SEARCH

SEARCH BY CITATION

Keywords:

  • tuberculous pleurisy;
  • QuantiFERON TB Gold In-Tube test;
  • nested-PCR

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

The conventional acid fast bacilli (AFB) smear and Mycobacterium tuberculosis (M.tb) culture of pleural effusion and tuberculin skin test (TST) in tuberculous pleurisy are unable to meet clinical needs because of their low sensitivities and specificities. To evaluate the diagnostic accuracies of QuantiFERON TB Gold In-Tube test (QFT-GIT) and nested-PCR in tuberculous pleurisy, we conducted a cross-sectional study in regions of China with a high tuberculosis (TB) epidemic. Seventy-eight participants were enrolled: 58 TB patients with diagnosis of confirmed or probable tuberculous pleurisy and 20 non-TB patients with a diagnosis of other non-TB diseases. The positive rates of AFB smear and M.tb culture in the pleural effusion were 5.8% (2/42) and 10.6% (5/47), respectively. The sensitivity and specificity of QFT-GIT were 93.1% (54/58) and 90.0% (18/20), whereas those of TST were 68.5% (37/54) and 86.7% (13/15), respectively; the sensitivity of QFT-GIT was significantly higher than TST (P = 0.013). The sensitivity and specificity of M.tb-specific nested-PCR in pleural effusion were 94.8% (55/58) and 90.0% (18/20), respectively, with a turnaround time of 7 h. Furthermore, combined QFT-GIT and nested-PCR detection improves the specificity to 100% with a sensitivity of up to 90.0%. This combination of immunoassay and molecular detection holds promise for the clinical diagnosis of tuberculous pleurisy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Tuberculous pleurisy is the most common extrapulmonary tuberculosis (TB), accounting for c. 10–20% of all tuberculous patients and c. 10–30% of disease causing pleural effusions (Porcel, 2009). The conventional acid fast bacilli (AFB) smear and Mycobacterium tuberculosis (M.tb) culture in pleural effusion are unable to meet clinical needs because of their low sensitivities (Light, 2007). There is an overriding need for the development of highly sensitive, specific and rapid tools to aid in the diagnosis of tuberculous pleurisy.

In recent years, the interferon-gamma (IFN-γ) release assay (IGRAs) has been considered a major breakthrough in TB diagnosis (Pai & Menzies, 2007; Stout & Menzies, 2008; Hardy et al., 2010; Rangaka et al., 2012). The QuantiFERON TB Gold In-Tube test (QFT-GIT) uses an ELISA to measure the amount of IFN-γ released in response to specific M.tb antigens compared with controls. The specific M.tb antigens are early secretory antigenic target-6 (ESAT-6), culture filtrate protein 10 (CFP-10) and TB 7.7, which are present in all M.tb and are able to stimulate the measurable release of IFN-γ in most infected persons, but which are absent from BCG vaccine strains and most nontuberculous mycobacteria (Andersen et al., 2000). Thus, as test antigens, these proteins offer improved test specificity compared with purified protein derivative (PPD). In August 2008, QFT-GIT became the second IGRA approved by the US Food and Drug Administration (FDA) as an aid for diagnosing M.tb infection (FDA, 2010). However, the usefulness of QFT-GIT in the diagnosis of tuberculous pleurisy in developing countries, especially in China and other regions with mandatory BCG-vaccinated coverage, remains unclear.

Research has shown that use of molecular biologic technology to detect M.tb-specific fragments in pleural effusion-specific fragments, could improve the diagnostic sensitivity and specificity for tuberculous pleurisy (Anie et al., 2007; Liu et al., 2007; Kumar et al., 2010). However, in previous studies, diverse methods with different primers were selected to detect M.tb in pleural fluid samples, demonstrating highly variable sensitivities (42.8–87.0%) and specificities (91–97%; Nagesh et al., 2001; Hasaneen et al., 2003; Chakravorty et al., 2005; Moon et al., 2005; Light, 2010).

