Analysis of an Interferon-Gamma Release Assay for Monitoring the Efficacy of Anti-Tuberculosis Chemotherapy

Transcription

1 Jpn. J. Infect. Dis., 64, , 2011 Original Article Analysis of an Interferon-Gamma Release Assay for Monitoring the Efficacy of Anti-Tuberculosis Chemotherapy Kiyoko Takayanagi, Misako Aoki, Kumiko Aman, Satoshi Mitarai 1 *, Nobuyuki Harada 1, Kazue Higuchi 1, Masao Okumura, Takashi Yoshiyama, Hideo Ogata, and Toru Mori 1 Department of Respiratory Medicine, Double-Barred Cross Hospital, and 1 Department of Mycobacterium Reference and Research, Research Institute of Tuberculosis, Japan Anti-Tuberculosis Association, Tokyo , Japan (Received October 27, Accepted January 31, 2011) SUMMARY: The reported effect of anti-tuberculosis chemotherapy on interferon-gamma (IFN- ) release in response to specific Mycobacterium tuberculosis antigens has been inconsistent. The objective of this study was therefore to determine the effect of anti-tuberculosis chemotherapy on IFN- response to ESAT-6 and CFP-10. QuantiFERON -TB Gold (QFT-G) was performed, and the IFN- response to ESAT-6 and CFP-10 were measured, for 50 people with culture-confirmed tuberculosis prior to initiating treatment and periodically for up to 120 weeks following initiation of said treatment. IFN- responses and bacteriological response were compared. The average IFN- response to ESAT-6 and CFP-10 and the proportion of QFT-G results that were positive decreased during chemotherapy and for several weeks thereafter, reaching lows at weeks 48 to 56. Furthermore, these measures were lower at 48 weeks for those with bacteriological reversion prior to the second monthly visit than for those with slower reversion. Although it was shown that anti-tuberculosis treatment generally reduced the specific release of IFN-, the effect is so variable that it could be used as a monitor of progress of chemotherapy with great care and reservation. INTRODUCTION The tuberculin skin test (TST) has been used as virtually the only method for diagnosing latent tuberculosis (TB) infection (LTBI). However, this test uses purified protein derivative (PPD), which contains many proteins with Bacille Calmette-Guérin (BCG), as an antigen, thus meaning that the diagnostic specificity of TST in BCG-vaccinated individuals may be affected (1). Two different Mycobacterium tuberculosis-specific antigens, namely early secreted antigenic target 6 kda protein (ESAT-6) and 10 kda culture filtrate protein (CFP-10), have been identified recently. QuantiFERON -TB Gold (QFT-G) (Cellestis Ltd., Carnegie, Australia) determines the interferon-gamma (IFN- ) released by an infected person s T lymphocytes in vitro in response to ESAT-6 and CFP-10. Indeed, a clinical trial for the diagnosis of TB in Japan revealed that QFT-G has a high sensitivity (89.0%) among active TB infections and a high specificity (98.1%) in the BCG-vaccinated population (1). A recent meta-analysis study reported a more conservative level, with a pooled sensitivity of 78% (2). This system has nevertheless been widely recognized as a new tool for the diagnosis of LTBI with higher specificity than that of the TST (3). A similar system that uses the ELISPOT technique to enumerate T-lymphocytes specifically producing IFN- has also been developed (4). A new version of the QuantiFERON test, known as QuantiFERON Gold in Tube (QFT-GIT), which includes TB 7.7 as an additional *Corresponding author: Mailing address: Bacteriology Division, Department of Mycobacterium Reference and Research, Research Institute of Tuberculosis, Japan Anti-Tuberculosis Association, Matsuyama, Kiyose, Tokyo , Japan. Tel: , Fax: , mitarai@jata. or.jp TB-specific antigen, has been developed recently (5). These tests, which are collectively known as IFN- release assays (IGRAs), are assumed to detect the potential activity of effector T cells that are maintained by the mycobacterial load of an infected host (1). If this is indeed the case, they could therefore be used to monitor the progress of chemotherapy together with bacteriological findings. This study was conducted in order to test this hypothesis by observing the long-term change in IFN- responses during and after a course of chemotherapy in a group of active TB patients. MATERIALS AND METHODS Patients: The study site was the Double-Barred Cross Hospital in Tokyo. All patients with active TB who were admitted to this hospital from May through November 2004 were enrolled in the study. The patients were