Yu2017 2
Yu2017 2
Yu2017 2
Department of Thoracic Surgery, Hangzhou Red Cross Hospital, Hangzhou, Zhejiang, China
a1111111111 ☯ These authors contributed equally to this work.
a1111111111 ‡ These authors also contributed equally to this work.
a1111111111 * xuxudong234@163.com
a1111111111
a1111111111
Abstract
Competing interests: The authors have declared against histopathology and CRS, respectively. The differences in sensitivity, PPV, and
that no competing interests exist. AUC of Xpert MTB/RIF and T-SPOT.TB were not statistically significant (P > 0.05), com-
pared to those of histopathology and CRS. However, the differences in specificity and NPV
of the two assays were significant (P < 0.05), compared to those of histopathology and
CRS.
Conclusions
Xpert MTB/RIF test is a valid diagnostic technique for TBP with higher sensitivity and speci-
ficity than T-SPOT.TB.
Introduction
Mycobacterium tuberculosis (MTB) infection causes tuberculosis (TB), one of the most serious
problems in public health globally [1]. The most common site of TB infection is the lung,
although the bacteria can also spread to extra-pulmonary sites, leading to extrapulmonary TB
(EPTB). Tuberculous pericarditis (TBP) is a type of EPTB. The incidence of TBP has been
increasing along with the acquired immune deficiency syndrome (AIDS) epidemic. TBP was
the most common reason for pericarditis in areas with high TB burden [2, 3]. TBP is known to
increase the risk of unfavorable outcomes, including cardiac tamponade, constrictive pericar-
ditis, and mortality [4]. The fatality rate of TBP is high (17–40%) above six months [3, 5].
Therefore, TBP should be diagnosed and treated as early as possible. However, the diagnosis of
TBP is challenging and is often postponed [6], because the MTB culture and positive smear
rates from pericardial effusions are very low [7]. Other tests are also used for diagnosing TBP;
for instance, adenosine deaminase (ADA) is commonly used in the diagnosis of TBP. A meta-
analysis showed that the overall sensitivity (95% confidence interval [CI]) of this test was 90%
(86–93%) and specificity was 86% (83–89%) [8]. Interferon-γ release (T-SPOT.TB) test using
pericardial effusion and blood can also be used for the diagnosis, and it has a sensitivity of 92%
(72–99%) and 83% (62–95%) and a specificity of 92% (78–98%) and 95% (81–99%), respec-
tively [9]. In addition, the lipoarabinomannan (LAM) assay using pericardial effusions and
urinary samples has been shown to have low sensitivity, but high specificity. For instance, a
previous study revealed that the sensitivity and specificity of the LAM assay using pericardial
infusion were 11.6% (6.0–21.3%) and 88% (70.0–95.8%), respectively, while those using uri-
nary samples were 17.4% (9.1–30.7%) and 93.8% (71.7–98.9%), respectively [10]. However, as
pericardial effusion is difficult to obtain in several cases and these tests lack scope, the diagno-
sis of TBP is still very difficult.
The Xpert MTB/RIF assay shows a high accuracy for PTB and EPTB detection, and it can
provide results within 2 hours [11, 12]. Previous studies have shown that this test has a high
sensitivity and specificity for PTB and EPTB [13, 14]. However, the diagnostic utility of Xpert
MTB/RIF on TBP has not been studied well. Only one relevant study has been published, and
it showed a sensitivity of 63.8% (52.4–75.1%) and a specificity of 100% (85.6–100%) compared
to those of the ADA and interferon-γ tests [15]. No studies have been conducted to evaluate
the diagnostic value of Xpert MTB/RIF assay for TBP using pericardial tissues.
This study aimed to assess the diagnostic performance of the Xpert MTB/RIF test for TBP
using the pericardial tissues. We also compared its performance with that of T-SPOT.TB
assay.
divided for histopathological examination and Xpert MTB/RIF assay and culture (including
MTB, bacteria, and fungi), after grinding.
Results
Thirty patients suspected to have TBP were registered. Among them, two patients refused
surgery, and one of them showed negative results in T-SPOT.TB and Xpert MTB/RIF assays.
