Annals of the American Thoracic Society

Rationale: IFN-γ release assays (IGRAs) including the QuantiFERON-TB gold in-tube test (QFT-GIT) are increasingly used in place of the tuberculin skin test (TST) in surveillance programs for Mycobacterium tuberculosis infection in the United States. However, data on conversions, reversions, and predictive value of QFT in such programs for health care workers (HCWs) are limited.

Objectives: The purpose of this study is to assess long-term reproducibility and conversion and reversion rates of QFT-GIT among HCWs who underwent serial testing at a tertiary care center in the United States.

Methods: Retrospective chart review of HCWs at the Central Arkansas Veterans Healthcare System (CAVHS) who underwent serial testing with QFT-GIT as a part of their employee screening between November 1, 2008 and January 31, 2011.

Measurements and Main Results: A total of 2,303 HCWs had at least 2 QFT-GITs 1 year apart. The initial QFT-GIT was positive for 69 and 2 were indeterminate. Of these 69 HCWs, 31 (45%) reverted on repeat testing, and 25 of 31 (80.6%) HCWs who reverted had a negative look-back TST. Of the 2,232 HCWs with an initial negative QFT-GIT, 71 (3.2%) converted on repeat testing. A third QFT-GIT assay was performed in 41 of the 71 converters and 90% (37 of 41) reverted back to negative. Only two HCWs had TST and QFT-GIT conversion.

Conclusions: Poor IGRA reproducibility and a low predictive value of QFT-GIT conversions indicate that QFT-GIT with current interpretation criteria should not be used for serial screening of U.S. HCWs. Negative TSTs have higher reproducibility than QFT-GIT for serial testing of HCWs in low tuberculosis incidence settings.

The most recent Centers for Disease Control and Prevention (CDC) guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care facilities (1) recommends periodic screening of health care workers (HCWs) for M. tuberculosis infection (MTBI) based on risk assessment. HCWs at risk for MTBI are usually screened with a tuberculin skin test (TST) and for symptoms suggestive of tuberculosis (TB). Latent M. tuberculosis infection (LTBI) is diagnosed on the basis of a positive TST result and the exclusion of active TB disease. IFN-γ release assays (IGRAs) are in vitro tests for MTBI that are increasingly used in place of TST to screen HCWs in the United States. Two IGRAs are commercially available for the detection of MTBI in the United States: the ELISA-based QuantiFERON-TB gold in-tube test (QFT-GIT; Cellestis, Carnegie, Australia) and the enzyme-linked immunospot–based T-SPOT.TB test (Oxford Immunotec, Oxfordshire, UK) (2). The most recent CDC guidelines for using IGRAs indicate that IGRAs, including QFT-GIT, may be used in surveillance programs for MTBI in HCWs (2).

The Central Arkansas Veterans Healthcare System (CAVHS) is a tertiary care academic Veterans Administration hospital with more than 3,000 employees and is affiliated with the University of Arkansas for Medical Sciences (UAMS) at Little Rock, Arkansas. Screening of HCWs for MTBI with TST, when hired and annually thereafter, in accordance with published guidelines (1), has been the standard practice at CAVHS for years. The annual TST conversion rate as reported by infection control in the 5 years before 2008 was less than 0.1%. In November 2008, CAVHS began to use QFT-GIT in place of TST for screening HCWs for MTBI. After this change, we were confronted with an unexpectedly high number of positive QFT-GIT test results. This resulted in a unique situation and the clinical dilemma of whether to offer LTBI treatment to HCWs who tested positive by QFT-GIT but had a negative TST history. We decided to immediately repeat QFT-GIT when it was positive and, to our surprise, many of the repeat QFT-GIT were negative (3, 4). Our reports of poor short-term reproducibility of QFT-GIT prompted the CAVHS infection control program to discontinue QFT-GIT for serial testing of its HCWs in 2011 and TST was reinstituted. We have subsequently identified a cohort of HCWs who underwent serial testing with QFT-GIT 1 year apart. The aim of the present study is to assess long-term reproducibility and conversion and reversion rates of QFT-GIT among HCWs who underwent serial testing at a tertiary care center in the United States.

