Purpose
To evaluate clinical manifestations of patients with uveitis and scleritis of unknown origin and positive QuantiFERON–TB Gold In-Tube test (quantiferon) in a country not endemic for tuberculosis.
Design
Multicenter retrospective cohort study.
Methods
Retrospective review of the clinical, laboratory, and imaging data of 77 patients. Main outcome measures consisted of ocular and systemic features as well as results of laboratory examinations.
Results
Out of all, 60 of 71 (85%) were living for at least 6 months in tuberculosis-endemic regions. Location of uveitis was variable; posterior uveitis (29/77; 38%) was the most frequent. Two clinical entities were commonly noted: retinal occlusive vasculitis (21/77; 27%) and serpiginoid choroiditis (11/77; 14%). Antituberculosis treatment was completed in 32 patients; 29 of them (91%) achieved complete remission. Mean quantiferon level was 7.5 U/mL; 71% had values above 2 U/mL and 41% above 10 U/mL. We observed no associations between quantiferon levels and clinical and/or imaging features. Previous tuberculosis infection was diagnosed in 5 of 77 patients (6.5%), while hilar/mediastinal lymphadenopathy was found in 25 of 76 patients (33%). Of these, 12 were consistent with the diagnosis of sarcoidosis, 9 were typical for (prior) tuberculosis, and 4 were compatible with both diagnoses.
Conclusions
Ocular features of patients with idiopathic uveitis and positive quantiferon were diverse, but retinal occlusive vasculitis and serpiginoid choroiditis were common. The quantiferon levels were usually highly elevated and 33% of patients exhibited lymphadenopathy, suggesting frequently the diagnosis of sarcoidosis. Ocular inflammation reacted favorably to antituberculosis treatment, although only a small minority had documented (prior) tuberculosis.
Infection with Mycobacterium tuberculosis may result in active disease, characterized by the active multiplication of mycobacteria with formation of granulomas and tissue destruction of the infected organs, preferentially the lungs. Ocular involvement during active tuberculosis (TB) can consist of formation of granulomas in various parts of the eye or orbit. However, the majority of infected people will not advance to active disease but live with a latent TB infection.
During the last decade, interferon gamma release assay tests have been developed, which allow identification of patients with latent TB infection with better specificity than the tuberculin skin test and differentiate between infection and prior vaccination. It has become apparent that a portion of the patients with uveitis of so far unidentified origins in fact suffer from TB infection–related uveitis. The presumed pathogenesis of uveitis related to latent TB infection consists of inflammation caused by presence of low numbers of bacteria within the eye with or without superimposed immune reaction to mycobacterial or ocular antigens. So far, limited data are available on the ocular manifestations in latent TB infection and most information is based on populations from TB-endemic areas. Various ocular manifestations have been reported, including features distinct from characteristic granulomatous inflammation. Herein we evaluate the clinical manifestations of patients with uveitis of unknown origin and positive QuantiFERON–TB Gold In-Tube test (quantiferon; Cellestis Limited, Carnegie, Victoria, Australia) living in an area not endemic for TB.
Methods
We identified 96 consecutive human immunodeficiency virus–negative patients with uveitis and scleritis and positive quantiferon results from 3 tertiary centers (Erasmus Medical Center Rotterdam, University Medical Center Utrecht, and Rotterdam Eye Hospital). In the Netherlands TB is not endemic and Bacillus Calmette–Guérin (BCG) vaccination is not routinely performed. The clinical, laboratory, and imaging data of these patients were reviewed. This study was performed in accordance with the Declaration of Helsinki and with the approval for retrospective review of patients’ data by the Institutional Review Boards of Erasmus Medical Center Rotterdam, University Medical Center Utrecht, and Rotterdam Eye Hospital.