To evaluate the diagnostic accuracies of QFT-GIT and nested-PCR in tuberculous pleurisy, we conducted a cross-sectional study in high TB epidemic regions of China. The aim was to provide evidence of the usefulness of QFT-GIT and nested-PCR in tuberculous pleurisy diagnosis in a BCG-vaccinated area and give clues as to the development of in-house M.tb-specific detection tools.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Participants

Seventy-eight patients with pleural effusion were enrolled consecutively in this cross-sectional study from 1 January 2011 to 31 October 2011 in Wuxi No. 5 People's Hospital. Confirmed tuberculous pleurisy was diagnosed with M.tb cultures positive in pleural effusion and/or confirmed TB infection by pleural biopsy. Probable tuberculous pleurisy was diagnosed using one of the following criteria: M.tb culture positive in sputum; M.tb culture positive in other biologic specimens; positive response to antituberculosis medication without other possible causes of pleural effusion (Moon et al., 2005). Twenty patients with pleural effusion who were diagnosed with diseases other than TB were also enrolled in this study as controls.

QFT-GIT test

The QFT-GIT was performed according to the manufacturer's instructions (QFT-GIT; Cellestis Ltd, Carnegie, Australia). Briefly, on the day of pleural effusion collection, a 3-mL venous blood sample was collected from each participant and aliquots placed into three tubes (TB-specific antigens, mitogen and nil tubes, respectively). The samples were incubated at 37 °C in a humidified 5% CO2 incubator for 24 h. On the second day, the tubes were centrifuged at 3000 rcf for 10 min and the plasma was collected and stored at 4 °C until IFN-γ assay was performed using ELISA. The optical density of each test was read using a 450-nm filter with a 620-nm reference filter with an ELISA plate reader.

The results were interpreted as positive, negative or indeterminate using QFT-GIT analysis software (QFT-GIT; Cellestis Ltd). If the IFN-γ secretion in response to TB antigen, after subtracting nil control IFN-γ, was ≥ 0.35 IU mL−1, it was considered positive for QFT-GIT; and if the value was < 0.35 IU mL−1, it was considered negative. If the negativity was associated with poor phytohaemagglutinin (PHA) response (i.e. IFN-γ secretion in response to mitogen was < 0.5 IU mL−1), it was considered as indeterminate or invalid result for QFT-GIT. The subjects with IFN-γ secretion > 8.0 IU mL−1 in the nil control samples were also considered indeterminate for QFT-GIT.

Tuberculin skin test

Immediately following blood collection from the right hand of each participant, 0.1 mL (2 T.U/0.1 mL) Tuberculin PPD RT23 (Statens Serum Institute, Copenhagen, Denmark) was administered intradermally in the middle third of the left forearm by an experienced nurse. The diameter induration transverse to the long axis of the forearm was measured between 48 and 72 h using a flexible plastic ruler. A diameter of skin induration ≥ 10 mm was considered positive for tuberculin skin test (TST).

Sample collections

In all, 40 mL pleural fluid was concentrated by centrifugation at 10 000 g at 4 °C for 20 min. Then the pellet was re-suspended in 1 mL sterile distilled water and stored at −20 °C for DNA extraction.

Three sputum specimens (spot-morning-spot) from each participant were collected and M.tb was detected with an AFB smear using the Ziehl–Neelsen method and mycobacterial culture in both Lowenstein Jensen (Biomerieux Inc., L'Etoile, France) and MGIT tubes (BD BACTC MGIT 960 system).

Nested PCR

The DNA from pleural fluid pellet suspension was extracted using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). Nested PCR was performed using the Seeplex® MTB Nested ACE Detection kit according to the manufacturer's instructions. This detection kit utilizes multi-target (IS6110 and MPB64) instead of single-target PCR for specific detection of M.tb. A mixture of bacterial clones and internal clones were used as positive controls. To eliminate any possibility of cross-contamination from the positive controls, the amplification sizes of the positive control PCR products (810 and 745 bp) were designed differently from those of the specimen PCR products (255 and 190 bp).

