Testing for Sarcoidosis, Syphilis, Tuberculosis, and Lyme Disease
Nearly every presentation of ocular inflammation has potential association with sarcoidosis, syphilis, tuberculosis, or Lyme disease. These conditions must be ruled out in all uveitis cases, but they are among the most challenging conditions to rule out. Definitive diagnostic tests are the exception and indirect testing dominates the workup.
Syphilis is caused by Treponema pallidum, a spirochete that cannot be cultured under normal conditions. It is rarely diagnosed by dark field microscopy or immunofluorescence of tissue biopsy. The mainstays of testing are direct and indirect treponemal antibody tests. The direct tests—the fluorescent treponemal antibody (FTA-ABS), Treponema pallidum particle agglutination assay (TPPA) and microhemagglutination assay (MHA-TP)—serve as seroconversion markers. Once a patient has been exposed to syphilis, these markers remain forever positive, even if the infection has been completely treated. Despite sensitivities and specificities near 99%, these tests are more useful for ruling out syphilis than establishing the diagnosis.
Given the relatively low pretest probability of syphilis in the United States (approximately 1/10,000), even with a 1% false-positive rate, only a small fraction of positive tests will indicate active disease. Many uveitis specialists therefore recommend a direct test as the initial screening test except in the highest-risk populations. A positive direct treponemal test should be followed by one of the indirect tests, such as the rapid plasmin reagent (RPR) or Venereal Disease Research Laboratory (VDRL) tests. The titers of these indirect antibodies vary with disease load and become undetectable in fully treated disease. Thus, a patient with positive FTA-ABS or TPPA and positive RPR or VDRL test results has active syphilis necessitating treatment. However, the patient who is FTA-ABS/TPPA positive and RPR or VDRL negative has a false-positive FTA-ABS result, latent syphilis, neurosyphilis, or previously treated syphilis.
False-positive FTA-ABS test results occur in about 1% to 2% of the population and are associated with pregnancy, collagen vascular diseases, exposure to other treponemes, and Lyme disease. Since false positives are rare, uveitis associated with positive FTA-ABS and negative RPR requires lumbar puncture and performance of VDRL on the cerebrospinal fluid to rule out neurosyphilis, typically in collaboration with a neurologist or infectious disease specialist.
Several newer modalities of testing for syphilis are entering clinical usage. Commercial testing for antisyphilis IgG, available in some hospitals, has lower false-negative and false-positive rates than indirect testing. Polymerase chain reaction (PCR) diagnostics are capable of detecting Treponema pallidum DNA in order to make a diagnosis from biopsy specimens (such as vitreous). This technique is particularly useful in posterior segment or panuveitic manifestations of syphilitic disease, but its true sensitivity and specificity are currently not known.
Tuberculosis (TB) is endemic in most parts of the world. More than 2 billion people worldwide have been infected with Mycobacterium tuberculosis (mTB). Like Treponema pallidum, mTB is very difficult to culture in the laboratory. Definitive diagnosis is through observation of mTB on biopsy material with acid-fast stains or by PCR amplification of mTB DNA. Neither one is commonly achieved in routine clinical practice, and so it is often necessary to resort to indirect means for diagnosis.
The tuberculin skin test (TST), including the Mantoux test and the purified protein derivative (PPD) test, is the mainstay of diagnosis worldwide. Forty-eight hours after intradermal administration of the protein derivative, the injected spot is read for extent of induration. Interpretation depends on the patient's pretest probability (Table 3). All patients with PPD reaction greater than 15 mm are positive for the test, and all patients with less than 5 mm induration are negative. Interpretation of intermediate values (5 to 10 mm) depends on factors such as whether the patient is an immigrant from an endemic area, has known TB contacts, is immunocompromised, or is a healthcare worker. Most positive PPD tests are not associated with active disease but will reflect exposure or latent disease.
