- Email: [email protected]
In vitro QuantiFERON-TB gold antigen specific interleukin-1beta to diagnose TB among HIV-positive subjects
Tuberculosis 96 (2016) 27e30
Contents lists available at ScienceDirect
Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube
In vitro QuantiFERON-TB gold antigen specific interleukin-1beta to diagnose TB among HIV-positive subjects Maddineni Prabhavathi a, Basirudeen Syed Ahamed Kabeer b, Anbarasu Deenadayalan a, Alamelu Raja a, * a Department of Immunology, National Institute for Research in Tuberculosis (ICMR), No. 1, Mayor Sathyamoorthy Road, Chetpet, Chennai, 600 031, Tamil Nadu, India b Cardiovascular Division, Sidra Medical and Research Center, Doha, Qatar
a r t i c l e i n f o
s u m m a r y
Article history: Received 17 April 2015 Received in revised form 4 October 2015 Accepted 7 October 2015
Background: The recently introduced IFN-g release assay (IGRA) has been reported to improve the diagnosis of TB. However, IGRA has suboptimal sensitivity to diagnose TB among HIV co-infected subjects. Apart from IFN-g, the pro inflammatory cytokines such as Interleukin-1beta (IL-1b), Tumor necrosis factor-alpha (TNF-a), IL-2, IL-6, IL-8 and IL-12 are also play a major role in mycobacterial infections. This study aimed to analyze these cytokines for detecting active TB among HIV sero positive subjects. Materials and methods: We had prospectively enrolled 53 HIV positive subjects and 55 HIV-TB coinfected patients from India. IGRA was performed by using QuantiFERON TB-Gold In tube (QFT-GIT) method. TB antigen specific IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL-12 levels were evaluated by ELISA in plasma harvested from QFT-GIT tubes. Results and conclusion: The TB antigen specific IL-1b levels were significantly elevated in HIV-TB coinfected patients compared to HIV positive subjects (p ¼ 0.0004). The specificity of both IL-1b (50.94%) and QFT-GIT (52.83%) remained similar in HIV positive subjects (p ¼ 0.24). However, IL-1b had shown higher sensitivity (72.73%) than QFT-GIT (54.55%) to diagnose TB among HIV co-infected patients. Moreover, in culture test positive HIV-TB patients, antigen specific IL-1b exhibited sensitivity of 84.21%; whereas QFT-GIT exhibited only 57.89% sensitivity. Unlike IFN-g (the read out marker of QFT-GIT), antigen specific IL-1b levels were not influenced by low CD4 counts. The other cytokine levels were not significantly differ between the 2 groups.From this study we concluded that TB antigen specific IL-1b may be an additional biomarker for active TB diagnosis among HIV positive subjects. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Tuberculosis (TB) caused by Mycobacterium tuberculosis (M. tb), remains one of the deadliest infectious diseases of mankind. According to WHO-2014 global TB report, an estimated 9 million people developed TB and 1.5 million died from the disease in 2013 [1]. This situation is further worsened by emergence of drug resistant strains of M. tb [2], human immuno deficiency virus (HIV) co-infection [3] and latent TB infection (LTBI) [4]. Individuals with HIV infection are at increased risk of re-activation of LTBI; as well as
* Corresponding author. Tel.: þ91 (044) 2836 9682; fax: þ91 (044) 2836 2528. E-mail addresses: [email protected] (M. Prabhavathi), bkabeer@ sidra.org (B.S. Ahamed Kabeer), [email protected] (A. Deenadayalan), [email protected] (A. Raja). http://dx.doi.org/10.1016/j.tube.2015.10.005 1472-9792/© 2015 Elsevier Ltd. All rights reserved.