Data analysis

The statistical analysis was performed with graphpad prism software (version 5.01; GraphPad Software, Inc.) and medcalc Software (Version 11.4.2; MedCalc Software bvba). The data were compared using the non-parametric Mann–Whitney test and Chi-square test or Fisher's exact test. P < 0.05 was considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Clinical characteristics of enrolled participants

Based on the final diagnosis, 78 enrolled participants were divided into two groups: a TB group (n = 58) with a diagnosis of confirmed or probable tuberculous pleurisy, and a non-TB group (n = 20) with diagnosis of other non-TB diseases. In the TB group, patients with confirmed tuberculous pleurisy (n = 17) were culture-positive for M.tb of pleural fluid (n = 5) and/or histologically confirmed to have TB by pleural biopsy under the thoracoscope (n = 14). Patients with probable tuberculous pleurisy (n = 41) were sputum culture-positive for M.tb (n = 11), or positively responded to anti-TB medications without other possible causes of pleural effusion (n = 30). The median age of enrolled patients was 49 years old and 20 of the 78 were men (25.6%). The etiologies of non-TB pleural effusion included pulmonary adenocarcinoma (n = 6, five males, 47–89 years old), small-cell lung cancer (n = 1, female, 52 years old), pulmonary low differentiated squamous cell carcinoma (n = 1, male, 76 years old), mesothelioma of pleura (n = 1, female, 56 years old), bacterial pneumonia (n = 6, six males, 33–91 years old), liver cirrhosis (n = 1, female, 46 years old), rheumatoid honeycombing (n = 1, female, 57 years old), pulmonary lymphangioleiomyomatosis (LAM; n = 1, female, 25 years old) and non-TB pleural effusion of an undetermined origin (n = 2, one male, 34–46 years old; Table 1).

Table 1. The demographic and clinical characteristics of study participants
 TotalTB patientsNon-TB patients
  1. AFB, acid fast bacilli; BCG, bacillus Calmette-Guerin; TB, tuberculosis.

  2. a

    Including one case of liver cirrhosis, one of mesothelioma of pleura, one of rheumatoid honeycombing, one of LAM and two of nontuberculous pleural effusion of undetermined origin.

n 785820
Male/Female58/2045/1313/7
Age, median (range)49 (12–91)50 (12–85)54 (25–91)
Diagnosis, n
Malignancy88
Bacterial pneumonia66
Othersa66
BCG vaccination history, n (%)5642 (72.4)14 (70.0)
Previous TB treatment, n (%)129 (15.5)3 (15.0)
Sputum AFB smear or culture positive, n (%)1717/51 (33.3)0/18 (0)
M.tb detection in pleural effusion
AFB smear positive, n (%)22/42 (5.8)0/9 (0)
Culture positive, n (%)55/47 (10.6)0/10 (0)
Confirmed TB by pleural biopsy, n (%)1414/16 (87.5)0/4 (0)

Novel immunoassay test improved diagnostic power of tuberculous pleurisy

All 78 enrolled participants were tested with QFT-GIT and TST. The positive rates of QFT-GIT and TST in the TB group were 93.1% (54/58) and 68.5% (37/54) (P = 0.013), respectively, whereas the negative rates of QFT-GIT and TST in the non-TB group (n = 20) were 90.0% (18/20) and 86.7% (13/15), respectively (P = 1.000; Fig. 1). Furthermore, the IFN-γ secretions in response to PHA were comparable in two groups, whereas that in response to TB antigen in the TB group were significantly higher than in the non-TB group (P < 0.0001; Fig. 2). The receiver operating curve (ROC) analysis showed that the area under the ROC (AUC) of QFT-GIT and TST for TB diagnosis was 0.913 and 0.812, respectively (P = 0.152, Fig. 3). Thus, QFT-GIT was more sensitive and specific than TST for diagnosing TB.

image

Figure 1. Comparison of QFT-GIT and TST results of 78 recruited subjects. The black column represents positive results and the grey column negative results. The data on the columns represent the patient numbers. The positive rates of QFT-GIT and TST in the TB group were 93.1% (54/58) and 68.5% (37/54) (P = 0.013), respectively, and the negative rates of QFT-GIT and TST in the non-TB group (n = 20) were 90% (18/20) and 86.7% (13/15), respectively (P = 1.000).

