CLINICAL INVESTIGATION
Evaluation of Cerebrospinal Fluid Adenosine Deaminase
Activity for the Differential Diagnosis of Tuberculous and
Nontuberculous Meningitis
Qin Sun, MM, Wei Sha, PhD, He-ping Xiao, BM, Qing Tian, BM and Hong Zhu, MM
Abstract: Introduction: The diagnosis value of adenosine deaminase
(ADA) activity in cerebrospinal fluid (CSF) of tuberculous meningitis
(TBM) has been well documented. However, the cutoff point of CSF
ADA has not been fully assessed. In the current study, the authors set
to calculate the cutoff points of ADA and monitor the changes of CSF
ADA activities in patients with TBM after antitubercular therapy.
Methods: CSF ADA activity in patients with different types of menin-
gitis was measured by Trinder enzyme-coupled assay. Results: The
mean CSF ADA values in the patients with TBM, bacterial meningitis,
viral meningitis, cryptococcal meningitis and noninfectious neurologic
disorders were 14.1 6 5.4, 9.6 6 5.5, 4.3 6 2.5, 7.8 6 3.4 and 2.6 6
1.3 U/L, respectively. CSF ADA activity was significantly higher in
TBM compared with patients with non-TBM (P , 0.05). Moreover,
the best cutoff point for differentiating between TBM and non-TBM
was 9.5 U/L. In addition, CSF ADA activity was decreased in patients
with TBM after antitubercular therapy in a time-dependent manner.
Conclusions: The determination of ADA with a cutoff value of 9.5 U/
L in CSF is a useful aid for the differential diagnosis of TBM and non-
TBM. Moreover, dynamic monitoring of CSF ADA activity may be an
indicator for evaluating antitubercular therapy in TBM.
Key Indexing Terms: Adenosine deaminase; Tuberculous meningitis;
Cerebrospinal fluid; Diagnosis. [Am J Med Sci 2012;344(2):116–121.]
T uberculous meningitis (TBM) is an infection of the menin-
ges caused by Mycobacterium tuberculosis. In developing
countries, TBM accounts for more than 20% of community-
acquired meningitis in adults.1 Delayed diagnosis and treatment
may be associated with serious central nervous system (CNS)
complications.2 The gold standard for the diagnosis of TBM is
the detection of M tuberculosis in cerebrospinal fluid (CSF) by
acid-fast bacilli smear or Mycobacteria culture. However, the
sensitivity of CSF acid-fast bacilli is less than 10%, and
Mycobacteria culture is only positive in 50% to 75% cases.3
Furthermore, conventional culture techniques may require 4 to
8 weeks,3–5 which is an unacceptable length of time for early
diagnosis and treatment. Thus, a rapid and reliable diagnostic
test that can be performed in standard laboratories is of great
importance to clinicians.
Adenosine deaminase (ADA) is a key enzyme, which
catalyzes adenosine deamination to inosine, and is found in
many tissues, particularly in T-lymphocytes from the lymphoid
From the Department of Tuberculosis (QS, WS, H-PX), Shanghai Pulmo-
nary Hospital, Tongji University School of Medicine, Shanghai, China; and
Emergency Department (QT, HZ), Xinhua Hospital, Shanghai Jiaotong Uni-
versity School of Medicine, Shanghai, China.
Submitted June 7, 2011; accepted in revised form September 16, 2011.
