- Email: [email protected]
Are Pleural Fluid Parameters Related to the Development of Residual Pleural Thickening in Tuberculosis?
Are Pleural Fluid Parameters Related to the Development of Residual Pleural Thickening in Tuberculosis?* Alicia de Pablo, MD; Victoria Villena, MD; Jose Echave-Sustaeta, MD; and Angel Lopez Encuentra, MD
Study objective: Identification of predictive factors for the development of residual pleural thickening (RPT). Design: Retrospective study. Location: A 1,500-bed tertiary hospital. Patients: Patients with pleural tuberculosis diagnosed between December 1991 and February 1995 in our Respiratory Disease Service. Interventions: The clinical and radiologic characteristics, and measurements of microbiological and biochemical parameters and markers in pleural fluid were studied. RPT was defined in a posteroanterior chest radiograph as a pleural space of > 2 mm measured in the lower lateral chest at the level of an imaginary line intersecting the diaphragmatic dome. Measurements and results: In 56 patients studied, 11 (19.6%) had RPT 10 mm and 24 (42.8%) had RPT > 2 mm. The pleural fluid of patients with RPT 10 mm had a significantly lower glucose concentration and pH and higher lysozyme and tumor necrosis factor-a levels than the other patients. The pleural fluid of patients with RPT >2 mm showed no significant differences. Conclusions: The development of RPT 10 mm was related to higher concentrations of lysozyme and tumor necrosis factor-a and lower glucose concentration and pH in pleural fluid compared (CHEST 1997; 112:129?-97) with development of lower measurements of RPT. Key words: lysozyme; pleural inflamm ation; pleural thickening; pleural tuberculosis; tumor necrosis factor-a Abbreviations: ADA =adenosine deaminase; IFN-')' =gamma interferon; LDH = Iactate dehydrogenase; PT= pleural tubercul osis; RPT = residual pleural thickening; TNF-a = tumor necrosis factor-a ; IT= the rapeutic thoracentesis
Tuberculosis treatment is based on th e administration of a combination of antituberculosis medications. The effectiveness of different therapeutic regimens of increasingly shorter duration has been demonstrated,1•2 but despite therapeutic success with tuberculosis, the treatment of pleural tuberculosis (PT) is followed b yradiographic residual pleural thickening (RPT) in about half of patients. 3 Corticoids4-7 and therapeutic thoracentesis (TT)8 have been added to pharmacologic antituberculosis treatment in an attempt to reduce RPT, but contradictory results have been reported.
*From th e Respiratory Disease Service, Hospital Universitario 12 de Octubre, Madr.id, Spain. Supported b y g rant FIS (Fondo de Investigacion Sanitaria) 92/0381. Manusc ript received D ecember 1 3, 9196; revis ion accepted May 17, 1997. Reprint requests: A licia de Pablo, MD, Seroicio de Neumologia, Hospital Universitario 12 de Octubre, Ctra de An
In PT, pleural fluid concentrations of biological parameters such as adenosine deaminase (ADA), interferon--y (IFN --y), lysozyme, and tumor necrosis factor (TNF -a) are increased as a r esult of inflammation produced b y tuberculosis.9 It is postulated that iny phocyte levels in the pleural space are creased T l m responsible for enhanced ADA activity in PT. 10 Likewise, it is thought that the tuberculosis mycobacterium enhances the release of TNF-a and IFN--y into the pleural space 11 by stimulating local immune response. In view of the participation of different biological parameters (ADA, lysozyme, IFN--y, and TNF-a) in the development of pleural inflammation and fibrosis secondary to PT, we stuched the relation between the concentrations of these parameters in pleural fluid and the development of RPT (which to our knowledge has not been studied), as well as clinical, biochemical, and microbiological characteristics of pleural fluid. MATERIALS AN D M E THODS
All patients d ai gnosed as havi ng PT in the Respiratory Disease tubre University H ospital of Mad rid, Setvice of the 12 de Oc CHEST I 112 I 5 I NOVEMBER, 1997
1293