Download figure to PowerPoint

image

Figure 2. IFN-γ levels of TB antigens stimulated (A) and mitogen-stimulated (M) results in TB (n = 58) and non-TB (n = 20) groups. The short transverse line represents median level. The dotted transverse line represents the cut-off value of QFT-GIT.

Download figure to PowerPoint

image

Figure 3. The ROC analysis of QFT-GIT and TST in TB vs. non-TB patients. The AUCs were 0.913 and 0.812, respectively (P = 0.152).

Download figure to PowerPoint

Molecular detection was rapid and sensitive in the diagnosis of tuberculous pleurisy

In addition, 78 samples of pleural fluid pellet suspension were amplified by nested-PCR for M.tb detection. Among 58 patients in the TB group, 55 (94.8%) were positive, whereas only two (10.0%) were positive among the 20 patients in the non-TB group; the sensitivity and specificity of nested-PCR were 94.8% and 90.0%, respectively. Compared with conventional AFB and M.tb culture, the specificity of nested-PCR was comparable with TST and QFT-GIT (90.0% vs. 86.7% and 90.0%, respectively), whereas the sensitivities of nested-PCR and QFT-GIT were comparable, and were much higher than TST, AFB and M.tb culture (Fig. 4). Moreover, the turnaround time was only 7 h for nested-PCR detection, whereas TST required 72 h and QFT-GIT 30 h.

image

Figure 4. The positive rates of AFB smear, M.tb culture, TST, QFT-GIT and nested-PCR tests in TB patients: 3.4%, 8.6%, 68.5%, 93.1% and 94.8%, respectively.

Download figure to PowerPoint

Combined immunoassay and molecular detection was highly specific in diagnosis of tuberculous pleurisy

We also compared the detection results of nested-PCR and QFT-GIT of the same patients and found that 52 (90.0%) were double-positive in the TB group and 16 (80.0%) were double-negative in the non-TB group. In the TB group, 3.0% of QFT-GIT were single-positive, and 5.0% of nested-PCR were single-positive and 2.0% double-negative. In contrast, in the non-TB group, 10.0% of QFT-GIT or nested-PCR were single-positive (Fig. 5). Importantly, in the non-TB group two nested-PCR positive patients who were QFT-GIT negative and two who were QFT-GIT positive were also nested-PCR negative. Thus, combined immunoassay and molecular detection would probably improve the detection accuracy. Detailed analysis showed that when both QFT-GIT and nested-PCR were positive, this increased the specificity to 100%, with the sensitivity up to 90.0% (Table 2). Thus, combined QFT-GIT and nested-PCR could improve the diagnosis of tuberculous pleurisy dramatically.

image

Figure 5. The results of QFT-GIT and nested-PCR detection of 78 subjects including TB patients (n = 58) and non-TB patients (n = 20).

Download figure to PowerPoint

Table 2. The diagnostic power of combined QFT-GIT and nested-PCR
 TB patientsNon-TB patients Sensitivity % (95% CI) Specificity % (95% CI) PPV % NPV %
  1. 95% CI, 95% confidence interval; NPV, negative predictive value; PPV, positive predictive value.

TST positive37268.5 (54.45–80.48)86.7 (59.54–98.34)94.943.3
QFT-GIT positive54293.1 (83.27–98.09)90.0 (68.30–98.77)96.481.8
Nested-PCR positive55294.8 (85.62–98.92)90.0 (68.30–98.77)96.585.7
Both QFT-GIT and nested-PCR positive52089.7 (78.83–96.11)100.0 (83.16–100.0)100.076.9

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Positive bacteriological examination is the gold standard for the diagnosis of TB. However, the immediate cause of the effusion is a delayed hypersensitivity response to mycobacterial antigens in the pleural space. For this reason, microbiological analyses were often negative and limited by the lengthy delay in obtaining results, and the rate of positive cultures for M.tb in pleural effusion is lower (1.7–24.5%; Edwards & Edwards, 1960; Light, 2011). In our study, the rate of culture positive for M.tb in pleural effusion is 10.6% (5/47), which is far from that required clinically. Diacon's study indicates that histopathological examination via thoracoscopy has an accuracy of almost 100% for the diagnosis of tuberculous pleurisy (Koegelenberg & Diacon, 2011). Sixteen of 58 patients in the TB group underwent thoracoscopy for biopsy of pleura, with the positive rate of 87.5%. Thus, thoracoscopy is highly sensitive and specific in diagnosis of tuberculous pleurisy. However, thoracoscopy is invasive procedure which is not suitable or available for all patients.