of rapid progression of a recently acquired tuberculous infection [5]. An estimated 1.1 million (13%) of the 9 million people who developed TB in 2013 were HIV-positive [1]. Delay in diagnosis and in initiation of appropriate treatment for TB lead to 45e85% deaths among HIV co-infected patients [6]. Thus early diagnosis and accurate treatment of TB are the key elements to control death rate of HIV co-infected patients. Due to non-specific clinical features, diagnosis of TB among HIV co-infected patients remains difficult. Decreased tuberculin reactivity, reduced sensitivity of acid fast staining and atypical radiographic presentations hinder the diagnosis of TB in HIV co-infected patients [7]. The recent exploration in diagnosis of TB infection is the development of Interferon-gamma (IFN-g) release assays (IGRAs). Although IGRAs are intended for LTBI and not active TB, there is concern about increasing use of IGRAs for active TB in highburden countries [8]. Studies from low- and middle-income
28
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
countries had demonstrated sensitivity of 69%e83% and specificity of 52%e61% for IGRAs in the diagnosis of active TB [9]. However, IGRA has suboptimal sensitivity (60%e70%) to detect active TB among the HIV co-infected patients, suggesting that ~1 in 3 HIV-TB co-infected patients will have negative IGRA results [9]. In HIV-TB co-infected patients, the reduced sensitivity of IGRA is often associated with indeterminate results [5]. In IGRA test, antigen specific T-cells secrete a plethora of cytokines and evaluation of these secreted cytokines other than IFN-g, may be a useful approach for diagnosis of active TB among HIV coinfected patients. Previously, we had measured 6 potential biomarkers, such as TB antigen specific interleukin-1beta (IL-1b), tumor necrosis factor-alpha (TNF-a), IL-2, IL-6 IL-8 and IL-12(p40) in 3 study groups such as TB patients (PTB), healthy household contacts (HHC) and healthy control subjects (HCS). Antigen-specific IL-1b and TNF-a levels were significantly higher in PTB than HHC and HCS. Moreover, antigen-specific IL-1b assay could differentiate between PTB and HHC; other cytokines levels did not differ among the study groups [10]. Thus the present study is extended to evaluate diagnostic ability of these 6 TB antigen specific cytokines in HIV and HIV-TB coinfected patients and to determine whether the diagnostic performance of IGRA could be improved by addition of these cytokines. 2. Materials and methods 2.1. Study subjects The present study was approved by Institutional Ethical Committee of National Institute for Research in Tuberculosis (NIRT), Chennai. All the study subjects were informed about the study procedure and written consent was obtained from all the volunteers. Individuals with previous history of TB, those who underwent TST in the past 16 months, those with silicosis, end stage renal disease and leukemia/lymphoma were excluded from this study. HIV positive subjects (N ¼ 53), were recruited from Government Hospital of Thoracic Medicine (GHTM), Chennai. All the subjects were apparently free of TB symptoms and did not have any close contact with TB patients. All were having normal chest X-ray (read by 2 clinicians) and they were negative for both sputum smear and culture test for TB; but sero positive for HIV. HIV-TB patients (N ¼ 55), were also recruited from GHTM, Chennai. This group included subjects who were having abnormal X-ray and at least one positive sputum smear and/or culture for TB, with sero positivity for HIV who had received <1 week ATT treatment. 2.2. HIV testing The HIV infection status was confirmed by 2 rapid tests (Retroquic Comb Aids-RS, Span Diagnostics, India and HIV TRI-DOT, J. Mitra & Co, India). When a serum was positive for both tests, it was considered as HIV positive. If a serum was positive for only one EIA, Western Blot was done as confirmatory test. 2.3. CD4 count The CD4 cell count was estimated in blood samples of all the study subjects by flow cytometry. As described previously [5], 100 ml of whole blood was labeled with saturating concentrations of anti CD3-FITC, anti CD4-PE and anti CD8-APC (BD Biosciences, CA, USA). The percentages of CD3, CD4 and CD8 cells among the total lymphocytes were obtained using Flowjo Software (Tree star, Inc., CA, USA). The absolute CD3, CD4 and CD8 counts were calculated by multiplying the percentage with the total lymphocyte count.
2.4. IGRA As described previously [10], IGRA was performed using QuantiFERON-TB Gold In-Tube (QFT-GIT) (Cellestis Ltd, Carnegie Victoria, Australia) kit as per manufacturer's instructions. The test results were interpreted using the software given by the manufacturer and the cut-off point for the diagnosis was determined as per the manufacturer's instructions. 2.5. Measurement of cytokines The levels of IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL-12p(40) in QFTGIT supernatants were measured by standard ELISA technique using commercially available BD opt-EIA Kit (BD Biosciences, Franklin Lakes, NJ, USA) as per the manufacturer's instructions. 2.6. Statistical analysis Statistical analysis was performed by using GraphPad Prism software version 5.0 (GraphPad software, CA, USA). The median value for each group was determined and compared using ManneWhitney U test. In all instances, a p < 0.05 was considered as significant. Receiver-operating-characteristic (ROC) curves were used to determine the cut-off points and discriminative ability was evaluated by the area under the ROC curve (AUC). Fisher exact test was used to compare the proportion of positivity between IL-1b and QFT-GIT assays. Spearman's rank correlation coefficient was used to measure the correlation between CD4þ counts and antigen specific cytokine levels. 3. Results and discussion Table 1 shows the demographic and baseline characteristics of the 2 study groups. Among 53 HIV and 55 HIV-TB patients, QFT-GIT had shown 7 and 14 indeterminate results respectively. In this study we had estimated the diagnostic accuracy of QFT-GIT with, and without the indeterminate results considered as negative [10e12].
Table 1 Demographic and baseline characteristics of the study groups.
No. of subjects (N) Age Sex Male Female HIV strain HIVeI HIVeI &II Smear test status Positive Negative Culture test (available for) Smear negative-culture positive, N Smear positive-culture positive, N Chest X-ray (available for) QFT-GIT Positive Negative Indeterminate CD4þ cell count <100 cells/ml 100 to 200 cells/ml >200 cells/ml % e Percentage.
