-thickness hyper-reflectivity and disruption of inner retinal architecture with no choroidal/RPE changes. TBRC lesions showed outer retinal disruption, RPE irregularity with focal defects, and choroidal shadowing. Indocyanine green angiography showed focal hypofluorescence in TBRC lesions, but was unremarkable in TBR. Following ATT, TBR lesions healed without pigmentation or scarring, whereas TBRC lesions resolved with discrete pigmented scars with retinal thinning and RPE disruption on OCT (32.4%) ( P =.028s).
Conclusions
The spectrum of posterior segment involvement in OTB includes intraretinal inflammation that can present as TBR and TBRC. These lesions are typically present in eyes with occlusive retinal vasculitis. Features on clinical examination and multimodal imaging help identify the tubercular etiology and differentiate between TBR and TBRC lesions.
INTRODUCTION
O cular tuberculosis (OTB) constitutes one of the most prevalent causes of infectious uveitis in both tuberculosis-endemic and nonendemic regions. The disease exhibits a remarkably diverse clinical spectrum, involving virtually all ocular structures and may present with scleritis, anterior uveitis, intermediate uveitis, retinal vasculitis, or varied forms of choroiditis. ,, Among these, choroiditis has traditionally been regarded as the most specific phenotypic manifestation, often correlated with evidence of active Mycobacterium tuberculosis ( Mtb ) infection either within the eye or at extraocular sites. Mtb reaches the choroid via hematogenous dissemination, primarily carried by Mtb-laden macrophages, and lodges preferentially in the perivascular choroidal stroma adjacent to the choriocapillaris. The high vascularity and rich perfusion of the choroid create an oxygen-rich milieu, conducive to the multiplication of Mtb bacilli. ,, Besides choroiditis, other ocular manifestations are generally regarded as immune-mediated reactions elicited by mycobacterial antigens rather than direct Mtb infection. This distinction between infectious and immune-mediated mechanisms critically shapes diagnostic criteria and guides therapeutic decision-making.
Although choroidal involvement remains the most frequently described manifestation of OTB, emerging evidence suggests that the retina can serve as an independent site of tubercular inflammation. , Intraretinal tubercular granulomas have also been demonstrated on histopathology. However, retinitis due to tuberculosis constitutes an under-recognized but clinically important subset within the OTB spectrum.
In this study, we describe a cohort of patients with OTB presenting with retinal involvement characterized by full-thickness retinitis lesions, occlusive retinal vasculitis, and intraretinal vascular changes.
MATERIALS AND METHODS
Patients with a diagnosis of OTB presenting with lesions mimicking retinitis were seen consecutively between 2021 and 2025 in the Uveitis Clinic of a tertiary care referral Institute in North India. This study adhered to the Declaration of Helsinki principles for human research. All participants provided written informed consent for inclusion in retrospective observational studies, which gained the approval of the Institutional Review Board and the Institutional Ethics Committee.
Study subjects
All patients had a detailed history recorded and underwent a complete ocular examination at presentation. The following demographic and clinical data were recorded: patient age, gender (male/female), race, medical history, laterality of the disease, initial visual acuity, presence or absence of anterior segment flare, cells, keratic precipitates, synechiae, associated vitritis (if present), morphologic description of retinal lesions, associated vasculitis, choroiditis, and additional features, if any. Fundus photography was performed using ultrawidefield imaging (California Optos) at presentation and follow-up. Ultrawidefield fluorescein angiography (FFA) was performed in all the patients, while selected patients underwent indocyanine green angiography (ICGA) (either Spectralis + HRA, Heidelberg Spectralis, or California Optos) at the discretion of the treating physician. Optical coherence tomography (OCT) scans (Spectralis + HRA; Heidelberg Engineering; or DRI OCT Triton, Topcon, Inc) were obtained passing through the retinal lesions.
The anatomical layer of retinochoroidal involvement was determined using multimodal imaging. The lesions were categorized as tubercular retinitis (TBR) or tubercular retinochoroiditis (TBRC) based on the features seen on OCT and other imaging techniques such as ICGA.
A. TBR: When the lesion was localized to the retina on OCT, presenting as full-thickness hyper-reflectivity and disorganization of the retinal layers, it was termed TBR. The TBR lesions had no involvement of the retinal pigment epithelium (RPE) and an intact choroid. These lesions did not show hypofluorescence on ICGA imaging and progressed to healing without any scarring.
B. TBRC: Lesions with outer retinal, RPE, and/or choroidal involvement on OCT, and localized hypofluorescence on ICGA were classified as TBRC.
All patients underwent tuberculin skin test, interferon gamma release assay (IGRA), radiological test (Chest X-ray or computed tomography), and enzyme-linked immunosorbent assay (ELISA) testing for HIVs. Additional targeted testing, including ELISA for Toxoplasma IgG/IgM, Treponema pallidum hemagglutination test, was carried out to exclude other causes of infectious retinitis, including syphilis, viral retinitis, and toxoplasmosis.
Diagnosis of OTB was established based on the Collaborative Ocular Tuberculosis Study (COTS) criteria, which included immunological and radiological evidence. , The immunological tests in the COTS criteria included a positive Mantoux test or IGRA, and radiological tests included chest imaging suggestive of pulmonary tuberculosis or latent TB infection. All patients received four-drug antitubercular therapy (ATT), including isoniazid 5 mg/kg daily, rifampicin 10 mg/kg/d, ethambutol 15 mg/kg daily, and pyrazinamide 25 to 30 mg/kg daily for the first 2 months. Thereafter, rifampicin and isoniazid were used for another 7 months. All patients received oral corticosteroids (initiated at prednisolone 1 mg/kg/d), which were tapered over a period of 8 to 12 weeks. All patients had a minimum follow-up of 6 months (range: 6-48 months). Follow-up data included visual acuity, details of the treatment, response to therapy, and recurrences, if any.