A retrospective chart review was conducted at CAVHS after obtaining the approval of the CAVHS Institutional Review Board. Electronic medical records were queried to identify HCWs who had at least two QFT-GIT 12 (±2) months apart as a part of their employee screening. Data were collected on age, sex, and QFT-GIT results. We also obtained information about (1) prior (look-back) TST results from tests performed between November 2006 and November 2008 (i.e., before implementation of QFT-GIT), and (2) follow-up TST results from January 2011 onward (i.e., after QFT-GIT was discontinued and TST was reinstituted for serial testing of HCWs).

QFT-GIT was performed and interpreted as per the manufacturer’s instructions (5) and has been described in detail previously (4). The test results were interpreted as positive if the TB response was at least 0.35 IU/ml and at least 25% of nil; and negative if the TB response was less than 0.35 IU/ml or less than 25% of nil, nil was less than 8.0 IU/ml, and the mitogen response was more than 0.5 IU/ml. Test results were considered indeterminate if nil concentrations were more than 8.0 IU/ml or mitogen responses were less than 0.5 IU/ml and the TB response was less than 0.35 IU/ml or less than 25% of nil (5). For this study, QFT-GIT conversion was defined as a change from negative to positive, and reversion was a change from positive to negative as proposed in the guidelines (2).

TSTs were performed and interpreted by experienced nurses in CAVHS employee health as per CDC guidelines. For this study, TST conversion was defined as an increase in induration of at least 10 mm from baseline induration size of less than 10 mm.

Statistical Analysis

Statistical analysis was conducted with SAS 9.3 version (SAS Institute, Cary, NC). All reported P values were two-tailed and calculated with statistical significance set to P less than or equal to 0.05. The chi-square test was used to compare categorical variables and independent group t tests were used to compare continuous variables. For comparison of TB response, IFN-γ values greater than 10 IU/ml were replaced with a value of 10 IU/ml.

A total of 2,303 HCWs underwent serial testing between November 1, 2008 and January 31, 2011 with at least 2 QFT-GIT approximately 1 year apart. The initial QFT-GIT (QFT-GIT 1) was interpreted as positive for 69 (3%) HCWs, negative for 2,232 (96.9%) HCWs, and indeterminate for 2 (0.1%) HCWs (Figure 1). Of the 2,232 HCWs who were negative on initial testing, 71 (3.2%) became positive (conversion) on repeat testing. Of the 69 HCWs whose initial QFT-GIT was positive, 31 (45%) had QFT-GIT reversion (Figure 1). The characteristics of the 102 HCWs who had QFT-GIT conversions or reversions on serial testing are given in Table 1. The mean age (±SD) of the 102 HCWs who tested positive with QFT-GIT was 47.6 (±9.1) years and 54 (53%) were male.

Table 1. Characteristics of 102 health care workers who converted or reverted on serial QuantiFERON-TB gold in-tube testing

CharacteristicNo. (%) of HCWs
Sex 
 Female48 (47)
 Male54 (53)
TST history 
 Negative94 (92)
 Positive7 (7)
 Unknown1 (1)
Mean age, yr (±SD)47.6 (±9.1)

Definition of abbreviations: HCWs, health care workers; TST = tuberculin skin test.

Note: Health care workers who converted, n = 71; health care workers who reverted, n = 31.

Among 31 HCWs whose QFT-GIT reverted to negative, the mean and median TB response from the initial QFT-GIT (QFT-GIT 1) were 1.28 and 0.76 IU/ml, respectively. Changes in TB responses for these 31 HCWs are shown in Figure 2A. The initial values were distributed over a range of 0.35–10 IU/ml whereas the TB responses obtained on repeat testing were clustered in a much smaller range (Figure 2A; mean of 0.05 and median of 0.03, respectively). TB response values were compared among HCWs whose repeat QFT-GIT test result reverted back to negative (n = 31) versus stable positives (n = 37). On comparing IFN-γ values, there was a significant difference (P = 0.005) in the mean levels between the two groups (Figure 3). However, when comparison was limited only to HCWs with negative baseline TST (i.e., those who undergo retesting in employee screening programs), there was no statistically significant difference (P = 0.97) in TB response values (Figure 4).

