Tube&e (1990) 71,121-126 0 Longman Group UK Ltd 1990
Adenosine deaminase (ADA) in peritoneal tuberculosis: diagnostic value in ascitic fluid and serum D. K. BHARGAVA*, M. GUPTA, S. NIJHAWAN, S. DASARATHY and A. K. S. KUSHWAHA Department
of Gastroenterology,
All India Institute of Medical
Sciences, New Delhi
Summsry-Simultaneous determination of ascitic fluid and serum adenosine deaminase (ADA) activity was evaluated as a diagnostic aid in peritoneal tuberculosis. The ascitea was due to peritoneal tuberculosis (group 1), cirrhosis of the liver (group 2), cirrhosis of the liver with spontaneous bacterial peritonitis (group 3), peritoneal malignancy (group 4), BuddChiari Syndrome (group 5) and miscellaneous conditions (group 6). Serum from patients of pulmonary tuberculosis and healthy volunteers was analysed for enzyme activity. In patients with peritoneal tuberculosis the ascitic fluid and serum ADA activity was significantly higher than for the other groups (P < 0.001). Levels above 36 u/l in ascitic fluid and above 54 u/l in the serum suggest tuberculosis. The ascitic fluid/serum ADA ratio was also higher in patients with peritoneal tuberculosis than with other causes of ascites (P < 0.01). A ratio of more than 0.994 was suggestive of tuberculosis.
Introduction Peritoneal tuberculosis is still a very important cause of ascites in India and other developing countries [l]. Its specific diagnosis requires histological or bacteriological confirmation and this is obtained either by tissue biopsied under direct vision (laparoscopic) or with a Cope’s needle. The needle biopsy yield is 30-50% and this reaches to 75-100% on laparoscopy [l-3]. Laparoscopy may not be available at all medical institutions. A ’ Correspondence to: Dr D. K. Bhargava, Additional Professor of Gastroenterology, All India Institute of Medical Sciences, New Delhi-110029, India.
positive culture and species identification may be possible in only 1040% 14-61 of patients and requires several weeks. Previous studies have confirmed the diagnostic value of adenosine deaminase activity in effusions due to peritoneal [7-91 meningeal [lo] pleural [ 111 and pericardial [12] tuberculosis. It has also been suggested that determination of pleural fluid/plasma ADA ratio improves the accuracy of the test [13]. This prospective study evaluates the diagnostic value of adenosine deaminase activity (ADA) in serum and ascitic fluid of patients with ascites due to various causes, including peritoneal tuberculo-
121
122
BHARGAVA
sis. The value of ascitic fluid/serum ADA ratio as an aid in the diagnosis is also determined. Materials and Methods The ADA activity was assessed in ascitic fluid of 87 patients with ascites due to various causes. We simultaneously evaluated enzyme activity in the serum of 43 of the 87 patients. Thirteen patients with sputum smear positive pulmonary tuberculosis and 13 healthy volunteers were also included as positive and negative controls for enzyme activity in serum. According to the final clinical diagnosis achieved by standard methods, patients were subdivided into five groups. Group I This included 17 patients with peritoneal tuberculosis. The tuberculin test was positive in 13 of 17 patients. Chest X-ray was normal in 15 patients and one each had pleural effusion and apical infiltration in the lung. The ascitic fluid was exudative and contained predominantly lymphocytes (> 250 cell&mm) in all patients. Laparoscopy was performed on all patients. This confirmed the diagnosis of tuberculosis. Histopathologic examination revealed caseating granulomas. Mycobacterium tuberculosis was isolated from four of the 17 patients.
AND
OTHERS
Group 6 This group consisted of one patient each with nephrotic syndrome, non-chirrhotic portal fibrosis and congestive heart failure. The adenosine deaminase activity was determined in ascites and serum by the- calorimetric method of Giusti [14]. This is based on the measurement of ammonia produced when adenosine deaminase acts on an excess of adenosine. Simultaneous ascitic fluid and serum ADA activity was determined to calculate ascitic fluid/serum ADA ratio as an aid in the diagnosis. Statistical methods
The results were compared by the Kruskal Wallis one way non-parametric analysis of variance [ 151. Results Adenosine
deaminase activity in asciticfluid
The ascitic fluid adenosine deaminase activity in different groups of patients is shown in Table 1. Individual results are given in Figure 1. The miscellaneous group consisted of one patient each with nephrotic syndrome, non-cirrhotic portal fibrosis and congestive heart failure and the corresponding ADA levels were 10.7,24.6 and 22.5 u/l respectively. Statistical analysis revealed that all
Group 2 This group included 31 patients with cirrhosis of the liver with transudative ascites. Cirrhosis was confirmed by clinical, biochemical and histological criteria.
ADA (U/I) 220 a
.:’ .
