Evaluation of QuantiFERON-TB Gold In-Tube in human immunodeficiency virus infection and in patient candidates for anti-tumour necrosis factor-alpha treatment

INT J TUBERC LUNG DIS 14(7):834–840© 2010 The Union

Evaluation of QuantiFERON®-TB Gold In-Tube in human immunodefi ciency virus infection and in patient candidates for anti-tumour necrosis factor-alpha treatment

I. Sauzullo,* F. Mengoni,* R. Scrivo,† G. Valesini,† C. Potenza,‡ N. Skroza,‡ R. Marocco,§ M. Lichtner,*§ V. Vullo,* C. M. Mastroianni*§

* Dipartimento di Malattie Infettive e Tropicali and † Dipartimento di Clinica e Terapia Medica, Reumatologia, Sapienza University, Rome, ‡ Unità Operativa Complessa (UOC) Dermatologia, Sapienza University, Polo Pontino, Terracina, § UOC Malattie Infettive, Sapienza University, Polo Pontino, Latina, Italy

Correspondence to: Claudio M Mastroianni, UOC Malattie Infettive, S M Goretti Hospital, Via Canova 1, Sapienza Univer-sity, Polo Pontino, Latina, Italy. Tel: (+39) 649 970 881. Fax: (+39) 649 972 625. e-mail: [email protected] submitted 4 November 2009. Final version accepted 14 January 2010.

S E T T I N G : Cross-sectional study at four out-patient

clinics in a single referral centre in Italy.

O B J E C T I V E : To evaluate the performance of Quanti-

FERON®-TB Gold In-Tube (QFT-GIT) in human im-

munodefi ciency virus (HIV) infected adults and in patients

with immune-mediated infl ammatory diseases (IMIDs)

who are candidates for anti-tumour necrosis factor-alpha

(TNF-α) treatment.

D E S I G N : A total of 402 immunocompromised patients

were enrolled, including 207 HIV-infected individuals

and 195 IMID patients scheduled for anti-TNF-α treat-

ment. Tuberculin skin test (TST) and QFT-GIT were

performed. For active tuberculosis (TB), test results were

compared with microbiological, histopathological and

clinical diagnoses.

R E S U LT S : In HIV-infected patients, the level of agree-

ment between the tests was 68% and QFT-GIT sensitiv-

ity was 66% (95%CI 47–82). We found a large propor-

tion of indeterminate QFT-GIT results (33.4%), which

correlated with CD4 count < 200 cells/μl (P < 0.0001).

The degree of agreement with TST was higher in IMID

patients (81.6%). Factors associated with discordant

positive TST and negative QFT-GIT results were bacille

Calmette-Guérin vaccination (P = 0.0001), previous TB

(P = 0.0001) and agricultural work (P = 0.0005).

C O N C L U S I O N : The performance of QFT-GIT varies

between different types of immunocompromised pa-

tients. Interferon-gamma release assays should not be

used to confi rm or rule out a diagnosis of active TB in

HIV-infected adults. As there were no cases of active TB

in the IMID subgroup, it was diffi cult to determine

which test performs better in this population.

K E Y W O R D S : tuberculosis; HIV infection; anti-TNF-α treatment; IFN-γ release assays

THE IDENTIFICATION of individuals infected with Mycobacterium tuberculosis at increased risk of tu-berculosis (TB) reactivation is crucial for TB con-trol.1–3 Until recently, the most widely used diagnostic tool to detect TB infection was the tuberculin skin test (TST). The main drawbacks of TST are the pres-ence of false-negative results (especially in immuno-compromised patients) or false-positive results (cross-reactivity with M. bovis bacille Calmette-Guérin [BCG] vaccine strains and non-tuberculous mycobacteria).4,5

The development of interferon-gamma release as-says (IGRAs) has opened new perspectives for the de-tection of TB infection.6–8 The QuantiFERON®-TB Gold In-Tube (QFT-GIT; Cellestis Ltd, Carnegie, VIC, Australia) and T-SPOT.TB™ assays, which mea-sure interferon-gamma (IFN-γ) released by sensitised T-cells after stimulation with M. tuberculosis-specifi c antigens, have higher specifi city than TST and better correlation with surrogate measures of exposure to M. tuberculosis; moreover, the tests are less affected

by BCG vaccination and could be useful for treat-ment monitoring.9–11

Despite a growing body of evidence to support the use of IGRAs in immunocompetent patients, data on the performance of these tests among immuno-compromised persons at high risk of TB remain lim-ited.7 In the present study, the authors evaluated the suitability of commercial QFT-GIT as a tool to identify TB infection in two categories of patients at increased risk of developing TB: 1) human immunodefi ciency virus (HIV) infected individuals and 2) immune-m ediated infl ammatory disease (IMID) patients who are candidates for treatment with tumour necrosis factor-alpha (TNF-α) blockers.