HIV
HIV-TB
53 19e56
55 21e52
29 (55%) 24 (45%)
38 (69%) 17 (31%)
46 (87%) 7 (13%)
51 (93%) 4 (7%)
e e e e e e
42 13 19 13 6 31
25 (47.16%) 21 (39.62%) 7 (13%)
30 (54.54%) 11 (20%) 14 (25.45%)
12 (22.64%) 14 (26.41%) 27 (50.94%)
22 (40%) 17 (30.90%) 16 (22.09%)
(76.36%) (23.63%) (34.54%)
(56.36%)
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
The levels of 6 cytokines (i.e. IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL12) in HIV and HIV-TB study groups were measured in both unstimulated and TB antigen stimulated plasma samples generated from QFT-GIT tubes. The antigen specific cytokine levels were obtained by subtracting the cytokines levels of unstimulated from antigen stimulated levels of cytokines. The TB antigen specific IL-1b levels ranged from 3.92 to 2198 pg/ml (median 294.2 pg/ml) in HIV and 14.82e9375 pg/ml (median 778.3 pg/ml) in HIV-TB patients respectively, indicting that significantly elevated levels of IL-1b was associated with HIV-TB compared to HIV group (p ¼ 0.0004) (Figure 1). To determine the diagnostic performance of antigen specific IL-1b assay, ROC curve analysis was done by considering all the HIV positive subjects without active TB as controls and all the HIV-TB co-infected patients as diseased; a cut-off point of 300 pg/ml (AUC ¼ 0.69) was used for sensitivity and specificity calculations. Sensitivity of IL-1b assay was calculated by considering the HIV-TB patients with IL-1b 300 pg/ml as positive for the test. Of the 55 HIV-TB patients tested, 40 patients were positive for IL-1b test and conferred 72.73% (95% CI: 59.04 to 83.86) sensitivity. In case of QFT-GIT, it was positive in 30 out of 55 HIV-TB patients; by considering indeterminate results as negative, QFT-GIT showed 54.54% (95% CI: 41.52 to 66.98) sensitivity and by excluding indeterminate results, QFT-GIT exhibited 73% (95% CI: 57.93 to 84.44) sensitivity respectively.
29
Of the 53 HIV positive subjects tested, IL-1b assay was negative in 27 subjects and yielded 50.94% specificity (95% CI: 36.84 to 64.94). QFT-GIT assay was negative in 21 out of 53 HIV subjects; hence yielded specificity of 45% (95% CI: 32.15 to 59.82) (by excluding indeterminate results) and 52.83% (95% CI: 39.66 to 65.66) (by considering indeterminate results as negative) respectively. 3.1. Combination of tests We further assessed whether combination of QFT-GIT assay and antigen specific IL-1b assay will improve their respective sensitivities to diagnose TB among HIV co-infected patients. Sensitivity was calculated as number of subjects positive for either test among HIVTB patients and specificity was calculated as number of subjects negative for both the two tests among HIV subjects. Though the sensitivity of combinational analysis had increased to 87.27%, the specificity had reduced to 22.64%, upon combining QFT-GIT and antigen specific IL-1b assays. We had also analyzed the sensitivity, by combining the conventional method of diagnosis (i.e., sputum smears) either with antigen specific IL-1b assay or QFT-GIT assay. Combination of smear test þ IL-1b assay and smear test þ QFT-GIT assay exhibited 94.54% (52/55) and 92.72% (51/55) sensitivity respectively. The overall results were further scrutinized and analyzed based on culture positivity of HIV-TB patients. Of 19 culture test positive
Figure 1. Diagnostic performance of antigen specific IL-1b in HIV and HIV-TB groups. This figure depicts (a) antigen specific IL-1b levels are significantly higher in HIV-TB compared to HIV group, (b) the diagnostic ability of antigen specific IL-1b is evaluated using ROC curve analysis.
Figure 2. Correlation of antigen specific IL-1b with CD4 levels of HIV-TB patients. Each dot in the graph represents an individual and the straight line represents the trend of correlation between antigen specific IL-1b with a) HIV-TB patients with <100 cells/ml CD4þ counts, b) HIV-TB patients with 100e200 cells/ml CD4þ counts, c) HIV-TB patients with >200 cells/ml CD4þ counts. r represents Spearman's rank correlation coefficient and p represents significance of correlation.
30
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
HIV-TB patients, antigen specific IL-1b was positive in 16 subjects and showed positivity of 84.21% (95% CI: 61.60 to 95.32); whereas QFT-GIT was positive in 11 subjects and exhibited 57.89% positivity (95% CI: 0.3624 to 0.7689). Altogether this data indicated that though the specificity of both TB antigen specific IL-1b and IFN-g (measured in QFT-GIT) remains similar, IL-1b has superior sensitivity than IFN-g to diagnose TB among HIV co-infected patients. Interestingly, out of 14 HIV-TB patients who gave indeterminate results in QFT-GIT, 10 subjects were positive for IL-1b assay.
major conclusion can be arrived upon the diagnostic utility of these cytokines. Acknowledgments The authors wish to thank all the patients participated in this study. The authors also wish to thank all the patients and clinicians of GHTM, Chennai. The help rendered by Mr. Selvakumar and Miss. Divya Raju in this study is gratefully acknowledged. Maddineni Prabhavathi expresses her gratitude to CSIR, India for providing Senior Research Fellowship.