Group 3 This consisted of seven patients with cirrhosis of the liver with spontaneous bacterial peritonitis. They were diagnosed on the basis of ascitic fluid polymorphonuclear cell count (> 250 cells/mm3) and an arterial-ascitic fluid pH gradient (> 0.1). Group
120 -
4 This group included
22 patients with malignant deposits in the peritoneum who presented with exudative ascites. On cytological examination of ascitic fluid 18 patients disclosed presence of malignant cells. Laparoscopic examination revealed metastatic lesions in the remaining four patients. Histologic examination confirmed metastatic adenocarcinoma in these patients.
’
loo-
.* 80 -
60 -
.
LO- .
.. c
20-
Group 5 This included seven patients of BuddChiari Syndrome with exudative ascites. All were diagnosed on clinical examination, laparoscopy, histology and angiography.
O-
Fig. 1
Puifoncal TB17
Ascitic fluid ADA
non Tubucubus Ascitrs70 activity (U/I)
in individual
patients.
ADENOSINE
Table 1
DEAMINASE
123
IN TUBERCULOSIS
Adenosine deaminase activity in ascitic fluid Patients Adhosine denminaret unitsllilre No
Diagnosis 1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Peritoneal malignancy 5. Budd-Chiari syndrome 6. Miscellaneous
17 31 7
141.03 + 61.5 10.0 +- 7.8 13.6 + 11.1
22 7 3
19.7 f 13.5 14.8 + 7.2 19.2 + 6.1
(P > 0.05). Hence non-tuberculous patients and healthy volunteers were combined to form the control group. Taking the 95th percentile value (Table 2) serum level of adenosine deaminase greater than 54 u/l had a specificity of 97.6% and sensitivity of 81.5%. However, with lower values of adenosine deaminase activity (85th and 90th percentile) the sensitivity increases but specificity decreases. After analysis, adenosine deaminase level of 54 u/l was taken as the cut-off point and values above this would be regarded as abnormal.
* Mean values and standard deviations.
Ascitic FluidlSerum
groups other than those with peritoneal tuberculosis had significantly lower ADA levels when compared with those in peritoneal tuberculosis group (P < 0.001). Differences among these other groups were not significant (P > 0.05) and all five were combined to form a control group. The 95th percentile value of the control group (Table 2) as cut-off level in this series had a sensitivity and a specificity of 100% and 97% respectively, with 36 u/l and higher values regarded as abnormal. Adenosine
deaminase activity in serum
ADA activity in serum was determined in 43 patients with ascites (Table 3 & Fig. 2). It was found that serum adenosine deaminase activity was significantly (p < 0.001) higher in patients with peritoneal and pulmonary tuberculosis than in non-tuberculous patients and healthy controls. Serum levels of adenosine deaminase were similar in pulmonary and peritoneal tuberculosis (P > 0.05). Serum levels of non-tuberculous patients were also similar to healthy controls
adenosine deaminase ratio
The ascitic fluid/serum ADA ratios obtained in different groups of patients are shown in Table 4 & Fig. 3. This was significantly higher in patients with peritoneal tuberculosis than those with other causes of ascites (P < 0.01). This ratio was also similar in the patients with different non-tuberculous causes of ascites (P > 0.05). The sensitivity and specificity of these values in detecting peritoneal tuberculosis are shown in Table 2. With the 85th percentile value, a ratio greater than 0.984 had a specificity of 86.2% and sensitivity of 75.6%. Thus a ratio of 0.984 was taken as cut-off point and values above this would be regarded as clearly abnormal. Correlation between Ascitic @id-A DA, Serum ADA and ratio
With the three methods of estimating the usefulness of ADA (exudative fluid, serum and the ratio
Table 3
Adenosine
deaminase
Table 2
Sensitivity and specificity of suggested upper limits of normal adenosine deaminase activity in the
detection
ADA activity Ull Sensitivity (%I Specificity (%I
of tuberculosis *Asciticfluid ADA 29 32 36
*Serum ADA 43
48
54
100 100 100 95.6 85.2 81.5 86.%91.3
97.1 94.9 92.9 97.6
fAscitic jluidl serum ratio 0.9740.984 1.26
85.7 75.6 57.1 96
86.2 87.7
*ADA levels at 85th, 90th and 95th percentile respectively. tFor ratio cut off values are 8Oth, 85th and 90th percentiles respectively.
Diagnosb 1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Peritoneal malignancy 5. Budd-Chiari syndrome 6. Nephrotic syndrome CONTROLS Pulmonary tuberculosis Healthy volunteers
activity in serum
Patients No
Adenosine deaminase” UII
14 17 3
100.5 + 50.7 27.6 + 21.1 24.4 + 26.5
4 4 1
25.0 + 10.9 30.3 zk 14.1 43.6
13
78.12 + 17.0
13
20.2 zk 13.1
*Mean values and standard deviations.
