Tuberculosis (TB) Infection Prevalence, Incidence and Risk Factors Among Child and Adult Household Contacts of Adult TB Cases in India

Abstract

Background. Household contacts (HHC) of adult pulmonary TB (PTB) cases are at high risk of developing TB infection (TBI) and TB disease. Understanding the prevalence, incidence and risk factors of TBI in the first year following known household exposure is important for discovery and validation of novel biomarkers of disease progression. Methods. We analyzed baseline and follow up data of HHC exposed to newly diagnosed adult PTB cases from August 2014 to March 2016 in CTRIUMPH, an ongoing cohort study conducted at Pune and Chennai.HHC were screened for active TB and those without evidence of TB underwent TBI testing using QGIT (Quantiferon Gold In Tube; BD) and TST (Tuberculin Skin Test PPD 2/5 TUs; SPAN). Follow-up TBI testing was done at 4-6 months and 12 months among those negative by both tests prior. TBI was defined as QGIT positive according to manufacturer instructions or TST ≥5 mm. Logistic regression was performed to identify risk factors for prevalent TBI. Poisson regression with person-time offset was used to assess risk factors associated with TBI incidence. Results. Of 666 HHC included, 55% were female, 23% were children <14 years. At baseline, 71% had TBI, 55% (350 of 638) by TST, 58% (299 of 514) by QGIT. In children <14 years, 56% (86 of 153) had TBI. Multivariate analysis showed, no BCG scar (p = 0.024), residing in rural area (p = 0.038) and high smear grading of index (p = 0.026) as factors significantly associated with prevalent TBI. No difference in TBI was found by gender, smoking, alcohol or HIV status. However, all HIV-infected (0.1%) HHC were infected with TB. Incidence of TBI was 65 (95% Poisson exact CI: 49-86) per 1000 PY between 4 and 6 months and 12 months was 40 (CI: 22-73) per 1000 PY. At 4-6 months, 47% (37 of 78) new infections were identified by TST and 17% (13 of 76) by QGIT; at 12 months, 9 new additional infections were identified only by TST. Incident TB infection was associated with older age, no BCG scar and area of residence using TST and associated with older age, non-diabetics, higher BMI and higher smear grading of index case using QGIT. Conclusion. TBI prevalence is high among HHC; however, prevalence and incidence varied by TBI test used. Incidence rates were highest during the first few months post-contact tracing. We plan to assess recently reported biosignature profiles among those who did and did not develop new TBI or TB disease during follow-up.