This study was supported by a grant from the Construction for Key
Public Health Branch of Infectious Disease in Shanghai (88GWZX0104).
Correspondence: He-ping Xiao, BM, Department of Tuberculosis,
Shanghai Pulmonary Hospital, Tongji University School of Medicine,
507 Zheng Min Road, Yangpu District, Shanghai 200433, China (E-mail:
xiaoheping_sars@163.com).
116 The American Journal of the Medical Sciences
tissue.6 The basic physiologic function of ADA is related to
proliferation, maturation and function of lymphoid cells.7
ADA activity increases in patients with impaired cellular
immunity and is therefore acts as a robust biochemical marker.8
Several previous studies have demonstrated the diagnostic
value of CSF ADA activity in differentiating TBM from other
forms of meningitis.1,9–11 However, the results are variable, and
there has been no consensus regarding the cutoff value of CSF
ADA activity.1,2,9,11 Compounding the difficulties in establish-
ing a cutoff value is the belief that CSF ADA activity may differ
from one human race to another.11 To our knowledge, few
studies have evaluated CSF ADA activity for the diagnosis of
TBM in China. Therefore, the aim of this study was to calculate
the cutoff point for ADA activity in the CSF of patients with
TBM and non-TBM in a local Chinese population. Moreover,
the time courses of CSF ADA activity in patients with TBM
after antitubercular therapy were investigated.
METHODS
Subjects
Our retrospective study involved a total of 334 patients.
One hundred twenty-one cases were positive for TBM, who
were recruited between October 2007 and September 2010 in
Shanghai Pulmonary Hospital affiliated to Tongji University, an
868-bed specialist hospital for tuberculosis. One hundred forty-
one cases of non-TBM meningitis and 72 cases of patients with
noninfectious neurologic disorders who were admitted during
the same period in Xinhua Hospital affiliated to Shanghai
Jiaotong University, a 1680-bed teaching hospital, were also
reviewed. Among patients with non-TBM, there were 51 cases
of bacterial meningitis, 66 cases of viral meningitis and 24 cases
of cryptococcal meningitis. All patients were older than 16
years. Detailed patient clinical characteristics and CSF param-
eters are shown in Tables 1 and 2. CSF samples for ADA
estimations were obtained before any treatment in all cases.
CSF ADA activity in 73 patients with TBM was measured
again 2, 6 and 10 weeks after antitubercular therapy. The
antimicrobial agents we used in the treatment of TBM included
the following 4 drugs: isoniazid, rifampin, pyrazinamide and
streptomycin. Some patients were also treated by some second-
line drugs, such as para-aminosalicylic acid, levofloxacin or
moxifloxacin, according to their conditions and the doctors’
experience. Corticosteroids were used in most cases with the
presence of increased intracranial pressure, altered conscious-
ness, focal neurologic findings, spinal block and tuberculous
encephalopathy.
Diagnostic Classification
Patients With TBM
1. Microbiologically diagnosed cases (n 5 37): confirmed by
the presence of M tuberculosis in CSF by culture.