3.2. Influence of CD4þ counts on test performance Since IFN-g response is influenced by CD4 counts, we had analyzed whether IL-1b assay was also influenced by CD4 counts. For that, we had regrouped HIV-TB co-infected patients into 3 subgroups based on CD4 counts i.e. <100, 100e200 and >200 cells/ ml. Among the 3 subgroups, QFT-GIT had shown highest negative results in patients who are having CD4 count <100 cells/ml (11 out of 22) compared to other 2 subgroups. The negativity of QFT-GIT was due to indeterminate results where, indeterminate results were higher in patients who were having CD4 counts <100 cells/ml (7 out of 22), followed by 100e200 cells/ml (4 out of 17) and >200 cells/ml (3 out of 16) respectively. However, antigen specific IL-1b assay was not influenced by CD4 counts and it showed a positivity of 77.27%, 70.58% and 75% among the HIV-TB patients who were having CD4 counts <100, 100e200 and >200 cells/ml respectively. This regroup analysis has shown that unlike IFN-g, IL1b levels were not influenced by low CD4 counts (Figure 2). The remaining cytokine levels did not differ significantly between HIV and HIV-TB study groups (Supplementary Figure 1). 4. Conclusion In the present study we have screened 6 potential biomarkers; out of which IL-1b has exhibited promising results for diagnosing TB in HIV co-infected patients. Though the specificity of IL-1b has remained similar to that of QFT-GIT, antigen specific IL-1b has showed higher sensitivity than IFN-g (read out marker of QFT-GIT). Of the 14 HIV-TB patients who gave indeterminate results, 10 subjects are shown to be positive by IL-1b assay. Moreover, IL-1b assay is not influenced by low CD4 counts. Thus, we suggest IL-1b could act as an additional biomarker along with the existing markers for diagnosis of TB among HIV co-infected patients. However, addition of IL-1b to the existing QFT-GIT did not improve its diagnostic accuracy. Moreover, the marginal increase in sensitivity of IL-1b assay over IFN-g has to be further evaluated in large number of study groups. Further studies need to analyze how IL-1b levels altered during TB treatment in HIV-TB patients. Also, serum IL-1b should also be measured to assess their diagnostic value. As this study has been done for a stipulated time point with small sample size and certain M. tb specific antigens, the remaining antigen specific cytokines such as TNF-a, IL-2, IL-6, IL-8 and IL-12 did not show significant difference between the study groups, hence no
Funding:
None.
Competing interests: competing interests. Ethical approval:
The authors declare that they have no
Not required.
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.tube.2015.10.005. References [1] World Health Organization (WHO). Global tuberculosis control e surveillance, planning, financing. Geneva: WHO Report; 2014 (WHO/HTM/TB/2014.08). [2] Daley CL. Global scale-up of the programmatic management of multidrugresistant tuberculosis. Indian J Tuberc 2014 April;61(2):108e15. [3] Lawn SD, Zumla AI. Tuberculosis. Lancet 2011 July;378(9785):57e72. [4] 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 August;149(3):177e84. [5] Syed Ahamed Kabeer B, Sikhamani R, Swaminathan S, Perumal V, Paramasivam P, Raja A. Role of interferon gamma release assay in active TB diagnosis among HIV infected individuals. PLoS One 2009 May;4(5):e5718. [6] Barnes PF, Block AB, Davidson PT, Snider DEJ. Tuberculosis in patients with human immuno deficiency virus infection. N Engl J Med 1991 June;324(23): 1644e50. [7] Shafer RW, Edlin BR. Tuberculosis in patients infected with human immunodeficiency virus: perspective on the past decade. Clin Infect Dis 1996 April;22(4):683e704. [8] Jiang W, Shao L, Zhang Y, Zhang S, Meng C, Xu Y, et al. High-sensitive and rapid detection of Mycobacterium tuberculosis infection by IFN-gamma release assay among HIV-infected individuals in BCG-vaccinated area. BMC Immunol 2009 May;10:31. [9] Metcalfe JZ, Everett CK, Steingart KR, Cattamanchi A, Huang L, Hopewell PC, Pai M. Interferon-g release assays for active pulmonary tuberculosis diagnosis in adults in low- and middle-income countries: systematic review and metaanalysis. J Infect Dis 2011 Nov;204(Suppl. 4):S1120e9. [10] Prabhavathi M, Kabeer BS, Deenadayalan A, Raja A. Role of QuantiFERON-TB gold antigen-specific IL-1b in diagnosis of active tuberculosis. Med Microbiol Immunol 2014 Dec. http://dx.doi.org/10.1007/s00430-014-0382-x. [11] Raby E, Moyo M, Devendra A, Banda J, De Haas P, Ayles H, et al. The effects of HIV on the sensitivity of a whole blood IFN-gamma release assay in Zambian adults with active tuberculosis. PLoS One 2008 Jun;3(6):e2489. [12] Portevin D, Moukambi F, Clowes P, Bauer A, Chachage M, Ntinginya NE, Mfinanga E, et al. Assessment of the novel T-cell activation markertuberculosis assay for diagnosis of active tuberculosis in children: a prospective proof-of-concept study. Lancet Infect Dis 2014 Oct;14(10):931e8.
