Comparative Results of QuantiFERON-TB Gold In-Tube and QuantiFERON-TB Gold Plus Assays for Detection of Tuberculosis Infection in Clinical Samples

The QuantiFERON-TB Gold plus (QFT-Plus) assay, an interferon gamma (IFN-γ) release assay (IGRA), was recently introduced as the next version of the QuantiFERON-TB Gold In-Tube (QFT-GIT) assay for diagnosing latent tuberculosis (TB). Whereas the QFT-GIT assay uses only one TB tube that induces a cell-mediated immune (CMI) response of CD4⁺ T cells, the QFT-Plus has an additional TB antigen 2 tube (TB2) for the CMI response of CD8⁺ T and CD4⁺ T cells, in addition to a TB antigen 1 (TB1) tube for the CMI response of CD4⁺ T cells only.

ABSTRACT

The QuantiFERON-TB Gold plus (QFT-Plus) assay, an interferon gamma (IFN-γ) release assay (IGRA), was recently introduced as the next version of the QuantiFERON-TB Gold In-Tube (QFT-GIT) assay for diagnosing latent tuberculosis (TB). Whereas the QFT-GIT assay uses only one TB tube that induces a cell-mediated immune (CMI) response of CD4⁺ T cells, the QFT-Plus has an additional TB antigen 2 tube (TB2) for the CMI response of CD8⁺ T and CD4⁺ T cells, in addition to a TB antigen 1 (TB1) tube for the CMI response of CD4⁺ T cells only. We compared the results of the QFT-Plus and QFT-GIT assays as routine clinical tests for diagnosing TB. Samples from 220 patients referred for routine IGRA in various clinical departments were used. Correlations between IFN-γ levels in the QFT-GIT and QFT-Plus assays were strong and showed good agreement (kappa value = 0.69). Seven cases with positive QFT-GIT assay results and negative QFT-Plus assay results showed IFN-γ values near the cutoff value. However, 10 cases with active TB, recent TB, or immune problems showed discordance with the positive results only in the TB2 tube in QFT-Plus, unlike the negative results in TB1 and TB tubes. In these cases, IFN-γ levels in the TB2 tube were significantly higher than those in other tubes. This is the first study to compare these assays as routine IGRAs in the clinical setting. The QFT-Plus assay showed good agreement with the QFT-GIT assay and is presumably advantageous for patients with active TB, recent TB, and specific immune conditions involving CD8⁺ T-cell responses.

INTRODUCTION

Tuberculosis (TB), caused by members of the Mycobacterium tuberculosis complex, including M. tuberculosis, M. bovis, and M. africanum, is one of the top 10 causes of death worldwide and is highly contagious, as the bacteria spread between people through airborne droplet nuclei (https://www.who.int/en/news-room/fact-sheets/detail/tuberculosis). Typical symptoms of early infection within weeks to months include those related to the lungs or, rarely, to other parts of the body; however, approximately 90% of TB infections are latent tuberculosis infection (LTBI), representing a noncontagious asymptomatic condition that affects approximately one-quarter of the global population. Only 5 to 10% of people with LTBI will develop active TB within 2 years after exposure (1), and the lifetime chance of progression to active TB is approximately 5 to 15% (https://www.who.int/en/news-room/fact-sheets/detail/tuberculosis). The tuberculin skin test (TST), which was developed by Koch in 1890, is occasionally used to assist in the diagnosis of active TB, but was mainly used for more than 100 years to diagnose LTBI based on the response of the skin to intradermal injection of a purified protein derivative of M. tuberculosis (2). However, the test cannot distinguish active from latent TB (3). Moreover, false-positive results can be obtained in the TST because of bacillus Calmette-Guerin (BCG) inoculation and nontuberculous mycobacteria (NTM) infection; these are likely to occur in patients who are immunocompromised, were administered a live virus vaccine in the 4 weeks before the TST, or recently suffered from a viral infection (4, 5).