Volume 344, Number 2, August 2012
 Diagnosis of Tuberculous and Nontuberculous Meningitis

2. Clinically diagnosed cases (n 5 24): this group were all

TABLE 1. Clinical characteristics of patients with meningitis culture-negative cases with the following observations—
Tuberculous meningitis 121 fever and features of meningeal irritation. Some patients
Male:Female 64:57 underwent cranial surgery. CSF samples showed pleocy-
Mean age (yr) with range 32.7 6 13.2(16–75) tosis with a predominance of polymorphonuclear cells,
Mycobacterium tuberculosis culture 37 increased proteins and decreased glucose (CSF: blood
Clinically diagnosed 84 glucose ratio ,0.2). All cases had a good clinical
Pulmonary tuberculosis 82 response to broad-spectrum antibiotics. None of the cases
Miliary tuberculosis 69 had clinical evidence for tuberculosis infection.
Pleural tuberculosis 5
Patients With Viral Meningitis (n 5 66)
Tuberculous lymphadenitis 11
This group was clinically diagnosed if the following
Nontuberculous meningitis 141 were present: acute onset of fever with signs of meningeal
Male:Female 81:60 irritation, CSF samples showed pleocytosis with lymphocytic
Mean age (yr) with range 30.3 6 12.7(16–63) predominance, mild increases in protein and normal glucose
Bacterial meningitis 51 levels. Nine patients were herpesvirus polymerase chain
Confirmed 27 reaction positive. All patients spontaneously improved without
Neisseria meningitides 9 specific treatment. None of the cases had clinical evidence for
Streptococcus pneumoniae 8 tuberculosis infection.
Staphylococcus aureus 6
Klebsiella pneumonia 4
Cryptococcal Meningitis Patients (n 5 24)
All cases tested positive for CSF latex agglutination
Clinically diagnosed 24
(titer .1:100), and 6 of them were positive for CSF India ink
Viral meningitis 66 smear. None of the cases had clinical evidence for tuberculosis
Herpesvirus PCR 9 infection.
Cryptococcal meningitis 24
Latex agglutination test (titer .1:100) 24 Patients With Noninfectious Neurologic Disorders (n 5 72)
India ink smear 6 This group included patients who had no evidence of
Noninfectious neurologic disorders 72 CNS or extra CNS infections and had undergone the lumbar
Male: Female 39: 33 puncture because of chronic intractable headache, vomit,
Mean age (yr) with range 37.5 6 14.7(18–72)
epilepsy, stroke, etc.
Headache and/or vomit 49 CSF ADA Activity Assay
Epilepsy 13 CSF specimens taken from patients were sent to the
Stroke 10 Shanghai Key Laboratory of Mycobacteria Tuberculosis at
PCR, polymerase chain reaction. Shanghai Pulmonary Hospital in a sterile plastic tube, immedi-
ately centrifuged and kept in –20°C. Samples were tested for
biochemical substances and enzyme activity on the same or
subsequent day. CSF ADA activity was measured by Trinder
2. Clinically diagnosed patients (n 5 84): this group had enzyme-coupled method. ADA catalyses the following
negative cultures with all of the following observations— reaction: Adenosine + H2O / Inosine + NH3. The ADA assay
subacute or chronic fever with signs of meningeal irritation is based on indirectly measuring the formation of NH3 produced
such as headache, neck stiffness and vomiting, with or when ADA acts in excess of adenosine. The release of ammonia
without other features of CNS involvement. CSF samples was determined colorimetrically at 546 nm after the develop-
showed pleocytosis with a predominance of lymphocytes, ment of an intense blue color with hypochlorite and phenol in
increased protein levels and decreased glucose (CSF:blood an alkaline solution. ADA activity is expressed in U/L at 37°C.
glucose ratio ,0.5). All cases improved with antitubercu- The lower limit of sensitivity of the assay was 0.1 U/L.
lar drugs.
Statistical Analysis
Bacterial Meningitis Patients Results are expressed as mean 6 standard deviation.
A Mann-Whitney U nonparametric test was used to compare
1. Confirmed cases (n 5 27): presence of pathogenic bacte- the means of each group. A paired sample t test was used to
ria in CSF by staining and/or culture. compare the means of CSF ADA before and after treatment. A

TABLE 2. CSF parameters of patients with TBM and Non-TBM (mean 6 SD, range)
Parameters Tuberculous (n 5 121) Bacterial (n 5 51) Viral (n 5 66) Cryptococcal (n 5 24)
White blood cells (3103/mL) 333 6 176 (50–890) 1867 6 1251 (150–4650) 278 6 46 (20–685) 254 6 139 (65–550)
Polymorphonuclear (%) 22.6 6 10.2 (2.0–55.0) 74.4 6 11.2 (54.0–95.0) 18.5 6 11.4 (2.0–48.0) 36.8 6 16.2 (15.0–78.0)
Lymphocyte (%) 68.3 6 23.6 (42.5–95.0) 21.4 6 19.7 (8.2–46.5) 77.3 6 26.2 (41.2–98.5) 42.6 6 15.7 (26.3–61.4)
Protein (g/L) 1.8 6 1.3 (0.8–10.3) 3.1 6 2.6 (1.1–18.8) 1.1 6 0.87 (0.3–5.6) 1.5 6 1.7 (0.5–13.9)
CSF, cerebrospinal fluid; TBM, tuberculous meningitis.