Contents lists available at ScienceDirect
Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube
In vitro QuantiFERON-TB gold antigen specific interleukin-1beta to diagnose TB among HIV-positive subjects Maddineni Prabhavathi a, Basirudeen Syed Ahamed Kabeer b, Anbarasu Deenadayalan a, Alamelu Raja a, * a Department of Immunology, National Institute for Research in Tuberculosis (ICMR), No. 1, Mayor Sathyamoorthy Road, Chetpet, Chennai, 600 031, Tamil Nadu, India b Cardiovascular Division, Sidra Medical and Research Center, Doha, Qatar
a r t i c l e i n f o
s u m m a r y
Article history: Received 17 April 2015 Received in revised form 4 October 2015 Accepted 7 October 2015
Background: The recently introduced IFN-g release assay (IGRA) has been reported to improve the diagnosis of TB. However, IGRA has suboptimal sensitivity to diagnose TB among HIV co-infected subjects. Apart from IFN-g, the pro inflammatory cytokines such as Interleukin-1beta (IL-1b), Tumor necrosis factor-alpha (TNF-a), IL-2, IL-6, IL-8 and IL-12 are also play a major role in mycobacterial infections. This study aimed to analyze these cytokines for detecting active TB among HIV sero positive subjects. Materials and methods: We had prospectively enrolled 53 HIV positive subjects and 55 HIV-TB coinfected patients from India. IGRA was performed by using QuantiFERON TB-Gold In tube (QFT-GIT) method. TB antigen specific IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL-12 levels were evaluated by ELISA in plasma harvested from QFT-GIT tubes. Results and conclusion: The TB antigen specific IL-1b levels were significantly elevated in HIV-TB coinfected patients compared to HIV positive subjects (p ¼ 0.0004). The specificity of both IL-1b (50.94%) and QFT-GIT (52.83%) remained similar in HIV positive subjects (p ¼ 0.24). However, IL-1b had shown higher sensitivity (72.73%) than QFT-GIT (54.55%) to diagnose TB among HIV co-infected patients. Moreover, in culture test positive HIV-TB patients, antigen specific IL-1b exhibited sensitivity of 84.21%; whereas QFT-GIT exhibited only 57.89% sensitivity. Unlike IFN-g (the read out marker of QFT-GIT), antigen specific IL-1b levels were not influenced by low CD4 counts. The other cytokine levels were not significantly differ between the 2 groups.From this study we concluded that TB antigen specific IL-1b may be an additional biomarker for active TB diagnosis among HIV positive subjects. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Tuberculosis (TB) caused by Mycobacterium tuberculosis (M. tb), remains one of the deadliest infectious diseases of mankind. According to WHO-2014 global TB report, an estimated 9 million people developed TB and 1.5 million died from the disease in 2013 [1]. This situation is further worsened by emergence of drug resistant strains of M. tb [2], human immuno deficiency virus (HIV) co-infection [3] and latent TB infection (LTBI) [4]. Individuals with HIV infection are at increased risk of re-activation of LTBI; as well as
* Corresponding author. Tel.: þ91 (044) 2836 9682; fax: þ91 (044) 2836 2528. E-mail addresses: [email protected] (M. Prabhavathi), bkabeer@ sidra.org (B.S. Ahamed Kabeer), [email protected] (A. Deenadayalan), [email protected] (A. Raja). http://dx.doi.org/10.1016/j.tube.2015.10.005 1472-9792/© 2015 Elsevier Ltd. All rights reserved.
of rapid progression of a recently acquired tuberculous infection [5]. An estimated 1.1 million (13%) of the 9 million people who developed TB in 2013 were HIV-positive [1]. Delay in diagnosis and in initiation of appropriate treatment for TB lead to 45e85% deaths among HIV co-infected patients [6]. Thus early diagnosis and accurate treatment of TB are the key elements to control death rate of HIV co-infected patients. Due to non-specific clinical features, diagnosis of TB among HIV co-infected patients remains difficult. Decreased tuberculin reactivity, reduced sensitivity of acid fast staining and atypical radiographic presentations hinder the diagnosis of TB in HIV co-infected patients [7]. The recent exploration in diagnosis of TB infection is the development of Interferon-gamma (IFN-g) release assays (IGRAs). Although IGRAs are intended for LTBI and not active TB, there is concern about increasing use of IGRAs for active TB in highburden countries [8]. Studies from low- and middle-income
28
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