confirmed as having active pulmonary or extra-pulmonary TB by smear and culture examinations. One patient with pleurisy was diagnosed as TB pleuritis as a result of the anti-tb treatment response. Those patients whom doctors considered would be difficult to track over a period of 2.5 years, mostly homeless and foreign-born patients and those who were identified as having multidrug-resistant strains, were excluded from the study. The study protocol was approved by the Ethical Committee of the hospital. Informed consent for this study was obtained from all patients at enrolment. Anti-TB chemotherapy and follow-up observation: All patients were treated with the standardized regimen according to the guidelines of the Japanese Society for Tuberculosis (6), i.e., basically a 6-month regimen of isoniazid and rifampicin, with pyrazinamide during the first 2 months. Treatment was extended beyond 6 months if pyrazinamide could not be used or if a patient had any underlying problem such as diabetes, drug resistance, treatment interrup- 133

2 tion due to side effects, or history of relapse. Acid-fast bacilli (AFB) smear and culture examinations were performed every month to monitor the efficacy of treatment. If the culture was positive on any follow-up examination, anti-tb drug susceptibility testing (DST) was performed on the M. tuberculosis isolates, and any change of treatment regimen was determined based on the DST results. A Directly Observed Treatment (DOT) strategy was adopted during hospitalization, and good compliance was secured by homebased DOT at the outpatient clinic. The patients were placed under follow-up for 2 years after successful completion of the anti-tb treatment. QFT-G test: Blood samples were collected once before the start of the anti-tb treatment and once every month during the 6-month period of the chemotherapy. After completion of the treatment, blood samples were collected on every follow-up visit at 3- to 6-month intervals. The whole blood drawn was sent to the Research Institute of Tuberculosis (RIT) at room temperature (22 ± 5 C) for testing, and processed within 12 h. The manufacturer s instructions were followed for subsequent stages of the QFT-G test. The cut-off value was set as 0.35 IU/mL of IFN- response to either ESAT-6 or CFP-10 (whichever was higher) for positive and non-positive tests. According to the official standards in Japan, a non-positive test may be subdivided into negative and doubtful, and the latter was used to indicate an intermediate level of response where the measurements ranged from 0.10 to 0.35 IU/mL (7). Conversion of QFT-G was defined as a change from non-positive (<0.35 IU/mL) to positive (.35 IU/mL), and reversion is vise versa. To analyze the relationship between the effect of bacteriological treatment and immunological response, all cases were sub-categorized into rapid and slow reversion groups according to the speed of bacteriological reversion after the start of treatment. Rapid reverters were considered to be those whose mycobacterial culture became negative within 2 months, whereas slow reverters showed no conversion after 2 months. When the changes in measurements were analyzed according to time after the start of treatment, some cases underwent examinations that were not conducted at the scheduled times. In such cases, the results of the examination conducted at the time nearest to the defined time within 2 weeks difference were used. Furthermore, for the qualitative (i.e., positive/negative) QFT-G test results, a missing result between two time points with known results, both of which were further than 2 weeks from the defined time, was substituted with a result interpolated as follows: if the results at both time points were the same (either positive or negative), the results between those time points were considered to be the same; if the results at the time points differed, the change from positive to negative, or vice versa, was assumed to have happened at the mid-point. Data management and analysis: QFT-G results were entered into a Microsoft Excel spreadsheet then transferred to Stata 8.2 (Stata Corp., College Road, Tex., USA) for statistical analysis. Analyses consisted of Student s t test for differences in means (unpaired, Welch s method for uneven variances), the chi-squared test for the difference in proportions and for significance of linear trends of proportions, and the Mann-Whitney rank-sum test for the difference in distribution of rank order