Skip to main content
Download PDF

Enhancing the interferon-γ release assay through omission of nil and mitogen values

Respiratory Research volume 24, Article number: 179 (2023) Cite this article

Abstract

Purpose

To address the limited utility of the interferon (IFN)-γ release assay (IGRA) caused by its variability and inconsistency.

Methods

This retrospective cohort study was based on data obtained between 2011 and 2019. QuantiFERON-TB Gold-In-Tube was used to measure IFN-γ levels in nil, tuberculosis (TB) antigen, and mitogen tubes.

Results

Of 9,378 cases, 431 had active TB. The non-TB group comprised 1,513 IGRA-positive, 7,202 IGRA-negative, and 232 IGRA-indeterminate cases. Nil-tube IFN-γ levels were significantly higher in the active TB group (median = 0.18 IU/mL; interquartile range: 0.09–0.45 IU/mL) than in the IGRA-positive non-TB (0.11 IU/mL; 0.06–0.23 IU/mL) and IGRA-negative non-TB (0.09 IU/mL; 0.05–0.15 IU/mL) groups (P < 0.0001). From receiver operating characteristic analysis, TB antigen tube IFN-γ levels had higher diagnostic utility for active TB than TB antigen minus nil values. In a logistic regression analysis, active TB was the main driver of higher nil values. In the active TB group, after reclassifying the results based on a TB antigen tube IFN-γ level of 0.48 IU/mL, 14/36 cases with negative results and 15/19 cases with indeterminate results became positive, while 1/376 cases with positive results became negative. Overall, the sensitivity for detecting active TB improved from 87.2 to 93.7%.

Conclusion

The results of our comprehensive assessment can aid in IGRA interpretation. Since nil values are governed by TB infection rather than reflecting background noise, TB antigen tube IFN-γ levels should be used without subtracting nil values. Despite indeterminate results, TB antigen tube IFN-γ levels can be informative.

Introduction

Tuberculosis (TB) is a major health concern despite World Health Organization (WHO)-led global control efforts [1]. Latent tuberculosis infection (LTBI) undermines control of TB, while its treatment can prevent development of active TB. However, diagnosis of LTBI is hampered by the lack of a method to identify dormant bacilli; instead, the host immunological response to Mycobacterium tuberculosis is evaluated [2].

To diagnose LTBI, the interferon (IFN)-γ release assay (IGRA) has superior performance compared to the tuberculin skin test (TST) [3] However, the multiple procedures and manipulations required for the IGRA lead to variability and inconsistency [4,5,6,7], although the utility of the IGRA is difficult to challenge given that LTBI cannot be confirmed using other currently available techniques.

The most widely used IGRA method, QuantiFERON-TB Gold-in-Tube (QFT-GIT), involves TB-antigen-stimulated whole-blood incubation, immunoassay, and subsequent interpretation. The IFN-γ level is measured in incubated blood in a TB-antigen tube (TBAg), a “Nil” tube (Nil; negative control), and a mitogen tube (Mitogen; positive control). To diagnose LTBI, a cutoff value is derived by comparing active TB patients and a population with a very-low likelihood of TB infection, on the premise that IFN-γ levels are identical in active TB and LTBI. The TBAg minus Nil (TBAg − Nil) IFN-γ level is used to eliminate background noise [2, 8].

Studies on blood markers of TB, particularly CXCR3 ligands (also known as IFN-γ-inducible chemokines), have reported several issues with the IGRA [9, 10]. TB-antigen-stimulated marker levels are significantly higher in patients with active TB than in IGRA-positive controls (reflecting LTBI), contradicting the above premise. Furthermore, the chemokine and IFN-γ Nil levels were higher in patients with active TB than in controls, implying that M. tuberculosis increases basal levels and subsequent release of markers, even without TB-antigen stimulation. Consequently, subtraction of the Nil IFN-γ level from the TBAg value could compromise the significance of the results. These phenomena were prominently found for chemokines, but were less marked for IFN-γ, which is typically present in samples at low levels. Such issues can lead to false results, particularly near the cutoff values. Furthermore, variability and inconsistency have been reported with regard to indeterminate IGRA results based on abnormally high Nil or low Mitogen levels of IFN-γ [4].

Given these uncertainties and inconsistencies, the IGRA interpretation criteria may have been applied too strictly, leading to inappropriate decision-making and treatment, particularly in countries with high TB prevalence [11, 12]. Therefore, more flexible strategies, such as those based on a zone of uncertainty, have been proposed (but not implemented) [6, 13,14,15,16].

To address potential caveats regarding the assay, we analyzed the IGRA results and clinical data from a large population in a university hospital with a relatively high prevalence of TB infection; we focused on actual measured values. Although limited to a single center, this study included prompt and precise processing in a well-controlled laboratory system, which should enhance the reliability of the results [17].

Methods

Subjects

We retrospectively evaluated all IGRA tests performed in Ajou University Hospital between November 2011 and July 2019. All valid QFT-GIT results were included in this study. We assessed their purpose and reviewed underlying disease and laboratory data. The data were obtained on the day of blood sampling for the IGRA, but data obtained within 1 week of the IGRA were also used if there had been no change in clinical status.