Thus, 27 patients were enrolled for the experiment (Fig 1).
Of the 27 patients, 14 (52%) were diagnosed as confirmed TBP (one of them was sputum
smear-positive), 3 (11%) were diagnosed with highly probable TBP, and 10 (37%) had no TBP
(one patient had connective tissue disease, and nine had non-specific pericarditis). Most of the
patients showed no obvious pericardial effusion and could not safely undergo pericardiocent-
esis. Only 7 (26%) patients showed pericardial effusion after pericardial puncture. Sixteen
(59%) patients showed unilateral or bilateral pleural effusion, and 9 (33%) patients showed
both pleural effusion and ascites. Among the patients with confirmed TBP, 4 (29%) had PTB
and 1 (17%) had both PTB and vertebral TB. Among the non-TBP patients, 3 (30%) had
Fig 1. Flowchart showing the classification of patients of included in the study. TB: Tuberculosis.
https://doi.org/10.1371/journal.pone.0188704.g001
obsolete PTB. HIV-positive patients were not included in the study. MTB, bacterial, and fungal
cultures were negative for all patients. The clinical features of patients with suspected TBP are
presented in Table 1.
T-SPOT.TB assay
Using histopathology as the reference standard, the T-SPOT.TB assay using peripheral blood
had sensitivity, specificity, PPV, NPV, PLR, NLR, and AUC values of 92.9% (66.1–99.8%),
15.4% (1.9–45.5%), 54.2% (32.8–74.5%), 66.7% (9.4–99.2%), 1.10 (0.83–1.44), 0.46 (0.05–4.53),
TB: Tuberculosis, HB: Hemoglobin, ESR: Erythrocyte sedimentation rate, CRP: C reactive protein, BNP: B-type natriuretic peptide, EF: Ejection fraction.
https://doi.org/10.1371/journal.pone.0188704.t001
and 0.541 (0.340–0.733), respectively. Using CRS as the reference standard, the assay possessed
sensitivity, specificity, PPV, NPV, PLR, NLR, and AUC values of 94.1% (71.3–99.9%), 20.0%
(2.5–55.6%), 66.7% (44.7–84.4%), 66.7% (9.4–99.2%), 1.18 (0.84–1.64), 0.29 (0.03–2.85), and
0.571 (0.367–0.758), respectively.
Table 2. Diagnostic efficiency of Xpert MTB/RIF and TSPOT.TB tests, using that of histopathology as a reference standard. P-values corresponded
to statistical comparisons between Xpert MTB/RIF and TSPOT.TB tests.
Sensitivity Specificity PPV NPV PLR NLR AUC
(95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI)
Xpert MTB/RIF 78.6% 92.3% 91.7% 80.0% 10.21 0.23 0.854
(49.2%–95.3%) (64.0%–99.8%) (61.5%–99.8%) (51.9%–95.7%) (1.52–68.49) (0.08–0.64) (0.666–0.959)
TSPOT.TB 92.9% 15.4% 54.2% 66.7% 1.10 0.46 0.541
(66.1%–99.8%) (1.9%–45.5%) (32.8%–74.5%) (9.4%–99.2%) (0.83–1.44) (0.05–4.53) (0.340–0.733)
P value P = 0.596 P < 0.001 P = 0.031 P = 1.000 P < 0.001 P = 0.027 P = 0.001
PPV: Positive predictive value, NPV: Negative predictive value, PLR: Positive likelihood ratio, NLR: Negative likelihood ratio, AUC: Area under curve.
https://doi.org/10.1371/journal.pone.0188704.t002
Table 3. Diagnostic value of Xpert MTB/RIF and TSPOT.TB tests, using that of CRS as a reference standard. P-values corresponded to statistical
comparisons between Xpert MTB/RIF and TSPOT.TB tests.