In this study, the participating centers previously included a quantiferon test as part of their screening for all new patients with uveitis and scleritis. In addition, several patients with unexplained chronic uveitis who underwent the screening before the interferon gamma release assay era were also tested. All our new uveitis patients received the quantiferon and tuberculin skin tests simultaneously during the screening; in several patients the tuberculin skin test was performed previously, but in all cases more than 6 months before the quantiferon tests were performed. The quantiferon cutoff value was 0.35 U/mL.
A specific cause of uveitis and/or associated systemic disorder was determined in 19 patients (ocular toxoplasmosis, n = 4; Behçet disease, n = 2; human leukocyte antigen B27–associated acute anterior uveitis, n = 2; viral anterior uveitis, n = 5; masquerade syndromes, n = 3; Vogt-Koyanagi-Harada disease, n = 1; Wegener disease n = 1; and presumed ocular histoplasmosis syndrome, n = 1). Uveitis in these patients was considered unrelated to latent TB infection and these cases were excluded from the final analysis. Four patients with biopsy-proven sarcoidosis and positive quantiferon results were included in our series because both diseases might share common pathogenic pathways and the differential diagnosis of sarcoidosis and TB can be difficult. All 77 included patients had been subjected to uveitis screening that included erythrocyte sedimentation rate, blood counts, serum angiotensin converting enzyme level, serology for syphilis, and, in those with anterior uveitis, also human leukocyte antigen B27 testing. Chest radiograph examination was performed in 69 patients; out of the remaining 8 patients, 7 had a computerized tomography (CT) scan primarily performed and in 1 patient the imaging data could not be retrieved.
The classification of uveitis was performed according to current criteria. In this study, the term intraocular inflammation included uveitis and scleritis. The diagnosis of serpiginoid chorioretinitis was performed in patients with multiple serpiginous-like chorioretinal lesions that exhibited multiple areas of activity and were not located adjacent to the optic disc. Immunosuppressive treatment was defined as systemic treatment with corticosteroids, sometimes in combination with methotrexate, azathioprine, or cyclosporine A. An anti-TB treatment was defined as a course of at least 3 antibiotics—isoniazid (INH), rifampin, pyrazinamide—in the initial 2-month phase of treatment followed by a 4-month consolidation phase using rifampin and INH (in total, 6 months). Three patients from one of the centers (Utrecht) received a longer consolidation phase (7 months, encompassing a total treatment of 9 months). In the absence of signs of pulmonary and/or systemic TB in the majority of patients, the decision to treat the patients with anti-TB drugs was based on the severity of the ocular disease and the patient’s preference in case of non-sight-threatening uveitis. The presence of macular edema was confirmed by optical coherence tomography in all clinically suspected patients.
Data were entered and analyzed using SPSS software version 20.0 (IBM Inc, Chicago, Illinois, USA). For all calculations with visual acuity (VA) data, Snellen VA was converted to the logarithm of the minimal angle of resolution (logMAR). For easier understanding, the logMAR results were converted back to Snellen VA and are presented as such in this report. In patients with unilateral involvement, visual acuity of the affected eye was included; to prevent the bias of bilateral disease, only the right eye was included in bilateral cases.
Results
The general characteristics of the 77 patients (44 male and 33 female) are given in Table 1 . The mean age at onset was 46 years and the median follow-up time was 3 years. Out of all patients, 24 of 71 (32%) were born in the Netherlands, of whom 13 stayed longer than 6 months in a country endemic for TB; 47 of 71 (66%) came from and stayed longer than 6 months in a country endemic for TB (demographic data could not be traced for 6 patients).