This study was approved by the Institutional Review Board (AJOUIRB-DB-2022-331) and complied with the Declaration of Helsinki. The requirement for informed consent was waived because this was a very-low-risk retrospective study.

Diagnostic procedures and criteria

Diagnosis of active pulmonary TB was made according to the WHO guidelines [18]. Bacteriologically diagnosed TB was defined as compatible clinical and bacteriological culture findings or rapid molecular tests, such as polymerase chain reaction (PCR). A clinical diagnosis of TB was defined as compatible clinical, radiological, and pathological findings and an adequate treatment response. The diagnosis of active TB was verified by some of the author-investigators (at least one radiologist and two respiratory medicine specialists).

Diseases were diagnosed according to the clinician’s judgment and the medical records. If necessary, we also referred to the International Classification of Diseases, Tenth Revision [19]. and the available guidelines for individual diseases. Non-tuberculous mycobacteria (NTM) pulmonary infection was diagnosed based on the ATS/IDSA guidelines [20]. Cardiac disease included previous heart attack, heart failure, arrhythmia, ischemic heart diseases, and heart-valve problems requiring maintenance treatment, except for well-controlled hypertension. Immunosuppressive treatment included corticosteroids (≥ 2 mg/kg or ≥ 20 mg of prednisone or equivalent, daily for > 14 days); cytotoxic immunosuppressants for autoimmune diseases, organ transplants, and malignancies; and radiotherapy for malignancies. IGRA positivity was defined according to established criteria [21].

Lymphopenia was defined as a peripheral blood lymphocyte count < 1,000/mm3 (or < 2,000/mm3 in children < 6 years of age). Neutropenia was defined as an absolute neutrophil count < 2,500/mm3, regardless of age. Hypoalbuminemia was defined as a serum albumin level < 3.5 mg/dL. The results of TSTs using RT23 SSI 2TU were included if performed within 1 month of IGRA in cases with no change in clinical status.

IGRA

Using QFT-GIT (Cellestis, Victoria, Australia; Qiagen, Hilden, Germany) kits, 1 mL of blood was added to each of three Vacutainer tubes precoated with saline (Nil), M. tuberculosis ESAT-6, CFP10, and TB 7.7 antigens (TBAg), and phytohemagglutinin (Mitogen). The tubes were incubated for 16–24 h at 37 °C, and plasma was harvested and frozen until further analysis.

IFN-γ levels were measured by enzyme-linked immunosorbent assay according to the manufacturer’s instructions. IFN-γ levels > 10 IU/mL are generally reported simply as > 10 IU/mL, due to the limit of linearity of the standard curve. In this study, the actual values, if available, were used to minimize underestimation of the concentration. However, based on the configuration of the standard curve and the measurements made following serial dilution from the previous study, the actual concentrations were still subject to underestimation because the relevant values were in the non-linear portion of the curve [9].

Statistical analysis

Data were analyzed using IBM SPSS Statistics for Windows (ver. 25.0; SPSS Inc., Chicago, IL, USA) and MedCalc (ver. 20.115; MedCalc Software, Ostend, Belgium). The data, being nonparametric, are presented as medians with the interquartile range (IQR). Intergroup comparisons were performed using the Mann–Whitney U test and the Kruskal–Wallis test, followed by Bonferroni post hoc pairwise comparison. The Fleiss kappa coefficient was used to analyze agreement for categorical data. Spearman’s rank correlation coefficient was used to analyze correlations. Receiver operating characteristic (ROC) analysis was performed to obtain areas under the curve (AUC) and optimal cutoff values according to the Youden index. Logistic regression analyses were performed to identify factors affecting the IGRA results. In all analyses, P < 0.05 was considered indicative of statistical significance.

Results

Characteristics of the study population

In total, 9,378 cases were included in this study, including 431 in the active TB group. The demographic data are listed in Table 1. The IGRA was performed as part of a general health examination (n = 4,307), for diagnosis of LTBI and contact investigation (n = 1,624), and to aid in diagnosis or exclusion of active TB (n = 3,447). Except for 18 Caucasians, all subjects were Asian.

Table 1 Characteristics of the study population
Full size table

There were some instances of missing data. History-of-smoking data were obtained in 5,231 cases. Laboratory data were available for C-reactive protein (CRP) in 2,776 cases, for albumin in 2,844 cases, for white blood cell (WBC) counts in 8266 cases, and for differential WBC counts in 5,530 cases.

In the active TB group, 267 had pulmonary TB, 112 had extrapulmonary TB, and 52 had concurrent extrapulmonary-pulmonary TB. M. tuberculosis culture was positive in 227 cases, and there were 314 cases of bacteriologically diagnosed TB. The sites of extrapulmonary TB were the pleura (n = 59), intestines (n = 48), lymph nodes (n = 23), peritoneum (n = 14), bones (n = 8), meninges (n = 4), and others (n = 8).

