Retinal Vasculitis in the Setting of Mycobacterium tuberculosis Exposure: Clinical Course and Management of Three Cases

Occlusive retinal vasculitis has diverse etiologies including systemic inflammatory or infectious disease or it can be isolated to the eye [1]. Mycobacterium tuberculosis (mTB) is a recognized etiology of retinal vasculitis, but diagnostic consensus on tubercular retinal vasculitis (TRV) is lacking as direct demonstration of ocular tissue mTB involvement is typically not performed [2, 3]. Historically, cases of peripheral, occlusive retinal vasculopathy with no identified cause were labeled Eales disease [4]. It is characterized by three stages of disease that may overlap [4, 5]. An initial peripheral retinal phlebitis (inflammatory stage) with subsequent occlusion of the retinal vessels (ischemic stage) can lead to neovascularization of the optic disc and retina (neovascular/proliferative stage) [4]. Complications include cystoid macular edema (CME), vitreous hemorrhage, retinal detachment, and anterior segment neovascularization [4]. Research later identified a potential role of mTB in Eales etiology, proposing a hypersensitivity reaction as an explanatory mechanism, but the precise pathophysiology is unclear [6, 7].

TRV can also present with an occlusive retinal vasculopathy nearly indistinguishable from Eales disease [3]. The clinical and potential microbiological overlap between TRV and Eales disease suggests the two entities may represent a spectrum of disease. This prompted the uveitis community to consider reserving “Eales disease” for cases of a true idiopathic nature without any systemic association [3]. Until recently, there was a lack of consensus on managing TRV. Recent guidelines published by the Collaborative Ocular Tuberculosis Study (COTS) provide clinical calculators and algorithms to guide clinicians in managing suspected TRV and considering antitubercular therapy (ATT) [8]. The present study describes the clinical course and management of patients diagnosed with retinal vasculitis in the setting of mTB exposure, with 1 case labeled as Eales disease and two cases labeled as TRV at that time.

A retrospective case series was performed by reviewing medical records of patients diagnosed with retinal vasculitis in the setting of positive mTB immunologic testing at Bascom Palmer Eye Institute between 1989 and 2022. Retinal vasculitis was diagnosed by the physician based on comprehensive ophthalmic examination, multimodal imaging, including fluorescein angiography (FA), fundus photography, and optical coherence tomography (OCT). Patients with diabetic retinopathy, hypertensive retinopathy, and retinal vascular occlusions were excluded given the potential for these conditions to cause concurrent abnormal findings on multimodal retinal imaging making it difficult to assess the intended primary condition of retinal vasculitis. Immunologic diagnostic testing for mTB was performed in all patients. Diagnostic testing for syphilis infection, sarcoidosis, sickle cell disease, and systemic vasculitides (anti-neutrophilic cytoplasmic antibody) was performed in all patients to rule out other recognized etiologies of retinal vasculitis. Diagnostic testing for systemic lupus erythematous was performed in case 1 and was negative. All patients presented and were treated prior to consensus guideline publishing by COTS. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000542075).

A 23-year-old healthy male from Honduras presented with blurry vision in the left eye. Visual acuity was 20/15 in the right eye and 20/20 in the left eye, and intraocular pressure (IOP) was normal. Anterior segment examination was unremarkable. Dilated fundus examination (DFE), shown in Figure 1a, of the right eye revealed neovascularization of the optic disc, intraretinal hemorrhages, and few hypopigmented and hyperpigmented choroidal lesions. DFE of the left eye, shown in Figure 1a, revealed diffuse preretinal hemorrhage and retinal perivenular sheathing. FA, shown in Figure 1b, showed perivascular leakage and hyperfluorescence outlining the choroidal lesions in the right eye. The left eye showed blockage from preretinal hemorrhage and hyperfluorescence at areas of leakage along the superior and inferior vascular arcade, shown in Figure 1b. There was vascular sheathing in the peripheral retina of both eyes, shown in Figure 2.

Diagnostic testing revealed a purified protein derivative (PPD) of greater than 10 mm induration. The patient was previously treated with isoniazid and pyridoxine for latent tuberculosis for an unknown duration at an outside institution. The patient’s retinal condition was ultimately labeled, at that time, as Eales disease because the latent mTB was considered sufficiently treated, and there was no cellular intraocular inflammation on examination.

Both eyes underwent pan retinal photocoagulation. One month after presentation, he underwent PPV and endolaser for recurrent vitreous hemorrhage in the left eye. One month following that, he underwent PPV and endolaser for recurrent vitreous hemorrhage in the right eye. Cultures of vitreous samples from both eyes were negative for mTB. At follow-up, 33 years after initial presentation, the visual acuity was 20/20 in the right eye and 20/15 in the left eye with no repeat vitreous hemorrhage. The examination showed stable fibrosis in areas of regressed neovascularization.