| Total N = 77 | Anterior Uveitis N = 19 | Occlusive Retinal Vasculitis N = 21 | Serpiginoid Chorioretinitis N = 11 | Lived in an Endemic Country >6 Months a N = 60 | Did Not Stay in Endemic Country >6 Months a N = 11 | |
|---|---|---|---|---|---|---|
| Mean age at onset of uveitis (y) | 46 | 49 | 40 c | 42 | 45.5 | 44.4 |
| Male-to-female ratio | 1.3:1 | 6:13 b | 20:1 c | 5:6 | 7:5 | 5:6 |
| Living in an endemic country >6 months | 60/71 (85%) | 15/17 (88%) | 19/21 (91%) | 8/11 (73%) | N.A. | N.A. |
| Contact with TB patient | 20/77 (26%) | 4/19 (21%) | 5/21 (24%) | 3/11 (27%) | 10/60 (17%) | 7/11 (64%) |
| Positive TST | 51/55 (93%) | 14/15 (93%) | 13/14 (93%) | 5/6 (83%) | 38/41 (93%) | 8/9 (89%) |
| Proven (previous) TB infection | 5/77 (7%) | 0/19 (0%) | 3/21 (14%) | 0/11 (0%) | 3/60 (5%) | 2/11 (18%) |
| Hilar and/or mediastinal adenopathy on radiologic and/or CT examination | 25/76 (33%) d | 5/19 (26%) | 7/21 (33%) | 4/11 (36%) | 21/60 (35%) | 4/11 (36%) |
| Mean quantiferon levels in U/mL (median) | 7.49 (5.75) | 7.89 (4.9) | 6.29 (2.2) | 5.55 (2.1) | 7.6 (7.1) | 5.5 (2.6) |
| Quantiferon levels >10 U/mL | 30/73 (41%) | 7/18 (39%) | 7/19 (37%) | 4/11 (36%) | 24/56 (43%) | 3/11 (27%) |
a Data were not available for 6 patients.
b Anterior uveitis occurred more frequently in female subjects ( P = .01).
c Occlusive retinal vasculitis occurred more frequently in male subjects ( P = .001). Patients with occlusive retinal vasculitis were younger ( P = .04) compared to the remainder.
d Of these 25 patients, 12 were considered compatible with the diagnosis of sarcoidosis (symmetrical configuration of lymph nodes and absence of old calcifications), 9 with tuberculosis (asymmetrical presentation, calcification, and old scars), and 4 were consistent with both diagnoses.
The anatomic location of intraocular inflammation was diverse, but posterior uveitis was most frequent ( Table 2 ). Anterior uveitis without posterior segment involvement was observed in 19 of 77 patients (25%). Anatomic location of uveitis did not differ between the patients born in the Netherlands and those born in an endemic country or between the patients who never visited an endemic country and those who were born in or stayed longer than 6 months in an endemic country.
| Characteristic | Total N = 77 | Unilateral N = 26 | Bilateral N = 51 |
|---|---|---|---|
| Unilateral-to-bilateral ratio | 1:2 | Not applicable | Not applicable |
| Location of inflammation | |||
| Anterior uveitis | 19 (25%) | 11 (42%) | 8 (16%) |
| Intermediate uveitis | 9 (12%) | 1 (4%) | 8 (16%) |
| Posterior uveitis | 29 (38%) | 10 (39%) | 19 (37%) |
| Panuveitis | 16 (21%) | 2 (8%) | 14 (28%) |
| Scleritis | 4 (5%) | 2 (8%) | 2 (4%) |
| Serpiginoid chorioretinitis | 11/77 (14%) | 4/26 (15%) | 7/51 (14%) |
| Retinal vasculitis | 34/77 (44%) | 10/26 (39%) | 24/51 (47%) |
| Occlusive retinal vasculitis | 21/77 (27%) | 6/26 (23%) | 15/51 (29%) |
| Papillitis a | 34/73 (47%) | 8/26 (31%) | 26/51 (51%) |
| Fibrovascular retinal lesions | 2/77 (3%) | 2/26 (8%) | 0 (0%) |
| Corneal opacities | 7/77 (9%) | 2/26 (8%) | 5/51 (10%) |
| Optic disc atrophy | 7/77 (9%) | 3/26 (12%) | 4/51 (8%) |
a Fluorescein angiography and/or appropriate clinical information were not available for 4 patients.