In the early 2000s, the interferon gamma (IFN-γ) release assay (IGRA) was developed as an alternative to the TST. This test stimulates T lymphocytes sensitized to M. tuberculosis from patients’ blood using antigens specific to M. tuberculosis that are distinct from those in BCG inoculation or most NTM infections and measures the level of IFN-γ secreted by T lymphocytes (6). As IGRA is a blood test, unlike the TST, patients do not have to visit the hospital twice to obtain the result. Moreover, this assay is less susceptible to false-positive results from BCG inoculation or NTM infection and is more sensitive than the TST for patients who are immunosuppressed and children (7, 8). However, the IGRA cannot distinguish latent from active TB (9). Neither the TST nor IGRA measures the infection, but rather the immunological memory, which can result from current or past active TB or from subclinical infection resulting in LTBI.

Two commercial IGRAs are commonly used for diagnosing LTBI, the QuantiFERON-TB Gold In-Tube (QFT-GIT) assay and the T-SPOT.TB assay. Recently, the QuantiFERON-TB Gold Plus (QFT-Plus) assay was developed as a new version of the QFT-GIT and was approved by the Food and Drug Administration in 2017. QFT-GIT contains a TB antigen tube with a peptide cocktail that stimulates the proteins ESAT-6, CFP-10, and TB7.7 to induce a cell-mediated immune (CMI) response from CD4⁺ T-helper lymphocytes (10). QFT-Plus has two distinct TB antigen tubes, the TB antigen 1 (TB1) tube, which contains peptides from ESAT-6 and CFP-10 but not TB7.7 to induce a CMI response from CD4⁺ T-helper lymphocytes, and the TB antigen 2 (TB2) tube, which contains an additional set of peptides along with the components in TB1 to induce a CMI response from CD8⁺ cytotoxic T lymphocytes (11–16). CD8⁺ T lymphocytes can reflect exposure to M. tuberculosis pathogens in situations where there are few CD4⁺ lymphocytes or no active TB and are thus considered important for detecting TB infection in immunosuppressed individuals, such as patients with human immunodeficiency virus infection or with active TB whose clinical symptoms are unclear (17, 18).

Several studies have compared QFT-GIT and QFT-Plus in various subjects, demonstrating high agreement between the two assays (11–16, 19, 20). QFT-Plus was as accurate as QFT-GIT, despite the absence of TB7.7 antigen for subjects with active TB and M. tuberculosis-uninfected volunteers (12). The agreement between QFT-GIT and QFT-Plus was 95.6% in health care workers (13), 93.7% in immunocompromised patients (14), and 96.6% in TB risk patients and health care workers (15). The main advantage of QFT-Plus is that it can provide information regarding the additional CD8⁺ cell response to distinguish between active and latent TB (16) and between recent and old infections (20), and it can improve TB detection in patients who are immunocompromised. However, the relative performance of these two assays in routine clinical practice has not been examined.

The Republic of Korea shows a high incidence of TB, with an estimated 77 cases per 100,000 people in 2016 (21); therefore, the government has increased the budget and strengthened patient management policies for controlling this disease. Furthermore, in the Republic of Korea, it is recommended that all newborn babies within 1 month of age are administered the BCG vaccine, so an IGRA not affected by BCG is important in the screening of latent TB. Significant discrepancies between TST and QFT-GIT have been reported in the Republic of Korea (22, 23), but there has been no comparison between TST and QFT-Plus. Therefore, in this cross-sectional study, we compared the results of the QFT-Plus and QFT-GIT assays, as well as those of TST and QFT-Plus, for samples requested as routine tests in clinical setting for the diagnosis of TB.

MATERIALS AND METHODS

Study design.