The TST has been used worldwide for more than a century as an aid in diagnosing TB infection but it is limited due to the cross-reaction with BCG vaccination, low sensitivity in immune-suppressed individuals, and inconvenience of administration. The advantages of QFT-GIT over the TST are that it requires only a single patient visit, results are available in 24 h, and the findings are not subject to reader bias. However, the data regarding QFT-GIT in the diagnosis of tuberculous pleurisy, especially in a BCG-vaccinated area, were limited (Diel et al., 2010; Zhang et al., 2010; Ates et al., 2011; Chung et al., 2011). In our study, the sensitivity and specificity of QFT-GIT were 93.1% and 90.0%, respectively, and the turnaround time was only 30 h. A previous study compared IGRA (T-SPOT.TB) performed with peripheral blood mononuclear cells (PBMCs) and pleural effusion mononuclear cells (PEMCs) in patients with suspect pleural effusion, and revealed higher sensitivities and specificities of the detection with PEMCs than PBMCs (95% vs. 90% and 76% vs. 67%, respectively; Losi et al., 2007). Therefore, QFT-GIT is more sensitive and rapid than conventional microbiological tests, and more specific than conventional TST for the diagnosis of tuberculous pleurisy.

Furthermore, we evaluated M.tb-specific nested-PCR to aid the diagnosis of tuberculous pleurisy. The sensitivity and specificity were 94.8% and 90.0%, respectively, both of which were comparable to QFT-GIT. The greatest concern with molecular biology techniques is false-positives due to cross-contamination during processing. To eliminate any possibility of cross-contamination from the positive controls, the Seeplex® MTB Nested ACE Detection kit used in this study designs the amplification sizes of the positive control PCR products differently from those of the specimen PCR products.

Two patients in the non-TB group were nested-PCR positive and QFT-GIT negative, which indicated that the combined immunoassay and molecular detection would improve the accuracy of diagnosis. The detailed analysis confirmed that both QFT-GIT and nested-PCR positive results increased the specificity to 100%, with the sensitivity of up to 90.0%.

In conclusion, QFT-GIT is more sensitive and rapid than conventional microbiological tests, and more specific than conventional TST in the diagnosis of tuberculous pleurisy. Thus, the combination of immunoassay and molecular detection holds promise in the clinical treatment of tuberculous pleurisy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