MATERIALS AND METHODS

Study populationThe study was conducted in four out-patient clinics of a single referral centre (Sapienza University, Rome,

S U M M A R Y

IGRA for TB in immunocompromised patients 835

Italy). From 1 June 2007 to 30 September 2009, 402 consecutive patients were enrolled from two different immunocompromised groups: 1) 207 HIV-infected persons showing active TB symptoms (i.e., persistent cough, fever or chills, night sweats, swollen lymph nodes, shortness of breath, fatigue and unexplained weight loss), or in whom TB was part of the differen-tial diagnosis; and 2) 195 non-HIV-infected IMID pa-tients who were tested for latent TB infection (LTBI), as they required anti-TNF-α treatment. Patients aged <18 years were excluded. To study real-life testing performance in routine clinical practice, no other ex-clusion criteria were defi ned. The median follow-up period was 7 months (range 1–15). The diagnosis of active TB was made on the basis of identifi cation of M. tuberculosis by microbiological and/or histologi-cal methods or clinical and radiological fi ndings and response to anti-tuberculosis treatment. Patients were diagnosed with LTBI on the basis of a positive TST or IGRA result.7,11 Patients without active TB were de-fi ned as those individuals whose culture or histology for M. tuberculosis was negative, with resolution of clinical symptoms and radiographic abnormalities fol-lowing treatment not involving anti-tuberculosis drugs or an eventual alternative diagnosis.

Screening at enrolment included clinical evaluation, chest X-ray (CXR), TST and QFT-GIT. All patients were questioned regarding their demographic details, BCG vaccination history, TB risk factor (household contact, confi rmed history of TB, birth or long-term residence in an area with a high prevalence of TB in-fection, CXR suggestive of LTBI) and current treat-ments. Sputum smear and culture were performed in all HIV-infected patients and in 16 IMID patients with clinical or radiological fi ndings suggestive of active TB. At least three sputum specimens were obtained for acid-fast bacilli smears and TB culture (conventional Löwenstein-Jensen medium and BACTEC Mycobac-teria Growth Indicator Tube 960 system, BD, Sparks, MD, USA). The reference standard for active TB (cul-ture positivity for M. tuberculosis and/or histology) and IGRA results was blinded. A blinded interpreta-tion of TST and QFT-GIT results was also performed.

The study was approved by the institutional review board of the Department of Infectious and Tropical Diseases, Sapienza University of Rome; informed con-sent was obtained from all participants before blood samples were taken.

QuantiFERON-TB Gold In-TubeThe QFT-GIT was performed and interpreted accord-ing to the manufacturer’s instructions.

Tuberculin skin test After blood was drawn for the QFT-GIT assay, a TST was performed according to the Mantoux method4

using 5 international units (IU) of purifi ed protein

d erivative (Biocine, Chiron, Siena, Italy) by a trained physician. TST induration ⩾5 mm was considered positive.11

Statistical analysisSPSS version 13.0 for Windows (Statistical Package for Social Science, Chicago, IL, USA) was used. The value differences between groups were analysed us-ing the non-parametric Mann-Whitney U-test. χ2 test or Fisher’s exact test were used for comparison of proportions between groups. The analysis of concor-dance between two tests was performed using the kappa statistics measure (κ, Cohen test).12 Odds ra-tios (ORs) and their 95% confi dence intervals (CIs) for factors associated with discordant results, and in-determinate QFT-GIT test results (lack of response to mitogen) were estimated by univariate analysis.

RESULTS

Population characteristics The characteristics of the 402 study patients are shown in Table 1. Among the 207 HIV-infected indi-viduals, 10 were excluded from the fi nal analysis due

Table 1 Baseline characteristics of the study population

Characteristics

Patients with HIV infection

n (%)

Patients with IMID

n (%)

Patients, n 207 195

Age, years, median [range] 39 [27–58] 52 [18–80]

Sex Male Female

119 (58) 88 (42)

115 (59) 80 (41)

Immigrant Low TB incidence High TB incidence

85 (41) 65 (76) 25 (24)

11 (6) 10 (90) 1 (10)

BCG-vaccinated* 29 (14) 17 (8.7)

Current treatments 150 (72)† 110 (56)‡

Active TB Pulmonary Extra-pulmonary Bone Brain Lymph nodes Abdomen

44 (21) 27 (61) 17 (39) 3 (18) 5 (29) 7 (41) 2 (12)

———————

Latent TB 6 (3) 27 (14)

No TB 140 (68)§ 168 (86)

