IFNgamma response to Mycobacterium tuberculosis, risk of infection and disease in household contacts of tuberculosis patients in Colombia

Abstract

Objectives: Household contacts (HHCs) of pulmonary tuberculosis patients are at high risk of Mycobacterium tuberculosis infection and early disease development. Identification of individuals at risk of tuberculosis disease is a desirable goal for tuberculosis control. Interferon-gamma release assays (IGRAs) using specific M. tuberculosis antigens provide an alternative to tuberculin skin testing (TST) for infection detection. Additionally, the levels of IFNgamma produced in response to these antigens may have prognostic value. We estimated the prevalence of M. tuberculosis infection by IGRA and TST in HHCs and their source population (SP), and assessed whether IFNgamma levels in HHCs correlate with tuberculosis development.

Methods: A cohort of 2060 HHCs was followed for 2-3 years after exposure to a tuberculosis case. Besides TST, IFNgamma responses to mycobacterial antigens: CFP, CFP-10, HspX and Ag85A were assessed in 7-days whole blood cultures and compared to 766 individuals from the SP in Medellín, Colombia. Isoniazid prophylaxis was not offered to child contacts because Colombian tuberculosis regulations consider it only in children under 5 years, TST positive without BCG vaccination.

Results: Using TST 65.9% of HHCs and 42.7% subjects from the SP were positive (OR 2.60, p<0.0001). IFNgamma response to CFP-10, a biomarker of M. tuberculosis infection, tested positive in 66.3% HHCs and 24.3% from the SP (OR = 6.07, p<0.0001). Tuberculosis incidence rate was 7.0/1000 person years. Children <5 years accounted for 21.6% of incident cases. No significant difference was found between positive and negative IFNgamma responders to CFP-10 (HR 1.82 95% CI 0.79-4.20 p = 0.16). However, a significant trend for tuberculosis development amongst high HHC IFNgamma producers was observed (trend Log rank p = 0.007).

Discussion: CFP-10-induced IFNgamma production is useful to establish tuberculosis infection prevalence amongst HHC and identify those at highest risk of disease. The high tuberculosis incidence amongst children supports administration of chemoprophylaxis to child contacts regardless of BCG vaccination.

Figure 1. Study profile.

HHCs = Household contacts, IGRA = IFNγ Release Assay, TST = Tuberculin Skin Test, SP = source population, CFP-10 = Culture Filtrate Protein-10.

Figure 2. IFNγ production by HHCs and SP.

IFNγ production in whole blood cultures stimulated with four mycobacterial antigens in household contacts and source population. A. IFNγ levels produced in non stimulated and CFP, CFP-10, HspX and Ag85A, stimulated cultures. B. Bar graphs depicting the percentage of positive responders and OR in HHCs and SP. C. Stacked bars represent a modification of the IFNγ production levels proposed by Andersen et al showing: Negative: <22pg/mL, Low: 22–99 pg/mL Medium: 100–999 pg/mL, High: ≥1000 pg/mL. HHCs: Household Contacts; SP: Source Population.

Figure 3. IFNγ production levels by age.

Geometric means of IFNγ production in response to CFP (A) and CFP-10 (B) by age in household contacts (HHCs) and individuals from source population (SP). Horizontal lines depict the cut-off value (22 pg/mL).

Figure 4. Hazard of TB development according to IFNγ production.

Hazard levels of tuberculosis development taking the modified version of IFNγ production categories proposed by Andersen et al as baseline predictors of disease. Negative: <22pg/mL, Low: 22–99 pg/mL Medium: 100–999 pg/mL, High: ≥1000 pg/mL. HHCs: Household Contacts SP: Source Population.