The non-TB group comprised 1,513 IGRA-positive, 7,202 IGRA-negative, and 232 IGRA-indeterminate cases. The IGRA-negative subjects were younger than the IGRA-positive subjects, likely because of the decline in the incidence and prevalence of TB in South Korea [22].

The IGRA-indeterminate non-TB group exhibited a significantly lower lymphocyte count compared with other groups. By contrast, they had higher leukocyte and neutrophil counts. Furthermore, this group showed a significantly higher prevalence of autoimmune diseases, hematologic malignancies, solid malignancies, acute infections, and the use of corticosteroids and immunosuppressants.

After the IGRA, 20 non-TB subjects developed active TB by September 2022; 10 of those cases were IGRA-positive.

Evaluation of IGRA results

For the intergroup comparison, cases with IGRA-indeterminate results were excluded to ensure validity. The IFN-γ level of TBAg was ≥ 10 IU/mL in 82 cases (18.9%) of active TB and in 132 cases (6.7%) in the IGRA-positive non-TB group. A level of 10 IU/mL was assigned in 20 cases (4.6%) of active TB and 30 (2.0%) IGRA-positive non-TB cases. No Nil values exceeded 10 IU/mL in all cases.

The Nil IFN-γ levels were significantly higher in the active TB group (median 0.18 [IQR 0.09–0.45] IU/mL) than in the IGRA-positive non-TB (0.11 [0.06–0.23] IU/mL) and IGRA-negative non-TB (0.09 [0.05–0.15] IU/mL) groups (P < 0.0001). The TBAg IFN-γ levels were significantly higher in the active TB group (4.66 [1.51–8.90] IU/mL) than in the IGRA-positive non-TB (2.35 [1.01–6.22] IU/mL) and IGRA-negative non-TB (0.10 [0.06–0.18] IU/mL) groups (P < 0.0001). The TBAg − Nil IFN-γ levels were significantly higher in the active TB group (4.07 [1.14–8.17] IU/mL) than in the IGRA-positive non-TB (2.19 [0.84–5.84] IU/mL) and IGRA-negative non-TB (0.01 [− 0.01–0.04] IU/mL) groups (P < 0.0001). Additionally, the IGRA-positive non-TB group had significantly higher IFN-γ levels in Nil, TBAg, and TBAg − Nil compared to the IGRA-negative non-TB group (Fig. 1).

Fig. 1
figure 1

Nil (a), TBAg (b), and TBAg − Nil (c) IFN-γ levels in the active-TB, IGRA-positive non-TB, and IGRA-negative non-TB groups. Cases with indeterminate IGRA results were excluded. Boxes span the 25th to 75th percentiles, whiskers indicate the 10th and 90th percentiles, and bullets represent outliers. Nil nil tube, TBAg tuberculosis antigen tube, TBAg − Nil TBAg minus Nil, IFN-γ interferon-γ, TB tuberculosis, IGRA interferon-γ release assay, non-TB non-tuberculosis, IGRA + IGRA-positive, IGRA- IGRA-negative ****P < 0.0001

Full size image

The Nil and TBAg IFN-γ levels correlated significantly in all cases and in the subgroups (Fig. 2). Due to the cases with TBAg IFN-γ levels ≥ 10 IU/mL in the active-TB and IGRA-positive non-TB groups, the correlations may have been underestimated.

Fig. 2
figure 2

Spearman’s rank correlation analysis of the IFN-γ levels of the Nil and TBAg in all subjects (a) and in the active TB (b), IGRA-positive non-TB (c), and IGRA-negative non-TB (d) groups. IFN-γ interferon-γ, Nil nil tube, TBAg tuberculosis antigen tube, TB tuberculosis, IGRA interferon-γ release assay, non-TB non-tuberculosis

Full size image

With regard to indeterminate results, eight cases had a high Nil IFN-γ level, two of which were active TB cases. Six of these cases were classified as having both high Nil and low Mitogen levels of IFN-γ.

TST was performed in 872 cases, and the agreement between IGRA and TST was significant (κ = 0.46, P < 0.001).

Diagnostic performance

For adequate ROC analysis, combined cases of active TB and LTBI must be differentiated from TB-naïve cases. However, because true LTBI cases could not be identified, active TB cases were differentiated from the non-TB group. ROC analysis revealed that the IFN-γ level of TBAg had a slightly greater AUC than the IFN-γ level of TBAg − Nil in various comparisons (Table 2). The diagnostic performance of TBAg IFN-γ levels was underestimated because both the LTBI cases among the non-TB group and the active TB cases were positively affected by omitting subtraction of the Nil values from the TBAg values. Also, the presence of the LTBI cases resulted in a wide variation of the cutoff value, depending on the different groupings.

Table 2 Diagnostic performance of Nil, TBAg, and TBAg–Nil IFN-γ levels for differentiation of active TB cases from those in other groups
Full size table

Factors affecting IGRA results

For binary logistic regression analyses, the subjects were divided into high- and low-IFN-γ level groups according to the median Nil, TBAg, and TBAg − Nil IFN-γ levels (0.09, 0.14, and 0.02 IU/mL, respectively). The groups and IGRA positivity were designated as dependent variables. Underlying and accompanying diseases, and laboratory results, were used as predictors. Predictors that were significant in univariate analyses were entered into the multivariate analysis. For missing data, multiple imputation was performed to maximize the use of all complete data [23].