124
BHARGAVA ADA ( U/l)
Table 4 Ascitic deaminase
fluid
ratio
OTHERS
of adenosine
activity
Diagnosis
1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Budd-Chiari syndrome 5. Peritoneal malignancy 6. Nephrotic syndrome
.
to serum
AND
Patients No
Ascitic j7uidlserum adenosine deuminuse ratio*
14 17 3
1.54 f 0.64 0.666 f 0.59 0.65 + 1.4
4
0.59 + 0.29
4
1.8 zk 2.0 0.244
1
*Mean values and standard deviations.
Tubereulos~s
Non Tubcrculous
27
DlsIxlscs L2 Fig. 2 Serum ADA activity (u/l) in individual patients with tuberculous diseases and non-tuberculous diseases and with healthy volunteers.
ADA
RATIO
... .
of ADA in ascitic fluid/serum), a comparative study of the sensitivity and specificity show that estimation of ADA in ascitic fluid was better than serum values. The least useful was that of the ratio of ascitic fluid/serum (Table 2). Correlation with Lymphocyte count Lymphocyte counts were performed in the peripheral blood, and ascitic fluid (Table 5). These showed significantly higher lymphocyte counts in both ascitic fluid and peripheral blood in patients of tuberculosis when compared to those with nontuberculous diseases (p < 0.05). However, no correlation was found between the absolute lymphocyte numbers and ADA activity (r = +0.21). Discussion
0’
Perileneal 1 BlL
Non Tubu~ubu~ AuttCI
29
3 Ascitic fluid/serum ADA ratio in patients with peritoneal tuberculosis and non-tubercuious causes of ascites. Fig.
Adenosine deaminase (ADA) is an aminohydrolase that catalyzes the deamination of adenosine to inosine [16]. Its biological activity is related to proliferation and differentiation of lymphocytes. Elevated enzyme levels have been detected in serum of patients having diseases in which cellular immunity is stimulated, e.g. typhoid fever [17], infectious mononucleosis, mediterranean spotted fever [18] and acute T-cell leukaemia [19]. However, in recent years, estimation of ADA in pleural, pericardial, meningeal and peritoneal effusions has gained importance in the diagnosis of tuberculosis. Studies have also shown the value of serosal effusion/plasma ADA ratio in the diagnosis of tuberculous pleural effusion [ 131 but none of the studies have evaluated the diagnostic value
125
ADENOSINE DEAMINASE IN TUBERCULOSIS
Table 5
Lymphocyte numbers* in a&tic fluid and peripheral blood.
Diagnosis
Patients
1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Budd Chiari syndrome 5. Peritoneal malignancy 6. Miscellaneous
Lymphocytes Peripheral blood Axitic fluid
17
366 + 365
2581 + 841
31
173 + 249
2017 + 1223
7
53 + 152
1826 f 1917
7
197 f 255
1785 + 517
22
62 + 55
1878 I!I 968
3
68 + 21
1979 If: 719
*Mean values and standard deviations.
of serum ADA or the ratio of ascitic fluid/plasma ADA. Our results suggest that the determination of ADA activity in ascitic fluid has a high sensitivity and specificity for the diagnosis of active peritoneal tuberculosis. A value of > 36 u/l in the ascitic fluid has been considered to be optimal in differentiating tuberculosis from other aetiologies of ascites. Simultaneous assessment of the serum ADA level showed an elevated level in patients of peritoneal and pulmonary tuberculosis. There was no statistically significant difference in the serum levels between peritoneal and pulmonary tuberculosis. However, the ADA values were significantly less in patients with ascites of non-tuberculous etiology and healthy volunteers. To attain maximum accuracy, 54 u/l was considered the cut-off point and values above this would clearly denote tuberculosis. A comparative study of the sensitivity and specificity showed that the ADA determination in ascitic fluid was better than serum. There was no correlation between ascitic fluid and serum ADA levels in patients of peritoneal tuberculosis. This may be due to the differential elevation
of T-lymphocytes in ascitic fluid and peripheral blood, as has been suggested in patients with pleural effusion [20]. Another study, however, failed to correlate the high percentage of Tlymphocytes in tuberculous pleural effusion. with the level of ADA [ll]. Even in our study no correlation was found between the absolute lymphocyte numbers and ADA activity. In our laboratory, the ratio of ascitic fluid/serum ADA was inferior to absolute serum or ascitic fluid levels of ADA in the prediction or exclusion of tuberculosis. This differs from the previous report that showed an improved accuracy of the ratio over the estimation of ADA in pleural fluid alone [13]. This is in agreement with another report that both serosal effusion and the ratio are useful in the diagnosis of tuberculous meningitis but the ratio had no significant advantage over estimation of ADA in CSF alone [21]. Thus the present study suggests that detection of adenosine deaminase activity is a useful test for the diagnosis of peritoneal tuberculosis and in differentiating it from other causes of ascites. Ascitic fluid/serum ADA ratio, and the absolute serum activity, were significantly raised in peritoneal tuberculosis in contrast to non-tuberculous patients. However, ascitic fluid/serum ADA ratio and absolute ADA level in serum do not appear to hold any marked advantages over conventional ascitic fluid criteria in the diagnosis of peritoneal tuberculosis.