A 44-year-old Haitian female presented with 3 months of flashing lights in both eyes. BCVA was 20/20 in both eyes, and the IOP was normal. The anterior segment examination was unremarkable. In the right eye, DFE showed sheathing of vessels in the temporal periphery, shown in Figure 3a. The left eye had vascular sheathing of temporal vessels, retinal neovascularization, and preretinal hemorrhage, shown in Figure 3a. FA of both eyes, shown in Figure 3b, demonstrated extensive peripheral capillary non-perfusion and leakage secondary to neovascularization.

Diagnostic testing revealed a positive interferon-gamma release assay. The patient was diagnosed with TRV but initially declined rifampin-isoniazid-pyrazinamide-ethambutol (RIPE) therapy recommended by the uveitis and infectious disease services.

She underwent scatter laser photocoagulation in the left eye and was lost to follow-up until she returned 8 months later with neovascularization in the temporal macula and vitreous hemorrhage, and there was severe vascular occlusion in the right eye. The patient underwent retinal laser photocoagulation in both eyes.

She took RIPE therapy for 4 months and continued isoniazid for an additional 5 months. At follow-up, 84 months from initial presentation, 1 month after completing RIPE therapy, BCVA was 20/15 in both eyes. DFE in the right and left eye showed stable diffuse sheathing of vessels. FA of the right eye showed stable neovascularization. The left eye showed stable neovascularization in the macula. There was less vascular leakage in the right eye at the final visit compared to baseline FA.

A 64-year-old African American male from an unknown country of origin presented with blurry vision in the right eye. The BCVA was 5′/200 E and 20/40 in the right and left eye, respectively. The IOP was normal in both eyes. The anterior segment examination showed trace anterior chamber cell in the right eye. There were intraretinal hemorrhages, sheathed vessels, and retina scars consistent with retinal laser photocoagulation in both eyes. FA revealed temporal capillary non-perfusion and vascular leakage in the right eye and vascular leakage, capillary non-perfusion, and a focus of leaking neovascularization in the left eye. OCT macula showed CME in the right eye. Treatment with topical prednisolone drops in the right eye was recommended while systemic workup was initiated.

Diagnostic testing showed a PPD skin test with 14 mm induration and a chest CT that revealed a cluster of nodular densities in the right lower lobe. He was diagnosed with TRV based on immunologic and chest imaging evidence and received rifampin and isoniazid for 4 months. Oral prednisone was started to treat the intraocular inflammation and vasculitis.

At follow-up 2 weeks later, the patient complained of blurry vision in both eyes. Uncorrected visual acuity was 20/400 and 20/70 in the right and left eye, respectively. Anterior segment examination now showed 3+ anterior chamber cells in both eyes. The DFE was stable. OCT retina showed continued CME in the right eye. He was treated with intravitreal triamcinolone without CME improvement. Subsequent treatment with intravitreal dexamethasone implant improved the CME in the right eye. After 60 months of follow-up, the BCVA improved to 20/80 and 20/30 in the right and left eye, respectively. DFE revealed stable diffuse retinal vessel sheathing. FA showed stable extent of capillary non-perfusion.

The 3 cases presented illustrate the clinical course of diverse patients treated for retinal vasculitis in the setting of mTB exposure. Historically, as illustrated in case 1, Eales disease was the diagnosis given to young men who presented with an apparently idiopathic obliterative vasculitis that predominantly affected the peripheral retina and evolved through three stages: inflammatory, occlusive/ischemic, and neovascular [5, 9]. Our patient’s history included treatment for latent mTB with two-drug regimen years before presenting to our institution. It was thus concluded that the retinal condition was likely not representative of true mTB infection given lack of intraocular inflammation and prior ATT. The ischemic sequelae of his disease were treated, and he maintained good vision for over 30 years. Eales disease etiopathogenesis, however, is poorly understood and a frequently reported systemic association is exposure to mTB [4]. Several studies using polymerase chain reaction detected mTB genome in ocular fluids and epiretinal membrane samples from Eales disease patients that was not detected in control samples [10‒12]. The role of the organism in the phenotype of Eales disease is unclear, with some studies suggesting it elicits a hypersensitivity reaction and does not represent active infection [7]. With this association, however, clinicians now must more critically consider the utility of ATT in cases of retinal vasculitis that have positive mTB exposure as active infection by mTB is not definitively excluded.

Traditional treatment of Eales disease includes peri- or intraocular corticosteroids and, rarely, systemic corticosteroids, but typically not ATT [4, 9]. Neovascularization is generally treated with retinal laser photocoagulation [4, 9]. A study by Biswas et al. [13] examined treatment outcomes in 898 eyes in 500 patients. Those treated with oral corticosteroid during the acute inflammatory stage and those receiving laser therapy had significantly better vision at the last visit [12, 13]. Poor visual outcome was associated with decreased visual acuity at presentation and stage of Eales at presentation [13]. Biswas et al. [13] also reported that visual acuity was lower in the 320 Mantoux-positive patients. In their study, only 3% of patients received ATT based on pulmonology evaluation. Immunologic assays and high-resolution chest imaging were not available in the years the patients presented [13].