Using a high IFN-γ level of TBAg as the dependent variable, active TB, a history of TB, age, and sex were significant independent predictors (Table S1). Using a high IFN-γ level of Nil as the dependent variable, active TB was the most significant independent predictor. A history of TB, a solid malignancy, and autoimmune diseases showed significant results, but the latter two had negative coefficients probably due to the effects of immunosuppressive treatment. None of the other immunologic, inflammatory, and infectious diseases were identified as significant predictors for high Nil values (Table 3; Fig. 3). Using a high IFN-γ level of TBAg − Nil as the dependent variable, active TB, age, a history of TB, and autoimmune diseases were independent predictors (Table S2). The odds ratio of active TB was higher for the IFN-γ level of TBAg (62.34) than for that of TBAg − Nil (39.09). Using a positive IGRA result as the dependent variable, active TB, a history of TB, age, sex, smoking, NTM infection, and hypoalbuminemia were independent predictors (Table S3). Chronic diseases were collectively evaluated to evaluate their impact on the Nil IFN-γ levels. The Nil IFN-γ levels were not different between groups with and without chronic diseases (Table S4). In addition, when a high IFN-γ level of Nil was used as the dependent variable and chronic disease collectively was used as an independent variable, the presence of chronic disease was not significant, and only active TB and a history of TB were significant independent predictors (Table S5; Fig. S1).

Table 3 Univariate and multivariate binary logistic regression analyses to identify factors associated with high Nil IFN-γ levels in the IGRA (QFT-GIT)
Full size table
Fig. 3
figure 3

Forest plot of significant factors in the multivariate logistic regression to predict high Nil IFN-γ levels in the IGRA (QFT-GIT). Cases were divided into high- and low-IFN-γ Nil groups based on the median value (0.09 IU/mL). Cases with indeterminate results were excluded. Nil nil tube, IFN-γ interferon-γ, IGRA interferon-γ release assay, QFT-GIT QuantiFERON-TB Gold-in-Tube, TB tuberculosis

Full size image

With regard to the indeterminate results, the demographic characteristics of the cases are presented in Table S6. When an indeterminate IGRA result was used as the dependent variable, age, a solid malignancy, autoimmune disease, corticosteroid use, acute infection, lymphopenia, CRP, and hypoalbuminemia were independent predictors (Table 4; Fig. 4).

Table 4 Univariate and multivariate binary logistic regression analyses to identify factors associated with indeterminate results due to low Mitogen IFN-γ levels in the IGRA (QFT-GIT)
Full size table
Fig. 4
figure 4

Forest plot of significant factors in the multivariate logistic regression to predict indeterminate results in the IGRA (QFT-GIT). Cases were divided according to the criterion for indeterminate results (Mitogen − Nil IFN-γ level < 0.5 IU/mL). For definitions of lymphopenia and hypoalbuminemia, refer to the Methods. IGRA interferon-γ release assay, QFT-GIT QuantiFERON-TB Gold-in-Tube, Mitogen mitogen tube, Nil nil tube, IFN-γ interferon-γ

Full size image

Use of TBAg IFN-γ levels to interpret IGRA results

Based on the median value of the TBAg IFN-γ levels of the cases that met the original IGRA cutoff (TBAg – Nil IFN-γ level of 0.35 IU/mL), we arbitrarily set the cutoff IFN-γ level for TBAg to 0.48 IU/mL. As a result, 14/36 cases with IGRA-negative active TB (38.9%) became positive, and 1/376 IGRA-positive active TB cases (0.003%) became negative. Among the non-TB subjects, 284/7202 IGRA-negative cases (3.9%) became positive and 42/1513 IGRA-positive cases (2.8%) became negative. With regard to the indeterminate results, 15/19 cases with IGRA-indeterminate active TB (78.9%) were considered positive using the new cutoff IFN-γ level of TBAg, excluding the Mitogen criterion. Therefore, reinterpretation with the new cutoff TBAg IFN-γ value improved overall sensitivity for the detection of active TB from 87.2 to 93.7%. The specificity for active TB slightly decreased from 80.5 to 79.9%, although the true specificity could not be determined due to the existence of LTBI. The sensitivity and specificity changes could not be assessed comprehensively because conversion from “true” to “false” could not be determined in the non-TB groups due to the inability to diagnose LTBI.

Discussion

We reinterpreted IGRA results by assessing the implications of the values for individual QFT-GIT kit tubes. Here, we propose a revised IGRA interpretation. First, measurement of the Nil level may be unnecessary, or even detrimental; interpretation based on the TBAg IFN-γ level is preferable. To use this strategy, identification of appropriate cutoff values and ranges is necessary. The TBAg IFN-γ level may be useful, even in cases with a low Mitogen value. A result should not be dismissed as indeterminate merely because the Mitogen level is suboptimal. Due to the variability in the IGRA, its results should be interpreted using a borderline range instead of a point estimate, and clinical interpretation based on risk assessment for TB exposure should be performed for borderline results.

