SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Objective

To evaluate the performance of an interferon-γ release assay (IGRA) versus the standard tuberculin skin test (TST) as a screening tool for latent tuberculosis (TB) infection prior to the initiation of anti–tumor necrosis factor therapy in patients with autoimmune inflammatory diseases.

Methods

This integrated analysis involved screening of patients with rheumatoid arthritis, those with psoriatic arthritis, and those with ankylosing spondylitis from phase III trials of golimumab. The IGRA used to screen for latent TB was the QuantiFERON-TB Gold In-Tube test.

Results

In this pooled analysis, 2,282 patients underwent both IGRA and TST screening prior to golimumab treatment. Among these patients, 13.8% had at least one test yielding positive findings for latent TB, including 9.4% with positive results by TST, 7.0% with positive results by IGRA, and 2.6% with positive results on both tests. The rate of indeterminate results for TB on IGRA was 1.8%. Agreement between the TST and IGRA results, measured by the kappa coefficient, was 0.22 (95% confidence interval 0.157–0.279; P = 0.021). Among the patients with positive IGRA findings, 36.9% had positive TST findings. Among the patients with positive TST findings, 27.4% had positive IGRA findings. Overall, 781 (34.2%) of the 2,282 patients had previously received the bacillus Calmette-Guérin (BCG) vaccine; among this vaccinated group, the rate of positivity for latent TB by TST was 15.2% (119 of 781), compared to a rate of positivity of 9.1% (71 of 781) by IGRA (P = 0.0002). Among patients who had not received the BCG vaccine, the rate of positivity by TST was 5.0% (62 of 1,248) and the rate of positivity by IGRA was 5.8% (72 of 1,248) (P = 0.3745). When the IGRA was repeated in patients whose results were initially indeterminate, the rate of indeterminate IGRA findings for latent TB was much lower than has been previously reported.

Conclusion

In the absence of a true gold standard test for latent TB infection, results of this comparison of IGRA and TST in a large cohort of patients with rheumatic diseases suggest that the IGRA provides greater specificity and possibly greater sensitivity than the TST.

Screening for latent tuberculosis (TB) infection prior to use of tumor necrosis factor (TNF) antagonists is a recommended approach to decrease the risk of reactivation of latent Mycobacterium tuberculosis infection (1). The standard screening tuberculin skin test (TST), in use for more than 100 years, has limitations, in that administration of the test can be difficult and there can be variability in interpretation of the results, such as a potential for false-positive results due to cross-reactivity with the bacillus Calmette-Guérin (BCG) vaccine and nontuberculous mycobacteria, and a potential for false-negative results, especially in the immunocompromised patient. These limitations could lead to unnecessary treatment of latent TB infection, with a possibility of treatment-related complications, or to the development of active TB that could have been prevented (2–4). Patients treated with anti-TNF therapy have been shown to have an increased risk of developing active TB; a substantial proportion of the TB cases are disseminated (i.e., miliary), extrapulmonary (5–10), and severe, with an increased mortality rate (11).

Recently developed blood assays that measure the release of interferon-γ (IFNγ) by TB-specific effector T cells, such as the QuantiFERON-TB Gold In-Tube (QFT-GIT) test (Cellestis) and the enzyme-linked immunospot–based T-SPOT TB assay (Oxford Immunotec), offer the possibility of improved detection of latent TB infection. However, data on the performance of these tests in patients with autoimmune inflammatory diseases are limited (12–18).

In the present study, we describe our experience, with the largest cohort to date, comparing the performance of an IFNγ release assay (IGRA) with that of the TST in screening for latent TB infection prior to the initiation of anti-TNF therapy. We performed an integrated analysis of TB screening data from 5 large phase III trials of golimumab, a human anti-TNFα monoclonal antibody, in patients with autoimmune inflammatory diseases. Golimumab was approved for use on the basis of the results of these prior studies (19–23), in which patients with rheumatoid arthritis (RA), patients with psoriatic arthritis (PsA), and patients with ankylosing spondylitis (AS) were also screened for latent TB prior to receiving the anti-TNF treatment.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

This integrated analysis included pooled data on patients who had undergone screening for TB with both an IGRA (i.e., the QFT-GIT) and the standard TST and who were enrolled in 1 of the 5 phase III, randomized, placebo-controlled clinical trials of subcutaneous golimumab for the treatment of rheumatic disease. The analysis included data from the following trials: Golimumab Before Employing Methotrexate as the First-Line Option in the Treatment of Rheumatoid Arthritis of Early Onset (GO-BEFORE; involving patients with active RA who had not received treatment with methotrexate [MTX]), Golimumab in Active Rheumatoid Arthritis Despite Methotrexate Therapy (GO-FORWARD; involving patients with RA whose disease remained active despite receiving treatment with MTX), Golimumab in Patients with Active Rheumatoid Arthritis After Treatment with Tumor Necrosis Factor α Inhibitors (GO-AFTER; involving patients with active RA who were previously treated with anti-TNF agent[s]), Golimumab in Active Psoriatic Arthritis Despite Prior Conventional Therapy (GO-REVEAL; involving patients with PsA whose disease remained active despite conventional treatment with disease-modifying antirheumatic drugs [DMARDs] or nonsteroidal antiinflammatory drugs [NSAIDs]), and Golimumab in Active Ankylosing Spondylitis Despite Prior Conventional Therapy (GO-RAISE; involving patients with AS whose disease remained active despite conventional treatment with DMARDs or NSAIDs). These trials were conducted worldwide at study sites in North and South America, Western and Eastern Europe, Australia, New Zealand, and Asia. Details of the study design, including study treatment regimen and the countries in which patients were enrolled, are presented in Table 1.