Sensitivity Specificity PPV NPV PLR NLR AUC
(95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI)
Xpert MTB/RIF 70.6% 100% 100% 66.7% undefined 0.29 0.853
(44.0–89.7%) (69.2–100%) (73.5–100%) (38.4–88.2%) (0.14–0.61) (0.664–0.959)
TSPOT.TB 94.1% 20.0% 66.7% 66.7% 1.18 0.29 0.571
(71.3–99.9%) (2.5–55.6%) (44.7–84.4%) (9.4–99.2%) (0.84–1.64) (0.03–2.85) (0.367–0.758)
P value P = 0.175 P = 0.001 P = 0.033 P = 1.000 - P = 1.000 P = 0.003
PPV: Positive predictive value, NPV: Negative predictive value, PLR: Positive likelihood ratio, NLR: Negative likelihood ratio, AUC: Area under curve.
https://doi.org/10.1371/journal.pone.0188704.t003
Discussion
The Xpert MTB/RIF assay is a rapid, automatic, and WHO-approved TB diagnostic test,
which can not only diagnose MTB complex DNA, but also confirm rifampicin resistance
caused by the rpoB gene mutations [11]. This assay is also the preferred diagnostic method for
detecting drug-resistant or HIV-associated TB [17]. It is widely used for diagnosing PTB and
EPTB. Several studies have reported the diagnostic efficiency of Xpert MTB/RIF for PTB and
EPTB. For instance, Chew MY et al. showed that Xpert MTB/RIF had a sensitivity of 75% and
a specificity of 99.5% in an intermediate-burden setting for diagnosing PTB [18]. A study in
Ethiopia showed that Xpert MTB/RIF had a sensitivity of 93.5% (78.3–98.9%) and a specificity
of 69.2% (66.4–70%) for tuberculous lymphadenitis [19]. In case of pleural TB, Xpert MTB/
RIF showed a sensitivity of 22.5% (12.4–37.6%) and a specificity of 98% (89.2–99.7%) [20].
T-SPOT.TB is another frequently used technique for diagnosing TB in various systems [21–
23]. The test can be performed on blood and other body fluids [9, 22] (such as pleural effusion
and pericardial effusion). However, in our hospital, all samples other than blood are tested by
the T-SPOT.TB assay. Therefore, this assay was not used on pericardial effusion or tissues.
TBP is a type of EPTB, which accounts for about 4% of pericarditis cases in the developed
countries [3]; in South Africa, this rate is as high as 60% [24]. In this study, TB was found to be
the cause of pericarditis in 63% (17/27) of the patients, which is a rate higher than that reported
previously [25]. This might be because our hospital is a diagnosis and treatment center for TB
in the Zhejiang province, with a naturally high proportion of patients with TB. We consider
that an imperfect reference standard may lead to misclassification of samples in diagnostic
validity studies. In case of EPTB, histopathology is an imperfect reference standard. This refer-
ence standard would lead to an underestimation of the true specificity of Xpert MTB/RIF. CRS
is a composite standard, which includes results of several tests or clinical conditions that may
sometimes be reclassified as false positive results of Xpert MTB/RIF as true positive results,
and thus, lead to an increase in Xpert MTB/RIF specificity. However, CRS itself may have
reduced specificity that could result in apparent false-negative Xpert MTB/RIF results, leading
to an underestimation of the true sensitivity of Xpert MTB/RIF. Therefore, a comparative
study with the two reference standards, histopathology and CRS, might provide a more credi-
ble range for sensitivity and specificity.
There have been very few reports on the diagnostic utility of Xpert MTB/RIF for TBP. The
Xpert MTB/RIF assay was performed on pericardial effusions in previously published studies
[15, 26]. The utility of Xpert MTB/RIF using pericardial tissues has never been reported. In
this study, the vast majority of patients (20/27) did not show adequate pericardial effusion
at the time of admission to safely perform pericardial puncture. Therefore, we assessed the
diagnostic value of Xpert MTB/RIF on pericardial tissues, which were obtained by pericardial
fenestration or resection. To the best of our knowledge, this is the first report to assess the diag-
nostic utility of Xpert MTB/RIF using pericardial tissue for TBP and compare it with T-SPOT.
TB test.