Identification of LTBI through IGRA is performed before organ transplantation before administering immunosuppressive agents, before recruiting a health check-up, and on visa issuance; if LTBI is predicted, an anti-TB drug is recommended. IGRA is also considered an adjuvant diagnostic tool for active TB infection, including in cases of pulmonary or extrapulmonary TB. At the Severance Hospital (Seoul, Republic of Korea), the QFT-GIT assay (Qiagen, Hilden, Germany) is used when routine IGRA is requested from each clinical department. For this study, we simultaneously used both the QFT-GIT and QFT-Plus (GFT-Plus; Qiagen) to compare 220 blood specimens for which IGRA was requested from October to November 2017 at the Severance Hospital. The two tests were performed at the same time and place by the same person to minimize variables. In general, an IGRA is referred to diagnose latent TB and sometimes to identify active infection, including cases of pulmonary or extrapulmonary TB. Among the 220 samples tested, 125 (56.8%) were used to check latent TB in patients with compromised immunity or those expected to undergo strong immunosuppressive treatment, and 25 (11.4%) were tested to detect latent TB in healthy individuals for visa or check-up (Table 1). A total of 63 (28.6%) samples were tested to rule out TB infection of organs, such as the lung, central nervous system, lymph nodes, and bowel, and the remaining 7 (3.2%) samples were tested for potential exposure to TB. Patient charts were reviewed, and the results of the TST were available from our hospital for 53 individuals. The TSTs and IGRAs were performed on the same day. If the diameter of the induration was greater than 10 mm within 48 to 72 h, the test was considered positive. The study was approved by the Yonsei Medical Center Ethics Review Committee (approval no. 4-2019-0876). If there were discrepancies between tests, the clinicians were informed to help determine TB preventive treatment.

TABLE 1.

Basic characteristics of the patients

Patient characteristic	Valuea
Sex, male (no. of patients [%])	123 (55.9)
Age, median (IQR)	47 (28–58)
Clinical department (no. of patients [%])
Gastroenterology	33 (15.0)
Rheumatology	30 (13.6)
Infectious internal medicine	26 (11.8)
Pulmonology	23 (10.5)
Ophthalmology	20 (9.1)
Transplantation surgery	15 (6.8)
Hematology	14 (6.4)
International clinic	13 (5.9)
Family medicine	11 (5.0)
Dermatology	8 (3.6)
Nephrology	7 (3.2)
Oncology	7 (3.2)
Other	13 (5.9)
Purpose of assay (no. of patients [%])
Check for latent TB in patients with compromised immunity or in those for whom immunosuppressants are being considered	125 (56.8)
Check of latent TB for visa	13 (5.9)
Check of latent TB for health checkup	12 (5.5)
TB infection involving lung	45 (20.4)
TB infection involving other organs besides lung	18 (8.2)
Exposure to TB	7 (3.2)

^aTotal N = 220. IQR, interquartile range.

QFT-GIT and QFT-Plus.

The QFT-GIT and GFT-Plus assays were performed according to the manufacturer’s instructions. First, the patient’s blood, collected through venipuncture, was collected in lithium heparin tubes, and then 1 ml of blood was transferred into each tube of the QFT-GIT and QFT-Plus kits within 4 h: a nil tube, a TB tube, and a mitogen tube for QFT-GIT or a nil tube, a TB1 tube, a TB2 tube, and a mitogen tube for GFT-Plus. The mitogen tube was used as a positive control to provide information regarding an individual’s immune response. The nil tube was used to adjust for background interference, such as excessive IFN-γ circulating in the blood or the presence of heterophile antibodies. Each tube was incubated for 16 to 24 h at 37°C. The tubes were then centrifuged for 15 min at 2,000 × g, and IFN-γ levels (IU/ml) in the plasma were measured by enzyme-linked immunosorbent assay.

IFN-γ values in TB, TB1, TB2, and mitogen tubes were subtracted from and adjusted to those in the nil tube. These corrected values were expressed as TB-Nil, TB1-Nil, TB2-Nil, and Mitogen-Nil, respectively. In the QFT-GIT assay, the results were considered positive when the TB-Nil value was ≥0.35 IU/ml and ≥25% of the nil value. Results showing a nil value of >8.0 IU/ml or a mitogen value of <0.5 IU/ml were considered indeterminate. The same interpretation was applied to QFT-Plus, but the result was considered positive if either TB1-Nil or TB2-Nil satisfied the positive condition.

Statistical analysis.