Table 1. Study details for clinical trials of golimumab in patients with rheumatoid arthritis (RA), psoriatic arthritis (PsA), and ankylosing spondylitis (AS) included in the integrated analysis*
StudySitesStudy design and population
  • *

    GO-BEFORE = Golimumab Before Employing Methotrexate as the First-Line Option in the Treatment of Rheumatoid Arthritis of Early Onset; MTX = methotrexate; GO-FORWARD = Golimumab in Active Rheumatoid Arthritis Despite Methotrexate Therapy; GO-AFTER = Golimumab in Patients with Active Rheumatoid Arthritis After Treatment with Tumor Necrosis Factor α Inhibitors; anti-TNFα = anti–tumor necrosis factor α; GO-REVEAL = Golimumab in Active Psoriatic Arthritis Despite Prior Conventional Therapy; DMARDs = disease-modifying antirheumatic drugs; NSAIDs = nonsteroidal antiinflammatory drugs; GO-RAISE = Golimumab in Active Ankylosing Spondylitis Despite Prior Conventional Therapy.

GO-BEFORE (n = 637)Total of 90 sites: 25 in Asia (India, Malaysia, Philippines, Singapore, Republic of Korea, Taiwan, Thailand); 34 in Europe/Australia/New Zealand (Australia, Austria, Belgium, Hungary, New Zealand, Poland, Italy, Russia, Spain, Ukraine, United Kingdom); 21 in North America (Canada, United States); 10 in Latin America (Argentina, Chile)Phase III, multicenter, randomized, double-blind, placebo-controlled, 4-arm parallel study of the efficacy and safety of golimumab administered in 2 doses multiple times subcutaneously as monotherapy or in combination with MTX in adult patients with active RA who were naive to treatment with MTX
GO-FORWARD (n = 444)Total of 60 sites: 7 in Asia (South Korea, Taiwan); 19 in Europe/Australia/New Zealand (Germany, Hungary, Poland, Australia, New Zealand); 19 in North America (Canada, United States); 15 in Latin America (Argentina, Chile, Mexico)Phase III, multicenter, randomized, double-blind, placebo-controlled, 4-arm parallel study of the efficacy and safety of golimumab administered in 2 doses multiple times subcutaneously as monotherapy or in combination with MTX in adult patients with active RA despite treatment with MTX
GO-AFTER (n = 461)Total of 86 sites: 0 in Asia; 33 in Europe/Australia/New Zealand (Australia, New Zealand, The Netherlands, Austria, Finland, Germany, Spain, United Kingdom); 53 in North America (Canada, United States); 0 in Latin AmericaPhase III, multicenter, randomized, double-blind, placebo-controlled, 3-arm parallel study of the efficacy and safety of golimumab administered multiple times subcutaneously in adult patients with active RA who had received previous treatment with ≥1 dose of biologic anti-TNFα agent(s)
GO-REVEAL (n = 405)Total of 58 sites: 0 in Asia; 22 in Europe/Australia/New Zealand (Belgium, Poland, Spain, United Kingdom); 36 in North America (Canada, United States); 0 in Latin AmericaPhase III, multicenter, randomized, double-blind, placebo-controlled, 3-arm study of the efficacy and safety of golimumab administered multiple times subcutaneously in adult patients with active PsA who had an inadequate response to current or previous treatment with DMARDs or NSAIDs
GO-RAISE (n = 356)Total of 46 sites: 9 in Asia (South Korea, Taiwan); 17 in Europe/Australia/New Zealand (Belgium, The Netherlands, Germany, Finland, France); 20 in North America (Canada, United States); 0 in Latin AmericaPhase III, multicenter, randomized, double-blind, placebo-controlled, 3-arm parallel study of the efficacy and safety of golimumab administered multiple times subcutaneously in adult patients with active AS despite previous treatment with DMARDs or NSAIDs

The patients were screened for eligibility using 3 TB screening tests: the standard TST, the QFT-GIT test, and a chest radiograph. Furthermore, eligibility criteria included having no history of latent/active TB prior to screening (except in GO-AFTER, which allowed the inclusion of patients with a history of latent TB who had been treated within the last 3 years) and having no signs or symptoms of active TB or no recent close contact with anyone with active TB (in cases of recent contact, the patient must have been referred to a TB specialist and, if warranted, treated for latent TB before or at the time of the first dose of study agent). Any patient with a newly identified positive finding for TB on a diagnostic test (positive on either the TST or IGRA) in whom there was no evidence of active TB was allowed to enter the trial as long as appropriate treatment for latent TB was initiated before or at the time of the first dose of study agent. All patients were required to have a chest radiograph, obtained within 3 months before the first dose of study agent, that showed no evidence of active TB or old, inactive TB.

The TST was performed according to the Mantoux method, using 5 tuberculin units (TU) of purified protein derivative (PPD) standard or 2 TU of PPD RT-23 (Statens Serum Institut). A trained health-care worker recorded each patient's reaction to the TST at 48–72 hours after placement. The TST was deemed positive for latent TB infection according to the local country guidelines for defining an immunosuppressed host or, in the absence of local guidelines, according to the presence of induration ≥5 mm.

The QFT-GIT test was the IGRA assay used. For this procedure, standard venipuncture is performed at a single visit to collect blood in tubes that contain the M tuberculosis–specific antigens. The QFT-GIT test also contains an extra antigen, TB7.7(p4), that was not present in the original version of this IGRA and is thought to improve sensitivity. In addition, this version of the IGRA shortens the manual processing time, since antigens are already present in the tubes (24). Initial IGRA sample-handling procedures were performed at investigational sites, and a central laboratory performed the enzyme-linked immunosorbent assay–based testing and reported the results for each patient according to the manufacturer's interpretation criteria. Positive results were confirmed by duplicate testing of the same sample. Any results initially indeterminate on the IGRA required a second sample to be drawn and tested, and the final results were used to determine study eligibility.

Concordance between the IGRA result and the TST result was determined using the kappa coefficient. Indeterminate results on the IGRA were considered to be discordant with the TST results. The IGRA and TST results were assessed overall, and also stratified by known BCG vaccination status and by geographic region. Analyses were also performed by defining positivity for latent TB according to the presence of induration ≥5 mm on the TST in all patients. Baseline demographic and disease characteristics were summarized using descriptive statistics. Statistical testing was performed using a chi-square test at a significance level of 0.05 (2-sided). Multivariate logistic regression analyses were performed to determine whether there were any factors (i.e., baseline MTX use, baseline steroid use, disease type, age, geographic region, and prior BCG vaccination) associated with the screening test results.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

Baseline characteristics and demographics.