This study was supportedby a grant from the Indian Council of Medical Research. The authors would Iike to thank Professor BN Tandon, Dr KR Sundaram and Dr MG Karmarkar for assistance.
References 1. Bhargava DK, Saraswat VA. Laparoscopy in patients with ascites (abstract). J Assoc Phys India 1988,36: 38. 2. Wolfe JHN, Behn AR, Jackson BT. TubercuIous peritonitis and role of diagnostic laparoscopy. Lancer 1979; 1: 852-853. 3. Jorge AD. Peritoneal tuberculosis. Endoscopy 1984; 16: 10-12. 4. Sochocky S. Tuberculous peritonitis: a review of 100 cases. Am Rev Respir Dis 1%7; 95: 398-401. 5. Dineen P, Homan WP, Grafe WR. TubercuIous peritonitis: 43 years experience in diagnosis and treatment. Am I Surg 1976; 1984: 717-722. 6. Sherman S, Rohwedder JJ, Ravi Kri&nan KP, Weg JG. Tuberculous enteritis and peritonitis. Report of 36 general hospital cases. Arch Intern Med 1980; 140: 506-508.
126 7. Martinez-Vazquez
8.
9.
10.
11.
12.
13.
14.
JM, Ocana I, Ribera E, Sequra RM, Pascual C. Adenosine deaminase activity in the diagnosis of tuberculous peritonitis. Cur 1986; 27: 1049-1053. Bhargava DK, Nijhawan S, Gupta M. Adenosine deaminase activity in the diagnosis of peritoneal tuberculosis. (letter). Lancet 1989; 1: 1261. Voigte MD, Kalvaria I, Trey C, Lombard BCL, Kirsh RE. Diagnostic value of ascitic adenosine deaminase in tuberculous peritonitis. Lancer 1989; 1: 751-753 Ribera E, Martinez-Vazquez JM, Ocana I, Segura RM, Pascual C. Activity of adenosine deaminase in cerebrospinal fluid for the diagnosis and follow up of tuberculous meningitis in adults. J Infect Di.s 1987; 155: 603-607. Ocana I, Martinez-vazquez JM, Segura RM, Fernandezde-Sevilla T, Capdevila JA: Adenosine deaminase in pleural fluids. Test for diagnosis of tuberculous pleural effusion. Chest 1983; 84: 51-53. Martinez-vazquez JM, Ribera E, Ocana I, Segura RM, Serrat M, Sagrista J. Adenosine deaminase activity in tuberculous pericarditis. Thorax 1986; 41: 88-89. Martiz FJ, Visser HS, Malan C et al. Nuwe benadering tot die diagnose van pleurale effusies. S Afr Med J 1981; 60: 217-218. Giusti G. Adenosine deaminase. In: Bergmeyer HV, ed.
BHAKGAVA
15. 16.
17.
18. 19.
AND OTHEKS
Methods of enzymatic analysis. Academic Press. New York 1974: 1092-1099. Colquhoun D: Lectures on Riostatisrics. Clarendon Press, Oxford, 1971. Van der Weyden MB, Kelley WN. Human adenosinc deaminase: distribution and properties. J Riol (‘hem 1076: 251: 54485456. Galanti B, Nardiello S, Russo M, Fiorentino F. Increased lymphocyte adenosine deaminase in typhoid fever. Stand J Infect Dis 1981; 13: 47-50. Piras MA, Garis C, Andreoni G. Immunological studies in Mediterranean spotted fever (letter). Lancet 1982: 1: 1249. Moriskaki T, Fuji H, Miwas G. Adenosine deaminase (ADA) in leukaemia: Clinical value of plasma ADA activity and characterisation of leukemic cell ADA. Am J Hematoll988;
19: 37-45.
20. Pattersson T, Klockars M, Hellstrom PE, Riska H, Wangel A. T and B lymphocytes in pleural effusions. Chest 1978; 73: 49-51. 21. Donald PR, Malan C, Walt AVD, Schueman JF. The simultaneous determination of cerebrospinal fluid and plasma adenosine deaminase activity as a diagnostic aid in tuberculous meningitis. S Afr Med J 1986; 69: 505-7.
