Intraocular Tuberculosis

Reema Bansal; Aman Sharma; Amod Gupta


Expert Rev Ophthalmol. 2012;7(4):341-349. 

In This Article


Microbiological or histopathological evidence of MTB from intraocular fluids or tissues constitutes the gold standard for diagnosing intraocular TB. However, the diagnosis of intraocular TB still remains largely presumptive as the ocular tissue is rarely sampled. The ocular specimens seldom provide definitive results, such as presence of AFB on smear or culture of MTB. Establishing a definitive diagnosis further becomes difficult due to absence of uniform diagnostic criteria, combined with difficulties encountered in conducting confirmatory laboratory tests. TST has been extensively used to assess immunological evidence of MTB infection either in patients with latent or active TB. Recent advances in TB detection by molecular diagnosis or IFN-g release assays (IGRAs) have improved the specificity of diagnosis. The current criteria of making a presumptive diagnosis of intraocular TB include presence of suggestive clinical symptoms and signs, corroborative evidence such as a positive TST, positive IGRAs, chest x-ray (CXR) findings, exclusion of known nontubercular uveitic entities, and a positive response to conventional ATT. An extraocular evidence of TB in a patient with uveitis also aids in diagnosing intraocular TB. We used arbitrary criteria for diagnosing tubercular uveitis in our population and found specific clinical signs as markers predicting TB as the cause of uveitis in a TB-endemic area.[24] The currently available laboratory methods provide indirect evidence of tubercular etiology of uveitis.

Indirect Evidence

TST The TST or the Mantoux test has been used for several years to detect latent TB, that is, a person who is infected with MTB but has no active TB disease. An intradermal injection of the purified protein derivative RT 23 (five tuberculin units) is performed on the ventral surface of the forearm. An induration (not erythema) of 10 mm or more after 48–72 h is considered positive, indicating latent TB.[101] However, in patients with HIV infection and those who are immunosuppressed, induration of 5 mm is taken as positive. Patients may show false-positive results if they have infection with atypical mycobacteria or have received Bacillus Calmette–Guérin (BCG) vaccination. In populations who receive BCG vaccination, the effect wanes with time.[51] The main advantages of this test are its low cost and easy availability. However its two-step process, subjective nature, nonreproducibility, booster effect, low sensitivity and specificity are some of the limitations of this test.

IGRAs The specificity of immunological diagnosis of TB has improved with the introduction of IGRAs, such as the QuantiFERON-TB Gold In-Tube (QFT; Cellestis Inc., VIC, Australia) and ELISpotPLUS (T-SPOT.TB, Oxford Immunotec, UK).[52] The QFT is approved by the US FDA and many other countries. It is an objective, single-visit blood test that quantifies the IFN-g response of T cells after in vitro stimulation of patientlymphocytes by MTB antigen. It is a more specific marker for MTB and previous exposure than TST, but has not been found superior to TST in sensitivity for use as a screening test or first-line study in TB-uveitis.[53] It has the advantage of not being affected by prior BCG vaccination and atypical mycobacteria. However, its high cost and technical difficulty are its disadvantages. Initial results of QFT testing in patients with uveitis suggested its use in uveitis patients with history of BCG vaccination and those in an immunocompromised state in place of TST.[54] A significantly high number of QFT-positive patients with SLC strengthened the long-suspected involvement of MTB in its pathogenesis.[34] A large study, however, found it only slightly superior to TST in diagnosing TB-uveitis.[55] It is important to interpret QFT together with TST. In a South Indian patient population, QFT alone was not found specific for intraocular TB.[56] The T-SPOT.TBtest is an overnight enzyme-linked immunospot (ELISpot) assay-based test. Approved in Europe, the test is under evaluation by the US FDA. It enumerates IFN-g-producing individual T cells after antigenic stimulation, and is more sensitive than TST for diagnosing latent TB infection. In the various studies done so far, the T-SPOT.TB test has been found to correlate more with MTB exposure than the TST. In patients with suspected TB-uveitis, Ang et al. found a higher sensitivity and a lower specificity of the TST than the T-SPOT.TB test.[57] The accuracy of diagnosing TB-uveitis increases when both tests are used in combination. Negative T-SPOT.TB or TST do not rule out TB-uveitis. An 'equivocal' T-SPOT.TB result in patients with TB-uveitis was associated with patients aged >55 years. Such patients were likely to have a negative QFT-TB Gold In-Tube result.[58] In HIV-infected and other immunocompromised patients, and in young children, a higher sensitivity of T-SPOT.TB test has been reported compared with the TST.