In this integrated analysis, data from 5 clinical trials of golimumab in patients with rheumatic disease were combined (Table 1). In this pooled cohort (total of 2,303 patients randomized to receive treatment), the majority of patients were female (65.8%), were white (80.8%) (self-reported race), had RA (67.0%), and had disease (RA, PsA, or AS) for at least 1 year (79.5%) (Table 2). The median age was 49 years. Geographically, patients were predominantly from North America (41.8%), with similar proportions of patients from Western and Eastern Europe (19.1% and 18.8%, respectively). Approximately 20% of the study population came from Asia and Latin America (11.6% and 8.8%, respectively).

Table 2. Baseline demographic and clinical characteristics of the patients in the combined cohort from the phase III golimumab trials*
  • *

    Except where indicated otherwise, values are the number (%) of patients. BCG = bacillus Calmette-Guérin; MTX = methotrexate; DMARDs = disease-modifying antirheumatic drugs; NSAIDs = nonsteroidal antiinflammatory drugs.

  • All patients receiving biologics were in the Golimumab in Patients with Active Rheumatoid Arthritis After Treatment with Tumor Necrosis Factor α Inhibitors (GO-AFTER) study.

  • Baseline use of MTX was determined in all patients in the Golimumab in Active Rheumatoid Arthritis Despite Methotrexate Therapy (GO-FORWARD) study and none of the patients in the Golimumab Before Employing Methotrexate as the First-Line Option in the Treatment of Rheumatoid Arthritis of Early Onset (GO-BEFORE) study (as per protocols). MTX use was allowed but not required in the GO-AFTER, Golimumab in Active Psoriatic Arthritis Despite Prior Conventional Therapy (GO-REVEAL), and Golimumab in Active Ankylosing Spondylitis Despite Prior Conventional Therapy (GO-RAISE) studies.

Total patients randomized2,303
Male788 (34.2)
Age, median (mean ± SD) years49.00 (48.58 ± 12.663)
Race 
 White1,860 (80.8)
 Black46 (2.0)
 Asian285 (12.4)
 Other112 (4.9)
Disease under investigation 
 Rheumatoid arthritis1,542 (67.0)
 Psoriatic arthritis405 (17.6)
 Ankylosing spondylitis356 (15.5)
Geographic region 
 North America962 (41.8)
 Western Europe440 (19.1)
 Eastern Europe432 (18.8)
 Latin America203 (8.8)
 Asia266 (11.6)
Disease duration 
 >3 years1,400 (60.8)
 >2 years to ≤3 years182 (7.9)
 >1 year to ≤2 years248 (10.8)
 ≤1 year468 (20.3)
Prior BCG vaccination788 (34.2)
Previous treatments 
 MTX1,269 (55.1)
 DMARDs1,771 (76.9)
  Hydroxychloroquine542 (23.5)
  Sulfasalazine809 (35.1)
  Leflunomide333 (14.5)
  Other DMARDs138 (6.0)
 Anakinra31 (1.3)
 Immunosuppressives56 (2.4)
 Systemic corticosteroids1,539 (66.8)
 NSAIDs2,247 (97.6)
 Biologics461 (20.0)
Required treatment for latent tuberculosis per protocol317 (13.8)
Treatments at baseline 
 MTX571 (24.8)
  Median (mean ± SD) dosage, mg/week15 (15.75 ± 5.177)
 Corticosteroids1,000 (43.4)
  Median (mean ± SD) oral dosage, mg/day5 (6.75 ± 2.655)

More than three-fourths (76.9%) of the patients had previously received treatment with one or more DMARDs. At baseline, almost one-half (43.4%) of the patients were receiving corticosteroids (median dosage 5 mg/day, mean ± SD 6.75 ± 2.655 mg/day), and approximately one-quarter (24.8%) of the patients were receiving MTX (median dosage 15 mg/week, mean ± SD 15.75 ± 5.177 mg/week) (Table 2). With regard to prior vaccination with the BCG vaccine, 788 patients had received the vaccine, 1,260 had not received the vaccine, and 255 had an unknown BCG vaccination status. As a result of the TB testing that was conducted at screening, 317 patients (13.8%) began treatment for latent TB infection prior to administration of the first study agent.

Across the 5 clinical trials included in this analysis, 99.1% of the patients (2,282 of 2,303) had both the IGRA and the TST performed at screening. All analyses of test results are based on this population of 2,282 patients (Table 3). Most of the patients who did not have both TB screening tests were from the GO-AFTER study (RA patients previously treated with anti-TNF agent[s]), in which a screening TST and IGRA were not required for patients with a history of latent TB infection who had documented treatment within the previous 3 years.

Table 3. Results of the tuberculin skin test (TST) and interferon-γ release assay (IGRA) in 2,282 patients with both tests performed at screening
Screening test resultNo. (%) of patients
IGRA/TST 
 IGRA positive 
  TST positive59
  TST negative101
 IGRA negative
  TST positive150
  TST negative1,931
 IGRA indeterminate
  TST positive6
  TST negative35
Overall 
 Positive by TST only215 (9.4)
 Positive by IGRA only160 (7.0)
 Positive by both TST and IGRA59 (2.6)
 Positive by either TST or IGRA316 (13.8)

Screening TST and IGRA results by BCG vaccination status.

Among the 2,282 patients in this pooled analysis, at least one TB screening test result was positive in 13.8% of the patients (316 of 2,282) (Table 3). Overall, 9.4% of the patients (215 of 2,282) had positive results on the TST, 7.0% (160 of 2,282) had positive results on the IGRA, and 2.6% (59 of 2,282) had positive results on both the IGRA and TST (Figure 1A and Table 3). The rate of indeterminate results on the IGRA (i.e., 2 sequential IGRAs with indeterminate results) was 1.8% (41 of 2,282).

thumbnail image

Figure 1. Overlap of screening test results and rates of positivity from the tuberculin skin test (TST) and an interferon-γ release assay (i.e., the QuantiFERON-TB Gold In-Tube [QFT-GIT] test) for the detection of latent tuberculosis infection in all patients (A), bacillus Calmette-Guérin (BCG)–vaccinated patients (B), and non–BCG-vaccinated patients (C).