All statistical analyses were performed using the Analyse-it Method Validation Edition version 3.5 (Analyse-it Software, Leeds, England). Agreements among the QFT-Plus, QFT-GIT, and TST assays were evaluated using Cohen’s kappa values. IFN-γ levels were compared between QFT-Plus and QFT-GIT by Spearman’s correlation coefficient and Passing-Bablok regression. The differences between continuous variables were analyzed using the Wilcoxon signed rank test for pairwise comparisons. A P value of <0.05 was considered statistically significant.

RESULTS AND DISCUSSION

Clinical characteristics.

The median age of patients included in this study was 47 years (interquartile range [IQR], 28 to 58 years), and 123 of the specimens (55.9%) were derived from male patients (Table 1). The major diseases of the patients included inflammatory bowel disease, autoimmune disease, Behcet’s uveitis, and psoriasis, based on samples sent from the departments of gastroenterology, rheumatology, ophthalmology, and dermatology, respectively, and the IGRA was conducted mainly to check for latent TB in patients recommended for immunosuppressant administration or in patients with compromised immunity (125/220, 56.8%). The IGRAs requested by transplant surgeons or the hematology department were mainly used to identify latent TB before transplantation, and those requested by the international health care center were for visa applications. In addition, the departments of infectious medicine, respiratory medicine, and family medicine requested IGRAs to check for cases of active TB. Overall, 68.2% (150/220) of the tests were performed to check for latent TB, and 31.8% (70/220) were to identify active TB infection.

Comparison of IFN-γ values in QFT-GIT and QFT-Plus.

The median IFN-γ values of TB-Nil, TB1-Nil, and TB2-Nil were 0.01 (IQR, 0 to 0.21), 0.025 (IQR, 0 to 0.20), and 0.04 (IQR, 0.01 to 0.28), respectively. There were no differences between the results for TB-Nil and TB1-Nil, whereas the IFN-γ level of TB2-Nil was higher than that in the other two tubes (P < 0.05). This indicated that the TB2 tube in QFT-Plus, with antigens for both CD8⁺ and CD4⁺ cells, had a higher IFN-γ level than the TB tube or the TB1 tube with antigens only for CD4⁺.

Spearman’s correlation coefficient (r_s) was 0.766 (95% confidence interval [CI], 0.704 to 0.817) between TB1-Nil and TB2-Nil, 0.627 (95% CI, 0.537 to 0.703) between TB-Nil and TB2-Nil, and 0.697 (95% CI, 0.619 to 0.761) between TB-Nil and TB1-Nil (Fig. 1), indicating a strong correlation between the two assays.

FIG 1.

Comparison between IFN-γ levels in the QFT-GIT and QFT-Plus assays. TB antigen-Nil versus TB1 antigen-Nil (A), TB antigen-Nil versus TB2 antigen-Nil (B), and TB1 antigen-Nil versus TB2 antigen-Nil (C). IFN-γ, interferon gamma; QFT-GIT, QuantiFERON-TB Gold In-Tube; QFT-Plus, QuantiFERON-TB Gold Plus; TB, tuberculosis.

Concordance between QFT-GIT and QFT-Plus.

There were 53 (24.1%) positive results in QFT-Plus and 45 (20.5%) positive results in QFT-GIT. There were 38 positive (17.3%), 144 negative (65.5%), and 9 (4.1%) indeterminate results in both assays at the same time (Table 2). The overall concordance rate between QFT-Plus and QFT-GIT was 86.8% (191/220), and the kappa value was 0.69 (95% CI, 0.59 to 0.80). Unlike previous in reports (12–15, 19), the concordance rate in this study did not exceed 90%, which was largely because of the large number of indeterminate results. However, the agreement between assays was substantial.

TABLE 2.