The clinical manifestations of inflammation varied widely; however, 2 explicit clinical posterior segment entities were commonly noted, specifically occlusive retinal vasculitis (21/77; 27%) and serpiginoid choroiditis (11/77, 14%). Occlusive retinal vasculitis manifested preferentially in male subjects ( P < .001) and was associated with younger age ( P = .04, Table 1 ). Intraocular granulomas were not observed. Corneal opacities (other than band keratopathy) were noted in 7 of 77 patients (9%). Granulomatous keratic precipitates were present in 10 of 77 patients (13%; 10/19 with anterior uveitis, of whom 4 were bilateral and 6 unilateral). Vitreous inflammation was present in 46 of 77 patients (60%) and vitreous hemorrhage in 6 of 77 (8%), and 2 patients exhibited retinal fibrovascular tumors. Vasculitis (arterial, venous, or both) was noted in 34 of 77 patients (44%) and 21 of the 34 (62%) exhibited vascular occlusions and retinal ischemic areas. The most common ocular complications included macular edema, glaucoma, and cataract ( Table 3 ). Vitreous hemorrhage and retinal neovascularizations were most common in retinal occlusive vasculitis compared to other types of uveitis ( P = .001, Table 3 ). The number of patients with 1 or more complications was 55 of 77 (71%). Out of 77 patients, 35 (45%) had to undergo 1 or more intraocular surgical procedures (n = 29) and/or laser treatment (n = 8).
| Total N = 77 | Anterior Uveitis N = 19 | Occlusive Retinal Vasculitis N = 21 | Serpiginoid Chorioretinitis N = 11 | Lived in an Endemic Country >6 Months a N = 60 | Did Not Stay in Endemic Country >6 Months a N = 11 | |
|---|---|---|---|---|---|---|
| Cystoid macular edema | 34 (45%) | 4 (21%) | 8 (38%) | 7 (64%) | 27 (45%) | 6 (55%) |
| Glaucoma | 18 (24%) | 5 (26%) | 7 (33%) | 2 (18%) | 14 (23%) | 4 (22%) |
| Cataract | 12 (16%) | 1 (5%) | 3 (14%) | 3 (27%) | 10 (17%) | 2 (17%) |
| Vitreous hemorrhage | 6 (8%) | 0 (0%) | 6 (28%) | 0 (0%) | 6 (10%) | 1 (9%) |
| Subretinal neovascularization | 8 (10%) | 0 (0%) | 7 (33%) | 1 (9%) | 6 (10%) | 1 (9%) |
| Other complications b | 12 (15%) | 1 (5%) | 7 (33%) | 3 (27%) | 13 (22%) | 0 (0%) |
a Data were not available for 6 patients.
b Includes hypotony, band keratopathy, hyphema, macular atrophy, macular pucker, tractional retinal detachment, and subretinal fibrosis.
Mean quantiferon value for the whole series was 7.49 U/mL (median 5.75 U/mL, Table 1 ); 52 of 73 patients (71%) had levels above 2 U/mL and 30 (41%) above 10 U/mL. No associations were found between the levels of quantiferon and various clinical manifestations, specifically when the following groups of patients were compared: patients with and without anterior uveitis (mean 7.9 U/mL vs 7.4 U/mL, P = .757), with and without occlusive retinal vasculitis (mean 6.3 U/mL vs 7.9 U/mL, P = .329), with and without serpiginoid choroiditis (mean 5.6 U/mL vs 7.8 U/mL, P = .261), and with and without hilar and/or mediastinal lymphadenopathy (mean 8.3 U/mL vs 7.1 U/mL, P = .437). An elevated serum angiotensin converting enzyme level was noted in 16 of 68 patients (24%). A positive association was noted between elevated serum angiotensin converting enzyme and quantiferon levels ( P = .04). A positive tuberculin skin test was present in 51 of 55 patients (93%), out of which 15 (15/51; 29%) had a test result exceeding 20 mm. An association between highly positive tuberculin skin test result and clinical and/or laboratory characteristics was not observed. There were 4 patients with negative tuberculin skin test (despite positive quantiferon results). Two patients were previously diagnosed with active TB, of whom 1 was still on anti-TB treatment, and 3 were diagnosed with TB during the assessment for uveitis. There was no difference in the levels of quantiferon for patients born in the Netherlands vs not born in the Netherlands ( P = .78), nor was there an association with age noted ( P = .095). Contact with a TB patient was recalled more often in the group who never visited an endemic country (7/11; 64% vs 10/60; 17%, P = .003).