TST available Positive Negative

197 (100) 81 (41)116 (59)

195 (100) 58 (30)137 (70)

Excluded 10 (4.8) —

* Vaccinated before the age of 18 years.† Antiretroviral treatment. ‡ One or more of the following: glucocorticoid (10 mg/day of prednisone or steroid equivalent) (n = 12); methotrexate (10 mg/week) + glucocorticoid (n = 88); cyclosporine (150 mg/daily) + methotrexate + glucocorticoid (n = 10).§ Includes another diagnosis: bacterial or viral pneumonia (n = 86), pneumo-cystosis (n = 9), fever of unknown origin (n = 33), atypical mycobacteriosis (n = 5), lymphoma (n = 4), lung cancer (n = 3).HIV = human immunodefi ciency virus; IMID = immune-mediated infl am-matory disease; TB = tuberculosis; BCG = bacille Calmette-Guérin; TST = tuberculin skin test.

836 The International Journal of Tuberculosis and Lung Disease

to high background values in negative-control wells based on the QFT-GIT assay criteria. The median CD4 count of the HIV-infected population was 219 cells/μl (range 4–995 cells/μl), and 45 (22%) patients had a CD4 count ⩽ 200 cells/μl; 150 (72%) patients were receiving antiretroviral treatment and 49 (23.6%) had a previous diagnosis of acquired immune-defi ciency syndrome.

Among IMID patients, the diagnoses were as fol-lows: rheumatoid arthritis (n = 72), psoriasis with or without arthritis (n = 87), ankylosing spondylitis (n = 10), Crohn’s disease (n = 18) and Behcet’s syn-drome (n = 8). A total of 110 patients (56%) were under treatment with corticosteroids and/or immuno-suppressive drugs.

Results in HIV-infected patientsPerformance of QFT-GITOf the 197 HIV-infected patients, 35 (17.7%) had a positive, 122 (62%) a negative and 40 (20.3%) an in-determinate QFT-GIT result. Among patients with a positive test result, 29 (83%) had active TB (20 with pulmonary TB [PTB] and nine with extra-pulmonary TB [EPTB]); LTBI was found in six (17%) TST-positive individuals who were household contacts of subjects with active TB. Of the QFT-GIT-negative patients, fi ve (4%) had EPTB and 117 (96%) were classifi ed as non-TB as another diagnosis was obtained. Among the 40 patients with an indeterminate QFT-GIT, 10 re-ceived a diagnosis of active TB (seven PTB and three EPTB) at the time of screening, whereas seven patients developed active TB within 2 months, and 23 sub jects were classifi ed as non-TB. The agreement between TST and QFT-GIT was 68% (κ = 0.30). The results of the two assays are shown in Table 2.

At enrolment, we found 44 patients with active TB (27 with PTB and 17 with EPTB). The diagnosis was confi rmed by culture and/or histopathology in 30 subjects, of whom 20 (66.6%) had a positive, three (10%) a negative and seven (23.4%) an inde-terminate QFT-GIT result. The overall sensitivity of the test (66%, 95%CI 47–82) was determined using culture/histopathology as reference standard.13 Sensi-tivity increased to 86% (95%CI 65–96) if indetermi-nate results were excluded.

Impact of CD4 count on performance of QFT-GIT Patients with positive QFT-GIT results had signifi -cantly lower CD4 counts than those with a negative QFT-GIT (median 177 vs. 467 cells/μl, P = 0.001). The lowest CD4 counts (median 97 cells/μl) were found in subjects with indeterminate QFT-GIT re-sults (P = 0.029 and P < 0.001 if compared with positive and negative results, respectively; Figure 1). The results showed that 27/40 (67.5%) patients with CD4 count ⩽ 200 cells/μl had an indeterminate QFT-GIT test, compared with 20% and 12.5% of individ-uals with respectively 200–350 and >350 cells/μl (Figure 2). The univariate analysis confi rmed that a CD4 count ⩽ 200 cells/μl was strongly associated with an indeterminate result (OR = 5.80, 95%CI 2.13–16.18, P < 0.001).

Results in IMID patient candidatesfor anti-TNF-α treatmentPerformance of QFT-GITOf the 195 IMID patients, 32 (16.4%) had a posi-tive, 137 (70.2%) a negative and 26 (13.4%) an

Table 2 QFT-GIT assay and TST results in the 402 study patients*

HIV-infected patients (n = 207)†QFT-GIT

Patients with IMID (n = 195)QFT-GIT

Positive(n = 35)

n (%)

Negative(n = 122)

n (%)

Indeterminate(n = 40)

n (%)

Positive(n = 35)

n (%)

Negative(n = 137)

n (%)

Indeterminate(n = 26)

n (%)

TST-positive 25 (71) 40 (33) 16 (40) 27 (84) 26 (19) 5 (19)TST-negative 10 (29) 82 (67) 24 (60) 5 (16) 111 (81) 21 (81)

* QFT-GIT test > 0.35 IU/ml was considered as positive; TST induration ⩾ 5 mm was considered as positive.† 10 HIV patients were excluded from the fi nal analysis because of high background values in negative control wells based on the QFT-GIT assay criteria.QFT-GIT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; HIV = human immunodefi ciency virus; IMID = immune-mediated infl ammatory disease.