Ó 2012 Lippincott Williams & Wilkins 117

Sun et al

TABLE 3. CSF ADA activity in the patients with TBM and

Non-TBM (mean 6 SD, U/L)
Patient Groups Number ADA Activity Range
Tuberculous meningitis 121 14.1 6 5.4 2.8–44.1
Nontuberculous meningitis 141 6.8 6 4.6 0.6–27.5
Bacterial meningitis 51 9.6 6 5.5 3.9–27.5
Viral meningitis 66 4.3 6 2.5 0.6–11.7
Cryptococcal meningitis 24 7.8 6 3.4 3.2–19.2
Noninfectious neurologic 72 2.6 6 1.3 0.2–6.4
disorders
CSF, cerebrospinal fluid; ADA, adenosine deaminase; TBM,
tuberculous meningitis.

receiver operating characteristic curve was plotted to identify

various cutoff points to determine the best level for CSF ADA
activity of patients with TBM. Statistical analysis was performed
using SPSS 13.0 for windows. A P value less than 0.05 (two
tailed) was considered statistically significant.

RESULTS
CSF ADA Activity Is Increased in Patients With TBM
and Non-TBM
The CSF ADA activity and individual distribution are
shown in Table 3 and Figure 1. The mean of CSF ADA activity
in patients with TBM was significantly higher than that in
patients with non-TBM (u 5 10.739, P , 0.001). CSF ADA
activity was also significantly increased relative to the patients
with bacterial meningitis (u 5 5.705, P , 0.001), viral menin-
gitis (u 5 10.702, P , 0.001), cryptococcal meningitis (u 5
5.747, P , 0.001) and patients with noninfectious neurologic
disorders (u 5 11.514, P , 0.001).
The Cutoff Values of CSF ADA Activity for
Differentiating TMB and Non-TBM
The optimum cutoff point for the total diagnostic index
to classify a patient as having TBM was plotted by use of
receiver operating characteristic curves (Figure 2). The test
FIGURE 2. ROC curve (I versus II): AUC 5 0.776 (95% CI: 0.689-
0.863), SE 5 0.044. ROC curve (I versus III): AUC 5 0.974 (95%
CI: 0.954-0.994), SE 5 0.010; ROC curve (I versus IV): AUC 5
0.872 (95% CI: 0.794-0.950), SE 5 0.040; ROC curve (I versus III
+IV): AUC 5 0.947 (95% CI: 0.917-0.977), SE 5 0.015; ROC
curve (I versus II+III+IV): AUC 5 0.885 (95% CI: 0.842-0.927),
SE 5 0.022. I, TBM; II, bacterial meningitis; III, viral meningitis; IV,
cryptococcal meningitis; II+III+IV, non-TBM; ROC, receiver
operating characteristic.
performances of CSF ADA with different cutoff points are
shown in Table 4. In particular, the best cutoff point for testing
TBM and non-TBM was 9.5 U/L, with a sensitivity and spec-
ificity of 87.6% and 80.1%, respectively. A value .27.5 U/L
was not observed in any case of the patients with non-TBM,
which means specificity was 100% when a cutoff value of 27.5
U/L was used, although sensitivity was low.

FIGURE 1. Box plot of CSF ADA activity (U/L) in patients with TBM, non-TBM and noninfectious neurologic disorders. The plots show
the ninetieth percentile (bars), 75th, 25th percentile (box), and median (bar in box), outliers (circle) and extremum (asterisk).