Concordance between QFT-GIT and QFT-Plus^a

QFT-GIT result	No. (%) of concordant QFT-Plus results
	Negative	Indeterminate	Positive	Total
Negative	144 (65.5)	5 (2.3)	15 (6.8)	164 (74.5)
Indeterminate	2 (0.9)	9 (4.1)	0 (0.0)	11 (5.0)
Positive	7 (3.2)	0 (0.0)	38 (17.3)	45 (20.5)
Total	153 (69.5)	14 (6.4)	53 (24.1)	220 (100.0)

^aQFT-GIT, QuantiFERON-TB Gold in-Tube; QFT-Plus, QuantiFERON-TB Gold Plus.

Discordant results were found in 29 subjects (13.2%), including two subjects (0.9%) who showed indeterminate results for the QFT-GIT assay and negative results for the QFT-Plus assay, and five subjects (2.3%) who showed indeterminate results for the QFT-Plus assay and negative results for the QFT-GIT assay. Seven subjects (3.2%) showed a positive result for the QFT-GIT assay but a negative result for the QFT-Plus assay. In these cases, the median TB-Nil value for the QFT-GIT assay was 0.57 (range, 0.36 to 0.67), which is close to the cutoff value of 0.35. Three cases were tested for IGRA as an adjuvant diagnostic tool for active TB infection when TB culture and PCR results were negative for the suspected samples. The other four cases who were expected to undergo immunosuppressive treatment were tested for IGRA to check latent TB, but the treatment caused autoimmune diseases in localized areas such as the skin and eyes. These patients did not have systemic autoimmune disease, chronic kidney disease, diabetes mellitus, or cancer and did not have immune-related conditions. Fifteen subjects (6.8%) showed negative results in the QFT-GIT assay but positive results in the QFT-Plus assay. Ten of these 15 cases were attributed to an additional TB2 antigen response in QFT-Plus (Table 3), as the TB2-Nil values were greater than the TB1-Nil and TB-Nil values (P = 0.002); however, there was no difference between the TB1-Nil and TB-Nil values (P = 0.8), indicating that the TB2-Nil was important in the positive determination. Strict discordances in which one result was <0.20 and the other was >0.70 IU/ml were not found in positive QFT-GIT assay results and negative QFT-Plus assay results, but there were three cases showing negative QFT-GIT assay results and positive QFT-Plus assay results.

TABLE 3.

Positive conversion by additional TB2 antigen response in QFT-Plus^d

Patient	Sexb	Age (yrs)	Clinical department	Diagnosisc	Classification	IFN-γ level (IU/ml) for:			TST resulte
						TB-Nil	TB1-Nil	TB2-Nil
P1	M	49	Pulmonology	Pulmonary TB	Tuberculosis	0.15	0.29	0.54
P2	F	25	Pulmonology	Pulmonary TB	Tuberculosis	−0.01	−0.05	0.37	P
P3	M	34	Transplantation surgery	ESRD, DM	Immune-related status	0.01	0.01	0.46	N
P4	F	32	Transplantation surgery	ESRD	Immune-related status	0.01	0.02	0.38	N
P5	F	50	Transplantation surgery	ESRD	Immune-related status	0.26	0.30	0.38	N
P6	M	62	Infectious internal medicine	Chronic heart failure, septic shock	Immune-related status	0.01	−0.04	0.37	N
P7	M	58	Infectious internal medicine	EBV infection	Immune-related status	0.03	0.28	1.35
P8	M	25	Rheumatology	Autoimmune disease	Immune-related status	−0.17	−0.17	0.64
P9	F	34	Gastroenterology	Thyroid cancer	Immune-related status	0.01	0.01	0.56	P
P10	F	26	Family medicine	Physical examination	Other	0.29	0.16	0.35	P
Mean ± SDa						0.059 ± 0.13 A	0.081 ± 0.16 A	0.54 ± 0.29 B

^aThe same letters in a row indicate no significant (P > 0.05) positive conversion, whereas different letters indicate significant positive conversion (P < 0.05) according to the TB2 results (Wilcoxon signed rank test).

^bM, male; F, female.

^cEBV, Epstein-Barr virus; ESRD, end-stage renal disease; DM, diabetes mellitus; TB, tuberculosis.

^dQFT-Plus, QuantiFERON-TB Gold Plus.