Adenosine deaminase (ADA) in peritoneal tuberculosis: diagnostic value in ascitic fluid and serum D. K. BHARGAVA*, M. GUPTA, S. NIJHAWAN, S. DASARATHY and A. K. S. KUSHWAHA Department
of Gastroenterology,
All India Institute of Medical
Sciences, New Delhi
Summsry-Simultaneous determination of ascitic fluid and serum adenosine deaminase (ADA) activity was evaluated as a diagnostic aid in peritoneal tuberculosis. The ascitea was due to peritoneal tuberculosis (group 1), cirrhosis of the liver (group 2), cirrhosis of the liver with spontaneous bacterial peritonitis (group 3), peritoneal malignancy (group 4), BuddChiari Syndrome (group 5) and miscellaneous conditions (group 6). Serum from patients of pulmonary tuberculosis and healthy volunteers was analysed for enzyme activity. In patients with peritoneal tuberculosis the ascitic fluid and serum ADA activity was significantly higher than for the other groups (P < 0.001). Levels above 36 u/l in ascitic fluid and above 54 u/l in the serum suggest tuberculosis. The ascitic fluid/serum ADA ratio was also higher in patients with peritoneal tuberculosis than with other causes of ascites (P < 0.01). A ratio of more than 0.994 was suggestive of tuberculosis.
Introduction Peritoneal tuberculosis is still a very important cause of ascites in India and other developing countries [l]. Its specific diagnosis requires histological or bacteriological confirmation and this is obtained either by tissue biopsied under direct vision (laparoscopic) or with a Cope’s needle. The needle biopsy yield is 30-50% and this reaches to 75-100% on laparoscopy [l-3]. Laparoscopy may not be available at all medical institutions. A ’ Correspondence to: Dr D. K. Bhargava, Additional Professor of Gastroenterology, All India Institute of Medical Sciences, New Delhi-110029, India.
positive culture and species identification may be possible in only 1040% 14-61 of patients and requires several weeks. Previous studies have confirmed the diagnostic value of adenosine deaminase activity in effusions due to peritoneal [7-91 meningeal [lo] pleural [ 111 and pericardial [12] tuberculosis. It has also been suggested that determination of pleural fluid/plasma ADA ratio improves the accuracy of the test [13]. This prospective study evaluates the diagnostic value of adenosine deaminase activity (ADA) in serum and ascitic fluid of patients with ascites due to various causes, including peritoneal tuberculo-
121
122
BHARGAVA
sis. The value of ascitic fluid/serum ADA ratio as an aid in the diagnosis is also determined. Materials and Methods The ADA activity was assessed in ascitic fluid of 87 patients with ascites due to various causes. We simultaneously evaluated enzyme activity in the serum of 43 of the 87 patients. Thirteen patients with sputum smear positive pulmonary tuberculosis and 13 healthy volunteers were also included as positive and negative controls for enzyme activity in serum. According to the final clinical diagnosis achieved by standard methods, patients were subdivided into five groups. Group I This included 17 patients with peritoneal tuberculosis. The tuberculin test was positive in 13 of 17 patients. Chest X-ray was normal in 15 patients and one each had pleural effusion and apical infiltration in the lung. The ascitic fluid was exudative and contained predominantly lymphocytes (> 250 cell&mm) in all patients. Laparoscopy was performed on all patients. This confirmed the diagnosis of tuberculosis. Histopathologic examination revealed caseating granulomas. Mycobacterium tuberculosis was isolated from four of the 17 patients.
AND
OTHERS
Group 6 This group consisted of one patient each with nephrotic syndrome, non-chirrhotic portal fibrosis and congestive heart failure. The adenosine deaminase activity was determined in ascites and serum by the- calorimetric method of Giusti [14]. This is based on the measurement of ammonia produced when adenosine deaminase acts on an excess of adenosine. Simultaneous ascitic fluid and serum ADA activity was determined to calculate ascitic fluid/serum ADA ratio as an aid in the diagnosis. Statistical methods
The results were compared by the Kruskal Wallis one way non-parametric analysis of variance [ 151. Results Adenosine
deaminase activity in asciticfluid
The ascitic fluid adenosine deaminase activity in different groups of patients is shown in Table 1. Individual results are given in Figure 1. The miscellaneous group consisted of one patient each with nephrotic syndrome, non-cirrhotic portal fibrosis and congestive heart failure and the corresponding ADA levels were 10.7,24.6 and 22.5 u/l respectively. Statistical analysis revealed that all
Group 2 This group included 31 patients with cirrhosis of the liver with transudative ascites. Cirrhosis was confirmed by clinical, biochemical and histological criteria.
ADA (U/I) 220 a
.:’ .