Download figure to PowerPoint

Agreement between the TST and IGRA results, measured by the kappa coefficient, was 0.22 (95% confidence interval [95% CI] 0.157–0.279; P = 0.021). Among the patients who were positive by IGRA, 36.9% (59 of 160) also were positive by TST, and of those who were positive by TST, only 27.4% (59 of 215) also were positive by IGRA (Figure 1A). As specified by the study protocol, the TST threshold that defined a positive result for TB varied according to local country guidelines on immunocompromised patients or, if no local country guidelines existed, United States guidelines were employed, requiring an induration of ≥5 mm. Most, but not all, study-site locations utilized the latter definition of positivity, a cutoff of 5 mm on the TST. Therefore, when the same analyses were performed with the use of a threshold of ≥5 mm to determine positivity for latent TB infection in all patients, the rate of positivity by TST increased a small amount (from 9.4% to 11.1%). Nonetheless, agreement between the IGRA and TST results remained fair (kappa coefficient of 0.22), and overall results were thus similar to the analyses of data based on local country guidelines.

Overall, 34.2% of all patients who had undergone both screening tests (781 of 2,282 patients) had previously received the BCG vaccine, while 54.7% (1,248 of 2,282) had no history of BCG vaccination and 11.1% (253 of 2,282) had an unknown BCG vaccination status (Table 4). Among patients who had previously undergone BCG vaccination, the rate of positivity for TB on the TST was 15.2% (119 of 781), compared with a rate of positivity of 9.1% on the IGRA (71 of 781) (P = 0.0002). Among patients who had not previously undergone BCG vaccination, the rate of positivity for TB on the TST was 5.0% (62 of 1,248) compared with a rate of positivity of 5.8% on the IGRA (72 of 1,248) (P = 0.3745).

Table 4. Results of the tuberculin skin test (TST) and interferon-γ release assay (IGRA) in patients with and those without prior bacillus Calmette-Guérin (BCG) vaccination*
Screening test resultPrior BCG vaccination (n = 781)No prior BCG vaccination (n = 1,248)
  • *

    Values are the number (%) of patients.

IGRA/TST  
 IGRA positive  
  TST positive2824
  TST negative4348
 IGRA negative  
  TST positive9132
  TST negative6101,120
 IGRA indeterminate  
  TST positive06
  TST negative918
Overall  
 Positive by TST only119 (15.2)62 (5.0)
 Positive by IGRA only71 (9.1)72 (5.8)
 Positive by both TST and IGRA28 (3.6)24 (1.9)
 Positive by either TST or IGRA162 (20.7)110 (8.8)

As was also observed in the overall population, the rates of positivity for TB on the IGRA and on the TST overlapped to a small extent in both the BCG-vaccinated and non–BCG-vaccinated patients (Figures 1B and C). Although this overlap was slightly greater in BCG-vaccinated patients, a large discordance between the tests remained, as demonstrated by a kappa coefficient of 0.27. Among the 110 non–BCG-vaccinated patients who had a finding of positivity for latent TB infection on either the TST or the IGRA, only 24 patients (21.8%) were positive by both tests. Among the patients with a positive finding for TB on the TST, only 38.7% were also positive by IGRA, and among the patients positive by IGRA, only 33.3% were also positive by TST.

Screening TST and IGRA results by geographic region.

At least one TB screening test result was positive in 9.3% of the patients (128 of 1,382) from sites in North America and Western Europe, 16.0% of the patients (69 of 432) from Eastern Europe, 15.8% of the patients (32 of 203) from Latin America, and 33.2% of the patients (88 of 265) from Asia (results not shown). The proportions of patients who had received prior BCG vaccination, by geographic region, were 11.8% in North America and Western Europe, 81.0% in Eastern Europe, 78.8% in Latin America, and 40.8% in Asia.

Screening TST and IGRA results by baseline corticosteroid use.

Patients were permitted to receive treatment with low-dose corticosteroids in these phase III golimumab studies, with the maximum dosage being no higher than a prednisone equivalent of 10 mg daily. The rate of corticosteroid use at baseline was similar between patients with positive results for latent TB infection on the TST and those with negative results by TST (48.8% versus 42.9%, respectively), and was somewhat higher in those with positive results on the IGRA compared with those with negative results on the IGRA (52.3% versus 42.3%, respectively). Among patients who had indeterminate results on the IGRA, 49.2% reported use of corticosteroids at baseline. The median daily dosage of corticosteroids at baseline was 5 mg in patients with either positive or negative results on the TST; among patients with positive results on the IGRA, those with negative results on the IGRA, and those with indeterminate results on the IGRA, the median daily dosage at baseline was 6.1 mg, 5 mg, and 7.8 mg, respectively.

Factors associated with TB test results, as determined by multivariate analysis.

To determine the presence of clinically important factors that may be associated with positive results for latent TB infection on the IGRA and TST, multivariate logistic regression analyses that included the covariates of baseline MTX use, baseline steroid use, disease type, age, geographic region, and prior BCG vaccination were also performed. As shown in Table 5, factors that were significantly associated with a positive result on the IGRA were age (≥65 years) and geographic regions outside of North America.

Table 5. Multivariate analyses of factors associated with positive findings of latent tuberculosis infection on screening tests in 2,282 patients with both tests performed*
 Positive findings on IGRA onlyIGRA and TST concordant resultsPositive findings on TST only, per local country guidelines
OR95% CIOR95% CIOR95% CI
  • *

    Values are the odds ratio (OR) and 95% confidence interval (95% CI) for the likelihood of a positive test result on the interferon-γ release assay (IGRA) and/or tuberculin skin test (TST) in relation to each factor. AS = ankylosing spondylitis; RA = rheumatoid arthritis; PsA = psoriatic arthritis; BCG = bacillus Calmette-Guérin.