Spain, between December 1991 and February 1995 were candidates for the study. Inclusion requirements were completion of antituberculosis treatment and absence of other radiologic pleural abnormalities. The diagnosis of PT was based on the presence of one or more of the following: (I) positive Ziehl stain or Lowenstein-Jensen culture of pleural fluid or tissue; (2) pleural granulomas in the absence of other granulomatous disease; and (3) pleural effusion associated with pulmonary tuberculosis, as defined by a positive sputum culture in LOwenstein-Jensen medium in the absence of any other cause for effusion. In the first thoracentesis performed, the pleural fluid biochemistry (lactate dehydrogenase [LDH], glucose, proteins, pH, cholesterol, and amylase), cell count, microbiology (culture in Lowenstein-Jensen medium of pleural fluid and pleural biopsy), and markers (ADA, IFN-')', lysozyme, and TNF-a) were studied. The radiologic size of the pleural effusion was defined as small when the costophrenic angle was obliterated but the hemidiaphragm was not covered, as medium when it occupied less than two thirds of the space between the mediastinum and chest wall at the height of the hilum, and as large when it was greater than medium . Between the lst to 20th days after concluding medical treatment, RPT was measured in the lower lateral hemithorax of a posteroanterior chest radiograph at the level of an imaginary line intersecting the diaphragmatic dome (Fig 1). Two pleural thickness were used to classify patients by the presence or absence of RPT. First, RPT was defined as a pleural thickness of :2:10 mm because this thickness may have important clinical repercussions. Second, analyses were repeated using RPT defined as a pleural thickness of >2 mm, a value considered radiologically abnormaJ.l2 Measurements were made by two independent observers and concurrence was evaluated using the kappa index. ADA was measured using an ultraviolet method with an NADH substrate adapted for an analyzer (Hitachi 717; Boehringer Mannheim; Mannheim , Germany). Lysozyme was measured using a turbidimetric kinetic method adapted for the above-mentioned analyzer. IFN-I' and TNF-a ware measured by radioimmunoanalysis (Centocor gamma interferon radioimmunoassay; Malvern, Pa; and Medgenix TNF-a; Brussels, Belgium). Statistical Analysis
The patients were divided into two groups based on the presence or absence of RPT after treatment concluded. Clinical, biochemical, and microbiological characteristics and pleural fluid markers were compared in the two groups of patients. Parameter distributions were compared with a normal distribution using the
Kolmogorov-Smirnov test. For nonnormal distributions, the groups were compared by the Mann-Whitney test, and for normal distributions, by Student's t test. Qualitative variables were compared with the x2 method and Yates' correction. The Fisher Exact Test was used as needed.
RESULTS
Fifty-six patients with PT were included in the study: 35 were male and 21 were female, with an age range of 15 to 77 years and a mean age of 28.5 years. RPT was measured after concluding medical treatment. The interobserver kappa index was 0.9. Eleven patients (19.6%) had RPT ::::::10 mm and 24 (42.8%) had RPT >2 mm. The antituberculosis regimens given were as follows : 45 patients received isoniazid and rifampin for 6 months with pyrazinamide for the first 2 months; 5 patients received the same treatment regimen for 9 months; 1 patient received rifampin and isoniazid for 6 months combined with pyrazinamide and ethambutol the first 2 months; and 5 patients received rifampin, pyrazinamide, and ethambutol for 12 months. In nine patients, medical treatment was complemented by thoracentesis performed to reduce dyspnea; four of them developed RPT >2 mm and two patients had it ::::::10 mm. The characteristics of the pleural fluid of these patients are shown in Table l. Only one patient with a positive human immunodeficiency serology had corticosteroid treatment because of high fever. One year after concluding treatment, 13 patients who had developed RPT >2 mm were reevaluated radiologically: in 10, pleural thickening had decreased by 3 to 28 mm (mean, 7.2 mm), and in 4, RPT disappeared. The clinical, radiologic, and microbiological characteristics of the pleural fluid of patients with and without RPT are shown in Table 2. Patients with RPT >2 mm were older and male subjects predom-
Table 1-Pleural Fluid Characteristics in Patients With or Without TT*
~
RPT>2mm.