The advantages of IGRA reinterpretation using TBAg IFN-γ levels were demonstrated for the diagnosis of active TB, then reinforced by other analyses including intergroup comparison and logistic regression. We suppose that that the advantages would be further enhanced if all TB infections (including true LTBI) could be analyzed in comparison to the TB-naïve group. Additionally, actual changes in positivity according to the cutoff value are important and not fully reflected in the AUC differences. These changes were documented in our results, although the analyses were also limited to the active TB group. Also, stable and definitive cutoff values could not be defined, particularly due to the presence of abundant LTBI cases, which could not be accurately determined. Further studies in well-defined populations are needed to address this drawback.

Previously, we reported the diagnostic utility of TB-antigen-stimulated IFN-γ-inducible chemokine levels [9]. In TB-antigen-stimulated whole-blood assays, the Nil level is usually subtracted from TBAg values, which may reduce significance. Therefore, we suggested not subtracting the Nil level; however, this concept may not have been widely adopted because it is somewhat counterintuitive.

This study showed that TB infection affects Nil values. First, Nil IFN-γ levels were significantly higher in the active TB group than in the non-TB group; they were also higher in the IGRA-positive non-TB group than in the IGRA-negative non-TB group. Second, Nil and TBAg IFN-γ levels were significantly correlated. Third, the predictive performance of TBAg IFN-γ levels was superior to those of TBAg − Nil in ROC analysis. Fourth, regression analysis showed that the factor with the greatest influence on Nil IFN-γ levels was active TB, followed by previous TB infection. The odds ratio of active TB was greater for a high TBAg IFN-γ level than for a high TBAg − Nil IFN-γ level. These issues have already been discussed [9]. The mechanisms of action of IFN-γ and IFN-γ-inducible chemokines are closely related; the Nil values of IFN-γ are reportedly increased in active TB, but to a lesser degree than the Nil values of IFN-γ-inducible chemokines due to lower concentrations [9]. In the present study, increased Nil IFN-γ levels in active TB were more clearly identified in a large population. Increased Nil IFN-γ levels in active TB have also been reported in other studies [24, 25].

Activated peripheral blood T lymphocytes produce IFN-γ in TB infection [26,27,28]. This may explain the greater IFN-γ production by lymphocytes in TB-infected subjects than in TB-naïve subjects without TB-antigen stimulation. IFN-γ and CXR3 ligands in peripheral blood are increased without incubation or stimulation, although IFN-γ may be more difficult to analyze relative to other chemokines due to its relatively low concentrations [29,30,31].

With regard to indeterminate results, our data confirmed the influence of the factors proposed in previous studies [32,33,34,35,36]. In the present study, the use of well-controlled procedures may have contributed to the low incidence of indeterminate results, although the rates of indeterminate results are reportedly high in Asian populations [37]. Notably, acute infection and increased CRP levels were identified as independent contributors to indeterminate results. This may be related to the role of inhibitory receptors in lymphocyte activation during acute infection [38]. The rates of indeterminate results were reportedly higher in patients with COVID-19 [39].

The IGRA-indeterminate non-TB group exhibited a higher lymphocyte and a lower neutrophil count compared to the other groups. In the multivariate analysis, the lymphocyte count remained a significant factor influencing the indeterminate results. This finding is plausible in that the lymphocytes are the main player in the IGRA process. However, the significance of the neutrophil count was lost in the multivariate analysis, possibly due to its association with secondary phenomena caused by infection and other underlying conditions.

Among cases with indeterminate results, many active TB cases had high TBAg IFN-γ levels despite the presence of low Mitogen IFN-γ levels. If the TBAg IFN-γ level is disproportionately high compared to that of Mitogen, this could be used to evaluate positivity. In some cases, low Mitogen IFN-γ levels may result from local issues in the mitogen tube and may not be due to suppressed immunity. Therefore, individual cases should be carefully reviewed. Also, in cases with low Mitogen IFN-γ levels due to the decreased immunity, the possibility of false negative results should be considered, particularly near the cutoff value.

In this study, another finding associated with the variability of the results was that the TBAg − Nil calculations yielded high rates of negative values (> 50% for IGRA-negative non-TB subjects). Typically, the TBAg level is greater than or equal to the Nil level.

TB subjects had significantly higher TBAg IFN-γ levels compared to IGRA-positive subjects. This contradicts the assumption that IFN-γ levels are indistinguishable between active TB and LTBI. Although there was little difference in discriminatory power and a sizable overlap between the two groups, values near the cutoff could be affected. Furthermore, TBAg IFN-γ levels were underestimated in the active TB group: subjects with IFN-γ levels > 10 IU/mL were more prevalent in the active TB group, and IGRA-negative subjects were included in the active TB group but not in the IGRA-positive non-TB group.