The T-SPOT.TB assay was found to be more sensitive than the Xpert MTB/RIF assay,
although the difference was not statistically significant, regardless of the reference standard
used. Xpert MTB/RIF assay had higher specificity (92.3% and 100%, with histopathology and
CRS as the reference standards, respectively) and PPV (91.7% and 100%) using pericardial tis-
sue than those of the T-SPOT.TB assay. This result was consistent with the result of a previous
study (specificity and PPV were both 100%) [15], in which Xpert MTB/RIF was performed
using pericardial effusion. However, it is still unknown whether the diagnostic value of Xpert
MTB/RIF using pericardial effusion and pericardial tissue is different. Further studies would
be needed to confirm this hypothesis.
In contrast, the specificity of the T-SPOT.TB assay using the peripheral blood was very low
(15.4% and 20.0%, respectively, with histopathology and CRS as the reference). These values
were significantly lower than the previously reported ones [9]. The PPV of T-SPOT.TB assay,
compared to histopathology and CRS, was also relatively low. This might be because China has
a high incidence of TB. Nearly half the population has been reported to be infected with MTB
[27], and patients with latent infection may show positive results.
The specificity and PPV of Xpert MTB/RIF and T-SPOT.TB assays were increased when
CRS was used as the reference. Although histopathology is considered the gold standard for
diagnosing TBP, it is not a perfect reference method. The results of pathological diagnosis are
closely related to the location of the biopsy site, the quality of equipment, and the experience
of the physician. CRS classifies TB based on positive results from one of the several criteria,
including culture, clinical manifestation, histopathology, imaging, and response to treatment,
resulting in the reclassification of false positives to true positives and increased specificity and
PPV [28].
Receiver operating curve (ROC) analysis showed that the AUC of Xpert MTB/RIF was sig-
nificantly higher than that of T-SPOT.TB, regardless of the reference standard. The T-SPOT.
TB assay was found to be not important for diagnosing TBP, while the Xpert MTB/RIF test
played a crucial role in it.
There were a number of limitations of the present study. The number of patients studied
was limited and included no positive cases of MTB culture. This might have caused a bias in
the evaluation of the effectiveness of the diagnosis. Furthermore, this was a single-center retro-
spective study, performed in the TB diagnosis and treatment center of the Zhejiang province.
This might have led to a selection bias, as the patients enrolled would always be suspected to
have TB. This might also explain the high prevalence rate of TBP in this study, compared to
the previous ones [9, 25].
Studies on the viability of the Xpert MTB/RIF assay using pericardial tissues for the diagno-
sis of TBP are rare; our study may have some clinical significance, although the applicability of
an invasive operation may be limited. In future studies, multi-center cooperation is needed to
assess the diagnostic validity of Xpert MTB/RIF in a wider population.
Conclusions
In summary, the diagnostic validity of Xpert MTB/RIF using pericardial tissue for TBP has not
been studied before, and to the best of our knowledge, this was the first such report. We con-
cluded that the Xpert MTB/RIF assay using pericardial tissue showed high specificity and PPV
for diagnosing TBP. The specificity and PPV of T-SPOT.TB were low, but the sensitivity was
high. The Xpert MTB/RIF assay had better diagnostic efficiency for TBP than T-SPOT.TB
assay had. Thus, Xpert MTB/RIF is a suitable diagnostic method for TBP, and the diagnostic
value of T-SPOT.TB is not exceptional.
Supporting information
S1 Supporting Information. Data generated or analyzed in this study, including patient
characteristics, raw data, and funding.
(ZIP)
Acknowledgments
We are thankful to all the patients who enrolled in the work and their families, as well as the
medical staff at Hangzhou Red Cross Hospital.
Author Contributions
Conceptualization: Xudong Xu.
Data curation: Guocan Yu.
Formal analysis: Guocan Yu, Bo Ye, Fangming Zhong.
Funding acquisition: Xudong Xu.
Investigation: Bo Ye, Da Chen, Fangming Zhong, Jun Yang.
Methodology: Guocan Yu, Xudong Xu.
Project administration: Xudong Xu.