FIGURE l. Method of radiologic measurement of RPT. 1294
LDH, U/L Glucose, mgldL Proteins, giL pH Cholesterol, mgldL Amylase, U/L Cells, J.LL ADA, U/L IFN--y, U/mL Lysozyme, nglmL TNF-n, nglmL
With TT (n=9)
Without TT (n=47)
p Value
400.4 65.1 4.98 7.27 87.7 121.7 3,481 61.6 76.7 22.4 415.6
463.5 69.5 5.08 7.28 87.4 105.7 3,831 90.2 90.1 17.2 322.35
NS NS NS NS NS NS NS NS NS NS NS
*Mean concentrations of pleural fluid biochemical components and markers in patients with or without TT. NS=nonsignificant. Clinical Investigations
Table 2-General Characteristics of the Study Population*
Age, yr Male Radiologic size 1 Small Moderate Large Positive PPD Positive Lw PF Positive Lw PB TT 6-mo treatment
RPT ? 10 mm (n=ll )
RPT < 10 mm (n=45)
p V alue
RPT > 2 mm (n= 24)
RPT :S2 mm (n=32)
p Value
39 11
25 24
NS 0.01
32 19
25 16
NS NS
2 6 3 5/10
8 22 14 20/37 6/32 12/42 7/45 37/45
NS NS NS NS NS NS NS NS
3 12 8 ll/21 3/16 8/21 4/24 22/24
7 16 9 14/26 5/23 8/32 5/32 24/32
NS NS NS NS NS NS NS NS
2n
4/ll 2/ll 9/ ll
*Clinical , radiologic, and microbiological charactetistics of the study population grouped by RPT thickness. PPD = tuberculin skin test; Lw= culture in Liiwenstein-Jensen medium; PF = pleural fluid; PB = pleural biopsy. 1 One patient was excluded b ecause of encapsulated pleural fluid.
inated. In patients with RPT :::::10 mm, there was a significant difference in sex: all the patients >vith RPT were male. There were no differences between the patients with and without RPT in the radiologic size of the pleural effusion or in the microbiology of pleural fluid or tissue. One patient with encapsulated pleural fluid was excluded from analysis of the radiologic size of the pleural effusion. The pleural-fluid biochemical values, cell count, and immunologic markers in patients with and without RPT are shown in Tables 3 and 4 We found that patients with RPT :::::10 mm had significantly lower glucose (p
DISCUSSION
The incidence of RPT after antituberculosis treatment varies from one study to another. Lee et al4
Table 3-Pleural Fluid Characteristics (RPT 10 mm)*
LDH, U!L Glucose, mgldL Proteins, giL pH Cholesterol , mgldL Amylase, U!L Cells, JJ.L ADA, U/L IFN--y, U/mL Lysozyme, nglmL TNF -a, nglmL
reported 10%, Soler et al 13 reported 72%, and Barbas et al 14 and Moudgil et al 5 reported intermediate values of 52% and 66%, respectively; we found an incidence of 42.8% . These variations can be attributed to the lack of a uniform concept of RPT; some define a pleural thickness of >2 mm as abnormal and others require a thickness of :::::10 mm. Still others do not clarify the definition of RPT. For this reason, we used two RPT values, :::::10 mm, which was considered more important because functional disturbances occur with this thickness, and >2 mm, which is radiologically abnormal. Studies of RPT in PT have examined the epidemiologic characteristics and biochemical and microbiological parameters in pleural fluid because of the following: RPT is a consequence of an inflammatory mechanism;l3 patients with RPT have higher LDH levels or a hypersensitivity reaction; 14 and there is a higher incidence of positive tuberculin test and fewer positive Lowenstein cultures of pleural fluid. However, these findings have not been confirmed in every series .15 Table 4-Pleural Fluid Characteristics (RPT >2 mm)*
RPT Present (n=ll )
RPT Absent (n=45 )
p V alue
627.7 51 5.01 7.20 82.7 96.2 4,467 74.1 91.7 26.4 577.2
410.8 73.2 5.07 7.30 88.5 111.6 3,606 87.3 87 16.7 275.5
NS 0.004 NS 0.02 NS NS NS NS NS 0.01 0.003
*Mean concentrations of pleural fluid biochemical components and markers in patients separated b ypresence or absence ofRPT ? 10 mm.
LDH, U/L Glucose, mgldL Proteins, giL pH Cholesterol, mgldL Amylase, U/L Cells, JJ.L ADA, U/L IFN --y, U/mL Lysozyme, nglmL TNF -a , nglmL
RPT Present (n=24 )
RPT Absent (n= 32)
p Value
490.9 63.5 5.08 7.25 89 104.3 3,961 117.3 75.2 20.1 424.2
425.3 72.8 5.04 7.30 86.4 111.8 3,635 62.8 97.8 16.7 267
NS NS NS NS NS NS NS NS NS NS NS
*Mean concentrations of pleural fluid biochemical components and markers in patients separated b ypresence or absence of RPT > 2 mm. CHEST I 112 I 5 I NOVEMBER, 1997
1295
144 i
I
I
7,5 !