countries had demonstrated sensitivity of 69%e83% and specificity of 52%e61% for IGRAs in the diagnosis of active TB [9]. However, IGRA has suboptimal sensitivity (60%e70%) to detect active TB among the HIV co-infected patients, suggesting that ~1 in 3 HIV-TB co-infected patients will have negative IGRA results [9]. In HIV-TB co-infected patients, the reduced sensitivity of IGRA is often associated with indeterminate results [5]. In IGRA test, antigen specific T-cells secrete a plethora of cytokines and evaluation of these secreted cytokines other than IFN-g, may be a useful approach for diagnosis of active TB among HIV coinfected patients. Previously, we had measured 6 potential biomarkers, such as TB antigen specific interleukin-1beta (IL-1b), tumor necrosis factor-alpha (TNF-a), IL-2, IL-6 IL-8 and IL-12(p40) in 3 study groups such as TB patients (PTB), healthy household contacts (HHC) and healthy control subjects (HCS). Antigen-specific IL-1b and TNF-a levels were significantly higher in PTB than HHC and HCS. Moreover, antigen-specific IL-1b assay could differentiate between PTB and HHC; other cytokines levels did not differ among the study groups [10]. Thus the present study is extended to evaluate diagnostic ability of these 6 TB antigen specific cytokines in HIV and HIV-TB coinfected patients and to determine whether the diagnostic performance of IGRA could be improved by addition of these cytokines. 2. Materials and methods 2.1. Study subjects The present study was approved by Institutional Ethical Committee of National Institute for Research in Tuberculosis (NIRT), Chennai. All the study subjects were informed about the study procedure and written consent was obtained from all the volunteers. Individuals with previous history of TB, those who underwent TST in the past 16 months, those with silicosis, end stage renal disease and leukemia/lymphoma were excluded from this study. HIV positive subjects (N ¼ 53), were recruited from Government Hospital of Thoracic Medicine (GHTM), Chennai. All the subjects were apparently free of TB symptoms and did not have any close contact with TB patients. All were having normal chest X-ray (read by 2 clinicians) and they were negative for both sputum smear and culture test for TB; but sero positive for HIV. HIV-TB patients (N ¼ 55), were also recruited from GHTM, Chennai. This group included subjects who were having abnormal X-ray and at least one positive sputum smear and/or culture for TB, with sero positivity for HIV who had received <1 week ATT treatment. 2.2. HIV testing The HIV infection status was confirmed by 2 rapid tests (Retroquic Comb Aids-RS, Span Diagnostics, India and HIV TRI-DOT, J. Mitra & Co, India). When a serum was positive for both tests, it was considered as HIV positive. If a serum was positive for only one EIA, Western Blot was done as confirmatory test. 2.3. CD4 count The CD4 cell count was estimated in blood samples of all the study subjects by flow cytometry. As described previously [5], 100 ml of whole blood was labeled with saturating concentrations of anti CD3-FITC, anti CD4-PE and anti CD8-APC (BD Biosciences, CA, USA). The percentages of CD3, CD4 and CD8 cells among the total lymphocytes were obtained using Flowjo Software (Tree star, Inc., CA, USA). The absolute CD3, CD4 and CD8 counts were calculated by multiplying the percentage with the total lymphocyte count.
2.4. IGRA As described previously [10], IGRA was performed using QuantiFERON-TB Gold In-Tube (QFT-GIT) (Cellestis Ltd, Carnegie Victoria, Australia) kit as per manufacturer's instructions. The test results were interpreted using the software given by the manufacturer and the cut-off point for the diagnosis was determined as per the manufacturer's instructions. 2.5. Measurement of cytokines The levels of IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL-12p(40) in QFTGIT supernatants were measured by standard ELISA technique using commercially available BD opt-EIA Kit (BD Biosciences, Franklin Lakes, NJ, USA) as per the manufacturer's instructions. 2.6. Statistical analysis Statistical analysis was performed by using GraphPad Prism software version 5.0 (GraphPad software, CA, USA). The median value for each group was determined and compared using ManneWhitney U test. In all instances, a p < 0.05 was considered as significant. Receiver-operating-characteristic (ROC) curves were used to determine the cut-off points and discriminative ability was evaluated by the area under the ROC curve (AUC). Fisher exact test was used to compare the proportion of positivity between IL-1b and QFT-GIT assays. Spearman's rank correlation coefficient was used to measure the correlation between CD4þ counts and antigen specific cytokine levels. 3. Results and discussion Table 1 shows the demographic and baseline characteristics of the 2 study groups. Among 53 HIV and 55 HIV-TB patients, QFT-GIT had shown 7 and 14 indeterminate results respectively. In this study we had estimated the diagnostic accuracy of QFT-GIT with, and without the indeterminate results considered as negative [10e12].
Table 1 Demographic and baseline characteristics of the study groups.
No. of subjects (N) Age Sex Male Female HIV strain HIVeI HIVeI &II Smear test status Positive Negative Culture test (available for) Smear negative-culture positive, N Smear positive-culture positive, N Chest X-ray (available for) QFT-GIT Positive Negative Indeterminate CD4þ cell count <100 cells/ml 100 to 200 cells/ml >200 cells/ml % e Percentage.
