Commentary: Tuberculosis in neuro ophthalmology - How different it is?

Ocular tuberculosis can affect all the ocular layers. It can either occur in isolation or with a pulmonary, or with another extra pulmonary focus. Tubercular bacterial infection involves 5 distinct stages in its pathogenesis. It can result from direct involvement or a manifestation of hypersensitivity reaction to the tubercular antigens. Confirming the etiology remains a challenge due to the paucibacillary nature of ocular tuberculosis. Early diagnosis and an appropriate treatment are the most essential elements for prompt regression of inflammation.

Anatomically, it may present as anterior, intermediate, posterior or pan uveitis more often a granulomatous than non-granulomatous uveitis[12] Posterior segment finding includes focal or multifocal yellow choroidal granulomas or multifocal chorioretinal pigmented scars – specifically along retinal vessels. Optic nerve involvement include optic neuritis, retrobulbar neuritis, neuroretinitis, optic disc granuloma, papilledema and disc edema.[1]

The retinitis may be seen with an associated choroiditis, although a direct retinal involvement is rare. Other signs include sub-retinal granuloma subretinal abscesses and vasculitis. Tuberculosis can cause Serpiginous-like choroiditis (SLC), which sometimes mimics a classic serpiginous choroidopathy (SC). Tubercular SLC show multifocal, scattered, highly pigmented lesions with significant vitreous cells. Unlike classic serpiginous choroiditis, SLC do not respond to corticosteroid therapy. Inflammatory control is hence achieved only after anti-tubercular treatment.[3] Rarely, an acute and robust inflammatory response may result in hypopyon in tubercular pan uveitis. Yellowish sub-retinal abscess can rupture into the vitreous and result in Endophthalmitis or Panophthalmitis.

Differential diagnosis of tubercular uveitis includes other granulomatous uveitis such as sarcoidosis, Vogt-Koyanagi-Harada disease sympathetic ophthalmia, herpetic infection, phaco antigenic uveitis, syphilis and leprosy. Other causes of choroidal granulomas include sarcoidosis, syphilis and fungal lesions.[3] In practice, delineating tuberculosis related optic neuropathy[4] includes individualized work up on all the differentials such as infectives (syphilis, toxoplasmosis, cat scratch disease, Lyme disease, leptospirosis, systemic fungal infection, HIV-associated disease), systemic inflammatory diseases (multiple sclerosis, neuromyelitis optica, sarcoidosis, Vogt-Koyanagi-Harada disease, Behçet's disease, other systemic vasculitidies) ocular inflammatory conditions (uveitis and posterior scleritis); and vascular, neoplastic, toxic and hereditary forms of optic neuropathy as well.

The criteria for diagnosing ocular tuberculosis[15] include: consistent ocular signs, positive Mantoux reaction or IFN-γ release assay, active or old tuberculous lesion on chest imaging; polymerase chain reaction (PCR) detection of M. tuberculosis DNA in ocular fluid samples; identification of acid-fast bacilli by microscopy or culture of ocular or other tissue samples; and/or positive response to four-drug anti-tuberculosis treatment (i.e., isoniazid, rifampicin, ethambutol and pyrazinamide). Information regarding the investigations undertaken to exclude other etiologies for the optic neuropathy—as well as optic nerve or brain imaging studies—are supportive.

Tuberculosis—a great masquerader apart from syphilis—is an enigma with several grey areas in its diagnosis and therapy. There are no standard diagnostic guidelines for diagnosing neuro-ophthalmic tuberculosis[5] and on the standard pattern of treatment. Although Mazurek et al. suggested an updated guidelines for Tubercular diagnostics by including interferon-gamma (IFN-γ) release assay (IGRA) as an alternative to the Mantoux skin test in the western world, the Index TB guidelines[6] (based on the Indian scenario) did not recommend replacing the Mantoux test, as the kits involved in IGRA are yet to be standardized and validated specific to the Indian context. However, Rupesh et al.[7] in their Collaborative Ocular Tuberculosis Study [COTS-1] suggested using a combination of these tests owing to the varying specificities.