1241 I
--"'0
!
I
'j
' 7,3 1
i
!
1041
I
.sw 841 0>
(j)
0
0
::J _J
<.9
!
..... or;"•
I 64 /
i
!
I
6,91
.!.
l
441 1
''I 1
RPT~10mm
RPT<10mm
65 ' -
,
RPT~10mm
RPT<10mm
~--~--------------~
I
1.545 1
...!....
I
I
45 L__------------~------~ RPT~10mm
RPT<10mm
FIGURE 2. Glucose concentration, pH, and TN F-a level in pleural fluid of patients with and without
RPT ~10 mm. Statistically significant differences were found in all of them (p=0.004, p=0.02, and p=0.003, respectively).
We found that neither age, treatment regimen, nor TT influenced the later development of RPT. Male subjects predominated among patients with RPT, as reported by Soler et al,l3 but in our study, this finding was significant only for RPT ~10 mm, whereas Soler et al 13 obtained significant sex differences for RPT >2 mm. 1296
'Ve found no relation between the development of RPT and either tuberculin test results or microbiological findings in pleural fluid or tissue, so our findings do not support the theory of increased hypersensitivity as the origin of RPT. Our biochemical results concur with the findings of Barbas et al 14 in that no difference in LDH Clinical Investigations
concentration was found in pleural fluid, in contrast with the findings of Soler et al. 13 To our knowledge, there have been no repmts of differences in glucose concentration in pleural fluid and only the study of Soler et aJl 3 analyzed pH, although no significant differences were observed. In contrast, we found significantly lower glucose concentrations and pH in the pleural fluid of patients with RPT ::::::10 mm. In patients with RPT ::::::10 mm, TNF-a and lysozyme levels were significantly higher than in those with lower levels of RPT. These findings and the biochemical findings described suggest enhanced inflammatory activity in the pleural space of patients who eventually developed intense RPT. Cytokines participate in the regulation of inflammation and fibrosis by activating or inhibiting fibroblast proliferation. The effects on fibroblasts of different cytokines, TNF-a, IFN-')', and interleukins vary depending on whether they act individually or synergistically.16 The participation of TNF -a in the development of pleural and pulmonary fibrosis has been demonstrated.17- 19 The fibrosis is characterized by a deposit of type I collagen as a result of fibroblast proliferation and stimulation. In conjunction \'Vith the cytokines, TNF -a stimulates fibroblast proliferation so that increased TNF -a in the pleural space leads to more intense localized fibrosis. In inflammatory processes, IFN -')' is one of the main macrophage activators,20 but it has been d emonstrated in vitro that it reduces collagen production by fibroblasts, thus impeding the development of fibrosis. 2 1 Therefore, the presence of a higher TNF -a concentration in pleural fluid may increase fibroblast proliferation and stimulation, thus enhancing the deposit of collagen in pleura and originating more pleural fibrosis. To conclude, in our region of central Spain, 42.8% of our patients vvith PT after antituberculosis treatment developed RPT >2 mm. Glucose, pH, TNF-a, and lysozyme values in pleural fluid differed significantly benveen patients who developed RPT ::::::10 mm and those who did not. REFERENCES Ame1ican Thoracic Society. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1994; 149:1359-74 2 Grupo de trabajo sobre tuberculosis. Consenso nacional para el control de Ia tuberculosis e n Espana. Med Clin (Bare) 1992; 98:24-31
3 Chan CHS, Arnold M, Chan CY, et al. Clinical and pathological features of tuberculous pleural effusion and its longterm consequences. Respiration 1991; 58:171-75 4 Lee CH , Wang WJ, Lan RS , et al. Corticosteroids in the treatment of tuberculous pleurisy: a double-blind, placebocontrolled, randomized study. Chest 1988; 94:1256-59 5 Moudgil H, Sridhar G, Leitch AG. Reactivation disease: the commonest form of tuberculous pleural effusion in Edinburgh, 1980-1991. Respir Med 1994; 88:301-04 6 Soler JJ, Gil Suay V, Cordero P, et al. Corticoides en Ia tuberculosis pleural i R educen el engrosamiento pleural residual [abstract]? Arch Bronconeumol1994; 30(suppl1):43 7 Galarza I, Caiiete C, Granados A, et la. Randomised trial of corti costeroids in the treatment of tuberculous pleurisy. Thorax 1995; 50:1305-07 8 Morrone N, Lombardi MC, Machado 0 . Prevention of pleural thickening through pleural aspiration in patients with tuberculous effusion. J Pneu mol (Sao Paulo) 1989; 15:180-84 9 Valdes L, San Jose E ,Alvarez D, e t al. Diagnosis of tuberculous pleurisy using the biologic parameters adenosine deaminase, lysozyme, and interfe ron gamma. Chest 1993; 103: 458-65 10 Sorribas A, Blanco F, Gomez JA, eta!. Adenosina desaminasa: caracterfsti cas bioqufmicas y significaci6n clfnica de una enzima clave para la inmunidad celular. Med Clin (Bare) 1988; 90:548-52 11 Barnes P, Fong S, Brennan P, et al. Local production of tumor necrosis factor and IFN-gam ma i n tuberculous pleuritis. J Immunol 1990; 145:149-54 12 Light RW. Pleural disease. 