Among the 71 HCWs with QFT-GIT conversion, only 2 had TST conversion. Changes in TB responses for these 71 HCWs are shown in Figure 2B. The mean and median TB response from the repeat QFT-GIT (QFT-GIT 2) were 1.33 and 0.70 IU/ml, respectively. A third test (QFT-GIT 3) was performed in 41 of these 71 new converters (offered to all new converters on the basis of our prior study results [3, 4]) and 90% (37 of 41) reverted back to negative (Figure 5). Changes in TB responses for these 37 HCWs are shown in Figure 2C. The mean and median TB response from the initial QFT-GIT were 1.46 and 0.73 IU/ml, respectively.

Among reverters, 25 of 31 (80%) had negative look-back TST status. Those with QFT-GIT reversion were more likely to have a negative look-back TST than those with stable positive QFT-GIT results (P = 0.01) (Figure 1). Of the 71 HCWs who had conversion, 64 (90%) had negative baseline TST status, 6 had a positive baseline TST, and the TST status was unknown for 1 individual (Figure 5). Only two HCWs had TST conversion and both of them had QFT-GIT conversion as well.

Our study results clearly demonstrate that QFT-GIT has poor long-term reproducibility in serial testing of U.S. health care workers undergoing screening for MTBI. Our results are in agreement with studies in low TB incidence countries (610). The variability of IGRAs is now well recognized and warrants careful consideration when IGRAs are used for serial testing (1115). We discontinued QFT-GIT in 2011 and reinstituted TST as a preferred test for MTBI serial testing in HCWs in our institution.

The reasons for IGRA variability are for the most part unexplained. Many hypotheses explaining poor long- and short-term reproducibility and thus high conversion and reversion rates have been proposed at the manufacturing, preanalytical, analytical, and immunological levels (1618). The first systematic review describing performance of IGRAs in HCWs by Zwerling and colleagues (19) as well as another review by Ringshausen and colleagues (20) have described the significant challenges associated with serial testing with IGRA. The reviews concluded that IGRAs have suboptimal reproducibility.

Our study is unique in that serial QFT-GIT results were compared with look-back TSTs as well as with subsequent TSTs performed after QFT-GIT discontinuation. The availability of test results for multiple years makes it clinically more relevant as compared with many other published studies that are cross-sectional or those that compare tests performed over a 1- to 2-year interval (616, 19, 20). This large study evaluated test results for more than 2,300 HCWs and was conducted in a real world setting. Therefore, our findings may be more generalizable and herald similar challenges with serial testing of HCWs with QFT-GIT at other institutions (2123).

The most clinically relevant question is whether QFT-GIT conversions in HCWs in low TB incidence settings are indicative of a recent M. tuberculosis infection and an increase in risk for active tuberculosis? Is the conversion due to a false positive test result? Is QFT-GIT conversion an indication for treatment of LTBI after active disease is excluded? Does it predict progression to tuberculosis and if yes, is the predictive value any better than TST? Many of these questions can be answered by the results of this longitudinal study. Almost all new QFT-GIT converters in our study were likely false positives. This is supported by the following findings:

1.

A 90% short-term reversion rate on repeat testing of new converters

2.

Most of the initial and repeat TB responses were not near the cut point separating positive and negative results and did not just “wobble” around the cutoff

3.

Most (90%) of these new converters had previous negative TST status and also remained negative on repeat TST

4.

None of these HCWs (new converters) were treated for LTBI (based on positive QFT-GIT results alone) and none were diagnosed with active disease after at least 2 years of follow-up

Of note, progression to active disease is approximately 5% in the first 2 years after TST conversions (1). If the predictive value of QFT-GIT for progression to active disease is as good as TST, we would expect few of the 71 HCWs with QFT-GIT conversion to develop active disease if indeed they were recently infected. This suggests that QFT-GIT conversion with serial testing of HCWs in low TB incidence settings has poor predictive value for progression to active disease, at least in the first 2 years.