^eTST, tuberculin skin test; P, positive; N, negative.

A total of 63 patients with suspected TB infection but vague symptoms were tested for IGRA as an adjuvant of diagnosis, and only three were confirmed as having active TB infection by TB culture and PCR. Two of three patients were positive in TB, TB1, and TB2 tubes, but one patient (P1) was negative in TB and TB1 tubes and positive in the TB2 tube. Patient P1 was the only case among 10 cases positive for only TB2 in QFT-Plus (Table 3). This patient could have missed out on proper TB treatment if TB infection were ruled out based on the QFT-GIT result only. Another patient (P2) among the 10 patients was diagnosed with active TB 3 months previously and showed a positive result in TST, which was conducted concurrently with IGRA. QFT-Plus results indicated that it may be more sensitive to recent TB infection than QFT-GIT. Among the remaining patients, seven (P3 to P9) had immune problems, and one (P10) was healthy but showed a positive result in TST. Those with immune problems had chronic kidney disease, diabetes mellitus, solid cancer, septic shock, systemic autoimmune disease, or Epstein-Barr virus infection. This suggests that QFT-Plus may be useful for detecting TB in patients suffering from immune disturbances.

Comparison of TST and IGRA.

A total of 53 patients had concurrent TST results available for comparison. Except for the samples with indeterminate results in IGRA, the overall concordance rate between TST and QFT-GIT was 63.5% (33/52), showing slight agreement (kappa = 0.039; 95% CI, −0.210 to 0.288). A total of 14 cases were negative in QFT-GIT and positive in TST, whereas five showed the opposite pattern. The agreement between TST and QFT-GIT was low, and most of these were cases were positive in the TST. This result was similar to those reported previously in the Republic of Korea regarding the comparison of the two assays (22, 23). The overall concordance rate between the TST and QFT-Plus was 57.1% (28/49), showing slight agreement (kappa = 0.041; 95% CI, −0.241 to 0.323), including 11 cases who were negative in QFT-Plus but positive in TST and 10 cases showing the opposite pattern. We evaluated whether QFT-Plus could provide reveal the reason for discordance between QFT-GIT and TST and compared the results of QFT-Plus and TST; however, these two methods also showed a low concordance. The ratio of negative conversion and positive conversion was similar. A comprehensive judgment, using clinical symptoms or other laboratory results, should be made when there is a discrepancy between the results of the IGRA and TST.

Conclusion and recommendations.

We compared the performance of QFT-Plus to that of QFT-GIT under the same conditions during testing of IFN-γ in the clinical setting. In particular, we included individuals for visa applications, health check-ups, preimmunosuppressive therapy, and pretransplantation, or for adjuvant diagnosis of current TB infection. QFT-Plus showed similar results as QFT-GIT and a good correlation with QFT-GIT as a routine IGRA. Considering that only the TB2 reaction of QFT-Plus, but not culture, PCR, or QFT-GIT, could identify a positive case with TB infection 3 months prior, QFT-Plus may be a more advantageous assay for active TB or recent infection because of the CD8⁺ cell response compared to QFT-GIT. The seven patients who were positive only from TB2 in QFT-Plus had immune conditions such as chronic kidney disease, diabetes mellitus, cancer, systemic autoimmune disease, and Epstein-Barr virus infection and were suspected to have latent TB. The fact that nearly all patients who were only TB2-positive except for active or recent TB infection were associated with immune conditions supported the conclusion that QFT-Plus may be more helpful than QFT-GIT for identifying latent TB in individuals with specific conditions such as immune disorders. However, our study was not designed or sufficiently powered to investigate the settings in which the QFT-Plus may perform better or worse than the QFT-GIT. The performances of QFT-Plus and QFT-GIT may differ because of differences in reasons for patient testing and in patient categories, which include a wide range of immunosuppressive conditions. Our study was based on patients who were referred for IGRA in the clinical setting, and only three patients had active TB. Therefore, prospective studies using larger sample sizes and involving multicenter subjects are needed.