  • Excludes 41 patients with indeterminate IGRA results.

  • For concordance analyses, indeterminate IGRA results were considered discordant with TST results.

  • §

    P < 0.05.

Methotrexate, yes vs. no1.050.75–1.470.950.73–1.230.970.71–1.31
Corticosteroids, yes vs. no1.200.83–1.731.080.82–1.440.910.65–1.28
Disease, AS vs. RA0.770.44–1.341.480.97–2.251.090.71–1.65
Disease, PsA vs. RA1.350.81–2.230.880.60–1.281.270.79–2.03
Age, ≥65 years vs. <65 years2.33§1.45–3.730.56§0.38–0.820.600.31–1.14
Region, Asia vs. North America8.48§4.78–15.030.21§0.14–0.327.47§4.61–12.08
Region, Eastern Europe vs. North America3.58§1.93–6.630.63§0.40–0.990.950.53–1.70
Region, Latin America vs. North America3.43§1.64–7.190.750.42–1.341.560.80–3.05
Region, Western Europe vs. North America3.41§1.99–5.830.44§0.30–0.642.10§1.30–3.38
Prior BCG vaccination, yes vs. no1.000.66–1.510.56§0.40–0.772.47§1.71–3.55

We also determined the probability of an association of baseline factors with a positive result on the TST, with positivity for latent TB defined on the basis of local country guidelines (Table 5) or according to the presence of induration ≥5 mm (results not shown). The results were consistent between these 2 interpretations of TST positivity for TB, with each interpretation showing that sites in Asia and Western Europe had statistically significantly higher odds of having patients with positive results on the TST compared with sites in North America. However, the magnitude of the odds ratio (OR) for the likelihood of having a positive TST result in patients from Asia compared to patients from North America was lower when positivity was determined on the basis of local guidelines (OR 7.47, 95% CI 4.61–12.08) than when it was based on induration ≥5 mm (OR 14.08, 95% CI 8.82–22.42).

In addition, patients who had previously received the BCG vaccine had higher odds of a positive result on the TST than those who had not received the BCG vaccine (OR 2.47, 95% CI 1.71–3.55 with positivity based on local country guidelines, versus OR 2.75, 95% CI 1.93–3.92 with positivity based on induration ≥5 mm). The magnitudes of the ORs for a positive TST result associated with other factors were similar when either TST interpretation method was used.

Analyses to determine the probability of an association of baseline factors with indeterminate results on the IGRA suggested that patients with AS were less likely than patients with RA to have an indeterminate IGRA result (OR 0.24, 95% CI 0.07–0.84 for AS versus RA). Patients in Asia were more likely than those in North America to have an indeterminate IGRA result (OR 4.24, 95% CI 1.91–9.43). The number of patients with indeterminate results was small (41 of 2,282) and therefore precluded us from making any conclusions regarding these data.

Finally, patients who had undergone prior BCG vaccination, were age 65 years or older, and generally those in regions outside of North America were less likely to have concordance of positive results between the IGRA (QFT-GIT) and the TST.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

In this pooled analysis of data from 5 large randomized controlled trials, we describe, in the largest cohort to date, a comparison of the use of TST and IGRA (i.e., QFT-GIT test) in screening for latent TB infection in an autoimmune inflammatory disease population prior to initiation of anti-TNF therapy. Patients with autoimmune inflammatory diseases who are typically receiving immunosuppressant treatment and may be in need of anti-TNF therapy are an important group for targeted screening of latent TB infection, given the high risk for progression to active TB and increased susceptibility to more severe forms of TB in such patients (8). Treatment of latent TB infection has been shown to be effective in reducing the risk of developing active TB (25). However, accurate determination of latent infection can be challenging.

The TST has been used for longer than a century as a screening test for latent TB infection. However, it is limited by false-positive results, which may be attributed to cross-reactivity with the BCG vaccine and environmental nontuberculous mycobacteria, and also limited by false-negative results, especially in immunocompromised individuals. Other limitations include variability in administration and interpretation of the TST findings, and the need for a second visit, for which many patients do not return. Such shortcomings may lead to inappropriate treatment with isoniazid, which has possible toxicity, or may lead to the development of active TB that could have been prevented. Immunocompromised patients are often anergic, and results of previous studies have suggested that the rate of TST false-negative results is ∼40% in patients with RA (3). Thus, more reliable screening methods are needed, especially for immunocompromised patients.

Sequencing of the M tuberculosis genome allowed the identification of M tuberculosis–specific antigens and the development of IGRAs, which offer the possibility of improved detection of latent TB infection. However, the data on the performance of these assays, such as the QFT-GIT test, in immunocompromised patients have been limited (26).

In our comparative analysis of the TST and IGRA in patients with autoimmune inflammatory arthritis, agreement, as measured by the kappa coefficient, was only fair (κ = 0.22), owing to a sizeable discordance in the rate of positive results between the 2 tests. Of the 316 patients who had a positive result on either the TST or the IGRA, only 59 patients (19%) were positive by both tests. Among the patients positive by TST, only 27% were positive by IGRA, and among patients positive by IGRA, only 37% were positive by TST. This large discordance has been observed in other studies and has been attributed to false-positive TST results, mainly due to BCG vaccination, and to false-negative TST results in immunosuppressed individuals (3, 15, 27–32).

In our analysis, the rate of positivity for latent TB infection by TST was significantly greater than the rate of positivity determined by IGRA among BCG-vaccinated patients, while among non–BCG-vaccinated patients, the rates of positivity by TST and IGRA were not significantly different. These findings support the idea that BCG vaccination could account for some of the positive results on the TST. The proportion of patients with positive results on the TST was 3-fold higher in BCG-vaccinated patients than in unvaccinated patients, and the proportion of patients with a positive result on the IGRA was only 1.5-fold higher in BCG-vaccinated patients compared with unvaccinated patients.