One of the limitations of our study is that it is retrospective. Second, not all the 71 new converters had a third QFT-GIT (QFT-GIT 3) performed. They were all offered repeat testing, but 30 of them declined for one of the following reasons: (1) baseline negative TST status (28 of 30); many of these HCWs were aware of the QFT-GIT variability that we presented in our previous study and hence believed that these results were likely false positives. Some expressed a lack of trust in this test and chose to wait until the next serial testing; (2) baseline positive TST status (2 of 30); QFT-GIT 1 was interpreted as a false negative result and the subsequent QFT-GIT (QFT-GIT 2) as a true positive result. Another limitation is that we did not have data on BCG vaccination for all employees.

In conclusion, QFT-GIT has poor long-term reproducibility in serial MTBI testing in the United States. The predictive value of QFT-GIT conversions for progression to tuberculosis is close to zero. Poor IGRA reproducibility and a low predictive value of QFT-GIT conversions indicate that QFT-GIT with current interpretation criteria should not be used for serial screening of U.S. HCWs. Negative TSTs have higher reproducibility than QFT-GIT for serial testing of HCWs in low TB incidence settings. We not only recommend TST as the preferred test for serial testing of HCWs until QFT-GIT reproducibility issues are addressed, but also to more rigorously avoid testing those HCWs who are at low risk for TB infection.

The authors thank Gerald H. Mazurek, M.D. (Division of Tuberculosis Elimination, CDC) for interpretation of study results and critical review of our manuscript.