Age was linked to high Nil and TBAg IFN-γ levels, as expected, given the decreasing incidence of TB via national control efforts [22]. Additionally, logistic regression analysis showed a higher incidence in men. Several significant predictors in univariate analyses lost significance in age-adjusted multivariate analysis.

The adoption of our interpretation strategy will require validation studies to determine the diagnostic performance in the most relevant populations and settings, including the determination of reference ranges and cut-offs, as these may differ in different population groups and settings.

This study had both limitations and strengths. First, since we included IGRA results from all age groups and clinical settings, there was considerable heterogeneity in the study population. However, the unselective inclusion of the IGRA results of a relatively large population may yield real-world results. Second, validity was reduced by the single-center design. However, because South Korea is a TB-endemic country, our population included active TB, LTBI, and younger TB-naïve subjects. Third, there were missing data for smoking and laboratory parameters because of the retrospective nature of the analysis. The large number of cases may overcome this shortcoming.

In conclusion, the results of our comprehensive assessment and reinterpretation strategy can facilitate tuberculosis control. Our results are a useful addition to the literature on the role of IGRA in the diagnosis of TB infection. Specifically, we suggest the use of TBAg IFN-γ levels for IGRA interpretation, because the use of Nil values may decrease the validity of the results. Mitogen levels may affect, but not absolutely determine, the validity of the results, because TBAg IFN-γ levels can be informative even in cases with low Mitogen IFN-γ levels.

Data Availability

The datasets generated during and/or analyzed during the current study are not publicly available due to the regulations and recommendations of the Institutional Review Board, but may be available from the corresponding author on reasonable request after obtaining the approval of the Institutional Review Board.

References

  1. Jeremiah C, Petersen E, Nantanda R, Mungai BN, Migliori GB, Amanullah F, et al. The WHO Global Tuberculosis 2021 Report - not so good news and turning the tide back to end TB. Int J Infect Dis. 2022;124:26–9.

    Article  Google Scholar 

  2. Shah M, Dorman SE. Latent tuberculosis infection. N Engl J Med. 2021;385:2271–80.

    Article  PubMed  Google Scholar 

  3. Lalvani A, Whitworth HS. Progress in interferon-gamma release assay development and applications: an unfolding story of translational research. Ann Transl Med. 2019;7:128.

    Article  Google Scholar 

  4. Banaei N, Gaur RL, Pai M. Interferon Gamma release assays for latent tuberculosis: what are the sources of variability? J Clin Microbiol. 2016;54:845–50.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Abubakar I, Stagg HR, Whitworth H, Lalvani A. How should I interpret an interferon gamma release assay result for tuberculosis infection? Thorax. 2013;68:298–301.

    Article  PubMed  Google Scholar 

  6. Moses MW, Zwerling A, Cattamanchi A, Denkinger CM, Banaei N, Kik SV, et al. Serial testing for latent tuberculosis using QuantiFERON-TB gold In-Tube: a Markov model. Sci Rep. 2016;6:30781.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Perry S, Sanchez L, Yang S, Agarwal Z, Hurst P, Parsonnet J. Reproducibility of QuantiFERON-TB gold in-tube assay. Clin Vaccine Immunol. 2008;15:425–32.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Pai M, Denkinger CM, Kik SV, Rangaka MX, Zwerling A, Oxlade O, et al. Gamma interferon release assays for detection of Mycobacterium tuberculosis infection. Clin Microbiol Rev. 2014;27:3–20.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Chung WY, Lee KS, Jung YJ, Lee HL, Kim YS, Park JH, et al. A TB antigen-stimulated CXCR3 ligand assay for the diagnosis of active pulmonary TB. Chest. 2014;146:283–91.

    Article  CAS  PubMed  Google Scholar 

  10. Chung WY, Lee KS, Park JH, Jung YJ, Sheen SS, Park JE, et al. TB Antigen-Stimulated CXCR3 ligand assay for diagnosis of Tuberculous Lymphadenitis. Int J Environ Res Public Health. 2021;18:8020.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Amour MA, Rees CA, Munseri PJ, Said J, Magohe AK, Matee M, et al. Serial T-SPOT.TB responses in Tanzanian adolescents: transient, persistent and irregular conversions. PLoS ONE. 2022;17:e0268685.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Yang Q, Ruan Q, Liu X, Shen Y, Jiang T, Wu J, et al. Preventive tuberculosis treatment effect on QuantiFERON TB-Gold in-tube testing in a high tuberculosis-endemic country: a clinical trial. Int J Infect Dis. 2020;91:182–7.

    Article  CAS  PubMed  Google Scholar 

  13. Nemes E, Rozot V, Geldenhuys H, Bilek N, Mabwe S, Abrahams D, et al. Optimization and interpretation of serial QuantiFERON testing to measure Acquisition of Mycobacterium tuberculosis infection. Am J Respir Crit Care Med. 2017;196:638–48.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Jonsson J, Westman A, Bruchfeld J, Sturegard E, Gaines H, Schon T. A borderline range for Quantiferon Gold In-Tube results. PLoS ONE. 2017;12:e0187313.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Fong KS, Tomford JW, Teixeira L, Fraser TG, van Duin D, Yen-Lieberman B, et al. Challenges of interferon-gamma release assay conversions in serial testing of health-care workers in a TB control program. Chest. 2012;142:55–62.