Figure 1 QFT-GIT results in HIV-infected patients according to CD4+ T-cell count. Box plots show the 10th, 25th, 50th (me-dian), 75th and 90th percentile and outlying values. Statistical signifi cance was analysed using the Mann-Whitney U-test. QFT-GIT = QuantiFERON-TB Gold In-Tube; HIV = human immuno-defi ciency virus.

IGRA for TB in immunocompromised patients 837

i ndeterminate QFT-GIT result. The occurrence of an indeterminate result was signifi cantly associated with immunosuppressive treatment (OR = 2.79, 95%CI 1.06–7.63, P < 0.02). No differences were found be-tween immunosuppressive drugs. The level of agree-ment with TST was 81.6% (κ = 0.52). The results of the two assays are shown in Table 2.

None of the subjects had active TB; 27 QFT-GIT-positive/TST-positive participants were considered to have LTBI and received isoniazid (INH) treatment 1 month before the fi rst dose of anti-TNF-α. The in-dividual IFN-γ responses of these patients toward mycobacterial antigens are shown in Figure 3. Results showed that the response (median 1.42 IU/ml, range 0.35–19.16) tended to be higher than the test cut-off of 0.35 IU/ml, and it was signifi cantly higher if com-pared to that reported in HIV-positive patients (me-dian 0.57 IU/ml, range 0.35–10.64, P < 0.001).

Analysis of discordance between QFT-GIT and TSTDiscrepant results between the two tests were found in 31 patients with IMID. In particular, 26 patients were QFT-GIT-negative/ TST-positive: 11 (42.3%) were BCG-vaccinated and four (15.3%) patients were farm-ers and potentially exposed to environmental non-t uberculous mycobacteria. Evidence of TB risk factors

was found in 11 (42.3%) patients: one had household contact with a TB case, fi ve had abnormal CXRs (cal-cifi ed granulomas or apical densities, and lymph node or pericardial calcifi cation) suggestive of LTBI and fi ve had a history of TB. All 11 subjects received INH before anti-TNF-α treatment. Active TB was ruled out by negative culture. In QFT-GIT-negative/TST-positive patients, the median TST reaction was 10 mm (range 5–20); this was signifi cantly lower than for pa-tients with concordant QFT-GIT and TST-positive re-sults (median 17 mm, range 6–35, P = 0.046). The me-dian IFN-γ response to specifi c antigens was 0.02 IU/ml (range 0–0.20). The only variables signifi cantly asso-ciated with discordant QFT-GIT-negative/TST-positive results were BCG vaccination, previous TB infection and agricultural work. Older age (>50 years) and immunosuppressive drugs were not associated with these discordant results (Table 3).

Figure 2 Percentage of HIV-infected patients with indetermi-nate QFT-GIT results according to CD4+ T-cell count. The per-centage of indeterminate results in patients with CD4 count <200 cells/μl was signifi cantly higher than those with CD4 count 200–350 and >350 cells/μl. No difference was found be-tween patients with CD4 count 200–350 cells/μl and those >350 cells/μl. Statistical signifi cance was analysed using χ2 test. HIV = human immunodefi ciency virus; QFT-GIT = Quanti-FERON-TB Gold In-Tube.

Figure 3 Individual IFN-γ response to antigens in 29 HIV-i nfected patients with active TB and in 27 individuals with IMID awaiting anti-TNF-α treatment and latently infected by M. tu-berculosis. The short solid lines represent the median value for each antigen. The black horizontal line shows the cut-off value for a positive response set at 0.35 IU/ml. IFN-γ = interferon-gamma; HIV = human immunodefi ciency virus; IMID = immune-mediated infl ammatory disease; TNF-α = tumour necrosis factor-alpha.