1 . Jensen PA, Lambert LA, Iademarco MF, Ridzon R; Centers for Disease Control and Prevention (CDC). Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings, 2005. MMWR Recomm Rep 2005;54:1141.
2 . Mazurek GH, Jereb J, Vernon A, LoBue P, Goldberg S, Castro K; IGRA Expert Committee; Centers for Disease Control and Prevention (CDC). Updated guidelines for using interferon gamma release assays to detect Mycobacterium tuberculosis infection—United States, 2010. MMWR Recomm Rep 2010;59:125.
3 . Joshi M, Monson T, Woods G. Practical experience with the QFT assay for LTBI annual testing among US health-care workers in a large tertiary setting [abstract]. Chest 2010;138:746A.
4 . Joshi M, Monson TP, Woods GL. Use of interferon-γ release assays in a health care worker screening program: experience from a tertiary care centre in the United States. Can Respir J 2012;19:8488.
5 . Cellestis. QuantiFERON-TB gold in-tube package insert [accessed 15 Jun 2013]. Available from: http://www.cellestis.com/IRM/content/PI/us.pdf
6 . Zwerling A, Benedetti A, Cojocariu M, McIntosh F, Pietrangelo F, Behr MA, Schwartzman K, Menzies D, Pai M. Repeat IGRA testing in Canadian health workers: conversions or unexplained variability? PLoS One 2013;8:e54748.
7 . Fong KS, Tomford JW, Teixeira L, Fraser TG, van Duin D, Yen-Lieberman B, Gordon SM, Miranda C. Challenges of interferon-γ release assay conversions in serial testing of health-care workers in a TB control program. Chest 2012;142:5562.
8 . Gandra S, Scott WS, Somaraju V, Wang H, Wilton S, Feigenbaum M. Questionable effectiveness of the QuantiFERON-TB gold test (Cellestis) as a screening tool in healthcare workers. Infect Control Hosp Epidemiol 2010;31:12791285.
9 . Park JS, Lee JS, Kim MY, Lee CH, Yoon HI, Lee SM, Yoo CG, Kim YW, Han SK, Yim JJ. Monthly follow-ups of interferon-γ release assays among health-care workers in contact with patients with TB. Chest 2012;142:14611468.
10 . Rangaka MX, Wilkinson KA, Glynn JR, Ling D, Menzies D, Mwansa-Kambafwile J, Fielding K, Wilkinson RJ, Pai M. Predictive value of interferon-γ release assays for incident active tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis 2012;12:4555.
11 . Ringshausen FC, Nienhaus A, Torres Costa J, Knoop H, Schlösser S, Schultze-Werninghaus G, Rohde G. Within-subject variability of Mycobacterium tuberculosis–specific γ interferon responses in German health care workers. Clin Vaccine Immunol 2011;18:11761182.
12 . Metcalfe JZ, Cattamanchi A, McCulloch CE, Lew JD, Ha NP, Graviss EA. Test variability of the QuantiFERON-TB gold in-tube assay in clinical practice. Am J Respir Crit Care Med 2013;187:206211.
13 . Whitworth WC, Hamilton LR, Goodwin DJ, Barrera C, West KB, Racster L, Daniels LJ, Chuke SO, Campbell BH, Bohanon J, et al. Within-subject interlaboratory variability of QuantiFERON-TB gold in-tube tests. PLoS One 2012;7:e43790.
14 . Herrera V, Perry S, Parsonnet J, Banaei N. Clinical application and limitations of interferon-γ release assays for the diagnosis of latent tuberculosis infection. Clin Infect Dis 2011;52:10311037.
15 . Mancuso JD, Bernardo J, Mazurek GH. The elusive “gold” standard for detecting Mycobacterium tuberculosis infection. Am J Respir Crit Care Med 2013;187:122124.
16 . Doberne D, Gaur RL, Banaei N. Preanalytical delay reduces sensitivity of QuantiFERON-TB gold in-tube assay for detection of latent tuberculosis infection. J Clin Microbiol 2011;49:30613064.
17 . Slater M, Parsonnet J, Banaei N. Investigation of false-positive results given by the QuantiFERON-TB gold in-tube assay. J Clin Microbiol 2012;50:31053107.
18 . Pai M, Elwood K. Interferon-γ release assays for screening of health care workers in low tuberculosis incidence settings: dynamic patterns and interpretational challenges. Can Respir J 2012;19:8183.
19 . Zwerling A, van den Hof S, Scholten J, Cobelens F, Menzies D, Pai M. Interferon-γ release assays for tuberculosis screening of healthcare workers: a systematic review. Thorax 2012;67:6270.
20 . Ringshausen FC, Schablon A, Nienhaus A. Interferon-γ release assays for the tuberculosis serial testing of health care workers: a systematic review. J Occup Med Toxicol 2012;7:6.
21 . Denkinger CM, Dheda K, Pai M. Guidelines on interferon-γ release assays for tuberculosis infection: concordance, discordance or confusion? Clin Microbiol Infect 2011;17:806814.
22 . Slater ML, Welland G, Pai M, Parsonnet J, Banaei N. Challenges with QuantiFERON-TB gold assay for large-scale, routine screening of U.S. healthcare workers. Am J Respir Crit Care Med 2013;188:10051010.
23 . Pollock NR, McAdam AJ, Pai M, Nardell EA, Bernardo J, Banaei N, Mobo J. Interferon γ-release assays for diagnosis of latent tuberculosis in healthcare workers in low-incidence settings: pros and cons. Clin Chem 2013;60:5.
Correspondence and requests for reprints should be addressed to Manish Joshi, M.D., Central Arkansas Veterans Healthcare System, 4300 W 7th street, 5C 144, Little Rock, AR 72205. E-mail:

Author Contributions: All authors made a contribution in the content and writing of the manuscript, including final revision of the manuscript for important intellectual content. M.J. was the principal investigator and contributed the original idea for the study and collected the data. T.P.M., G.LW., and A.J. had input into the study design and data interpretation.

Institution at which study was performed: Central Arkansas Veterans Healthcare System, Little Rock, AR.

Author disclosures are available with the text of this article at www.atsjournals.org.

Comments Post a Comment




New User Registration

Not Yet Registered?
Benefits of Registration Include:
 •  A Unique User Profile that will allow you to manage your current subscriptions (including online access)
 •  The ability to create favorites lists down to the article level
 •  The ability to customize email alerts to receive specific notifications about the topics you care most about and special offers
Annals of the American Thoracic Society
11
3

Click to see any corrections or updates and to confirm this is the authentic version of record