Group 3 This consisted of seven patients with cirrhosis of the liver with spontaneous bacterial peritonitis. They were diagnosed on the basis of ascitic fluid polymorphonuclear cell count (> 250 cells/mm3) and an arterial-ascitic fluid pH gradient (> 0.1). Group
120 -
4 This group included
22 patients with malignant deposits in the peritoneum who presented with exudative ascites. On cytological examination of ascitic fluid 18 patients disclosed presence of malignant cells. Laparoscopic examination revealed metastatic lesions in the remaining four patients. Histologic examination confirmed metastatic adenocarcinoma in these patients.
’
loo-
.* 80 -
60 -
.
LO- .
.. c
20-
Group 5 This included seven patients of BuddChiari Syndrome with exudative ascites. All were diagnosed on clinical examination, laparoscopy, histology and angiography.
O-
Fig. 1
Puifoncal TB17
Ascitic fluid ADA
non Tubucubus Ascitrs70 activity (U/I)
in individual
patients.
ADENOSINE
Table 1
DEAMINASE
123
IN TUBERCULOSIS
Adenosine deaminase activity in ascitic fluid Patients Adhosine denminaret unitsllilre No
Diagnosis 1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Peritoneal malignancy 5. Budd-Chiari syndrome 6. Miscellaneous
17 31 7
141.03 + 61.5 10.0 +- 7.8 13.6 + 11.1
22 7 3
19.7 f 13.5 14.8 + 7.2 19.2 + 6.1
(P > 0.05). Hence non-tuberculous patients and healthy volunteers were combined to form the control group. Taking the 95th percentile value (Table 2) serum level of adenosine deaminase greater than 54 u/l had a specificity of 97.6% and sensitivity of 81.5%. However, with lower values of adenosine deaminase activity (85th and 90th percentile) the sensitivity increases but specificity decreases. After analysis, adenosine deaminase level of 54 u/l was taken as the cut-off point and values above this would be regarded as abnormal.
* Mean values and standard deviations.
Ascitic FluidlSerum
groups other than those with peritoneal tuberculosis had significantly lower ADA levels when compared with those in peritoneal tuberculosis group (P < 0.001). Differences among these other groups were not significant (P > 0.05) and all five were combined to form a control group. The 95th percentile value of the control group (Table 2) as cut-off level in this series had a sensitivity and a specificity of 100% and 97% respectively, with 36 u/l and higher values regarded as abnormal. Adenosine
deaminase activity in serum
ADA activity in serum was determined in 43 patients with ascites (Table 3 & Fig. 2). It was found that serum adenosine deaminase activity was significantly (p < 0.001) higher in patients with peritoneal and pulmonary tuberculosis than in non-tuberculous patients and healthy controls. Serum levels of adenosine deaminase were similar in pulmonary and peritoneal tuberculosis (P > 0.05). Serum levels of non-tuberculous patients were also similar to healthy controls
adenosine deaminase ratio
The ascitic fluid/serum ADA ratios obtained in different groups of patients are shown in Table 4 & Fig. 3. This was significantly higher in patients with peritoneal tuberculosis than those with other causes of ascites (P < 0.01). This ratio was also similar in the patients with different non-tuberculous causes of ascites (P > 0.05). The sensitivity and specificity of these values in detecting peritoneal tuberculosis are shown in Table 2. With the 85th percentile value, a ratio greater than 0.984 had a specificity of 86.2% and sensitivity of 75.6%. Thus a ratio of 0.984 was taken as cut-off point and values above this would be regarded as clearly abnormal. Correlation between Ascitic @id-A DA, Serum ADA and ratio
With the three methods of estimating the usefulness of ADA (exudative fluid, serum and the ratio
Table 3
Adenosine
deaminase
Table 2
Sensitivity and specificity of suggested upper limits of normal adenosine deaminase activity in the
detection
ADA activity Ull Sensitivity (%I Specificity (%I
of tuberculosis *Asciticfluid ADA 29 32 36
*Serum ADA 43
48
54
100 100 100 95.6 85.2 81.5 86.%91.3
97.1 94.9 92.9 97.6
fAscitic jluidl serum ratio 0.9740.984 1.26
85.7 75.6 57.1 96
86.2 87.7
*ADA levels at 85th, 90th and 95th percentile respectively. tFor ratio cut off values are 8Oth, 85th and 90th percentiles respectively.
Diagnosb 1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Peritoneal malignancy 5. Budd-Chiari syndrome 6. Nephrotic syndrome CONTROLS Pulmonary tuberculosis Healthy volunteers
activity in serum
Patients No
Adenosine deaminase” UII
14 17 3
100.5 + 50.7 27.6 + 21.1 24.4 + 26.5
4 4 1
25.0 + 10.9 30.3 zk 14.1 43.6
13
78.12 + 17.0
13
20.2 zk 13.1
*Mean values and standard deviations.