Table 3 Risk factors associated with QFT-GIT and TST discordant results in subjects with immune-mediated infl ammatory diseases: univariate analysis

Risk factors

QFT-GIT-negative/ TST-positive

OR (95%CI) P value

BCG vaccination 9.16 (2.3–37.1) 0.0001Previous TB 26 (2.7–62.7) 0.0001Older age (>50 years) 0.66 (0.2–1.8) 0.3Agricultural work 5.30 (1.8–47.1) 0.0005Immunosuppressive treatment 2.25 (0.8–5.6) 0.4

QFT-GIT = QuantiFERON-TB Gold In-Tube; TST = tuberculin skin test; OR = odds ratio; CI = confi dence interval; BCG = bacille Calmette-Guérin; TB = tuberculosis.

838 The International Journal of Tuberculosis and Lung Disease

Discrepant results were also detected in fi ve pa-tients who were QFT-GIT-positive but TST-negative. The median IFN-γ response to specifi c antigens was 0.78 IU/ml (range 0.55–12.2). A sputum culture was performed on these patients to rule out active TB, and INH was prescribed regardless of CXR fi ndings. The likelihood of a false-negative TST result could be increased by previous long-term immunosuppressive treatment, but no association was found between immunosuppressive treatment and discordant QFT-GIT-positive/ TST-negative results (OR = 0.16, 95%CI 0.01–1.81, P < 0.09).

DISCUSSION

The development of optimal screening methods for TB identifi cation in immunosuppressed patients repre-sents an area of evolving investigation. IGRAs perform better than TSTs in immunocompromised patients, although a relatively high frequency of false-positive and indeterminate results were found and the test cannot distinguish between active TB and LTBI.14,15

We report the results of using the QFT-GIT for the detection of TB infection in immuno compromised patients at risk of TB reactivation. This is the fi rst comparative study to assess the performance of QFT-GIT in HIV-infected patients and candidates for anti-TNF-α treatment in a country with low TB preva-lence, Italy. In HIV-infected adults, the sensitivity of QFT-GIT for TB infection compared with culture was lower than that reported in our previous study in an HIV-negative immunocompetent population (66% vs. 72.9%).16 These fi ndings are due to the large pro-portion of indeterminate QFT-GIT results among the HIV-infected population with severe immuno-suppression. A strong correlation between low CD4 count and the occurrence of indeterminate results was found in previous studies using QFT-GIT as an IGRA.17–19 In contrast, Dheda et al. did not fi nd any impact of a low CD4 count on the performance of the IFN-γ response using T-SPOT.TB.20 In this study, the performance of the QFT-GIT assay was impaired in patients with advanced stages of HIV-associated immunosuppression, and indeterminate QFT-GIT r esults were strongly associated with a CD4 count < 200 cells/μl.

When excluding the indeterminate results, the QFT-GIT sensitivity increased to 86.6%. The quantitative IFN-γ response to mycobacterial antigens tended to be low and closer to the test cut-off of 0.35 IU/ml. The assay sensitivity could be improved using differ-ential cut-off points related to the degree of immuno-suppression; when we applied a value of ⩾0.25 IU/ml as cut-off, fi ve HIV-infected subjects with indetermi-nate results could be reclassifi ed as positive.

In this study, seven HIV-infected patients with inde-terminate results developed active TB within 2 months of the IFN-γ response analysis. These fi ndings sug-

gest that clinicians should watch for the development of active TB in HIV-infected patients with indetermi-nate QFT-GIT results. The infl uence of CD4 count on the performance of QFT-GIT could be a limitation of this test in HIV infection. Some authors suggest that, in advanced HIV disease, the other commercial IGRA, the T-SPOT.TB test, should be preferred because it is less dependent on CD4 count.21 A combination of dif-ferent immunological tests might represent the optimal approach to improve the effi cacy of screening proce-dures in HIV-infected patients.22 Given the low sensi-tivity of QFT-GIT in HIV-positive persons with active TB, and the high rate of indeterminate results in this subgroup of patients, IGRAs should not be used to confi rm or rule out a diagnosis of active TB in HIV-infected adults.

In the study, the performance of QFT-GIT was also evaluated in immunosuppressed patients with IMID awaiting anti-TNF-α treatment. The use of drugs that block TNF-α has emerged as effective treatment for patients with IMID.23 However, several studies have shown that the inhibition of TNF-α, a key cytokine in the host defence against M. tuberculosis, has been associated with increased risk of disease reactivation in LTBI patients.23–25 Screening patients for TB be-fore starting TNF-α blocking treatment has become mandatory for clinicians.26–32 This study suggests good diagnostic accuracy of QFT-GIT for LTBI in IMID patients taking immunosuppressive drugs. The IFN-γ quantitative response was superior to the test cut-off value of 0.35 IU/ml, suggesting that QFT-GIT has a high LTBI detection level. The degree of agreement with TST was found to be higher than in the HIV-infected population.