variables. The mean values of the IFN- response measurements (in IU/mL) were computed as geometric means. Logistic regression analysis was applied to examine the combined effects of related variables on QFT-G results after the treatment. Statistical significance was judged at a risk of 5%. Regression analysis was applied to test the difference in the slopes of the regression coefficients for variables from two groups (8). RESULTS The subjects were 39 men and 11 women, with an average age of 53.4 (range, 20 85) years. Only one subject had undergone anti-tb chemotherapy; all the others were primary treatment cases. A total of 46 patients had pulmonary disease only, and two patients also had pleurisy. Two had both pulmonary TB and pleurisy. As far as the chest X-ray findings were concerned, 10, 28, and 8 subjects had minimal, medium, and far-advanced pulmonary involvement, respectively. The amounts of bacteria discharged were classified as 2 + or 3 + by AFB smear microscopy in most patients (Table 1). Ten patients were found to have underlying diseases (8 diabetes mellitus, 1 Addison s disease, and 1 pemphigus). The latter two received oral steroid therapy (predonisolone 10 and 5 mg/day, respectively). Both these patients were QFT-G positive at the time of enrollment. No patient was positive for human immunodeficiency virus infection. One of the M. tuberculosis strains isolated was resistant to isoniazid, three were resistant to streptomycin, and one was resistant to both isoniazid and streptomycin. A treatment regimen of four drugs, including pyrazinamide in the first 2 months, was administered to 45 patients, and a treatment regimen comprising three drugs was administered to the remaining 5 patients. The average duration of treatment was 35 (range, 23 57) weeks. The proportion of patients with positive QFT-G test results at the time of admission was 76.0% (38/50). Of the remaining 12, 7 were categorized as doubtful, with an IFN- level of more than 0.10 but less than 0.35, and another 5 were negative. Of these 12 subjects, 7 (67%) exhibited positive tests after the start of treatment, but 5 remained nonpositive throughout the study. Two patients died during observation at 6 months, and another at 15 months. Two patients left the program by 24 weeks of observation (4%), 11 (22%) by 48 weeks, and 35 (70%) by 120 weeks, for a variety of reasons. Bacteriological reversion (from positive to negative) was confirmed in all these truncated cases, and no treatment failure was observed. Reversion was observed within 8 weeks of treatment in 27 (54%) cases, and thereafter in the remaining 23 (46%) cases. Table 2 presents the change in the QFT-G test in terms of positive and non-positive (doubtful/negative) categorization throughout the 120-week observation period for those subjects with an actual observation or with an interpolated surrogate result each time. After the start of chemotherapy, a clear statistically significant (chi-squared for trend = 20.23, P = ) decrease in the positive rate was observed until 48 weeks. After 48 weeks, a slight and irregular increase in the positive rate was observed until 120 weeks, although this variation failed to reach statistical significance (chi-squared = 0.851, P = ). Focusing on the status at 48 weeks, the transitions from the pretreatment test status are presented in Table 3. A total of 66% of subjects with a positive pretreatment test had become non-positive by 48 weeks, whereas only a small fraction of those with a non-positive pretreatment test be- 134

3 Table 1. Characteristics of patients at enrolment Chest X-ray findings 4) Drug susceptibility 6) No. Age Sex Sample 1) Diagnosis 2) Complication 3) Laterality Cavity Extent AFB 5) H R S E Z 1 28 M SP PTB b + ME 1+ S S S S S 2 60 M GA PTB b + ME S S S S S 3 60 M SP PTB b + ME 3+ S S S S S 4 32 F SP PL r NA NA NA NA NA 5 78 F SP PTB HT b ME 1+ S S S S S 6 58 M SP PTB DM r + ME 3+ S S S S S 7 62 F SP PTB DM r MI S S S S S 8 62 M SP PTB DM b + ME 2+ S S S S S 9 46 M SP PTB b + FA 2+ S S S S S F SP PTB l + ME 2+ S S S S S M SP PTB DM b + ME 3+ S S S S S M SP PTB b + ME 3+ S S S S S M SP PTB r + ME 2+ S S S S S M SP PTB DM, HT r ME 2+ S S S S S M SP PTB HT b + FA 3+ S S S S S M SP PTB b + ME 2+ S S S S S M SP PTB DM b + FA 3+ S S S S S M SP PTB Pemphigus r + ME 2 + NA NA NA NA NA M SP PTB DM l + MI 3 + S S S S S M SP PTB COPD r ME 3+ S S S S S M SP PTB b MI 3+ S S S S S M SP PTB b + FA 3+ S S S S S M SP PTB r + ME 2+ R S R S S M SP PTB b + ME 3+ S S S