Given that the antigens in the QFT-GIT test are not found in the BCG vaccine (24) and that the QFT-GIT test maintains high specificity in BCG-vaccinated patients (33), the somewhat higher rate of positivity for TB by IGRA in BCG-vaccinated patients compared with unvaccinated patients is likely a reflection of the higher prevalence of latent TB infection in countries that have BCG vaccination programs, as opposed to false-positive IGRA results. Supporting this concept, the proportions of patients with prior BCG vaccination were much higher in Eastern Europe, Latin America, and Asia, where the prevalence of TB is also much higher, when compared with the proportions of patients with prior BCG vaccination in North America and Western Europe. Multivariate analyses also indicated that prior BCG vaccination was associated with positive TST results, but not positive QFT-GIT test results, after controlling for geographic region, and concordance between the 2 tests was less likely in those patients with prior BCG vaccination.

Because BCG vaccination can confound TST results, we examined the agreement between TST and IGRA results in patients who had not undergone BCG vaccination. Although there was slightly more overlap in the rates of positivity by the TST and IGRA among BCG-vaccinated patients, a large discordance remained. Thus, in our analysis, BCG vaccination does not appear to be a main cause for the observed large discordance between the TST and IGRA positive test results. Although the TST is known to react with more types of nontuberculous mycobacteria than the IGRA, in non–BCG-vaccinated patients, the rate of positivity for TB by IGRA was slightly higher than that by TST (5.8% versus 5.0%, respectively), suggesting that the IGRA has a slightly greater sensitivity than the TST for the detection of latent TB in this population.

Other studies have also demonstrated a potentially greater sensitivity of the IGRA than the TST for the detection of latent TB infection in immunocompromised patients. Some of these studies have reported a similar rate of positive IGRA results between an immunocompromised patient group and a healthy control group, whereas the rate of positive results on the TST was much lower in the immunocompromised group than in the healthy control group (3, 15, 30, 34). Others have found that positive results on the IGRA have a greater association with TB risk factors than positive results on the TST, suggesting that the IGRA has better sensitivity and specificity than the TST (31). However, because there is no gold standard test for latent TB infection, the accuracy of the IGRA and TST cannot be directly determined, and the number of patients with discordant test results who truly have latent TB infection is unknown.

This was a global study that allowed for the evaluation of the IGRA in comparison with the TST in geographic areas with a high incidence versus low incidence of TB, as well as in large numbers of BCG-vaccinated and unvaccinated patients. The rates of latent TB infection determined by both the IGRA and TST were lowest in North America and Western Europe, and highest in Asia. In multivariate analyses, patients in all regions outside of North America were more likely to have a positive IGRA (QFT-GIT) test result, whereas patients in Asia and Western Europe were more likely to have a positive TST result, compared with those in North America. A very high proportion of patients in Eastern Europe and Latin America had prior BCG vaccination, and this may explain why the increased likelihood of a positive TST result in those regions was not observed after controlling for prior BCG vaccination. Most, but not all, of the countries with sites participating in the studies used a 5-mm cutoff on the TST to determine the threshold of positivity for TB, and therefore we performed analyses using a 5-mm TST cutoff in all patients. Results of these analyses were similar to those performed using local country guidelines to define positivity for TB.

The rate of IGRA-indeterminate results in our study was much lower than in some previous reports. Indeterminate rates in previous studies have generally ranged from 2% to more than 20% (3, 15, 17, 27–30, 35, 36), and some have cited the high rate of indeterminate results as a source of concern in using the IGRA in immunocompromised individuals (27). Unlike the protocols in other studies, we repeated the IGRA in those patients whose initial results were indeterminate, and many of these repeat tests yielded negative results. The rate of initially indeterminate IGRA results was ∼5%, which was reduced to 1.8% after a second test. Because poor sample handling can be a cause of indeterminate IGRA results (24), repeating the IGRA likely helped to address this issue, thereby lowering the indeterminate rate. The majority of persistently indeterminate IGRA results could be attributed to the low response to mitogen.

The lower rate of indeterminate results in this study compared with other studies is likely also related to the inclusion of a less severely immunocompromised population compared with higher rates of indeterminate results reported in other studies. Indeterminate IGRA results have been associated with more significant levels of immunocompromise in patients, characterized by either more intensive immunosuppressive treatment or lower CD4 counts in patients with human immunodeficiency virus (HIV) infection (30–32, 35, 37). Although a large proportion of the patients in our analysis were receiving MTX and/or other DMARDs at the time of study enrollment and initial TB testing, lower doses of oral corticosteroids were used, no patients were receiving anti-TNF therapies or other biologics, and patients could not have severe comorbid conditions or HIV infection. In our analysis, in which the maximum corticosteroid dosage was 10 mg/day prednisone or equivalent, use of corticosteroids or MTX did not appear to have a substantial effect on the TST or IGRA results.

Since there is no gold standard test for latent TB infection, it is not known with certainty which patients in the study truly had latent TB infection. Therefore, we cannot ascertain the true performance of the IGRA and can only infer it from comparison with results of the TST and factors associated with a false-positive versus true-positive result. Another limitation of the study is that all patients with a positive TST or IGRA result were treated for latent TB infection prior to initiating anti-TNF therapy. Therefore, patients could not be followed up to determine whether a positive result on the TST or IGRA increased the risk of developing active TB. Two previously published studies that followed up patients with a positive TST or IGRA (QFT-Gold/QFT-GIT test) result who were not treated for latent TB found that the IGRA was more predictive of the development of active TB than the TST (38, 39). In our analysis, none of the patients who were positive by either the TST or the IGRA and who had received isoniazid prophylaxis developed active TB during the 1 year followup period after initiation of golimumab therapy. This is consistent with previous observations in an infliximab study, in which all patients (n = 80) who were positive by TST received isoniazid and none of them developed active TB (40). Among the patients who had a negative finding on both the TST and the IGRA at screening, 5 developed active TB through 1 year of followup.

Thus, in the absence of a true gold standard test to screen for latent TB infection, results of this large-cohort comparison of an IGRA (the QFT-GIT test) and the TST in patients with rheumatic disease suggest that the IGRA provides greater specificity and possibly greater sensitivity than the TST.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Hsia had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Hsia, Schluger, Chaisson, Rahman.

Acquisition of data. Hsia, Beutler, Doyle, Hsu, Rahman.