118 Volume 344, Number 2, August 2012


TABLE 4. Test performance characteristics of CSF ADA with different cutoff points
Patient Groups Cutoff (U/L) Accuracy (%) Sensitivity (%) Specificity (%)
I vs. II 10.8 74.8 75.2
I vs. III 8.3 93.6 92.6
I vs. IV 9.8 80.0 79.3
I vs. III+IV 9.0 90.0 90.1
I vs. II+III+IV 9.5 83.6 87.6
I, TBM; II, bacterial meningitis; III, viral meningitis; IV, cryptococcal meningitis; II+III+IV, non-TBM; PPV, positive predictive value; NPV,
negative predictive value; LR, likelihood ratio.
Antitubercular Therapy Decreases CSF ADA Activity
The CSF ADA activity was also determined in 73
patients with TBM after 2, 6 and 10 weeks of antitubercular
therapy. The mean of CSF ADA activity in these patients was
14.6 6 6.9 U/L before receiving antitubercular treatment. After
antitubercular treatment for 2, 6 and 10 weeks, the mean of CSF
ADA activity was significantly decreased to 8.6 6 4.8, 3.9 6
2.8 and 2.7 6 1.5 U/L (all P , 0.05), respectively, indicating
that antitubercular therapy reduced the CSF ADA activity in
a time-dependent manner. Figure 3 shows the dynamic changes
of CSF ADA activity before and after treatment.
DISCUSSION
TBM remains a major global health problem. The
growing number of new cases, especially in patients with
immunodeficiency disorders, clearly illustrates the importance
of the problem not only in developing countries but also in
developed countries.2,12 China has the second highest TB bur-
dened, with an HIV prevalence that is increasing rapidly.13
TBM is one of the most serious manifestations of extrapulmo-
nary tuberculosis. The diagnosis of TBM is complicated
because it causes various clinical manifestations, which overlap
with those of other acute or subacute infectious diseases of the
CNS, such as bacterial and viral meningitis.14 The initiation of
antituberculosis medication in suspected patients with TBM is
often delayed because of a lack of confidence in the currently
available laboratory tests.14,15 Some newer methods such as
nucleic acid amplification tests and rapid serologic methods
using antigen-antibody interaction have not been fully evalu-
ated or are too costly to perform as a routine test in most
countries with poor resources and shortage of advanced medical
FIGURE 3. Dynamic changes of CSF ADA activity in patients with TBM after antitubercular therapy.
Ó 2012 Lippincott Williams & Wilkins
Diagnosis of Tuberculous and Nontuberculous Meningitis
PPV (%) NPV (%) LR+ LR–
74.5 87.5 55.9 2.84 0.33
95.5 97.4 87.5 20.58 0.08
83.3 96.0 44.4 4.75 0.25
90.0 92.3 87.1 9.01 0.11
80.1 79.1 87.6 4.40 0.15
technology.3,16–18 Thus, a simple, rapid and reliable diagnostic
test that can be performed in standard laboratories is of great
necessity to clinical physicians in developing countries.
ADA has been considered a marker of cell-mediated
immunity and its activity has been observed in various
infections, including TBM.1,9–11 The laboratory method for
measuring ADA is inexpensive and relatively simple to per-
form. Thus, it may be useful in laboratories with limited resour-
ces, particularly in developing countries such as China where
the incidence of tuberculosis is high. In tuberculous pleural
effusion, tuberculous pericarditis and tuberculous peritonitis,
the high activity of ADA has been documented and has been
proven to be beneficial in the diagnosis of tuberculosis.19–22
Supporting evidence indicates that both humoral and cell-
mediated immunity play major roles in the pathophysiology
of TBM, so it has been suggested that ADA activity in CSF
could be used as a tool in differentiating TBM from non-TBM
infectious meningitis.1,9–11
Kashyap et al2 previously reported that the mean CSF
ADA activity of 117 patients with TBM was 14.3 6 3.87 U/L,
which is significantly higher than that of patients with non-
TBM. The current study demonstrated that the mean CSF
ADA activity of 121 patients with TBM was 14.1 6 5.4 U/L.