S S M SP PTB HT b + ME 3 + S S S S S M SP PTB b ME S S S S S M SP PTB + PL b + FA 2+ S S S S S F SP PTB b FA S S S S S M SP PTB CVA b MI S S S S S M SP PTB HT l + ME 2 + R S S S S F PE PL r S S S S S M SP PTB DM b + FA 3+ S S S S S M SP PTB b + ME 2+ S S S S S M SP PTB l MI 2+ S S S S S M SP PTB DM l + ME 3 + S S R S S M SP PTB b + FA 3+ S S S S S M SP PTB l ME 3+ S S S S S M SP PTB b ME 2+ S S S S S F SP PTB r MI 2+ S S S S S F SP PTB l + ME 2+ R S S S S M SP PTB Hemorrhoid b + ME 3 + S S S S S F SP PTB CVA r ME 2 + S S R S S F SP PTB Hyperlipidemia b + ME 2 + S S S S S M SP PTB + PL r + MI S S S S S M SP PTB Hepatitis C b + ME 3 + S S S S S F SP PTB Perkinsonism b FA 1+ S S S S S M SP PTB b + ME 3 + S S R S S M SP PTB Renal dysfunction b + FA 3 + R S S S S M SP PTB DM l MI 1+ S S S S S M SP PTB Addison s disease b ME 3+ S S S S S 1) : SP, sputum; GA, gastric aspirate; PE, pleaura effusion. 2) : PTB, pulmonary tuberculosis; PL, tuberculous pleuritis. 3) : HT, hypertension; DM, diabetes mellitus; COPD, chronic obstructive pulmonary disease; CVA, cerebro-vascular accident. 4) : b, bilateral; r, right; l, left; MI, minimal; ME, medium; FA, far-advanced. 5) : AFB, smear examination of acid-fast bacilli. 6) : H, isoniazid; R, rifampicin; s: streptomycin; e, ethambutol; z, pyrazinamide; s, susceptible; r, resistant; NA, data unavailable. came positive at 48 weeks. A similar trend was observed for the measured IFN- responses to ESAT-6 and CFP-10. Thus, the lowest geometric mean was seen at 48 weeks for CFP-10 and at 56 weeks 135

4 for ESAT-6. A comparison of the means at 48 weeks for both ESAT-6 and CFP-10, as well as the higher value of the ESAT-6/CFP-10 ratio, indicated that the decrease in the response to antigens during the 48 weeks of observation was statistically significant. This finding was further confirmed by a regression analysis of measurements with time (the downward slope of the regression line was significantly negative for both ESAT-6 and CFP-10) (Table 4). The same was true for the higher value of either CFP-10 or ESAT-6. Fig. 1 reveals a clear upward trend in the mean values of response, although this fluctuates somewhat due to the small number of observations. The slope for the period weeks was significantly upward (positive) for CFP-10, but was not significant for ESAT-6 or for the higher of either ESAT-6 or CFP-10. The relationship between positivity at 48 weeks and speed of bacteriological reversion after the start of treatment was analyzed (Table 5). Although not tabulated, the pretreatment QFT positivity of rapid reverters (78% or 21/27) and that of slow reverters (74% or 17/23) were comparable. Likewise, cases with rapid reversion of sputum culture (within 8 weeks after the start of treatment) tended to exhibit a higher proportion of non-positive tests at 48 weeks than those with Table 2. Changes in QFT-G positivity after the start of treatment Weeks after start of treatment Before Positive % 1) Doubtful Negative Not tested Total ) : Chi-test for slope between Before and 48 weeks, P = 0.00; between 48 and 120 weeks, P = Table 3. Transition during 48 weeks from pretreatment test QFT-G at 48 weeks Pretreatment Not tested Negative Doubtful Positive Total Positive (%) 11 (38) 8 (28) 10 (34) 29 (100) 9 Doubtful (%) 5 (83) 0 (0) 1 (17) 6 (100) 1 Negative (%) 1 (25) 2 (50) 1 (25) 4 (100) 1 Total (%) 17 (44) 10 (26) 12 (31) 39 (100) 11 Table 4. Statistical tests for trends of IFN- response (log-transformed) after the start of treatment ESAT-6 CFP-10 ESAT-6 or CFP-10 1) Mean (95% confidence interval) (IU/mL) Before (n = 50) (5.634, 0.113) (5.753, 0.027) (6.141, 0.139) After 48 weeks (n = 18) (5.582, 0.015) (3.432, 0.005) (6.423, 0.030) t test for comparison P = P = P = Regression on weeks of treatment during 0 48 weeks Slope P 2) Regression on weeks of treatment during weeks Slope P 2) ) : Higher value of either ESAT-6 or CFP-10. 2) : P (probability) of the slope (regression) coefficient being zero. 1 ESAT-6 CFP-10 Response value (IU/mL) Weeks after Start of Treatment Fig. 1. Trends of mean values of IFN- response after the start of treatment. Arrows at the 2, 6, and 9 month time points indicate the end of the intensive phase, the end of the standard course of treatment, and the end of treatment for most of those with extended treatment, respectively. 136