124
BHARGAVA ADA ( U/l)
Table 4 Ascitic deaminase
fluid
ratio
OTHERS
of adenosine
activity
Diagnosis
1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Budd-Chiari syndrome 5. Peritoneal malignancy 6. Nephrotic syndrome
.
to serum
AND
Patients No
Ascitic j7uidlserum adenosine deuminuse ratio*
14 17 3
1.54 f 0.64 0.666 f 0.59 0.65 + 1.4
4
0.59 + 0.29
4
1.8 zk 2.0 0.244
1
*Mean values and standard deviations.
Tubereulos~s
Non Tubcrculous
27
DlsIxlscs L2 Fig. 2 Serum ADA activity (u/l) in individual patients with tuberculous diseases and non-tuberculous diseases and with healthy volunteers.
ADA
RATIO
... .
of ADA in ascitic fluid/serum), a comparative study of the sensitivity and specificity show that estimation of ADA in ascitic fluid was better than serum values. The least useful was that of the ratio of ascitic fluid/serum (Table 2). Correlation with Lymphocyte count Lymphocyte counts were performed in the peripheral blood, and ascitic fluid (Table 5). These showed significantly higher lymphocyte counts in both ascitic fluid and peripheral blood in patients of tuberculosis when compared to those with nontuberculous diseases (p < 0.05). However, no correlation was found between the absolute lymphocyte numbers and ADA activity (r = +0.21). Discussion
0’
Perileneal 1 BlL
Non Tubu~ubu~ AuttCI
29
3 Ascitic fluid/serum ADA ratio in patients with peritoneal tuberculosis and non-tubercuious causes of ascites. Fig.
Adenosine deaminase (ADA) is an aminohydrolase that catalyzes the deamination of adenosine to inosine [16]. Its biological activity is related to proliferation and differentiation of lymphocytes. Elevated enzyme levels have been detected in serum of patients having diseases in which cellular immunity is stimulated, e.g. typhoid fever [17], infectious mononucleosis, mediterranean spotted fever [18] and acute T-cell leukaemia [19]. However, in recent years, estimation of ADA in pleural, pericardial, meningeal and peritoneal effusions has gained importance in the diagnosis of tuberculosis. Studies have also shown the value of serosal effusion/plasma ADA ratio in the diagnosis of tuberculous pleural effusion [ 131 but none of the studies have evaluated the diagnostic value
125
ADENOSINE DEAMINASE IN TUBERCULOSIS
Table 5
Lymphocyte numbers* in a&tic fluid and peripheral blood.
Diagnosis
Patients
1. Peritoneal tuberculosis 2. Cirrhosis of the liver 3. Cirrhosis of the liver with spontaneous bacterial peritonitis 4. Budd Chiari syndrome 5. Peritoneal malignancy 6. Miscellaneous
Lymphocytes Peripheral blood Axitic fluid
17
366 + 365
2581 + 841
31
173 + 249
2017 + 1223
7
53 + 152
1826 f 1917
7
197 f 255
1785 + 517
22
62 + 55
1878 I!I 968
3
68 + 21
1979 If: 719
*Mean values and standard deviations.
of serum ADA or the ratio of ascitic fluid/plasma ADA. Our results suggest that the determination of ADA activity in ascitic fluid has a high sensitivity and specificity for the diagnosis of active peritoneal tuberculosis. A value of > 36 u/l in the ascitic fluid has been considered to be optimal in differentiating tuberculosis from other aetiologies of ascites. Simultaneous assessment of the serum ADA level showed an elevated level in patients of peritoneal and pulmonary tuberculosis. There was no statistically significant difference in the serum levels between peritoneal and pulmonary tuberculosis. However, the ADA values were significantly less in patients with ascites of non-tuberculous etiology and healthy volunteers. To attain maximum accuracy, 54 u/l was considered the cut-off point and values above this would clearly denote tuberculosis. A comparative study of the sensitivity and specificity showed that the ADA determination in ascitic fluid was better than serum. There was no correlation between ascitic fluid and serum ADA levels in patients of peritoneal tuberculosis. This may be due to the differential elevation
of T-lymphocytes in ascitic fluid and peripheral blood, as has been suggested in patients with pleural effusion [20]. Another study, however, failed to correlate the high percentage of Tlymphocytes in tuberculous pleural effusion. with the level of ADA [ll]. Even in our study no correlation was found between the absolute lymphocyte numbers and ADA activity. In our laboratory, the ratio of ascitic fluid/serum ADA was inferior to absolute serum or ascitic fluid levels of ADA in the prediction or exclusion of tuberculosis. This differs from the previous report that showed an improved accuracy of the ratio over the estimation of ADA in pleural fluid alone [13]. This is in agreement with another report that both serosal effusion and the ratio are useful in the diagnosis of tuberculous meningitis but the ratio had no significant advantage over estimation of ADA in CSF alone [21]. Thus the present study suggests that detection of adenosine deaminase activity is a useful test for the diagnosis of peritoneal tuberculosis and in differentiating it from other causes of ascites. Ascitic fluid/serum ADA ratio, and the absolute serum activity, were significantly raised in peritoneal tuberculosis in contrast to non-tuberculous patients. However, ascitic fluid/serum ADA ratio and absolute ADA level in serum do not appear to hold any marked advantages over conventional ascitic fluid criteria in the diagnosis of peritoneal tuberculosis.