    Article  CAS  PubMed  Google Scholar 

  16. Whitworth WC, Hamilton LR, Goodwin DJ, Barrera C, West KB, Racster L, et al. Within-subject interlaboratory variability of QuantiFERON-TB gold in-tube tests. PLoS ONE. 2012;7:e43790.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Yun J, Chung H-S, Koh W-J, Chung DR, Kim Y-J, Kang E-S. Significant reduction in rate of Indeterminate results of the QuantiFERON-TB gold In-Tube test by shortening incubation Delay. J Clin Microbiol. 2013;52:90–4.

    Article  PubMed  Google Scholar 

  18. Eurosurveillance editorial t. WHO revised definitions and reporting framework for tuberculosis. Euro Surveill. 2013;18:20455.

    Google Scholar 

  19. Harrison JE, Weber S, Jakob R, Chute CG. ICD-11: an international classification of diseases for the twenty-first century. BMC Med Inform Decis Mak. 2021;21:206.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Griffith DE, Aksamit T, Brown-Elliott BA, Catanzaro A, Daley C, Gordin F, et al. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med. 2007;175:367–416.

    Article  CAS  PubMed  Google Scholar 

  21. Mazurek GH, Jereb J, Vernon A, LoBue P, Goldberg S, Castro K, et al. Updated guidelines for using Interferon Gamma release assays to detect Mycobacterium tuberculosis infection - United States, 2010. MMWR Recomm Rep. 2010;59:1–25.

    PubMed  Google Scholar 

  22. Song JH, Huh K, Chung DR. Modern history of tuberculosis in Korea. Infect Chemother. 2019;51:414–26.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Jeon Y, Kim MJ, Lee WI, Kim MH, Kang SY. Diagnostic utility of New equation for active tuberculosis based on parameters of Interferon-gamma release assay. Lab Med. 2017;48:214–9.

    Article  PubMed  Google Scholar 

  25. Lombardi G, Petrucci R, Corsini I, Bacchi Reggiani ML, Visciotti F, Bernardi F, et al. Quantitative analysis of Gamma Interferon Release Assay response in children with latent and active tuberculosis. J Clin Microbiol. 2018;56:e01360–17.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Andrade Junior DR, Santos SA, Andrade DR. Measurement of peripheral blood mononuclear cells producing IFN-gamma in patients with tuberculosis. Braz J Infect Dis. 2008;12:123–7.

    Article  PubMed  Google Scholar 

  27. Tambunan BA, Priyanto H, Nugraha J. Soedarsono. Cd4 + and Cd8 + T-Cells expressing Interferon Gamma in active pulmonary tuberculosis patients. Afr J Infect Dis. 2018;12:49–53.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Veenstra H, Crous I, Brahmbhatt S, Lukey P, Beyers N, van Helden PD, et al. Changes in the kinetics of intracellular IFN-gamma production in TB patients during treatment. Clin Immunol. 2007;124:336–44.

    Article  CAS  PubMed  Google Scholar 

  29. Chung W, Lee K, Jung Y, Kim Y, Park J, Sheen S, et al. Serum CXCR3 ligands as biomarkers for the diagnosis and treatment monitoring of tuberculosis. Int J Tuberc Lung Dis. 2015;19:1476–84.

    Article  CAS  PubMed  Google Scholar 

  30. Chung WY, Yoon D, Lee KS, Jung YJ, Kim YS, Sheen SS, et al. The usefulness of serum CXCR3 ligands for evaluating the early treatment response in tuberculosis: a longitudinal cohort study. Med (Baltim). 2016;95:e3575.

    Article  CAS  Google Scholar 

  31. Lee K, Chung W, Jung Y, Kim Y, Park J, Sheen S, et al. CXCR3 ligands as clinical markers for pulmonary tuberculosis. Int J Tuberc Lung Dis. 2015;19:191–9.

    Article  CAS  PubMed  Google Scholar 

  32. Jeong SJ, Han SH, Kim CO, Baek JH, Jin SJ, Ku NS, et al. Predictive factors for indeterminate result on the QuantiFERON test in an intermediate tuberculosis-burden country. J Infect. 2011;62:347–54.

    Article  PubMed  Google Scholar 

  33. Ahn SS, Kim HW, Park Y. Frequency and factors of Indeterminate QuantiFERON-TB Gold In-Tube and QuantiFERON-TB gold PLUS test results in Rheumatic Diseases. J Clin Med. 2021;10:4357.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Kaur M, Singapura P, Kalakota N, Cruz G, Shukla R, Ahsan S, et al. Factors that contribute to Indeterminate results from the QuantiFERON-TB gold In-Tube test in patients with inflammatory bowel disease. Clin Gastroenterol Hepatol. 2018;16:1616–21.