Studies by Moghtaderi et al9 and Chotmongkol et al1 reported
a higher mean CSF ADA activity of 23.05 6 13.1 and 39.44 6
41.46 U/L, respectively, but there were only 22 and 16 cases of
patients with TBM enrolled in their studies. Moreover, different
sensitivities and specificities of CSF ADA activity were
reported by using different cutoff values. Rana et al23 calculated
a cutoff value of 10 U/L with a sensitivity of 92.5% and a spec-
ificity of 97%. Kashyap et al2 reported a sensitivity of 82% and
a specificity of 83% when a cutoff value of 11.4 U/L was used.
119
Sun et al
Choi et al11 proposed a cutoff value of 7 U/L and showed
a sensitivity of 83% and a specificity of 95%. In the current
study, we calculated an ADA cutoff value of 9.5 U/L with
a sensitivity of 87.6% and a specificity of 80.1% in CSF for
the differential diagnosis of TBM and non-TBM infectious
meningitis.
However, some studies have reported a lower efficacy
of CSF ADA activity test and an overlap between TBM and
bacterial meningitis. A recent meta-analysis by Tuon et al24
concluded that ADA could not distinguish between bacterial
meningitis and TBM, but using ranges of ADA values could
be important to improve TBM diagnosis, particularly after
bacterial meningitis has been ruled out. In contrast, Thwaites
et al25 demonstrated the efficacy of the combination of clinical
and laboratory features in the diagnosis of adult TBM, using
a newly proposed score and a classification-tree method.
However, ADA was not included in their score because the
diagnosis of TBM was compared with the diagnosis of bacte-
rial meningitis, a common problem in developing Asian coun-
tries. Our study proposed a cutoff value of 10.8 U/L to
differentiate between TMB and bacterial meningitis, but this
was associated with low sensitivity (75.2%) and specificity
(74.5%). Therefore, our study suggests that ADA is not sen-
sitive and specific enough to distinguish TMB from bacterial
meningitis. However, CSF in patients with bacterial meningi-
tis often shows pleocytosis with a predominance of polymor-
phonuclear cells. If CSF analysis shows polymorphonuclear
leukocyte dominant marked pleocytosis (CSF white blood
cells .1000 3 103/mL), it is almost always bacterial menin-
gitis (in the current study, none of the patients with TBM
showed CSF white blood cells .1000 3 103/mL with poly-
morphonuclear leukocyte dominant pattern). After bacterial
meningitis is ruled out, a cutoff value of 9.0 U/L to identify
TBM from viral and cryptococcal meningitis is shown in
Table 4 with a high sensitivity (90.1%) and specificity
(90.0%) in the current study.
As for cryptococcal meningitis, only a small amount of
data is available. In the study performed by Rohani et al,26 the
mean of ADA in 14 patients with TBM (16.33 6 5.66 U/L) was
significantly higher than that in 12 patients with cryptococcal
meningitis (7.26 6 3.8 U/L). Although Martinez et al27 reported
that CSF ADA activity was not significantly different between
the group with TBM and the group with cryptococcal menin-
gitis, in the current study, we calculated a cutoff point of
9.8 U/L for the differential diagnosis of TBM and cryptococcal
meningitis.
We propose that the diagnostic value of CSF ADA
in different countries depends on the different incidence of
other non-TBM infectious meningitis. For example, in Brazil,
tuberculosis accounts for 3% of meningitis cases in the
emergency room, and the main alternative diagnosis is viral
meningitis.24 Thus, ADA could be useful in countries where the
incidence of viral meningitis is relatively high. Nevertheless, in
some developing Asian countries, TBM accounts for more than
20% of all meningitis cases,1 and the main alternative diagnosis
is bacterial meningitis, which decreases the usefulness of ADA
in those countries.
The current study also demonstrated dynamic changes of