3rd ed. Baltimore: Williams & Wilkins, 1990 13 Soler JJ, Gil Suay V, Cordero P, et al. Engrosamiento pleural en el d errame pleural tuberculoso: factores asociados . Arch Bronconeumol 1995; 31:157-61 14 Barbas CS, Cukier A, Varvalho CR, et al. The relationship between pleural fluid findings and the development of pleural thickening in patients with pleural tuberculosis. Chest 1991; 100:1264-67 15 Flores J, Perez-Rod1iguez E, Gaud6 J, et al. Valor de La presencia de micobacterias en Ia tuberculosis pleural. Arch Bronconeurnol 1994; 30(suppl 1):38-39 16 Elias JA, Freundlich B , Kern JA, e t la. Cytokine networks in the regulation of inflammation and fibrosis in the lung. Chest 1990; 97:1439-45 17 Owens MW, Grimes SR. Pleural mesothelial cell response to inflammation: tumor necrosis factor-induced mitogenesis and collagen synthesis. Am J Physiol 1993; 265(pt 1):L382-88 18 Law K, Weiden M, Harkin T, et al. Increased r elease of interleukin-1 b eta, interleukin-6, and tumor necrosis factoralpha by bronchoalveolar cells al vaged from involved sites in pulmonary tuberculosis. Am J Respir Crit Care Med 1996; 153:799-804 19 Piguet P, Collart M, Grau G, et la. Tumor necrosis factor/ cachectin plays a key role in bleomycin-induced pneumopathy and fibrosis. J Exp Med 1989; 170:655-63 20 Rook GA, Attiyah R. Cytokines and the Koch phenomenon. Tubercle 1991; 72:13-20 21 Goldstein RH, Fine A. Potential th erapeutic initiatives for fibrogenic lung diseases. Chest 1995; 108:848-55
CHEST I 112 I 5 I NOVEMBER, 1997
1297
Study objective: Identification of predictive factors for the development of residual pleural thickening (RPT). Design: Retrospective study. Location: A 1,500-bed tertiary hospital. Patients: Patients with pleural tuberculosis diagnosed between December 1991 and February 1995 in our Respiratory Disease Service. Interventions: The clinical and radiologic characteristics, and measurements of microbiological and biochemical parameters and markers in pleural fluid were studied. RPT was defined in a posteroanterior chest radiograph as a pleural space of > 2 mm measured in the lower lateral chest at the level of an imaginary line intersecting the diaphragmatic dome. Measurements and results: In 56 patients studied, 11 (19.6%) had RPT 10 mm and 24 (42.8%) had RPT > 2 mm. The pleural fluid of patients with RPT 10 mm had a significantly lower glucose concentration and pH and higher lysozyme and tumor necrosis factor-a levels than the other patients. The pleural fluid of patients with RPT >2 mm showed no significant differences. Conclusions: The development of RPT 10 mm was related to higher concentrations of lysozyme and tumor necrosis factor-a and lower glucose concentration and pH in pleural fluid compared (CHEST 1997; 112:129?-97) with development of lower measurements of RPT. Key words: lysozyme; pleural inflamm ation; pleural thickening; pleural tuberculosis; tumor necrosis factor-a Abbreviations: ADA =adenosine deaminase; IFN-')' =gamma interferon; LDH = Iactate dehydrogenase; PT= pleural tubercul osis; RPT = residual pleural thickening; TNF-a = tumor necrosis factor-a ; IT= the rapeutic thoracentesis
Tuberculosis treatment is based on th e administration of a combination of antituberculosis medications. The effectiveness of different therapeutic regimens of increasingly shorter duration has been demonstrated,1•2 but despite therapeutic success with tuberculosis, the treatment of pleural tuberculosis (PT) is followed b yradiographic residual pleural thickening (RPT) in about half of patients. 3 Corticoids4-7 and therapeutic thoracentesis (TT)8 have been added to pharmacologic antituberculosis treatment in an attempt to reduce RPT, but contradictory results have been reported.