Hilar and/or mediastinal adenopathy was diagnosed in 25 of 76 patients (33%) (10 by chest radiograph, 13 by CT scan and 2 by both imaging methods; Table 1 ) and was characterized as compatible with the diagnosis of sarcoidosis in 12 of 25 (48%); in 9 of 25 (36%) the adenopathy was compatible with the diagnosis of TB (asymmetrical configuration, calcifications, and/or presence of lung parenchyma lesions suggestive of TB) and in 4 (16%) was compatible with both diagnoses. Out of 12 patients with bilateral adenopathy on chest radiograph, 4 also had CT scans done, confirming the chest findings in 2; in 2 patients, however, the CT scan revealed no abnormalities. Out of 15 patients with hilar adenopathy on CT scan, 9 had concurrent chest radiographs available, of which 2 were concordant, but in 7 no abnormalities were seen on chest radiograph. No correlation was observed between the presence of hilar/mediastinal adenopathy and general characteristics such as sex, age, BCG status, quantiferon levels, diameter of tuberculin skin test induration, serum angiotensin converting enzyme level, location of uveitis, and presence of specific ocular characteristics. Biopsy of enlarged chest lymph nodes was performed in 12 of 25 patients (48%) with hilar and/or mediastinal adenopathy. Microscopic evaluation revealed only 1 patient with granuloma and associated necrosis; 2 showed nonspecific inflammatory infiltrates and the remainder exhibited granulomas without necrosis consistent with the diagnosis of sarcoidosis. The patient with necrotic granuloma in the lung showed negative results in culture, polymerase chain reaction (PCR), and Ziehl-Neelsen stain, but 3 (out of 12) other biopsies were positive for M tuberculosis (2 by culture, of which 1 was also positive by PCR and 1 by Ziehl-Neelsen stain only); these 3 patients had occlusive retinal vasculitis. Two out of these 3 patients were previously treated with immunosuppressive agents compared to 1 out of 9 with negative biopsies. Four additional patients underwent bronchoalveolar lavage, but none tested positive for M tuberculosis by staining, PCR, or culture. In addition, 4 patients had no thoracic lymphadenopathy but had enlarged lymph nodes elsewhere (in the pancreas, liver, spleen, and neck).
Anti-TB treatment was initiated in 42 patients. Eight patients stopped the medication prematurely (2 patients because of liver toxicity, 1 because of general malaise, and 5 for nonmedical reasons); 2 were still on treatment at the time of data collection ( Table 4 ). Thirty-two patients finished a full course of anti-TB treatment (10 in combination with corticosteroids, an additional 12 patients after a course with corticosteroids and/or other immunosuppressive drugs) and had an average follow-up after anti-TB treatment of 2.1 years (standard deviation: 3.2 months). Out of 32 patients who received a full anti-TB treatment course, 29 (91%) exhibited gradual decrease of inflammatory activity with complete disappearance of inflammation even after withdrawing all other (local and/or systemic) treatment. Two showed persisting intraocular inflammation; in 1 patient intraocular inflammation recurred 3 years after completion of anti-TB treatment. Development of VA during the various treatment regimens is shown in Table 4 . Mean VA at 1-year follow-up (after the initiation of anti-TB treatment) was improved in patients who finished a full course of anti-TB treatment (whether or not in combination with immunosuppressive medication) ( P = .004; Table 4 ). Patients who did not receive anti-TB treatment had a better visual acuity at onset compared to those with anti-TB treatment, but also this group showed an improved VA at 1-year follow-up ( P < .001, Table 4 ).