References
1. World Health Organization. Global tuberculosis report. 2016.
2. Ntsekhe M, Mayosi BM. Tuberculous pericarditis with and without HIV. Heart Fail Rev. 2013; 18(3):
367–373. https://doi.org/10.1007/s10741-012-9310-6 PMID: 22427006
3. Syed FF, Mayosi BM. A modern approach to tuberculous pericarditis. Prog Cardiovasc Dis. 2007; 50
(3): 218–236. https://doi.org/10.1016/j.pcad.2007.03.002 PMID: 17976506
4. Jung IY, Song YG, Choi JY, Kim MH, Jeong WY, Oh DH, et al. Predictive factors for unfavorable out-
comes of tuberculous pericarditis in human immunodeficiency virus-uninfected patients in an intermedi-
ate tuberculosis burden country. BMC Infect Dis. 2016; 16(1): 719. https://doi.org/10.1186/s12879-016-
2062-5 PMID: 27899066
5. Mayosi BM, Wiysonge CS, Ntsekhe M, Gumedze F, Volmink JA, Maartens G, et al. Mortality in patients
treated for tuberculous pericarditis in sub-Saharan Africa. S Afr Med J. (Suid-Afrikaanse tydskrif vir gen-
eeskunde). 2008; 98(1): 36–40.
6. Strang G, Latouf S, Commerford P, Roditi D, Duncan-Traill G, Barlow D, et al. Bedside culture to con-
firm tuberculous pericarditis. Lancet. 1991; 338(8782–8783): 1600–1601.
7. Mayosi BM, Burgess LJ, Doubell AF. Tuberculous pericarditis. Circulation. 2005; 112(23): 3608–3616.
https://doi.org/10.1161/CIRCULATIONAHA.105.543066 PMID: 16330703
8. Xie DL, Cheng B, Sheng Y, Jin J. Diagnostic accuracy of adenosine deaminase for tuberculous pericar-
ditis: a meta-analysis. Eur Rev Med Pharmacol Sci. 2015; 19(22): 4411–4418. PMID: 26636531
9. Bian S, Zhang Y, Zhang L, Shi X, Liu X. Diagnostic value of interferon-gamma release assays on peri-
cardial effusion for diagnosis of tuberculous pericarditis. PLoS One. 2016; 11(10): e0165008. https://
doi.org/10.1371/journal.pone.0165008 PMID: 27755587
10. Pandie S, Peter JG, Kerbelker ZS, Meldau R, Theron G, Govender U, et al. The diagnostic accuracy of
pericardial and urinary lipoarabinomannan (LAM) assays in patients with suspected tuberculous peri-
carditis. Sci Rep. 2016; 6: 32924. https://doi.org/10.1038/srep32924 PMID: 27633798
11. WHO guidelines approved by the guidelines review committee. In: Automated real-time nucleic acid
amplification technology for rapid and simultaneous detection of tuberculosis and rifampicin resistance:
Xpert MTB/RIF assay for the diagnosis of pulmonary and extrapulmonary TB in adults and children: Pol-
icy Update. 2013. edn. Geneva: World Health Organization. Copyright (c) World Health Organization.
12. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al. Rapid molecular detection of
tuberculosis and rifampin resistance. N Engl J Med. 2010; 363(11): 1005–1015. https://doi.org/10.1056/
NEJMoa0907847 PMID: 20825313
13. Steingart KR, Schiller I, Horne DJ, Pai M, Boehme CC, Dendukuri N. Xpert(R) MTB/RIF assay for pul-
monary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014; 1:
Cd009593.
14. Mazzola E, Arosio M, Nava A, Fanti D, Gesu G, Farina C. Performance of real-time PCR Xpert (R)MTB/
RIF in diagnosing extrapulmonary tuberculosis. Infez Med. 2016; 24(4): 304–309. PMID: 28011966
15. Pandie S, Peter JG, Kerbelker ZS, Meldau R, Theron G, Govender U, et al. Diagnostic accuracy of
quantitative PCR (Xpert MTB/RIF) for tuberculous pericarditis compared to adenosine deaminase and
unstimulated interferon-gamma in a high burden setting: a prospective study. BMC Med. 2014; 12: 101.