5 Table 5. QFT-G status at 48 weeks according to time of bacteriological reversion QFT-G at 48 weeks Reversion Not tested Negative Doubtful Positive Total Rapid (within 2 months) (%) 12 (55) 5 (28) 5 (23) 22 (100) 5 Slow (after 2 months) (%) 5 (29) 5 (29) 7 (41) 17 (100) 6 Total 17 (44) 10 (26) 12 (31) 39 (100) 11 Mann-Whitney test P = slower reversion, although the difference was not statistically significant. In terms of QFT-G reversion from an initially positive test, a similar difference was observed between rapid bacteriological reverters (71%) and slow bacteriological reverters (58%) at 48 weeks, although the difference was not statistically significant (chi-squared = 0.468, P = 0.484). No such difference was observed in the QFT-G reversion rate at earlier time points: at 12 weeks, QFT-G reversion rates were 24% for rapid reverters and 35% for slow reverters; at 24 weeks, they were 29% for rapid reverters and 33% for slow reverters. A comparison of the geometric means of the IFN- responses revealed that the mean value was significantly lower for rapid bacteriological reverters ( versus , P = 0.032) for ESAT-6, but not for CFP-10 ( versus , P = 0.293). A similar weekly comparison was performed for the IFN- measurement change between rapid and slow bacteriological reverters in terms of the regression slope from the start of treatment up until 48 weeks. The regression coefficient for ESAT-6 was for rapid reverters and for slow reverters, thereby suggesting that the IFN- responses decreased faster in those subjects exhibiting a quicker response to chemotherapy. However, this difference did not reach statistical significance (P for F test = 0.115). No difference in the regression coefficient was observed for CFP-10 ( versus , P = 0.437). Finally, the patient s age, sex, and pretreatment QFT-G test result (positive versus non-positive) were found to have no significant relationship with the QFT-G change by 48 weeks. Thus, a multivariate analysis with logistic regression using these factors as independent variables to predict the QFT-G status (positive/non-positive) at 48 weeks revealed no significant variable contributing to QFT-G reversion after the start of treatment. DISCUSSION Lalvani et al. (9) reported that the number of ESAT-6 peptide-specific IFN- -secreting T cells declined progressively during 20 weeks of anti-tb treatment based on their observation of five TB patients. These authors assumed that the killing of M. tuberculosis with chemotherapy reduced the secretion of antigens such as ESAT-6 and CFP-10, which in turn resulted in the contraction of the IFN- -secreting effector T cells and thus the decline of frequency of those cells, or reversion of the patients IGRAs test. Similar reports confirmed the decrease in IFN- response to M. tuberculosisspecific antigen(s) with the use of ELISPOT during treatment of active TB (10 12) and LTBI (13,14). At the same time, however, other studied reported contradictory findings whereby no significant change (15) or even an increase in IGRA response was obsreved during treatment (16 18). Furthermore, using the ELISPOT system, Wilkinson et al. reported a transient rise in the specific T-cell responses and a decline thereafter, during treatment of LTBI (19). Nicol et al. reported a similar behavior in pediatric TB patients (20). Dheda et al. (12) assumed that this inconsistency was due to the different antigens concerned (protein or peptides) and the incubation time or stimulation of blood cells with antigen. However, the experiments conducted by these authors to confirm their hypothesis failed to show a significant contribution of the different test procedures to the inconsistency of the results. The majority of such studies used the ELISPOT system, and very few, with the exception of our study on LTBI treatment given to QFT-G-positive nosocomial contacts (14) and a study of Pai et al. on LTBI treatment in health care workers (16), have used QFT-G thus far. In the former study, Higuchi et al. (14) found a significant reduction in the mean IFN- response values, accompanied a decrease in the positive rate after 6 months of treatment. No such change was observed in those contacts who declined treatment. The current study deals with a group of active, bacteriologically confirmed, and satisfactorily treated TB that was monitored using QFT-G frequently, and for a long period of time, both during and after the course of treatment in a relatively lowprevalence setting. This study found that the IFN- response decreased significantly during chemotherapy; the QFT-G positive rate also decreased. No significant transient rise of response similar