HIV
HIV-TB
53 19e56
55 21e52
29 (55%) 24 (45%)
38 (69%) 17 (31%)
46 (87%) 7 (13%)
51 (93%) 4 (7%)
e e e e e e
42 13 19 13 6 31
25 (47.16%) 21 (39.62%) 7 (13%)
30 (54.54%) 11 (20%) 14 (25.45%)
12 (22.64%) 14 (26.41%) 27 (50.94%)
22 (40%) 17 (30.90%) 16 (22.09%)
(76.36%) (23.63%) (34.54%)
(56.36%)
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
The levels of 6 cytokines (i.e. IL-1b, TNF-a, IL-2, IL-6, IL-8 and IL12) in HIV and HIV-TB study groups were measured in both unstimulated and TB antigen stimulated plasma samples generated from QFT-GIT tubes. The antigen specific cytokine levels were obtained by subtracting the cytokines levels of unstimulated from antigen stimulated levels of cytokines. The TB antigen specific IL-1b levels ranged from 3.92 to 2198 pg/ml (median 294.2 pg/ml) in HIV and 14.82e9375 pg/ml (median 778.3 pg/ml) in HIV-TB patients respectively, indicting that significantly elevated levels of IL-1b was associated with HIV-TB compared to HIV group (p ¼ 0.0004) (Figure 1). To determine the diagnostic performance of antigen specific IL-1b assay, ROC curve analysis was done by considering all the HIV positive subjects without active TB as controls and all the HIV-TB co-infected patients as diseased; a cut-off point of 300 pg/ml (AUC ¼ 0.69) was used for sensitivity and specificity calculations. Sensitivity of IL-1b assay was calculated by considering the HIV-TB patients with IL-1b 300 pg/ml as positive for the test. Of the 55 HIV-TB patients tested, 40 patients were positive for IL-1b test and conferred 72.73% (95% CI: 59.04 to 83.86) sensitivity. In case of QFT-GIT, it was positive in 30 out of 55 HIV-TB patients; by considering indeterminate results as negative, QFT-GIT showed 54.54% (95% CI: 41.52 to 66.98) sensitivity and by excluding indeterminate results, QFT-GIT exhibited 73% (95% CI: 57.93 to 84.44) sensitivity respectively.
29
Of the 53 HIV positive subjects tested, IL-1b assay was negative in 27 subjects and yielded 50.94% specificity (95% CI: 36.84 to 64.94). QFT-GIT assay was negative in 21 out of 53 HIV subjects; hence yielded specificity of 45% (95% CI: 32.15 to 59.82) (by excluding indeterminate results) and 52.83% (95% CI: 39.66 to 65.66) (by considering indeterminate results as negative) respectively. 3.1. Combination of tests We further assessed whether combination of QFT-GIT assay and antigen specific IL-1b assay will improve their respective sensitivities to diagnose TB among HIV co-infected patients. Sensitivity was calculated as number of subjects positive for either test among HIVTB patients and specificity was calculated as number of subjects negative for both the two tests among HIV subjects. Though the sensitivity of combinational analysis had increased to 87.27%, the specificity had reduced to 22.64%, upon combining QFT-GIT and antigen specific IL-1b assays. We had also analyzed the sensitivity, by combining the conventional method of diagnosis (i.e., sputum smears) either with antigen specific IL-1b assay or QFT-GIT assay. Combination of smear test þ IL-1b assay and smear test þ QFT-GIT assay exhibited 94.54% (52/55) and 92.72% (51/55) sensitivity respectively. The overall results were further scrutinized and analyzed based on culture positivity of HIV-TB patients. Of 19 culture test positive
Figure 1. Diagnostic performance of antigen specific IL-1b in HIV and HIV-TB groups. This figure depicts (a) antigen specific IL-1b levels are significantly higher in HIV-TB compared to HIV group, (b) the diagnostic ability of antigen specific IL-1b is evaluated using ROC curve analysis.
Figure 2. Correlation of antigen specific IL-1b with CD4 levels of HIV-TB patients. Each dot in the graph represents an individual and the straight line represents the trend of correlation between antigen specific IL-1b with a) HIV-TB patients with <100 cells/ml CD4þ counts, b) HIV-TB patients with 100e200 cells/ml CD4þ counts, c) HIV-TB patients with >200 cells/ml CD4þ counts. r represents Spearman's rank correlation coefficient and p represents significance of correlation.
30
M. Prabhavathi et al. / Tuberculosis 96 (2016) 27e30
HIV-TB patients, antigen specific IL-1b was positive in 16 subjects and showed positivity of 84.21% (95% CI: 61.60 to 95.32); whereas QFT-GIT was positive in 11 subjects and exhibited 57.89% positivity (95% CI: 0.3624 to 0.7689). Altogether this data indicated that though the specificity of both TB antigen specific IL-1b and IFN-g (measured in QFT-GIT) remains similar, IL-1b has superior sensitivity than IFN-g to diagnose TB among HIV co-infected patients. Interestingly, out of 14 HIV-TB patients who gave indeterminate results in QFT-GIT, 10 subjects were positive for IL-1b assay.
major conclusion can be arrived upon the diagnostic utility of these cytokines. Acknowledgments The authors wish to thank all the patients participated in this study. The authors also wish to thank all the patients and clinicians of GHTM, Chennai. The help rendered by Mr. Selvakumar and Miss. Divya Raju in this study is gratefully acknowledged. Maddineni Prabhavathi expresses her gratitude to CSIR, India for providing Senior Research Fellowship.