In this current series, all patients had evidence of mTB exposure, with case 2- and case 3-positive mTB tests discovered during workup of the retinal vasculitis. All three cases, however, had a wide range of severity. Case 1 maintained good BCVA with control of the neovascularization. Case 2 maintained nearly 20/20 BCVA throughout follow-up of 84 months with treatment of neovascular complications. Case 3 demonstrated improvement in BCVA with treatment of the intraocular inflammation and CME. Each patient received ATT, but different regimens. The heterogenous clinical course and ATT regimens further emphasizes the unclear role of mTB in our cases. At the time of our patients’ presentation, decision to use ATT was made based on patient origin from endemic countries and positive mTB exposure either immunologic or radiologic or both. No consensus guidelines existed pertaining to TRV diagnosis or management.

TRV is one distinct manifestation of ocular tuberculosis, but a diagnosis that is often presumptive in the setting of indirect evidence of ocular mTB [2, 3]. It is an occlusive retinal phlebitis that can also have neovascular complications making it nearly indistinguishable from Eales disease [3]. One study of patients diagnosed with TRV and mTB PCR-positive ocular fluids showed that the most common features of TRV in descending order were vitritis, vitreous snowball opacities, retinal hemorrhages, neuroretinitis, focal choroiditis patches, and serous retinal detachment [14]. These features may assist in attempting to distinguish Eales from TRV. A salient reason to distinguish between Eales disease and TRV is to determine which patients may benefit from ATT. The 3 cases here emphasize the importance of considering mTB in patients presenting with retinal vasculitis, regardless of the clinical label used. The cases also demonstrate that favorable vision outcomes are possible with management of the vasculitis sequelae, despite the differing ATT regimens and the uncertain value of ATT. Further research into optimal ATT regimen for suspected TRV is needed.

The COTS group published algorithms for initiations of ATT and studied the utility of ATT in patients with TRV [2, 15]. The COTS algorithm for initiation of ATT considers patient country of origin, PPD, interferon-gamma release assay, and chest imaging. Consensus agreement to treat was weak in non-endemic regions and was stronger for patients from an endemic region with a positive immunologic test and a positive radiologic test [2]. Cases 1 and 2 in this series were born outside of the USA, and consideration of country of origin was important in determining need for ATT.

Regarding utility of ATT in TRV, only a trend toward less frequent treatment failure in patients with occlusive-type vasculitis emerged [2]. In the current series, case 2 demonstrated improved, but continued, vascular leakage on FA despite ATT. It is possible that this case was more consistent with hypersensitivity to mTB rather than active tubercular infection, therefore failing to have resolution of the vasculitis with ATT alone. Cases 1 and 3 maintained stable to improved BCVA with treatment of neovascular and inflammatory complications that occurred despite treatment with ATT.

The 3 cases presented demonstrate the clinical course of patients with retinal vasculitis and mTB exposure. All three cases received ATT, albeit of differing regimens, and treatment for neovascular complications of the disease. The clinical distinction between Eales disease and TRV is evolving as research demonstrates a potential role of mTB in Eales disease. Clinicians should treat the ischemic complications of retinal vasculitis promptly. Clinicians should also use new consensus-driven algorithms and risk calculators to guide initiation of ATT in cases of retinal vasculitis with mTB exposure.

Limitations of the current study include the retrospective nature. Other limitations are that the diagnosis of TRV is presumed as direct confirmation of intraocular mTB is not routinely performed and that ATT regimens varied.

The study was approved by the Institutional Review Board of the University of Miami School of Medicine Medical Sciences Subcommittee for the Protection of Human Subjects, IRB identification number 20220953, approval date January 19, 2024. The study was conducted in accordance with the provisions of the Declaration of Helsinki. Written informed consent was obtained from the patients for publication of the details of their medical case and any accompanying images in accordance with the University of Miami protocol.

The authors have no conflicts of interest to declare.

The sponsor or funding organization had no role in the design or conduct of this research. This study was supported in part by an unrestricted grant from Research to Prevent Blindness (New York, NY), NIH Center Core Grant P30EY014801 (Bethesda, MD), and the Department of Defense (DOG Grant #W81XWH-09-1-0675) (Washington, DC).

L.C.K.: project design, manuscript writing, and primary data collection. J.D.S., M.P.F., and A.P.: manuscript revisions. T.A.A. and H.W.F.: project design and manuscript revisions.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of the patients but are available from L.C.K. or H.W.F. upon reasonable request.