CSF ADA activity in patients with TBM after antitubercular
therapy. We found that CSF ADA activity decreased gradually
after antitubercular therapy, indicating that dynamic changes of
CSF ADA activity may be used for monitoring response to
therapy. A decrease in CSF ADA levels may indicate a good
response to treatment, whereas resistance to drugs may be
suspected if ADA activity remains unchanged or increases. Our
120
data showed that all of the patients had a decline in CSF ADA
activity except 3 patients for whom ADA levels increased
slightly after antitubercular therapy. Two of them were
confirmed as drug-resistant pulmonary tuberculosis compli-
cated with TBM by tubercle bacillus drug sensitivity test
approximately 2 months later. Another one showed continuous
increases in CSF ADA, which coincided with the development
of complications and indicated a poor response to treatment.
In addition, the current study has some limitations. A
portion of the TBM cases were diagnosed clinically, which is
controversial and might lead to biased results. For example,
viral meningitis is often self-limiting with or without treatment
and may show similar laboratory findings in CSF as those of
TBM.
In conclusion, the determination of ADA with a cutoff
value of 9.5 U/L in CSF can be a useful aid for the differential
diagnosis of TBM in China. The incidence of other infectious
diseases of CNS, such as bacterial and viral meningitis, has an
important influence on the test performance characteristics of
CSF ADA activity. Considering its relatively low sensitivity
and specificity in distinguishing from bacterial meningitis, CSF
ADA should be interpreted judiciously in light of the patient’s
clinical condition and other laboratory tests. In regions where
the incidence of bacterial meningitis is low or after bacterial
meningitis has been ruled out, we can use a cutoff value of 9.5
U/L to aid the diagnosis of TBM. It may help the physicians to
initiate early diagnostic antitubercular therapy in suspected
patients before receiving confirmatory diagnostic evidence.
ACKNOWLEDGMENTS
We acknowledge the outstanding contributions from the
technicians in Shanghai Key Laboratory of Mycobacteria Tu-
berculosis at Shanghai Pulmonary Hospital, and we are grateful
to all study participants. We also thank Medjaden Bioscience
Limited for assisting in the preparation of this manuscript.
REFERENCES
1. Chotmongkol V, Teerajetgul Y, Yodwut C. Cerebrospinal fluid aden-
osine deaminase activity for the diagnosis of tuberculous meningitis in
adults. Southeast Asian J Trop Med Public Health 2006;37:948–52.
2. Kashyap RS, Kainthla RP, Mudaliar AV, et al. Cerebrospinal fluid
adenosine deaminase activity: a complimentary tool in the early diag-
nosis of tuberculous meningitis. Cerebrospinal Fluid Res 2006;3:5.
3. Thwaites GE, Chau TT, Farrar JJ. Improving the bacteriological
diagnosis of tuberculous meningitis. J Clin Microbiol 2004;42:378–9.
4. Thwaites G, Chau TT, Mai NT, et al. Tuberculous meningitis.
J Neurol Neurosurg Psychiatry 2000;68:289–99.
5. Mathai A, Radhakrishnan VV, Sarada C, et al. Detection of heat
stable mycobacterial antigen in cerebrospinal fluid by Dot-Immunobind-
ing assay. Neurol India 2003;51:52–4.
6. Lee YC, Rogers JT, Rodriguez RM, et al. Adenosine deaminase
levels in nontuberculous lymphocytic pleural effusions. Chest 2001;
120:356–61.
7. Erel O, Kocyigit A, Gurel MS, et al. Adenosine deaminase activities
in sera, lymphocytes and granulocytes in patients with cutaneous leish-
maniasis. Mem Inst Oswaldo Cruz 1998;93:491–4.
8. Andreasyan NA, Hairapetyan HL, Sargisova YG, et al. ADA2 iso-
form of adenosine deaminase from pleural fluid. FEBS Lett 2005;579:
643–7.
9. Moghtaderi A, Niazi A, Alavi-Naini R, et al. Comparative analysis of
cerebrospinal fluid adenosine deaminase in tuberculous and non-tuberculous
meningitis. Clin Neurol Neurosurg 2010;112:459–62.
Volume 344, Number 2, August 2012