3.2. Influence of CD4þ counts on test performance Since IFN-g response is influenced by CD4 counts, we had analyzed whether IL-1b assay was also influenced by CD4 counts. For that, we had regrouped HIV-TB co-infected patients into 3 subgroups based on CD4 counts i.e. <100, 100e200 and >200 cells/ ml. Among the 3 subgroups, QFT-GIT had shown highest negative results in patients who are having CD4 count <100 cells/ml (11 out of 22) compared to other 2 subgroups. The negativity of QFT-GIT was due to indeterminate results where, indeterminate results were higher in patients who were having CD4 counts <100 cells/ml (7 out of 22), followed by 100e200 cells/ml (4 out of 17) and >200 cells/ml (3 out of 16) respectively. However, antigen specific IL-1b assay was not influenced by CD4 counts and it showed a positivity of 77.27%, 70.58% and 75% among the HIV-TB patients who were having CD4 counts <100, 100e200 and >200 cells/ml respectively. This regroup analysis has shown that unlike IFN-g, IL1b levels were not influenced by low CD4 counts (Figure 2). The remaining cytokine levels did not differ significantly between HIV and HIV-TB study groups (Supplementary Figure 1). 4. Conclusion In the present study we have screened 6 potential biomarkers; out of which IL-1b has exhibited promising results for diagnosing TB in HIV co-infected patients. Though the specificity of IL-1b has remained similar to that of QFT-GIT, antigen specific IL-1b has showed higher sensitivity than IFN-g (read out marker of QFT-GIT). Of the 14 HIV-TB patients who gave indeterminate results, 10 subjects are shown to be positive by IL-1b assay. Moreover, IL-1b assay is not influenced by low CD4 counts. Thus, we suggest IL-1b could act as an additional biomarker along with the existing markers for diagnosis of TB among HIV co-infected patients. However, addition of IL-1b to the existing QFT-GIT did not improve its diagnostic accuracy. Moreover, the marginal increase in sensitivity of IL-1b assay over IFN-g has to be further evaluated in large number of study groups. Further studies need to analyze how IL-1b levels altered during TB treatment in HIV-TB patients. Also, serum IL-1b should also be measured to assess their diagnostic value. As this study has been done for a stipulated time point with small sample size and certain M. tb specific antigens, the remaining antigen specific cytokines such as TNF-a, IL-2, IL-6, IL-8 and IL-12 did not show significant difference between the study groups, hence no
Funding:
None.
Competing interests: competing interests. Ethical approval:
The authors declare that they have no
Not required.
Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.tube.2015.10.005. References [1] World Health Organization (WHO). Global tuberculosis control e surveillance, planning, financing. Geneva: WHO Report; 2014 (WHO/HTM/TB/2014.08). [2] Daley CL. Global scale-up of the programmatic management of multidrugresistant tuberculosis. Indian J Tuberc 2014 April;61(2):108e15. [3] Lawn SD, Zumla AI. Tuberculosis. Lancet 2011 July;378(9785):57e72. [4] 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 August;149(3):177e84. [5] Syed Ahamed Kabeer B, Sikhamani R, Swaminathan S, Perumal V, Paramasivam P, Raja A. Role of interferon gamma release assay in active TB diagnosis among HIV infected individuals. PLoS One 2009 May;4(5):e5718. [6] Barnes PF, Block AB, Davidson PT, Snider DEJ. Tuberculosis in patients with human immuno deficiency virus infection. N Engl J Med 1991 June;324(23): 1644e50. [7] Shafer RW, Edlin BR. Tuberculosis in patients infected with human immunodeficiency virus: perspective on the past decade. Clin Infect Dis 1996 April;22(4):683e704. [8] Jiang W, Shao L, Zhang Y, Zhang S, Meng C, Xu Y, et al. High-sensitive and rapid detection of Mycobacterium tuberculosis infection by IFN-gamma release assay among HIV-infected individuals in BCG-vaccinated area. BMC Immunol 2009 May;10:31. [9] Metcalfe JZ, Everett CK, Steingart KR, Cattamanchi A, Huang L, Hopewell PC, Pai M. Interferon-g release assays for active pulmonary tuberculosis diagnosis in adults in low- and middle-income countries: systematic review and metaanalysis. J Infect Dis 2011 Nov;204(Suppl. 4):S1120e9. [10] Prabhavathi M, Kabeer BS, Deenadayalan A, Raja A. Role of QuantiFERON-TB gold antigen-specific IL-1b in diagnosis of active tuberculosis. Med Microbiol Immunol 2014 Dec. http://dx.doi.org/10.1007/s00430-014-0382-x. [11] Raby E, Moyo M, Devendra A, Banda J, De Haas P, Ayles H, et al. The effects of HIV on the sensitivity of a whole blood IFN-gamma release assay in Zambian adults with active tuberculosis. PLoS One 2008 Jun;3(6):e2489. [12] Portevin D, Moukambi F, Clowes P, Bauer A, Chachage M, Ntinginya NE, Mfinanga E, et al. Assessment of the novel T-cell activation markertuberculosis assay for diagnosis of active tuberculosis in children: a prospective proof-of-concept study. Lancet Infect Dis 2014 Oct;14(10):931e8.