to that reported by Wilkinson et al. occurred in the early period of treatment (19). The decrease in response continued until 48 weeks after the start of treatment, whereas most patients had completed chemotherapy by 36 weeks. After 48 weeks, the response remained at a constant level, or possibly exhibited a slight increase or recovery, until 120 weeks. These findings are consistent with the assumption that mycobacterial load is reflected in the IGRA results. In light of the long division time of M. tuberculosis, clearance of bacterial load could continue even after chemotherapy, thus resulting in a continued decrease in the response. The possible increase in response after completion of chemotherapy could indicate regrowth of surviving bacilli after termination of chemotherapy during control of the host, or a condition similar to LTBI. The division of patients into two groups on the basis of their time of bacteriological negative reversion (i.e., rapid reverters whose culture reverted to negative within 2 months and the remaining slow reverters) showed that the specific IFN- response tended to decline more quickly in the rapid reverters than in the slow reverters, although the difference was not statistically significant. A difference in the change of IGRAs during treatment according to the response to chemotherapy has been reported by Carrara et al. (11). Our results, although only suggestive, support this finding. The QFT-G guideline in Japan introduced the gray zone 137

6 (from 0.1 to <0.35 IU/mL) for diagnosing TB infection (7), and recommends to diagnose persons within the gray zone as infected if they are at high risk of TB infection, as described previously (21). Seven TB patients in this study were within the gray zone, thereby suggesting that the introduction of this gray zone for high risk groups could reduce the likelihood of such infections being overlooked. A similar idea was also introduced for the interpretation of TST results in order to increase the positive predictive value of TST in different risk populations (22). Other studies have proposed the gray zone idea to define conversion and reversion for QFT-GIT (23). The analyses of rapid and slow bacterial responses undertaken in this study demonstrate the change of IFN- response in QFT-G along with the bacteriological effect of chemotherapy in TB patients and after the completion of treatment, thereby presumably reflecting a change in mycobacterial load in the disease lesion. The contrast between this finding and several previous reports mainly based on ELISPOT (16 18) should be investigated further, including a comparison between ELISPOT and QFT (24,25) in terms of both their diagnostic performance for LTBI as well as such dynamic aspects as time course during treatment and any other clinicopathological characteristics. It is clear from these findings that medical interventions, mainly effective anti-tb treatment, can decrease the TBspecific IFN- response, although the extent of this change varies markedly from one patient to another. At the same time, this decrease may occur spontaneously, presumably due to an interaction between the bacilli and the host s immune mechanism in an early phase of infection (26) and in its late phase (27). Such acute resolution and waning of infection could also be relevant to the investigation into the dynamics of IFN- response in TB infection. Thus, the change in IFN- response after introduction of the anti-tb treatment may reflect the progress of the chemotherapy, but it may be difficult to use this change as a reliable monitor of the treatment. Conflict of interest None to declare. REFERENCES 1. Mori, T., Sakatani, M., Yamagishi, F., et al. (2004): Specific detection of tuberculosis infection: an interferon-gamma-based assay using new antigens. Am. J. Respir. Crit. Care Med., 170, Pai, M., Zwerling, A. and Menzies, D. (2008): T-cell-based assays for the diagnosis of latent tuberculosis infection: an update. Ann. Intern. Med., 149, Mazurek, G.H., Jereb, J., Lobue. P., et al. (2005): Division of Tuberculosis Elimination, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention. Guidelines for using the QuantiFERON-TB Gold test for detecting Mycobacterium tuberculosis infection, United States. Morbid. Mortal. Wkly. Rep., 54 (RR 15), Lalvani, A. (2004): Counting antigen-specific T cells: a new approach for monitoring response to tuberculosis treatment? Clin. Infect. Dis., 38, Harada, N., Higuchi, K., Yoshiyama, T., et al. (2008): Comparison of the sensitivity and specificity of two whole blood interferon-gamma assays for M. tuberculosis infection. J. Infect., 56, Recommendation of Treatment Committee (2002): Japanese Society for Tuberculosis. Revision of standards for treatment of tuberculosis. Kekkaku, 77, (in Japanese). 