    Article  PubMed  Google Scholar 

  35. Banach DB, Harris TG. Indeterminate QuantiFERON(R)-TB gold results in a public health clinic setting. Int J Tuberc Lung Dis. 2011;15:1623–30.

    Article  CAS  PubMed  Google Scholar 

  36. Kim KH, Kang JM, Ahn JG. Low-dose steroids are associated with indeterminate QuantiFERON-TB gold In-Tube assay results in immunocompetent children. Sci Rep. 2021;11:6468.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Sharninghausen JC, Shapiro AE, Koelle DM, Kim HN. Risk factors for Indeterminate Outcome on Interferon Gamma Release Assay in Non-US-Born Persons screened for latent tuberculosis infection. Open Forum Infect Dis. 2018;5:ofy184.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Carmo AM, Love PE, Marshall AJ. Editorial: inhibitory receptors and pathways of lymphocytes. Front Immunol. 2020;11:1552.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Ward JD, Cornaby C, Schmitz JL. Indeterminate QuantiFERON gold plus results reveal deficient Interferon Gamma responses in severely ill COVID-19 patients. J Clin Microbiol. 2021;59:e0081121.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

None.

Funding

The authors did not receive support from any organization for the submitted work.

Author information

Author notes

  1. Yun Jung Jung and Ji Eun Park contributed equally to this work.

Authors and Affiliations

  1. Department of Pulmonary and Critical Care Medicine, Ajou University School of Medicine, 164 World cup-ro, Suwon, 16499, Gyeonggi-do, South Korea

    Yun Jung Jung, Ji Eun Park, Ji Won Park, Keu Sung Lee, Wou Young Chung, Joo Hun Park, Seung Soo Sheen & Kwang Joo Park

  2. Department of Radiology, Ajou University School of Medicine, Suwon, South Korea

    Seulgi You, Joo Sung Sun & Kyung Joo Park

  3. Department of Health and Medical Information, Ajou University Hospital, Suwon, South Korea

    Youn Jung Kim

Authors
  1. Yun Jung Jung
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ji Eun Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Ji Won Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Keu Sung Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Wou Young Chung
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Joo Hun Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Seung Soo Sheen
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Seulgi You
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Joo Sung Sun
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Kyung Joo Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Youn Jung Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Kwang Joo Park
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis (Kw. J. P., Y. J. J., J. E. P.). Study design, data collection, and drafting of the manuscript (Y. J. J., J. E. P.). Collection and review of the clinical and laboratory data, and the statistical analysis (J. W. P., K. S. L., W. Y. C., J. H. P., S. S. S.). Reviewing the radiologic findings (S. Y., J. S. S., Ky. J. P.). Extraction and organisation of data from electronic medical records (Y. J. K.). Main contributor to the study design, clinical evaluation, data collection, data analysis, and manuscript drafting (Kw. J. P.). All authors contributed to data interpretation and critical review of the manuscript.

Corresponding author

Correspondence to Kwang Joo Park.

Ethics declarations

Ethics approval and consent to participate

Ethics approval was obtained from the Institutional Review Board of Ajou University Hospital (the ethical committee of the institute). Patient consent for publication was waived by the Institutional Review Board of Ajou University Hospital due to the retrospective nature of this study and the fact that patients were de-identified.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

12931_2023_2485_MOESM1_ESM.docx

Additional file 1: table S1. Univariate and multivariate binary logistic regression analyses to identify factors associated with high TBAg IFN-γ levels in the IGRA (QFT-GIT).

12931_2023_2485_MOESM2_ESM.docx

Additional file 2: table S2. Univariate and multivariate binary logistic regression analyses to determine factors associated with high TBAg − Nil IFN-γ levels in the IGRA (QFT-GIT).

12931_2023_2485_MOESM3_ESM.docx

Additional file 3: table S3. Univariate and multivariate binary logistic regression analyses to identify factors associated with IGRA (QFT-GIT) positivity.

Additional file 4: table S4. Comparison of Nil IFN-γ levels according to the presence of chronic disease.

12931_2023_2485_MOESM5_ESM.docx

Additional file 5: table S5. Univariate and multivariate binary logistic regression analyses to identify factors associated with high Nil IFN-γ levels in the IGRA (QFT-GIT), with chronic disease collectively serving as an independent variable.

Additional file 6: table S6. Characteristics of the cases with indeterminate results of IGRA (QFT-GIT).

12931_2023_2485_MOESM7_ESM.jpg

Additional file 7: figure S1. Forest plot of the significant factors in the multivariate logistic regression to predict high Nil IFN-γ levels in the IGRA (QFT-GIT) with chronic disease collectively serving as an independent variable. Nil nil tube, IFN-γ interferon-γ, IGRA interferon-γ release assay, QFT-GIT QuantiFERON-TB Gold-in-Tube.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Jung, Y.J., Park, J.E., Park, J.W. et al. Enhancing the interferon-γ release assay through omission of nil and mitogen values. Respir Res 24, 179 (2023). https://doi.org/10.1186/s12931-023-02485-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12931-023-02485-4

Keywords

Respiratory Research

ISSN: 1465-993X

Contact us