https://doi.org/10.1186/1741-7015-12-101 PMID: 24942470
16. Banaei N, Gaur RL, Pai M. Interferon gamma release assays for latent tuberculosis: What are the
sources of variability? J Clin Microbiol. 2016; 54(4): 845–850. https://doi.org/10.1128/JCM.02803-15
PMID: 26763969
17. Lawn SD, Mwaba P, Bates M, Piatek A, Alexander H, Marais BJ, et al. Advances in tuberculosis diag-
nostics: the Xpert MTB/RIF assay and future prospects for a point-of-care test. Lancet Infect Dis. 2013;
13(4): 349–361. https://doi.org/10.1016/S1473-3099(13)70008-2 PMID: 23531388
18. Chew MY, Ng J, Cai HM, Lim TG, Lim TK. The clinical utility of Xpert(R) MTB/RIF testing in induced spu-
tum. Int J Tuberc Lung Dis. 2016; 20(12): 1668–1670.
19. Biadglegne F, Mulu A, Rodloff AC, Sack U. Diagnostic performance of the Xpert MTB/RIF assay for
tuberculous lymphadenitis on fine needle aspirates from Ethiopia. Tuberculosis. 2014; 94(5): 502–505.
https://doi.org/10.1016/j.tube.2014.05.002 PMID: 24931451
20. Meldau R, Peter J, Theron G, Calligaro G, Allwood B, Symons G, et al. Comparison of same day diag-
nostic tools including Gene Xpert and unstimulated IFN-gamma for the evaluation of pleural tuberculo-
sis: a prospective cohort study. BMC Pulm Med. 2014; 14: 58. https://doi.org/10.1186/1471-2466-14-58
PMID: 24708530
21. Li G, Li F, Zhao HM, Wen HL, Li HC, Li CL, et al. Evaluation of a new IFN-gamma release assay for
rapid diagnosis of active tuberculosis in a high-incidence setting. Front Cell Infect Microbiol. 2017; 7:
117. https://doi.org/10.3389/fcimb.2017.00117 PMID: 28443247
22. Kang JY, Rhee CK, Kang NH, Kim JS, Yoon HK, Song JS. Clinical utility of two interferon-gamma
release assays on pleural fluid for the diagnosis of tuberculous pleurisy. Tuberc Respir Dis. 2012; 73(3):
143–150.
23. Feng Y, Diao N, Shao L, Wu J, Zhang S, Jin J, et al. Interferon-gamma release assay performance in
pulmonary and extrapulmonary tuberculosis. PLoS One. 2012; 7(3): e32652. https://doi.org/10.1371/
journal.pone.0032652 PMID: 22427859
24. Bathoorn E, Limburg A, Bouwman JJ, Bossink AW, Thijsen SF. Diagnostic potential of an enzyme-
linked immunospot assay in tuberculous pericarditis. Clin Vaccine Immunol. 2011; 18(5): 874–877.
https://doi.org/10.1128/CVI.00031-11 PMID: 21450973
25. Ma W, Liu J, Zeng Y, Chen S, Zheng Y, Ye S, et al. Causes of moderate to large pericardial effusion
requiring pericardiocentesis in 140 Han Chinese patients. Herz. 2012; 37(2): 183–187. https://doi.org/
10.1007/s00059-011-3428-5 PMID: 21301790
26. Dhana AV, Howell P, Sanne I, Spencer D. Identification of Mycobacterium tuberculosis from pericardial
fluid using the new Xpert MTB/RIF assay. BMJ Case Rep. 2013; https://doi.org/10.1136/bcr-2013-
200615 PMID: 23986128
27. Gao L, Lu W, Bai L, Wang X, Xu J, Catanzaro A, et al. Latent tuberculosis infection in rural China: base-
line results of a population-based, multicentre, prospective cohort study. Lancet Infect Dis. 2015; 15(3):
310–319. https://doi.org/10.1016/S1473-3099(14)71085-0 PMID: 25681063
28. Suzana S, Ninan MM, Gowri M, Venkatesh K, Rupali P, Michael JS. Xpert MTB/Rif for the diagnosis of
extrapulmonary tuberculosis—an experience from a tertiary care centre in South India. Trop Med Int
Health. 2016; 21(3): 385–392. https://doi.org/10.1111/tmi.12655 PMID: 26671654