The present study was partly supported by the National Natural Science Foundation of China (30901277), the US–China Biomedical Collaborative Research Foundation (81161120426) and Wuxi Social Development Guiding Program (CSZ00N1229). All authors have stated that they do not have any conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References
  • Andersen P, Munk ME, Pollock JM & Doherty TM (2000) Specific immune-based diagnosis of tuberculosis. Lancet 356: 10991104.
  • Anie Y, Sumi S, Varghese P, Madhavi LG, Sathish M & Radhakrishnan VV (2007) Diagnostic approaches in patients with tuberculous pleural effusion. Diagn Microbiol Infect Dis 59: 389394.
  • Ates G, Yildiz T, Ortakoylu MG, Ozekinci T, Erturk B, Akyildiz L & Caglar E (2011) Adapted T cell interferon-gamma release assay for the diagnosis of pleural tuberculosis. Respiration 82: 351357.
  • Chakravorty S, Sen MK & Tyagi JS (2005) Diagnosis of extrapulmonary tuberculosis by smear, culture, and PCR using universal sample processing technology. J Clin Microbiol 43: 43574362.
  • Chung JH, Han CH, Kim CJ & Lee SM (2011) Clinical utility of QuantiFERON-TB GOLD In-Tube and tuberculin skin test in patients with tuberculous pleural effusions. Diagn Microbiol Infect Dis 71: 263266.
  • Diel R, Loddenkemper R & Nienhaus A (2010) Evidence-based comparison of commercial interferon-gamma release assays for detecting active TB: a metaanalysis. Chest 137: 952968.
  • Edwards PQ & Edwards LB (1960) Story of the tuberculin test from an epidemiologic viewpoint. Am Rev Respir Dis 81(1,Pt 2):147.
  • FDA (2010) QuantiFERON-TB Gold In-Tube – P010033/S011. Available at http://www.fda.gov/Medical Devices/products and Medical Procedures/Device Approvals and Clearances/PMAApprovals/ucm106548.htm.
  • Hardy AB, Varma R, Collyns T, Moffitt SJ, Mullarkey C & Watson JP (2010) Cost-effectiveness of the NICE guidelines for screening for latent tuberculosis infection: the QuantiFERON-TB Gold IGRA alone is more cost-effective for immigrants from high burden countries. Thorax 65: 178180.
  • Hasaneen NA, Zaki ME, Shalaby HM & El-Morsi AS (2003) Polymerase chain reaction of pleural biopsy is a rapid and sensitive method for the diagnosis of tuberculous pleural effusion. Chest 124: 21052111.
  • Koegelenberg CF & Diacon AH (2011) Pleural controversy: close needle pleural biopsy or thoracoscopy – which first? Respirology 16: 738746.
  • Kumar P, Sen MK, Chauhan DS, Katoch VM, Singh S & Prasad HK (2010) Assessment of the N-PCR assay in diagnosis of pleural tuberculosis: detection of M. tuberculosis in pleural fluid and sputum collected in tandem. PLoS ONE 5: e10220.
  • Light R (2007) Pleural Diseases. Tuberculous Pleural Effusions, 5th edn. pp. 211224. Lippincott, Williams and Wilkins, Baltimore.
  • Light RW (2010) Update on tuberculous pleural effusion. Respirology 15: 451458.
  • Light RW (2011) Pleural effusions. Med Clin North Am 95: 10551070.
  • Liu KT, Su WJ & Perng RP (2007) Clinical utility of polymerase chain reaction for diagnosis of smear-negative pleural tuberculosis. J Chin Med Assoc 70: 148151; discussion 146–147.
  • Losi M, Bossink A, Codecasa L et al. (2007) Use of a T-cell interferon-gamma release assay for the diagnosis of tuberculous pleurisy. Eur Respir J 30: 11731179.
  • Moon JW, Chang YS, Kim SK, Kim YS, Lee HM & Chang J (2005) The clinical utility of polymerase chain reaction for the diagnosis of pleural tuberculosis. Clin Infect Dis 41: 660666.
  • Nagesh BS, Sehgal S, Jindal SK & Arora SK (2001) Evaluation of polymerase chain reaction for detection of Mycobacterium tuberculosis in pleural fluid. Chest 119: 17371741.
  • Pai M & Menzies D (2007) The new IGRA and the old TST: making good use of disagreement. Am J Respir Crit Care Med 175: 529531.
  • Porcel JM (2009) Tuberculous pleural effusion. Lung 187: 263270.
  • Rangaka MX, Wilkinson KA, Glynn JR, Ling D, Menzies D, Mwansa-Kambafwile J, Fielding K, Wilkinson RJ & Pai M (2012) Predictive value of interferon-gamma release assays for incident active tuberculosis: a systematic review and meta-analysis. Lancet Infect Dis 12: 4555.
  • Stout JE & Menzies D (2008) Predicting tuberculosis: does the IGRA tell the tale? Am J Respir Crit Care Med 177: 10551057.
  • Zhang S, Shao L, Mo L et al. (2010) Evaluation of gamma interferon release assays using Mycobacterium tuberculosis antigens for diagnosis of latent and active tuberculosis in Mycobacterium bovis BCG-vaccinated populations. Clin Vaccine Immunol 17: 19851990.