Statistical analysis
Statistical analysis was performed using Statistical Package for Social Sciences (SPSS version 30 for Windows). Quantitative variables were described using the mean with SD. Fisher’s exact test was used to cf the healing pattern of lesions in TBRC vs TBR. A P value of <.05 was considered significant in all tests.
RESULTS
The study included 8 patients (six males), with an age range of 21 to 54 years (mean: 36.62 ± 13.24). All included patients were Asian Indians. The disease was bilateral in four patients. The demographic profile and clinical description of these patients are detailed in Table 1 . Mean presenting visual acuity at presentation was 0.72 ± 0.53, range: 0 to 1.78 (LogMAR chart). Slit-lamp biomicroscopy revealed anterior chamber cells in nine eyes (75%) and vitritis in all the eyes ( Table 1 ). Across 12 eyes of 8 patients, a total of 37 lesions were identified (mean: 3.1 ± 2.3 lesions per eye; median: 2 lesions per eye). Lesions involved both the macula and the peripheral retina in five patients, were confined to the macula in one patient, and were limited to the peripheral retina in two patients ( Table 2 ).
TABLE 1
Demographic Profile and Clinical Presentation
| Patient No. | Age (Y) | Gender | Eyes Affected | Visual Acuity at Presentation | Anterior Segment Inflammation | Vitritis |
|---|---|---|---|---|---|---|
| 1 | 37 | M | OS | 20/200 | 1+ cells, 1+ flare | + |
| 2 | 50 | F | OD | <20/2000 | 2+ cells, 1+ flare | + |
| 3 | 21 | M | OU |
OD-20/200
OS-20/20 |
OD-1+ cells, 1+ flare, OS-quiet | OU+ |
| 4 | 25 | M | OU | OD-20/200 OS-<20/2000 | OU-2+ cells, granulomatous keratic precipitates+, iris nodules+ | OU+ |
| 5 | 54 | M | OS | OS-<20/2000 | 2+ cells, 1+ flare | + |
| 6 | 41 | M | OU | OD-20/200 | Cells 1+, flare 1+ | OU+ |
| 7 | 36 | M | OS | OS-20/120 | Cells 2+, flare 2+ | + |
| 8 | 21 | F | OU |
OD-20/40
OS-20/80 |
OU-quiet | OU+ |
TABLE 2
Clinical Presentation, Distribution of Lesions, and FFA Findings
| Patient No. | Lesion Type | Location of Lesions | Concomitant Retinal Vasculitis | FFA |
|---|---|---|---|---|
| 1 | TBR, TBRC | Macular + peripheral lesions | + | Retinitis lesions showed central hypofluorescence with adjacent retinal vessels surrounding the lesion like a Moat around a castle (moat sign); and Staining of peripheral healed choroiditis and laser scars |
| 2 | TBR, TBRC | Macular + peripheral lesions | + | Moat sign+, periphlebitis with peripheral CNP areas+, leakage through neovascularisation+, staining of choroiditis scars |
| 3 | TBR, TBRC | Macular + peripheral lesions | + | Moat sign+, peripheral CNP areas, and staining of peripheral healed choroiditis scars |
| 4 | TBR | Peripheral (multifocal Inferior lesions) | + | Moat sign+ |
| 5 | TBR, TBRC | Macular | + | Moat sign+, peripheral CNP areas and late leakage through neovascularisation |
| 6 | TBR | Macular + peripheral lesions | + | Moat sign+, staining of peripheral healed choroiditis scars |
| 7 | TBR | Multifocal, peripheral lesions | + | Moat sign+ |
| 8 | TBR, TBRC | Macular + peripheral lesions | + | Moat sign+, and staining of peripheral healed choroiditis scars |
CNP = capillary nonperfusion; FFA = fundus fluorescein angiography; TBR = tubercular retinitis; TBRC = tubercular retinochoroiditis.
Pattern 1: TBR
Twenty-five lesions (67.6%) were confined to the neurosensory retina and were classified as TBR. These lesions appeared as discrete, multifocal, yellow-white perivascular lesions with sharply demarcated margins, measuring approximately 200 to 500 μm in diameter. The lesions showed a predilection for retinal veins and typically abutted or partially encircled the vessel wall. Adjacent intraretinal hemorrhages and exudation were observed in eyes with concurrent vasculitis ( Figure 1 ).
Multimodal imaging features of tubercular retinitis (TBR) with associated perivascular choroiditis. Color fundus photograph (A) shows multiple yellow-white intraretinal retinitis lesions (double arrow) with adjacent pigmented perivascular choroiditis scars (arrowheads) and nasal segmental retinal vasculitis (arrow). Indocyanine green angiography (ICGA) early (B) and late phases (C) demonstrate hypofluorescent areas corresponding to choroiditis scars (arrowheads), while the active TBR lesions show no ICGA abnormality, confirming absence of primary choroidal involvement. Ultrawidefield fluorescein angiography (FFA) early (D), mid (E), and late phases (F) reveal perivenular staining and leakage corresponding to vasculitis, with associated areas of capillary nonperfusion and peripheral laser scars.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