Analysis and interpretation of data. Hsia, Schluger, Cush, Chaisson, Matteson, Xu, Beutler, Doyle, Hsu, Rahman.

ROLE OF THE STUDY SPONSOR

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

This study was funded by Janssen Research & Development, LLC. The analyses were conducted by Janssen statisticians and programmers. All authors reviewed and approved the content of the manuscript before submission and jointly agreed to submit the final version of the manuscript. The individual studies included in the integrated analysis were partially funded by Merck (formerly, Schering-Plough). Dr. Rahman is currently an employee of Pfizer Inc., but was employed by Janssen Research & Development, LLC at the time this integrated analysis was conducted.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES

The authors greatly appreciate the medical writing support of Mary Ann Thomas, RN, BSN and Mary Whitman, PhD (Janssen Services, LLC).

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. ROLE OF THE STUDY SPONSOR
  8. Acknowledgements
  9. REFERENCES
  • 1
    Furst DE, Keystone EC, Fleischmann R, Mease P, Breedveld FC, Smolen JS, et al. Updated consensus statement on biological agents for the treatment of rheumatic diseases, 2009. Ann Rheum Dis 2010; 69 Suppl I: i229.
  • 2
    Diel R, Loddenkemper R, Meywald-Walter K, Gottschalk R, Nienhaus A. Comparative performance of tuberculin skin test, QuantiFERON-TB-Gold In Tube Assay, and T Spot TB test in contact investigations for tuberculosis. Chest 2009; 135: 10108.
  • 3
    Ponce de Leon D, Acevedo-Vasquez E, Alvizuri S, Gutierrez C, Cucho M, Alfaro J, et al. Comparison of an interferon-γ assay with tuberculin skin testing for detection of tuberculosis (TB) infection in patients with rheumatoid arthritis in a TB-endemic population. J Rheumatol 2008; 35: 77681.
  • 4
    Sellam J, Hamdi H, Roy C, Baron G, Lemann M, Puechal X, et al, for the RATIO (Research Axed on Tolerance of Biotherapies) Study Group. Comparison of in vitro-specific blood tests with tuberculin skin test for diagnosis of latent tuberculosis before anti-TNF therapy. Ann Rheum Dis 2007; 66: 16105.
  • 5
    Ahmad K, Rogers S. Two years of experience with etanercept in recalcitrant psoriasis. Br J Dermatol 2007; 156: 10104.
  • 6
    Sterry W, Barker J, Boehncke WH, Bos JD, Chimenti S, Christophers E, et al. Biological therapies in the systemic management of psoriasis: International Consensus Conference. Br J Dermatol 2004; 151 Suppl 69: 317.
  • 7
    Wolfe F, Michaud K, Anderson J, Urbansky K. Tuberculosis infection in patients with rheumatoid arthritis and the effect of infliximab therapy. Arthritis Rheum 2004; 50: 3729.
  • 8
    Gardam MA, Keystone EC, Menzies R, Manners S, Skamene E, Long R, et al. Anti-tumor necrosis factor agents and tuberculosis risk: mechanisms of action and clinical management. Lancet Infect Dis 2003; 3: 14855.
  • 9
    Keane J, Gershon S, Wise RP, Mirabile-Levens E, Kasznica J, Schwieterman WD, et al. Tuberculosis associated with infliximab, a tumor necrosis factor α-neutralizing agent. N Engl J Med 2001; 345: 1098104.
  • 10
    Keane J. TNF-blocking agents and tuberculosis: new drugs illuminate an old topic. Rheumatology (Oxford) 2005; 44: 71420.
  • 11
    Gomez-Reino JJ, Carmona L, Valverde VR, Mola EM, Montero MD, on behalf of the BIOBADASER Group. Treatment of rheumatoid arthritis with tumor necrosis factor inhibitors may predispose to significant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis Rheum 2003; 48: 21227.
  • 12
    Inanc N, Aydin SZ, Karakurt S, Atagunduz P, Yavuz S, Direskeneli H. Agreement between Quantiferon-TB gold test and tuberculin skin test in the identification of latent tuberculosis infection in patients with rheumatoid arthritis and ankylosing spondylitis. J Rheumatol 2009; 36: 267581.
  • 13
    Dinser R, Fousse M, Sester U, Albrecht K, Singh M, Kohler H, et al. Evaluation of latent tuberculosis infection in patients with inflammatory arthropathies before treatment with TNF-α blocking drugs using a novel flow-cytometric interferon-γ release assay. Rheumatology (Oxford) 2008; 47: 2128.
  • 14
    Greenberg JD, Reddy SM, Schloss SG, Kurucz OS, Bartlett SJ, Abramson SB, et al. Comparison of an in vitro tuberculosis interferon-γ assay with delayed-type hypersensitivity testing for detection of latent Mycobacterium tuberculosis: a pilot study in rheumatoid arthritis. J Rheumatol 2008; 35: 7705.
  • 15
    Matulis G, Juni P, Villiger PM, Gadola SD. Detection of latent tuberculosis in immunosuppressed patients with autoimmune diseases: performance of a Mycobacterium tuberculosis antigen-specific interferon γ assay. Ann Rheum Dis 2008; 67: 8490.
  • 16
    Kobashi Y, Mouri K, Obase Y, Fukuda M, Miyashita N, Oka M. Clinical evaluation of QuantiFERON TB-2G test for immunocompromised patients. Eur Respir J 2007; 30: 94550.
  • 17
    Ferrara G, Losi M, Meacci M, Meccugni B, Piro R, Roversi P, et al. Routine hospital use of a new commercial whole blood interferon-γ assay for the diagnosis of tuberculosis infection. Am J Respir Crit Care Med 2005; 172: 6315.
  • 18
    Meier T, Eulenbruch HP, Wrighton-Smith P, Enders G, Regnath T. Sensitivity of a new commercial enzyme-linked immunospot assay (T SPOT-TB) for diagnosis of tuberculosis in clinical practice. Eur J Clin Microbiol Infect Dis 2005; 24: 52936.
  • 19