A discordance between QFT-GIT and TST was ex-pected in IMID patients. BCG vaccination, previous TB and agricultural work were the only factors asso-ciated with discordant QFT-GIT-negative/TST-positive results. We did not fi nd any association between i mmunosuppressive treatment and discordant QFT-GIT-positive/ TST-negative results. Immunosuppres sive treatment was signifi cantly associated with indeter-minate QFT-GIT results. The rate of indeterminate results was similar to that reported in previous stud-ies,27–29 but lower than that seen in this study’s HIV-infected population. Most indeterminate QFT-GIT results were overrepresented in TST-negative patients; these fi ndings suggest that the availability of an inter-nal positive and negative control is a distinct advan-tage for considering the possibility of false-negative TST results in immunosuppressed patients. A possible limitation of this study was the different selection cri-teria for the enrolment of the two groups of immuno-suppressed patients. Approximately 25% of patients in the HIV-positive group had active TB; on the contrary, as there were no cases of active TB in the IMID sub-group, it is diffi cult to determine which test performs better in this population. This is the main difference

IGRA for TB in immunocompromised patients 839

between the two groups and may have contributed to the varying IGRA fi ndings. Although the QFT-GIT identifi es more patients infected with M. tuberculosis than TST, the authors believe that the varying per-formance of IGRAs does not lead to different clini-cal recommendations for the two immunosuppressed categories.

CONCLUSION

The QFT-GIT performance varies across the differ-ent immunosuppressed populations. For a number of reasons, we suggest that in immunocompromised pa-tients all available data should be used to demonstrate or exclude LTBI.

AcknowledgementsThe authors gratefully acknowledge the contributions to this re-search by the study participants and staff. The study was supported in part by grants from the Sapienza University of Rome (2007–2008) and Fondazione Eleonora Lorillard Spencer Cenci (2008).

References 1 World Health Organization. Global tuberculosis control: epi-

demiology, strategy, fi nancing. WHO report 2009. WHO/HTM/ TB/2009.411. Geneva, Switzerland: WHO, 2009.

2 Jensen P A, Lambert L A, Iademarco M F, Ridzon R. Guide-lines for preventing the transmission of Mycobacterium tuber-culosis in health-care settings. MMWR Recomm Rep 2005; 54: 1–141.

3 Horsburgh C R Jr. Priorities for the treatment of latent tuber-culosis infection in the United States. N Engl J Med 2004; 350: 2060–2067.

4 Huebner R E, Schein M F, Bass J B. The tuberculin skin test. Clin Infect Dis 1993; 17: 968–975.

5 Tissot F, Zanetti G, Francioli P, Zellweger J P, Zysset F. Infl u-ence of bacille Calmette-Guérin vaccination on size of tubercu-lin skin test reaction: to what size? Clin Infect Dis 2005; 40: 211–217.

6 Mazurek G H, Jereb J, Lobue P, et al. Guidelines for using the QuantiFERON-TB Gold test for detecting Mycobacterium tu-berculosis infection, United States. MMWR Recomm Rep 2005; 54 (RR-15): 49–55.

7 Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for the diagnosis of latent tuberculosis infection: an up-date. Ann Intern Med 2008; 149: 177–184.

8 Lalvani A. Diagnosing tuberculosis infection in the 21st cen-tury: new tools to tackle an old enemy. Chest 2007; 131: 1898–1906.

9 Sauzullo I, Mengoni F, Lichtner M, et al. In vivo and in vitro ef-fects of antituberculosis treatment on mycobacterial interferon-gamma T-cell response. PLoS One 2009; 4: e5187.

10 Menzies D, Pai M, Comstock G. Meta-analysis: new tests for the diagnosis of latent tuberculosis infection: areas of uncer-tainty and recommendations for research. Ann Intern Med 2007; 146: 340–354.

11 Guidelines for prevention and treatment of opportunistic in-fections in HIV-infected adults and adolescents: recommenda-tions from CDC, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR Recomm Rep 2009; 58 (RR-4): 1–207.

12 Landis J R, Koch G G. The measurement of observer agree-ment for categorical data. Biometrics 1977; 33: 159–174.

13 Health Protection Agency Tuberculosis Programme Board. Health protection agency position statement on the use of in-

terferon gamma release assay (IGRA) tests for tuberculosis (TB): interim HPA guidance. London, UK: Health Protection Agency, 2008. http://www.hpa.org.uk/HPA/Topics/Infectious Diseases/InfectionsAZ/1191942150139 Accessed February 2010.

14 Domínguez J, Latorre I, Altet N, et al. IFN-γ-release assays to diagnose TB infection in the immunocompromised individual. Expert Rev Respir Med 2009; 3: 309–327.

15 Lange C, Pai M, Drobniewski F, Migliori G B. Interferon-gamma release assays for the diagnosis of active tuberculosis: sensible or silly? Eur Respir J 2009; 33: 1250–1253.

