Study Of Clinical Characteristic And Diagnosis Of Tuberculous Posterior Uveitis

PurposeTo determine the clinical characteristics of posterior tuberculous uveitis in Chinesepatients and analyze related risk factors of the disease.MethodsThe hospital-based observational case series study included patients who attended athird-referral hospital and presented with chronic and recurrent posterior uveitiswithout detected etiology in the period from March2011to March2013. The patientsunderwent the purified protein derivative test （PPD） and/or interferon gammarelease test （IGRA）. Patients with positive test results（induration≥20mm on thetuberculin skin test and/or a positive interferon gamma release test） and patientswith systemic manifest tuberculosis received anti-tuberculous therapy（ATT）consisting of isoniazid, rifampicin, pyrazinamide and ethambutol. If patientsshowed a positive response to the ATT after4weeks （defined as resolution or atleast remission of ocular inflammation）, they continued receiving the standard ATTaccording to the Centers for Disease Control guidelines for a minimum of6months.Patients who responded to the ATT and did not show recurrence of uveitis in thefollow-up period were diagnosed as posterior tuberculous uveitis and formed thestudy group. The remaining patients were diagnosed as non-tuberculous posterioruveitis and formed the control group.The clinical characteristics were comparedbetween the two groups. Univariate and multivariate logistic regression analyseswere carried out to determine the clinical predictors of tubercular posterior uveitis.The odds ratio （OR） and the relative95%confidence interval （95%CI） werecalculated. Sensitivity, specificity, positive predictive value, negative predictivevalue, positive likelihood ratio, negative likelihood ratio, and the area under the receiver operator curve for each clinical signs and for various combinations ofparameters were calculated.Results85patients were eligible for inclusion into the study. None of these patients had anunderlying human immunodeficiency virus infection or were under systemicimmunosuppression. The study group consisted of46patients （27men,58.7%）receiving, and showing a positive response to, the anti-tuberculous therapy. Themean age in the study group was45.7±16.0years. The control group included38patients （22men,57.9%） who had a non-tuberculous uveitis. The mean age was39.3±11.3years. Bilateral uveitis was detected in20（43.4%） patients in the studygroup and in14（36.8%） patients in the control group. Both groups did not varysignificantly in gender, age and laterality of the uveitis. In the study group ascompared with the control group, significantly （P=0.02） more participants lived inthe rural region （26（57%） versus12（32%） patients; P=0.02）. The meanbody mass index was significantly （P=0.038） lower in the patients of the studygroup. Multifocal choroiditis （n=9（20%） versus n=1（3%）; P=0.04） and retinalvasculitis （n=25（54%） versus n=8（21.1%）; P=0.002） were significantly morecommon in the study group. Out of25patients with retinal vasculitis in the studygroup,11patients （44%） additionally showed choroiditis lesions, compared withonly one （13%） out of8patients in the control group （P=0.014）. In univariatelogistic regression analysis, presence of multifocal choroiditis, retinal vasculitis, andvasculitis associated with choroiditis were significantly associated with tuberculosis.In multivariate regression analysis, multifocal choroiditis（Odds Ratio（OR）:32.1）,choroidal granuloma （OR:21.4） and retinal vasculitis （OR:11.2） wereindependent predictors of posterior tubercular uveitis. Sensitivity and specificitywere20%and97%,13%and97%,54%and79%for multifocal choroiditis,choroidal granuloma, and retinal vasculitis, respectively.Combinations of clinicalsigns and investigations as compared to the single clinical parameter showed a higher diagnostic precision with an area under the receiver operator curve of0.95at asensitivity of96%and a specificity of94%.ConclusionsOur findings suggest that intraocular tuberculosis is a major cause for uveitis to beconsidered in patients with “idiopathic” uveitis in China. Patients with primarilyunexplained posterior uveitis and showed multifocal choroiditis, choroidal granulomaand retinal vasculitis had a high predictive value for the diagnosis of tuberculousuveitis. The presence of any one of the clinical signs of multifocal choroiditis,choroidal granuloma, or retinal vasculitis in patients with “idiopathic” uveitis may bea strong indicator to perform a tuberculin skin test, an interferon gamma release assay,or chest radiograph and computed tomography to detect a potential tuberculous cause.Tuberculosis is an important part in the differential diagnosis of unexplained uveitisin China.