    Emery P, Fleischmann RM, Moreland LW, Hsia EC, Strusberg I, Durez P, et al. Golimumab, a human anti–tumor necrosis factor α monoclonal antibody, injected subcutaneously every four weeks in methotrexate-naive patients with active rheumatoid arthritis: twenty-four–week results of a phase III, multicenter, randomized, double-blind, placebo-controlled study of golimumab before methotrexate as first-line therapy for early-onset rheumatoid arthritis [published erratum appears in Arthritis Rheum 2010;62:2812]. Arthritis Rheum 2009: 60: 227283.
  • 20
    Keystone EC, Genovese MC, Klareskog L, Hsia EC, Hall ST, Miranda PC, et al. Golimumab, a human antibody to tumour necrosis factor α given by monthly subcutaneous injections, in active rheumatoid arthritis despite methotrexate therapy: the GO-FORWARD Study. Ann Rheum Dis 2009; 68: 78996.
  • 21
    Kavanaugh A, McInnes I, Mease P, Krueger GG, Gladman D, Gomez-Reino J, et al. Golimumab, a new human tumor necrosis factor α antibody, administered every four weeks as a subcutaneous injection in psoriatic arthritis: twenty-four–week efficacy and safety results of a randomized, placebo-controlled study [published erratum appears in Arthritis Rheum 2010;62:2555]. Arthritis Rheum 2009; 60: 97686.
  • 22
    Inman RD, Davis JC Jr, Heijde D, Diekman L, Sieper J, Kim SI, et al. Efficacy and safety of golimumab in patients with ankylosing spondylitis: results of a randomized, double-blind, placebo-controlled, phase III trial. Arthritis Rheum 2008; 58: 340212.
  • 23
    Smolen JS, Kay J, Doyle MK, Landewe R, Matteson EL, Wollenhaupt J, et al, for the GO-AFTER study investigators. Golimumab in patients with active rheumatoid arthritis after treatment with tumour necrosis factor α inhibitors (GO-AFTER study): a multicentre, randomised, double-blind, placebo-controlled, phase III trial. Lancet 2009; 374: 21021.
  • 24
    QuantiFERON-TB Gold clinicians guide and QuantiFERON-TB Gold In-Tube Method [package insert]. Valencia (CA): Cellestis; 2010.
  • 25
    Centers for Disease Control and Prevention / American Thoracic Society. Targeted tuberculin testing and treatment of latent tuberculosis infection. MMWR 2000; 49: 151.
  • 26
    Centers for Disease Control and Prevention. Updated guidelines for using interferon gamma release assays to detect Mycobacterium tuberculosis infection—United States, 2010. MMWR 2010; 59: 125.
  • 27
    Richeldi L, Losi M, D'Amico R, Luppi M, Ferrari A, Mussini C, et al. Performance of tests for latent tuberculosis in different groups of immunocompromised patients. Chest 2009; 136: 198204.
  • 28
    Cobanoglu N, Ozcelik U, Kalyoncu U, Ozen S, Kiraz S, Gurcan N, et al. Interferon-gamma assays for the diagnosis of tuberculosis infection before using tumour necrosis factor-alpha blockers. Int J Tuberc Lung Dis 2007; 11: 117782.
  • 29
    Bocchino M, Matarese A, Bellofiore B, Giacomelli P, Santoro G, Balato N, 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: 90713.
  • 30
    Kim EY, Lim JE, Jung JY, Son JY, Lee KJ, Yoon YW, et al. Performance of the tuberculin skin test and interferon-γ release assay for detection of tuberculosis infection in immunocompromised patients in a BCG-vaccinated population. BMC Infect Dis 2009; 9: 207.
  • 31
    Jones S, de Gijsel D, Wallach FR, Gurtman AC, 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: 11905.
  • 32
    Luetkemeyer AF, Charlebois ED, Flores LL, Bangsberg DR, Deeks SG, Martin JN, 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: 73742.
  • 33
    Pai M, Zwerling A, Menzies D. Systematic review: T-cell-based assays for the diagnosis of latent tuberculosis infection: an update. Ann Intern Med 2008; 149: 17784.
  • 34
    Schoepfer AM, Flogerzi B, Fallegger S, Schaffer T, Mueller S, Nicod L, et al. Comparison of interferon-gamma release assay versus tuberculin skin test for tuberculosis screening in inflammatory bowel disease. Am J Gastroenterol 2008; 103: 2799806.
    Direct Link:
  • 35
    Brock I, Ruhwald M, Lundgren B, Westh H, Mathiesen LR, Ravn P. Latent tuberculosis in HIV positive, diagnosed by the M. tuberculosis specific interferon-γ test. Respir Res 2006; 7: 56.
  • 36
    Ferrara G, Losi M, D'Amico R, Roversi P, Piro R, Meacci M, et al. Use in routine clinical practice of two commercial blood tests for diagnosis of infection with Mycobacterium tuberculosis: a prospective study. Lancet 2006; 367: 132834.
  • 37
    Aichelburg MC, Rieger A, Breitenecker F, Pfistershammer K, Tittes J, Eltz S, et al. Detection and prediction of active tuberculosis disease by a whole-blood interferon-γ release assay in HIV-1-infected individuals. Clin Infect Dis 2009; 48: 95462.
  • 38
    Higuchi K, Harada N, Fukazawa K, Mori T. Relationship between whole-blood interferon-gamma responses and the risk of active tuberculosis. Tuberculosis (Edinb) 2008; 88: 2448.
  • 39
    Diel R, Loddenkemper R, Meywald-Walter K, Niemann S, Nienhaus A. Predictive value of a whole blood IFN-γ assay for the development of active tuberculosis disease after recent infection with Mycobacterium tuberculosis. Am J Respir Crit Care Med 2008; 177: 116470.
  • 40
    Westhovens R, Yocum D, Han J, Berman A, Strusberg I, Geusens P, et al, for the START Study Group. The safety of infliximab, combined with background treatments, among patients with rheumatoid arthritis and various comorbidities: a large, randomized, placebo-controlled trial. Arthritis Rheum 2006; 54: 107586.