7. Committee for Prevention of Tuberculosis (2006): Japan Tuberculosis Society. Guidelines for the use of QuantiFERON TB 2G. Kekkaku, 81, (in Japanese). 8. Snedecor, G.W. and Cochran, W.G. (1967): Statistical Methods. 6th ed. Iowa Univ. Press, p Lalvani, A., Nagvenkar, P., Udwadia, Z., et al. (2001): Enumeration of T cells for RD1 encoded antigens suggests a high prevalence of latent Mycobacterium tuberculosis infection in healthy urban Indians. J. Infect. Dis., 183, Pathan, A.A., Wilkinson, K.A., Kleneman, P., et al. (2001): Direct ex vivo analysis of antigen specific IFN-gamma secreting CD4 T cells in Mycobacterium tuberculosis infected individuals: associations with clinical disease state and effect of treatment. J. Immunol., 167, Carrrara, S., Vincenti, D., Petrosillo, N., et al. (2004): T cell-based assay for monitoring the course of antituberculosis therapy. Clin. Infect. Dis., 38, Dheda, K., Pooran, A., Pai, M., et al. (2007): Interpretation of Mycobacterium tuberculosis antigen-specific IFN- release assays (T- SPOT.TB) and factors that may modulate test results. J. Infect., 55, Ewer, K., Millington, K.A., Deeks, J.J., et al. (2006): Dynamic antigenspecific T-cell responses after point-source exposure to Mycobacterium tuberculosis. Am. J. Respir. Crit. Care Med., 174, Higuchi, K., Harada, N. and Mori, T. (2008): Interferon- responses after isoniazid chemotherapy for latent tuberculosis. Respirology, 13, Pai, M., Joshi, R., Bandyopadhyay, M., et al. (2007): Sensitivity of a whole-blood interferon-gamma assay among patients with pulmonary tuberculosis and variations in T-cell responses during anti-tuberculosis treatment. Infection, 35, Pai, M., Joshi, R., Dogra, S., et al. (2006): Persistently elevated T cell interferon- responses after treatment for latent tuberculosis infection among health care workers in India: a preliminary report. J. Occup. Med. Toxicol., 1:7. doi: / Ferrand, R.A., Bothamley, G.H., Whelan, A., et al. (2005): Interferongamma responses to ESAT-6 in tuberculosis patients early into and after anti-tuberculosis treatment. Int. J. Tuberc. Lung Dis., 9, Ulrichs, T., Anding, P., Kaufmann, S.H.E., et al. (2000): Numbers of IFN-gamma-producing cells against ESAT-6 increase in tuberculosis patients during chemotherapy. Int. J. Tuberc. Lung Dis., 4, Wilkinson, K.A., Kon, O.M. and Newton, S.M. (2006): Effect of treatment of latent tuberculosis infection on the T cell response to Mycobacterium tuberculosis antigens. J. Infect. Dis., 193, Nicol, M.P., Pienaar, D., Wood, K., et al. (2005): Enzyme-linked immunospot assay responses to early secretory antigenic target 6, culture filtrate protein 10, and purified protein derivative among children with tuberculosis: implications for diagnosis and monitoring of therapy. Clin. Infect. Dis., 40, Harada, N., Higuchi, K., Sekiya, Y., et al. (2004): Basic characteristics of a novel diagnostic method (QuantiFERON TB 2G) for latent tuberculosis infection with the use of Mycobacterium tuberculosis-specific antigens, ESAT-6 and CFP-10. Kekkaku, 79, (article in Japanese with English summary). 22. Jensen, P.A., Lambert, L.A., Iademarco, M.F., et al. (2005): Centers for Disease, Control and Prevention. Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care settings, Morbid. Mortal. Wkly. Rep., 54 (RR 17), Pai, M., Joshi, R., Dogra, S., et al. (2009): T-cell assay conversions and reversions among household contacts of tuberculosis patients in rural India. Int. J. Tuberc. Lung Dis., 13, Menzies, D., Pai, M. and Comstock, G. (2007): Meta-analysis: new tests for the diagnosis of latent tuberculosis infection: areas of uncertainty and recommendations for research. Ann. Intern. Med., 146, Lee, J.Y., Choi, H.J., Park, I.N., et al. (2006): Comparison of two commercial interferon-gamma assays for diagnosing Mycobacterium tuberculosis infection. Eur. Respir. J., 28, Hill, P.C., Brookes, R.H., Fox, A., et al. (2007): Longitudinal assessment of an ELISPOT test for Mycobacterium tuberculosis infection. PLoS Med., 4, e Mori, T., Harada, N., Higuchi, K., et al. (2007): Waning of the specific interferon-gamma response after years of tuberculosis infection. Int. J. Tuberc. Lung Dis., 11,