Diagnosis of active tuberculosis using flow cytometry in HIV positive individuals in a tuberculosis endemic setting
The diagnosis of tuberculosis (TB) in Human Immunodeficiency Virus (HIV) infected individuals is a challenge due to atypical presentations including negative sputum smear and culture results, which renders traditional microbiological tests inadequate in this population. Immune based assays are an attractive alternative, but current commercial blood interferon gamma (IFNγ ) release assays have limited utility in areas of high Mycobacterium tuberculosis (M. tuberculosis) prevalence. Immune responses in the lung differ from those in peripheral blood, hence a better understanding of lung compartment-specific immune response to TB may assist in developing new diagnostic assays.We evaluated the IFNγ production from induced sputum (ISp) and blood samples of 31 HIV positive, smear-negative TB suspects who were being investigated for active TB. We analysed the IFNγ responses of total lymphocytes in ISp samples; and total, CD4+ T lymphocytes and memory CD27- CD4+ T lymphocytes in blood. IFNγ was quantitated after stimulation with two M. tuberculosis antigens, PPD (purified peptide derivative) and ESAT6 (early secretory antigen 6). Eleven sputum samples were excluded due to poor sample quality.We observed that IFNγ secretion from total lymphocytes in sputum was, significantly higher than blood after stimulation with either ESAT6 (0.64% vs 0.10%, p = 0.03) or PPD (2.04% vs 0.21%, p = 006), and the PPD specific IFNγ secreting lymphocytes in sputum could be used to differentiate between the M. tuberculosis negative and positive groups. None of the lymphocyte populations analysed in blood differentiated these two M. tuberculosis groups. The counterintuitive finding is that, median PPD specific IFNγ secreting lymphocytes in sputum were significantly lower (p = 0.04) in the M. tuberculosis positive (0.23%) cohort when compared to the M. tuberculosis negative group (2.04%). The sensitivity and specificity of this finding to discriminate active and latent TB at a diagnostic threshold of 1.25% PPD-specific IFNγ secreting total sputum lymphocytes is 78% and 70% respectively. The routine diagnostic utility of measuring T lymphocyte immune responses in sputum by flow cytometry however is not practical due to limitations imposed by the required cellular content and viability, as well as the need to process these samples within two hours of collection. These results however, are valuable and add insight into the lung compartment specific immune response to M. tuberculosis antigens.