This study was supportedby a grant from the Indian Council of Medical Research. The authors would Iike to thank Professor BN Tandon, Dr KR Sundaram and Dr MG Karmarkar for assistance.
References 1. Bhargava DK, Saraswat VA. Laparoscopy in patients with ascites (abstract). J Assoc Phys India 1988,36: 38. 2. Wolfe JHN, Behn AR, Jackson BT. TubercuIous peritonitis and role of diagnostic laparoscopy. Lancer 1979; 1: 852-853. 3. Jorge AD. Peritoneal tuberculosis. Endoscopy 1984; 16: 10-12. 4. Sochocky S. Tuberculous peritonitis: a review of 100 cases. Am Rev Respir Dis 1%7; 95: 398-401. 5. Dineen P, Homan WP, Grafe WR. TubercuIous peritonitis: 43 years experience in diagnosis and treatment. Am I Surg 1976; 1984: 717-722. 6. Sherman S, Rohwedder JJ, Ravi Kri&nan KP, Weg JG. Tuberculous enteritis and peritonitis. Report of 36 general hospital cases. Arch Intern Med 1980; 140: 506-508.
126 7. Martinez-Vazquez
8.
9.
10.
11.
12.
13.
14.
JM, Ocana I, Ribera E, Sequra RM, Pascual C. Adenosine deaminase activity in the diagnosis of tuberculous peritonitis. Cur 1986; 27: 1049-1053. Bhargava DK, Nijhawan S, Gupta M. Adenosine deaminase activity in the diagnosis of peritoneal tuberculosis. (letter). Lancet 1989; 1: 1261. Voigte MD, Kalvaria I, Trey C, Lombard BCL, Kirsh RE. Diagnostic value of ascitic adenosine deaminase in tuberculous peritonitis. Lancer 1989; 1: 751-753 Ribera E, Martinez-Vazquez JM, Ocana I, Segura RM, Pascual C. Activity of adenosine deaminase in cerebrospinal fluid for the diagnosis and follow up of tuberculous meningitis in adults. J Infect Di.s 1987; 155: 603-607. Ocana I, Martinez-vazquez JM, Segura RM, Fernandezde-Sevilla T, Capdevila JA: Adenosine deaminase in pleural fluids. Test for diagnosis of tuberculous pleural effusion. Chest 1983; 84: 51-53. Martinez-vazquez JM, Ribera E, Ocana I, Segura RM, Serrat M, Sagrista J. Adenosine deaminase activity in tuberculous pericarditis. Thorax 1986; 41: 88-89. Martiz FJ, Visser HS, Malan C et al. Nuwe benadering tot die diagnose van pleurale effusies. S Afr Med J 1981; 60: 217-218. Giusti G. Adenosine deaminase. In: Bergmeyer HV, ed.
BHAKGAVA
15. 16.
17.
18. 19.
AND OTHEKS
Methods of enzymatic analysis. Academic Press. New York 1974: 1092-1099. Colquhoun D: Lectures on Riostatisrics. Clarendon Press, Oxford, 1971. Van der Weyden MB, Kelley WN. Human adenosinc deaminase: distribution and properties. J Riol (‘hem 1076: 251: 54485456. Galanti B, Nardiello S, Russo M, Fiorentino F. Increased lymphocyte adenosine deaminase in typhoid fever. Stand J Infect Dis 1981; 13: 47-50. Piras MA, Garis C, Andreoni G. Immunological studies in Mediterranean spotted fever (letter). Lancet 1982: 1: 1249. Moriskaki T, Fuji H, Miwas G. Adenosine deaminase (ADA) in leukaemia: Clinical value of plasma ADA activity and characterisation of leukemic cell ADA. Am J Hematoll988;
19: 37-45.
20. Pattersson T, Klockars M, Hellstrom PE, Riska H, Wangel A. T and B lymphocytes in pleural effusions. Chest 1978; 73: 49-51. 21. Donald PR, Malan C, Walt AVD, Schueman JF. The simultaneous determination of cerebrospinal fluid and plasma adenosine deaminase activity as a diagnostic aid in tuberculous meningitis. S Afr Med J 1986; 69: 505-7.
