P092 Risk factors for active tuberculosis in patients with inflammatory bowel disease: A case-control study
S. Riestra1, R. de Francisco*2, M. Arias-Guillén3, C. Saro4, M. García-Alvarado5, J.M. Duque6, L. Blanco2, O. Castaño2, F. Muñoz5, I. Pérez-Martínez2, P. Martínez-Camblor7, D. Pérez-Hernández8
1Hospital Universitario Central de Asturias , Gastroenterology, Oviedo, Spain, 2Hospital Universitario Central de Asturias, Gastroenterology, Oviedo, Spain, 3Hospital Universitario Central de Asturias, Respiratory, Oviedo, Spain, 4Hospital de Cabueñes, Gastroenterology, Gijón, Spain, 5Hospital de León, Gastroenterology, León, Spain, 6Hospital San Agustín, Gastroenterology, Avilés, Spain, 7Hospital Universitario Central de Asturias, Oficina de Investigación Biosanitaria, Oviedo, Spain, 8Consejería de Salud, Servicio de Vigilancia Epidemiológica, Oviedo, Spain
Patients with inflammatory bowel disease (IBD) who receive anti-tumor necrosis factor (antiTNF) therapy are at increased risk of active tuberculosis (TB), mainly by reactivation of latent TB infection. Other risk factors for active TB in IBD patients are still poorly understood.
A retrospective, case-control, multicenter study was conducted. Patients who developed active TB after the IBD diagnosis were identified from the IBD databases of four hospitals in Northern Spain. As controls, 3 IBD patients per TB case were randomly selected from the same databases, matched on sex, age and year of IBD diagnosis. Both in cases and controls, IBD characteristics and risk factors for active TB were collected. Features of TB episode were described.
A total of 34 cases and 102 matched controls were included, of whom 50% were women; the mean age at IBD diagnosis was 36 years. 9 out of 34 cases were diagnosed between 1989-1999, and 25 cases from 2000 to 2012 (15 cases associated to antiTNF therapy). The cases were current or former smokers, and had received immunomodulator (IMM) or antiTNF treatment, more often than controls; also, the penetrating pattern was more common among cases with Crohn's disease than among controls.
In the univariate analysis, hospitalization and exposure to corticosteroids, IMM or anti TNF in the previous 3, 6 or 12 months, were associated to higher risk for active TB; on the other hand, higher levels of hemoglobin and albumin were associated to lower risk for TB. In the multivariate analysis, only antiTNF therapy in the previous 12 months (OR 8.34 [2.46-28.22], p: 0.001), hospitalization in the previous 3 months (OR 6.25 [1.45-26.90], p: 0.014), and albumin level at the TB diagnosis (OR 0.90 [0.82-0.98], p: 0.013), were significantly associated to active TB. A case of nosocomial infection was demonstrated by the genotyping of Mycobacterium tuberculosis.
Mean age at TB diagnosis was 43 ± 16 years; IMM therapy in the previous 12 months and extrapulmonary presentation were more frequent among TB cases associated to antiTNF treatment (80% vs 37%, p: 0.017; 63% vs 32%, p: 0.03, respectively). Active TB was diagnosed an average of 13 months after starting antiTNF therapy, and only 47% took place after 12 months of starting antiTNF.
In addition to the antiTNF treatment, hospitalization is associated with increased risk of active TB in IBD patients; a mechanism of nosocomial transmission of TB is possible in patients with IBD. Less than 50% of active TB associated with antiTNF occur in the first 12 months after starting this treatment; TB screening should be mandatory after starting treatment with antiTNF.