In the largest published series on neuroophthalmic tuberculosis by Davis et al.,[5] majority of the treated cases worldwide did include steroids in varied forms and dosing. At the same time, 36.7% were treated devoid of it, although the authors recognized their value in incorporating them, but were skeptical on periocular long-term steroids in lieu of the fact that its duration of action may extend beyond the course of anti-tuberculosis treatment. Pharmacologically, addition of Rifampicin resulted in increased plasma clearance of prednisolone[8] by 45% and reduced areas under curves[8] by 66% and merits adding steroids at full dosing. In lesions involving papillomacular bundle, juxtapapillary location and at optic nerve related involvement the full dosing possibly help in reducing anticipated Jarish-Herxeimer reactions as well.

The crux of the controversy in neuroophthalmic tuberculosis is centered on the usage of Ethambutol and Isoniazid in the management due to the concern on of drug-induced neuropathy. What is practical is to look for the potential toxicity with colour vision, visual field defects (for enlarged blind spot, central scotoma, altitudinal defect, peripheral defect, paracentral scotoma).

The pointers to the potential drug toxicity include deterioration in color vision, visual acuity, and/or visual field following initial improvement. The article “Ocular tuberculosis: Position paper on diagnosis and treatment management” by Figueira L et al. is an eye opener with different levels of recommendations for monitoring patients on anti-tubercular medicines that an astute clinician can ponder over in real time practice. In practice adapt the details derived from Optical coherence Tomogram (OCT), Fundus auto fluorescence, Colour fundus photogram to deliver the best possible patient centered care.[6] In cases of diagnostic dilemmas, correlating the clinical findings from the maximum feasible investigations in addition to the basic mandatory tests could help the clinician to potentially unravel the clinical enigma. The beauty of OCT findings (visible as a resolution of the outer plexiform edema and submacular detachment hand-in-hand with improved visual acuity) helps in proving the Gass's postulate on the pathogenesis of neuroretinitis. It states that the diffusion of fluid from the optic disc to the outer plexiform layer and through the outer limiting membrane to the subretinal space. Gass differed from the macular origin[9] proposed by Leber as “stellate maculopathy” (using fluorescein angiogram).

Although tuberculosis can cause neuroretinitis, the potential differentials of neuroretinitis are intriguing from its broad 3 variants[9] that aids the clinician to plan the treatment differently. Schmalfuss et al.[10] described the enhancement of the optic disc extending up to 4 mm posteriorly along the optic nerve and postulated them as highly specific for neuroretinitis related to cat scratch disease (CSD). However, subsequent authors have proven this sign in other causes as well.[9] On the contrary, the extensive enhancement of the retrobulbar optic nerve (beyond 4 mm behind the globe) is suggestive of idiopathic neuroretinitis, particularly the recurrent variety.[9] Unlike posterior scleritis, classically, neuroretinitis is not associated with pain with ocular movement.[9] Purvin et al.[9] stated that its presence should point a clinician to look for alternate etiology beyond the 3 recognized variants of neuroretinitis.

For an astute clinician, the pointers for CSD[9] apart from history of cat exposure include age less than 16 years, preceding systemic symptoms and poor visual acuity despite small/no Relative Affarent Pupillary Defect (RAPD). In contrast, features that suggest a low likelihood of Bartonella henselea are the ones which recur with no systemic symptoms, visual field defect outside the central field, preserved acuity with a large RAPD, poor recovery and evidence of a previous episode in either eye when steroid treatment should be a strong consideration. Simultaneous bilateral involvement[9] is suggestive of an underlying infectious etiology. This unique pattern is observed in diverse etiologies like Bartonella infection, Lyme Disease, mumps, secondary syphilis, toxoplasmosis rabies vaccine, chikungunya and dengue. As a rule of thumb, bilateral simultaneous neuroretinitis is uncommon and should suggest the clinician to rule out the possibility of malignant hypertension or increased intra cranial pressure before working up further on the lines of the above list.

In practice, the optic disc edema could be a manifestation of related entity called tubercular posterior scleritis, and merits specific clinical work up with adjunctive tests on those lines. More importantly, treating such cases with steroids alone without anti-tubercular drugs will obviously worsen the condition.

The million dollar question that is not answered is the end-point of anti-tubercular treatment in the eye. What is the general consensus is that ocular tuberculosis being an extrapulmonary, merits the duration beyond 6 months with an average of 9 months.