10. Xu HB, Jiang RH, Li L, et al. Diagnostic value of adenosine deam-
inase in cerebrospinal fluid for tuberculous meningitis: a meta-analysis.
Int J Tuberc Lung Dis 2010;14:1382–7.
11. Choi SH, Kim YS, Bae IG, et al. The possible role of cerebrospinal
fluid adenosine deaminase activity in the diagnosis of tuberculous men-
ingitis in adults. Clin Neurol Neurosurg 2002;104:10–5.
12. Fujiwara PI, Clevenbergh P, Dlodlo RA. Management of adults
living with HIV/AIDS in low-income, high-burden settings, with spe-
cial reference to persons with tuberculosis. Int J Tuberc Lung Dis 2005;
9:946–58.
13. Zhang LF, Liu XQ, Zuo LY, et al. Longitudinal observation of
an interferon gamma-released assay (T-SPOT.TB) for Mycobacterium
tuberculosis infection in AIDS patients on highly active antiretroviral
therapy. Chin Med J (Engl) 2010;123:1117–21.
14. Kashyap RS, Kainthla RP, Biswas SK, et al. Rapid diagnosis of
tuberculous meningitis using the Simple Dot ELISA method. Med Sci
Monit 2003;9:MT123–6.
15. Kashyap RS, Agarwal NP, Chandak NH, et al. The application of the
Mancini technique as a diagnostic test in the CSF of tuberculous men-
ingitis patients. Med Sci Monit 2002;8:MT95–8.
16. Rafi W, Venkataswamy MM, Ravi V, et al. Rapid diagnosis of tuber-
culous meningitis: a comparative evaluation of in-house PCR assays
involving three mycobacterial DNA sequences, IS6110, MPB-64 and
65 kDa antigen. J Neurol Sci 2007;252:163–8.
17. Kashyap RS, Ramteke SP, Deshpande PS, et al. Comparison of an
adenosine deaminase assay with ELISA for the diagnosis of tuberculous
meningitis infection. Med Sci Monit 2007;13:BR200–4.
18. Kashyap RS, Kainthla RP, Satpute RM, et al. Demonstration of IgG
antibodies to 30 Kd protein antigen in CSF for diagnosis of tuberculous
Ó 2012 Lippincott Williams & Wilkins
Diagnosis of Tuberculous and Nontuberculous Meningitis
meningitis by antibody-capturing ELISA. Neurol India 2004;52:359–
62.
19. Chen ML, Yu WC, Lam CW, et al. Diagnostic value of pleural fluid
adenosine deaminase activity in tuberculous pleurisy. Clin Chim Acta
2004;341:101–7.
20. Martinez-Vazquez JM, Ribera E, Ocana I, et al. Adenosine deami-
nase activity in tuberculous pericarditis. Thorax 1986;41:888–9.
21. Tuon FF, Litvoc MN, Lopes MI. Adenosine deaminase and tubercu-
lous pericarditis—a systematic review with meta-analysis. Acta Trop
2006;99:67–74.
22. Voigt MD, Kalvaria I, Trey C, et al. Diagnostic value of
ascites adenosine deaminase in tuberculous peritonitis. Lancet 1989;1:
751–4.
23. Rana SV, Chacko F, Lal V, et al. To compare CSF adenosine de-
aminase levels and CSF-PCR for tuberculous meningitis. Clin Neurol
Neurosurg 2010;112:424–30.
24. Tuon FF, Higashino HR, Lopes MI, et al. Adenosine deaminase and
tuberculous meningitis—a systematic review with meta-analysis. Scand
J Infect Dis 2010;42:198–207.
25. Thwaites GE, Chau TT, Stepniewska K, et al. Diagnosis of adult
tuberculous meningitis by use of clinical and laboratory features. Lancet
2002;360:1287–92.
26. Rohani MY, Cheong YM, Rani JM. The use of adenosine deaminase
activity as a biochemical marker for the diagnosis of tuberculous men-
ingitis. Malays J Pathol 1995;17:67–71.
27. Martinez E, Domingo P, Ris J, et al. Cerebrospinal fluid adenosine
deaminase levels in a patient with cryptococcal meningitis. Clin Infect
Dis 1992;15:1061–2.
121