One of the remarkable advances in the diagnosis of tuberculosis (TB) in the last few decades has been the development and wide usage of interferon (IFN)-γ release assays (IGRAs). These assays quantify the in vitro release of IFN-γ from T-cells stimulated with Mycobacterium tuberculosis (M. tuberculosis) antigens. The innovation, together with their prevailing employment, facilitates the early identification of at-risk individuals allowing for early intervention treatment for latent TB infection, thereby preventing the spread of TB in health care settings and communities where the TB patients reside [1,2].

Two major IGRAs are widely utilized;one is QuantiFERON^®-TB Gold In-Tube (QFT-GIT) (Cellestis, Victoria, Australia) and the other is T-SPOT.TB assay (Oxford Immunotec, Abingdon, United Kingdom). IGRAs, although not recommended for diagnosis of active TB in low- income countries [3], have been used to rule out latent TB infection especially for individuals who have received BCG or screening of health care workers [2,4].QFT-GIT simultaneously employs three M. tuberculosis antigens, ESAT-6, CFP-10, and TB7.7, to which cell- mediated immune responses are measured in a single tube [5].QFT- GIT replaces QuantiFERON^®TB-Gold (QFT-G), which was the former generation of QuantiFERON applying two M. tuberculosis antigens with a separate reaction. QFT-GIT, therefore, is more sensitive and specific relative to QFT-G due to the additive effect [6]. With those features in mind, we hypothesized that, in patients with TB, the QFT- GIT should detect TB positivity for a longer period after the initiation of treatment. To this end, the change of IFN-γ values in patients with active TB serially registered to the study, and receiving anti-TB drugs, was examined for a period of one year after treatment.

Patients

Adult patients aged 18 and over and diagnosed with active pulmonary tuberculosis, based on a positive culture result for M. tuberculosis from respiratory samples [7], were prospectively enrolled to the study at the Saitama Medical University Hospital (Saitama, Japan) between June 6, 2011 and August 31, 2014. Patients were treated with the anti-TB drugs: isoniazid (INH), rifampicin (RFP), ethambutol (EB) and pyrazinamid (PZA). Treatment regimens for TB treatment were based on established guidelines [8] according to the drug susceptibility result and patients’ tolerance.

Ethical Consideration

The research protocol for the study was approved by the institutional review boards of Saitama Medical University Hospital (institutional review board [IRB] number 11-005, 6 June 2011). Informed consent was obtained from participants before enrollment to the study.

QuantiFERON^®-TB Gold In-Tube

Blood samples for the QFT-GIT were collected from active TB patients. The QFT-GIT test was performed according to the manufacturer's instructions. The test results with an IFN-γ level ≥ 0.35 IU/ml in TB antigens (greater than the nil control value) were considered positive, and those with < 0.10 IU/ml were considered negative. Those with IFN-γ levels between 0.10 and 0.35 were considered indeterminate. Using those cutoffs, the sensitivity and specificity of the QFT-GIT test were reported at 84% and 99%, respectively [9].

Fourteen patients with active pulmonary TB, culture-confirmed M. tuberculosis isolated from respiratory samples, were enrolled to this study (Table 1). Subsequently they were treated with anti-TB drugs [8] and blood samples were collected for QFT-GIT at the time of diagnosis (1^st QFT), completion of treatment (2^nd QFT) and 12 months after (3^rd QFT) (Figure 1). Overall, the IFN-γ values, although showing a downward trend, were far above the cutoff value to be diagnosed as a negative result even after a year of treatment. At the 1st QFT, 12 out of 14 patients (85.7%) had a positive QFT, with a mean IFN-γ level of 2.37 IU/ml. Similarly, 75% of those who completed TB treatment for either 6 months or 9 months also had a positive QFT at the 2nd QFT, in which the mean IFN-γ level was 1.74 IU/ml. One year after treatment, 88.9% of QFT still remained steadily positive with mean IFN-γ levels of 1.70 IU/ml. No patients have relapsed after completing TB treatment.

Table 1: Characteristics of pulmonary tuberculosis patients enrolled in the study (n=14) AFB, acid-fast bacilli; INH, isoniazid; RFP, rifampicin; PZA, pyrazinamide; EB, ethambutol, SM: streptomycin. Data were at the time of diagnosis.

Figure 1: The change in IFN-γ level in patients with active pulmonary tuberculosis.Blood sampling for the QFT-GIT was performed before initiating anti-TB drugs (1st QFT), after completion of treatment (2nd QFT) and 12 months after treatment (3rd QFT). Two patients who missed 2nd QFT are depicted as the dashed lines. Shaded area indicates IFN-γ levels below the cutoff value for the positive QFT 0.35 IU/ml (greater than the nil control value). Data at the bottom were means ± SD of the IFN-γ levels using Prism 6 (GraphPad Software, Inc. La Jolla, CA). Statistical analysis was calculated using unpaired t-tests (Welch's method for uneven variances). N.S. = not significant.

Although a variety of studies using QFT-G and QFT-GIT have been conducted to assess the conversion ratio of IFN- γ release after initiating TB treatment, the literature focusing on the long-lasting potency of IGRAs is limited. In this paper, we have therefore examined the period of IFN-γ release assessed by QFT-GIT in response to M. tuberculosis antigens in serially registered TB patients who were followed up until one year after treatment.

In previous studies in which the former version QFT-G was used, it was concluded that treatment of active TB disease resulted in reduced IFN-γ release in response to the M. tuberculosis antigens [10]. Despite a decline in IFN-γ release in IGRAs by TB treatment, some literature indicated that the conversion ratio was heterogeneous, supposedly due to different genetic characteristics and the immune competence in the host who carried TB [10]. To minimize genetic influence, only studies conducted in Japan were compared with our data. Patients with active TB who were 83% positive by QFT-G reduced to 58% after treatment, followed by 50% 6 months later [11]. Likewise, treatment of active TB disease led 76% of positive patients to 62%, which subsequently dropped to 36% in a year after [12].These data, together with our results, indicated that the incubation of lymphocytes with three M. tuberculosis antigens simultaneously results in apparent TB positivity for a longer period after the onset of treatment, which also demonstrates that QFT- GIT is the more sensitive examination as opposed to QFT-G.

This raises the concern that TB reinfection, although rarely experienced in the general public, could occur under specific circumstances, such as for health-care workers (HCWs) in TB wards or residents in highly TB prevalent areas [13-15].IGRAs, such as QFT- GIT or T-SPOT.TB, are not only used as a diagnostic tool for latent TBinfection but also implemented for the investigation of TB transmission among HCWs [2-4,15]. With their high sensitivity and specificity, applying QFT-GIT to screening for latent TB infections is clearly advantageous, yet it will be harder to distinguish the TB reinfection from those whose TB has previously been treated due to its long lasting potency.

To sum up, assessing the status of being infected with TB is easily implementable owing to the latest advancements in IGRAs, such as QFT-GIT; however there are limitations to evaluating the infectious condition of M. tuberculosis in specific environments. Additionally, appraising long-lasting potency of IGRA is difficult based on our study with small sample size. Thus, further investigation with a large scale is required.

Authors have no conflict of interest.

We would like to thank my colleague Johnathan Canton, Ph.D., Program in Cell Biology, Hospital for Sick Children, Toronto, Canada, for proofreading of manuscript draft.

This work was supported by the Saitama Medical University Hospital. No additional external funding received for this study.