16 Sauzullo I, Mengoni F, Lichtner M, et al. QuantiFERON-TB Gold and selected region of difference 1 peptide-based assays for the detection of Mycobacterium tuberculosis infection in a cohort of patients enrolled with suspected tuberculosis. Diagn Microbiol Infect Dis 2008; 62: 395–401.

17 Brock I, Ruhwald M, Lundgren B, Westh H, Mathiesen L, Ravn P. Latent tuberculosis in HIV positive, diagnosed by the M. tuberculosis specifi c interferon-gamma test. Respir Res 2006; 7: 56–65.

18 Luetkemeyer A, Charlebois E, Flores L, et al. Comparison of an interferon-γ release assay with tuberculin skin testing in HIV-infected individuals. Am J Respir Crit Care Med 2007; 175: 737–742.

19 Jones S, de Gijsel D, Wallach F R, Gurtman A C, Shi Q, Sacks H. Utility of QuantiFERON-TB Gold In-Tube testing for latent TB infection in HIV-infected individuals. Int J Tuberc Lung Dis 2007; 11: 1190–1195.

20 Dheda K, Lalvani A, Miller R F, et al. Performance of a T-cell–based diagnostic test for tuberculosis infection in HIV-infected individuals is independent of CD4 cell count. AIDS 2005; 19: 2038–2041.

21 Stephan C, Wolf T, Goetsch U, et al. Comparing QuantiFERON-TB Gold, T-SPOT tuberculosis and tuberculin skin test in HIV-infected individuals from a low prevalence tuberculosis coun-try. AIDS 2008; 22: 2471–2479.

22 Richeldi L, Losi M, D’Amico R, et al. Performance of tests for latent tuberculosis in different groups of immunocompromised patients. Chest 2009; 136: 198–204.

23 Moiton M P, Richez C, Dumoulin C, Mehsen N, Dehais J, Schaeverbeke T. Role of anti-tumour necrosis factor-alpha ther-apeutic agents in the emergence of infections. Clin Microbiol Infect 2006; 12: 1151–1153.

24 Gómez-Reino J J, Carmona L, Valverde V R, Mola E M, Montero M D. Treatment of rheumatoid arthritis with tumor necrosis factor inhibitors may predispose to signifi cant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis Rheum 2003; 48: 2122–2127.

25 Keane J, Breshinan B. Tuberculosis reactivation during immuno-suppressive therapy in rheumatic diseases: diagnostic and ther-apeutic strategies. Curr Opin Rheumatol 2008; 20: 443–449.

26 Bartalesi F, Vicidomini S, Goletti D, et al. QuantiFERON-TB Gold and TST are both useful for latent TB screening in auto-immune diseases. Eur Respir J 2009; 33: 586–593.

27 Ponce de Leon D, Acevedo-Vasquez E, Alvizuri S, et al. Com-parison of an interferon-γ assay with tuberculin skin testing for detection of tuberculosis infection in patients with rheuma-toid arthritis in a TB-endemic population. J Rheumatol 2008; 35: 776–781.

28 Bocchino M, Matarese A, Bellofi ore B, et al. Performance of two commercial blood IFN-γ release assays for the detection of Mycobacterium tuberculosis infection in patient candidates for anti-TNF-α treatment. Eur J Clin Microbiol Infect Dis 2008; 27: 907–913.

29 Greenberg J D, Reddy S M, Schloss S G, et al. Comparison of an in vitro tuberculosis interferon-gamma assay with delayed type hypersensitivity testing for detection of latent Myco-bacterium tuberculosis: a pilot study in rheumatoid arthritis. J Rheumatol 2008; 35: 770–775.

840 The International Journal of Tuberculosis and Lung Disease

30 Matulis G, Juni P, Villiger P M, Gadola S D. Detection of latent tuberculosis in immunosuppressed patients with autoimmune diseases: performance of a Mycobacterium tuberculosis anti-gen specifi c interferon-γ assay. Ann Rheum Dis 2008; 67: 84–90.

31 Vassilopoulos D, Stamoulis N, Hadziyannis E, Archimandritis A J. Usefulness of enzyme-linked immunospot assay (ELISPOT)

compared to tuberculin skin testing for latent tuberculosis screening in rheumatic patients scheduled for anti-tumor necro-sis factor treatment. J Rheumatol 2008; 35: 1271–1276.

32 Shovman O, Anouk M, Vinnitsky N, et al. QuantiFERON-TB Gold in the identifi cation of latent tuberculosis infection in rheumatoid arthritis: a pilot study. Int J Tuberc Lung Dis 2009; 13: 1427–1432.