The TST is cost-effective and easy to administer and read, but it requires the patient to return to a healthcare professional for reading of the test, and it can be falsely positive in the millions of individuals worldwide who have been immunized with Bacille Calmette–Guérin (BCG) vaccine. Both drawbacks can be overcome with interferon gamma release assays (IGRAs). In the United States, the QuantiFERON TB-Gold test (Cellistis, Valencia, California) and the T-SPOT TB test (Oxford Immunotec, Marlborough, Massachusetts) are FDA-approved IGRAs. In these tests, the patient's peripheral blood leukocytes are purified and mixed with a very specific set of mTB peptides that do not include BCG cross-reacting proteins. Resulting lymphocyte gamma interferon production is quantified. Sensitivity is comparable to the TST and specificity is higher due to loss of BCG cross-reactivity. However, the test is considerably more expensive than TST testing. There is controversy as to whether the QuantiFERON test should be first line in TB testing or reserved for indeterminate cases.
Chest x-ray is a useful adjunct to skin testing. However, it is important to recognize that most cases of uveitic TB from paucibacillary or miliary disease are not accompanied by pulmonary disease; thus a negative chest x-ray in the setting of positive PPD and suspicious disease (eg, serpiginoid uveitis or panuveitis) does not rule out TB. Such cases can be problematic since treatment commits the patient to 9 months or more of a multidrug regimen with significant potential toxicity. In such cases, it is worth obtaining vitreous biopsy for PCR testing. Consultation with an infectious disease specialist is essential in these cases.
Definitive diagnosis of sarcoidosis requires histopathologic demonstration of noncaseating granulomata in biopsied tissues. Suspicious conjunctival nodules should be biopsied; this is the easiest way to make a definitive diagnosis.
Lacking definitive histopathology, there is considerable disagreement as to the optimal means for testing for sarcoidosis. The core diagnostic test remains the chest x-ray. Radiographic evidence of hilar adenopathy or interstitial fibrosis necessitates referral to a pulmonologist for pulmonary function tests, bronchoscopy, or mediastinoscopy.
In the absence of x-ray findings, workup becomes more problematic. High-resolution chest computed tomography (CT) scanning with contrast enhancement has high sensitivity for detection of pulmonary sarcoidosis, but it should be followed up with either mediastinal biopsy or bronchoscopy if the clinician can justify that these will alter management. Similarly, positive gallium scan or even 18F-labeled fluorodeoxyglucose positron emission tomography scans can provide supportive evidence for tissue biopsy.
Several serum enzyme or electrolyte tests have limited utility in the diagnosis of sarcoidosis. The angiotensin-converting enzyme (ACE) test measures serum concentration of this enzyme produced by macrophages and is present in higher levels in patients with a high systemic granuloma load. When studied for ocular sarcoidosis in a population of uveitis patients, despite 84% sensitivity and 95% specificity, the positive predictive value of ACE testing was only 47%, and the negative predictive value even lower. Because ACE levels are not specific for sarcoidosis, a positive test never makes the diagnosis of sarcoidosis. ACE levels can also be depressed by treatment (eg, systemic corticosteroids) and will be near zero in patients taking ACE-inhibitor antihypertensive drugs. Similar arguments apply to lysozyme (positive predictive value of 12%) and urinary 24-hour calcium testing. Although the utility of these indirect tests for sarcoidosis is extremely limited, their diagnostic accuracy improves substantially when combined with positive imaging studies such as chest radiography or high-resolution chest CT.
Lyme disease, caused by the spirochete bacterium Borrelia burgdorferi, can also have protean manifestations, ranging from unilateral or bilateral anterior uveitis to intermediate uveitis to polyneuropathy to scleritis. As with all clinical tests, pretest probabilities should guide testing for Lyme disease. In this case, geography is a dominant feature; Lyme risk varies tremendously throughout the United States, corresponding to the habitat of its vector (ticks of the Ixodes genus), with highest probabilities in the Northeast and northern Midwest and minimal disease in the mountain states (Figure 2). Serologic testing, with screening ELISA followed by confirmatory Western blot screening, is the mainstay of diagnosis. A significant proportion of certain populations have previous exposure to Lyme (eg, ~6% of outdoor workers in New Jersey are seropositive for Lyme exposure). PCR diagnostics are capable of detecting Borrelia burgdorferi, but sensitivity and specificity of this test for ocular disease remain unknown.