Albini et al. reported on the available clinical data that QuantiFERON yielded the highest positive post-test probability (11.3%) compared with TST (1.0%) and T-SPOT.TB (5.2%).[59] However, IGRAs have certain other limitations such as their inability to differentiate latent TB infection from active disease. While there is a known risk of progression of active disease in TST positive patients,[60] such information is still lacking for patients with a positive IGRA test. The IGRAs provide only a supportive evidence for diagnosing intraocular TB.[59] According to the guidelines by NICE, the use of IGRAs in developed countries with a low TB prevalence is recommended in patients who have a positive TST and in immunocompromised patients.[102] In areas with an intermediate-to-high prevalence of TB, TB might be more frequently associated with chronic uveitis than reported earlier. QFT was found to be useful for initiating ATT in immunosuppressed patients with chronic posterior uveitis.[61] It is important that if planning both TSTs and IGRAs, a blood sample is drawn first for IGRA before performing the intradermal injection of tuberculin to prevent possible prior sensitization and thereby increasing the false positivity of the IGRA test. Patients with autoimmune diseases who require anti-TNF agents are routinely subjected to IGRA tests to rule out latent TB, as administrations of these agents may lead to reactivation and dissemination of TB.

Chest Radiography & Computed Tomography Chest radiography (CXR) is an important investigation for evaluating patients with suspected intraocular TB as lungs are the primary sites of TB infection. Any evidence of active/healed or reactivated TB on chest radiographs with relevant clinical signs corroborates the presumtive diagnosis of intraocular TB. However, since nearly 60% of the patients with extrapulmonary TB do not have pulmonary TB, intraocular TB is often associated with the absence of any radiological evidence of pulmonary TB.[61] As many as 10–14% of patients with primary TB may have normal CXR findings.[62,63]

Radiographic lesions of active pulmonary TB can be detected on computed tomography (CT) scanning in some people with no abnormalities on CXR.[64] CT scans of the chest provide increased sensitivity and specificity for diagnosing active TB, and may be considered in patients with inconclusive CXRs.[65] However, radiation hazards and costs should be carefully considered.

PET/CT Doycheva et al. evaluated the use of [18F]fluorodeoxyglucose (FDG) PET/CT in 20 patients with uveitis and positive QFT test.[66,67] An increased FDG uptake in mediastinal or hilar lymph nodes was detected in nine patients (45%). PET/CT-guided lymph node biopsy enabled MTB detection in culture in two patients. Remission of uveitis was achieved in nine of 11 (82%) patients treated with ATT. It was found to be an effective tool for identifying lesions appropriate for biopsy. Mehta et al. reported increased FDG uptake in a patient with recurrent posterior uveitis due to presumed ocular TB (positive QFT test) and a normal CT chest scan.[68] Increased FDG uptake was seen in right pararacheal, precarinal and bilateral hilar nodes, and in the left choroid. However, the availability and high cost of PET/CT limit its use as a routine diagnostic tool in patients of uveitis with suspected TB.

Direct Evidence

Direct evidence is based on evidence of bacilli by Ziehl–Neelsen staining or culture from ocular fluids (aqueous, vitreous or subretinal fluid), or a positive PCR.

Microscopic & Histopathological Examination Demonstration of AFB on direct smear or culture of MTB from ocular samples constitutes a definitive diagnosis. The low sensitivity of AFB-positive smear and the minute quantity of intraocular fluid available at a time (not more than 50–100 µl) are major drawbacks. Moreover, diagnosis by culture reports is delayed by 6–8 weeks due to the slow-growing nature of bacteria. Early reports of microbiologically proven intraocular TB in immunocompetent patients included cases of iritis, vitritis, multifocal choroiditis and choroidal tuberculoma.[26,28,30,35] TB as a cause of uveitis has been proven by histopathologic or microbiologic examination in some reports. However, these reports suggest difficulty in arriving at a proper diagnosis before a major intervention such as enucleation.

PCR The limitations of conventional diagnostic tools have encouraged the use of PCR for MTB detection in intraocular fluids. It is a powerful tool for rapidly detecting the mycobacterial genome in clinical and research specimens. For the first time, Kotake et al. detected MTB in aqueous fluids by PCR in two cases with active retinal vasculitis.[42] In patients with presumed tubercular granulomatous uveitis 37.7% of aqueous samples were positive.[41] Any of the intraocular specimens (aqueous or vitreous humor or subretinal fluid, or tissue specimen) may be subjected to PCR.[32,41,42,69–74]

Real-time or quantitative PCR (qPCR) allows real-time monitoring of the progress of a PCR reaction. It also allows fast detection and quantifies the pathogen load in the sample. The other added advantages include minimal risk of carryover and cross-contamination. Real-time PCR has been used in detecting the mycobacterial load in tuberculous uveitis with promising results.[18,75,76]