R É S U M É

C O N T E X T E : Etude transversale dans quatre polycliniques

d’un seul centre de référence en Italie.

O B J E C T I F : Evaluer les performances du test Quanti-

FERON®-TB Gold In Tube (QFT-GIT) chez les adultes

infectés par le virus de l’immunodéfi cience humaine

(VIH) et chez les patients atteints d’une maladie infl am-

matoire à médiation immunitaire (IMID), candidats pour

un traitement contre le facteur de nécrose tumorale-alpha

(TNF-α).

S C H É M A : Nous avons enrôlé 402 patients en état d’im-

munodépression comportant 207 individus infectés par

le VIH et 195 patients atteints d’IMID chez qui un traite-

ment anti-TNF-α était prévu. On a pratiqué simultané-

ment le test cutané tuberculinique (TST) et le QFT-GIT.

Pour la tuberculose (TB) active, les résultats des tests

ont été comparés avec les diagnostics microbiologique,

histopathologique et clinique.

R É S U LTAT S : Chez les patients infectés par le VIH, le

niveau de concordance entre les deux tests est de 68%

et la sensibilité globale du test QFT-GIT est de 66%

(IC95% 47–82). Nous avons observé une forte propor-

tion de résultats QFT-GIT indéterminés (33,4%), qui

ont été en étroite association avec un décompte de cel-

lules CD4 < 200/μl (P < 0,0001). Chez les patients at-

teints d’IMID, le niveau de concordance avec le TST est

plus élevé (81,6%). Les facteurs en association avec un

TST positif accompagné de résultats négatifs discordants

du QFT-GIT sont la vaccination par le bacille Calmette-

Guérin (P = 0,0001), une maladie TB antérieure (P =

0,0001) ainsi que le fait de travailler comme fermier

(P = 0,0005).

C O N C L U S I O N : Les performances du QFT-GIT varient

en fonction des divers types de patients en immuno-

dépression. Il ne faut pas utiliser des tests de libération

d’interféron-gamma pour affi rmer ou exclure une TB

active chez les adultes infectés par le VIH. Comme aucun

cas de TB active n’existe dans le sous-groupe atteint

d’IMID, nous ne pouvons pas déterminer quel test a une

meilleure performance dans cette population.

R E S U M E N

M A R C O D E R E F E R E N C I A : Estudio transversal de cuatro

servicios de consulta externa en un centro de referencia

en Italia.

O B J E T I V O : Evaluar el rendimiento diagnóstico de la

prueba QuantiFERON®-TB Gold en Tubo (QFT-GIT)

en pacientes adultos infectados por el virus de la inmuno-

defi ciencia humana (VIH) y en pacientes con enferme-

dades infl amatorias mediadas por el sistema inmunita-

rio (IMID), que precisan un tratamiento dirigido contra

el factor de necrosis tumoral alfa (TNF-α).

M É T O D O : Se incorporaron al estudio 402 pacientes con

inmunodepresión, entre los cuales había 207 con infec-

ción por el VIH y 195 pacientes con enfermedades IMID

con indicación de un tratamiento contra el TNF-α. Se

practicaron en forma simultánea la prueba cutánea de la

tuberculina (TST) y la QFT-GIT. En caso de tuberculo-

sis (TB) activa, los resultados de las pruebas se com-

pararon con el diagnóstico microbiológico, histopa-

tológico y clínico.

R E S U LTA D O S : En los pacientes infectados por el VIH el

grado de concordancia de ambas pruebas fue 68% y la

sensibilidad global de la prueba QFT-GIT fue 66%

(IC95% 47–82). Se observó una gran proporción de re-

sultados no concluyentes con esta prueba (33,4%), los

cuales presentaron una fuerte relación con un recuento

de células CD4 < 200/μl (P < 0,0001). En las enferme-

dades IMID, el grado de concordancia entre ambas

pruebas fue más alto (81,6%). Los factores asociados

con resultados discordantes, es decir TST y QFT-GIT

negativa, fueron el antecedente de vacunación con el

bacille Calmette-Guérin (P = 0,0001), el antecedente de

TB (P = 0,0001) y el trabajo como agricultor (P =

0,0005).

C O N C L U S I Ó N : El rendimiento diagnóstico de la prueba

QFT-GIT varía en las distintas categorías de pacientes

inmunodeprimidos. No se deben usar las pruebas de libe-

ración de interferón gama con el objeto de descartar el

diagnóstico de TB activa en los pacientes con VIH.

Puesto que no se presentaron casos de TB activa en el

subgrupo de pacientes con enfermedad IMID, no es po-

sible determinar cuál de las pruebas ofrecería un mejor

rendimiento.