*From th e Respiratory Disease Service, Hospital Universitario 12 de Octubre, Madr.id, Spain. Supported b y g rant FIS (Fondo de Investigacion Sanitaria) 92/0381. Manusc ript received D ecember 1 3, 9196; revis ion accepted May 17, 1997. Reprint requests: A licia de Pablo, MD, Seroicio de Neumologia, Hospital Universitario 12 de Octubre, Ctra de An
In PT, pleural fluid concentrations of biological parameters such as adenosine deaminase (ADA), interferon--y (IFN --y), lysozyme, and tumor necrosis factor (TNF -a) are increased as a r esult of inflammation produced b y tuberculosis.9 It is postulated that iny phocyte levels in the pleural space are creased T l m responsible for enhanced ADA activity in PT. 10 Likewise, it is thought that the tuberculosis mycobacterium enhances the release of TNF-a and IFN--y into the pleural space 11 by stimulating local immune response. In view of the participation of different biological parameters (ADA, lysozyme, IFN--y, and TNF-a) in the development of pleural inflammation and fibrosis secondary to PT, we stuched the relation between the concentrations of these parameters in pleural fluid and the development of RPT (which to our knowledge has not been studied), as well as clinical, biochemical, and microbiological characteristics of pleural fluid. MATERIALS AN D M E THODS
All patients d ai gnosed as havi ng PT in the Respiratory Disease tubre University H ospital of Mad rid, Setvice of the 12 de Oc CHEST I 112 I 5 I NOVEMBER, 1997
1293
Spain, between December 1991 and February 1995 were candidates for the study. Inclusion requirements were completion of antituberculosis treatment and absence of other radiologic pleural abnormalities. The diagnosis of PT was based on the presence of one or more of the following: (I) positive Ziehl stain or Lowenstein-Jensen culture of pleural fluid or tissue; (2) pleural granulomas in the absence of other granulomatous disease; and (3) pleural effusion associated with pulmonary tuberculosis, as defined by a positive sputum culture in LOwenstein-Jensen medium in the absence of any other cause for effusion. In the first thoracentesis performed, the pleural fluid biochemistry (lactate dehydrogenase [LDH], glucose, proteins, pH, cholesterol, and amylase), cell count, microbiology (culture in Lowenstein-Jensen medium of pleural fluid and pleural biopsy), and markers (ADA, IFN-')', lysozyme, and TNF-a) were studied. The radiologic size of the pleural effusion was defined as small when the costophrenic angle was obliterated but the hemidiaphragm was not covered, as medium when it occupied less than two thirds of the space between the mediastinum and chest wall at the height of the hilum, and as large when it was greater than medium . Between the lst to 20th days after concluding medical treatment, RPT was measured in the lower lateral hemithorax of a posteroanterior chest radiograph at the level of an imaginary line intersecting the diaphragmatic dome (Fig 1). Two pleural thickness were used to classify patients by the presence or absence of RPT. First, RPT was defined as a pleural thickness of :2:10 mm because this thickness may have important clinical repercussions. Second, analyses were repeated using RPT defined as a pleural thickness of >2 mm, a value considered radiologically abnormaJ.l2 Measurements were made by two independent observers and concurrence was evaluated using the kappa index. ADA was measured using an ultraviolet method with an NADH substrate adapted for an analyzer (Hitachi 717; Boehringer Mannheim; Mannheim , Germany). Lysozyme was measured using a turbidimetric kinetic method adapted for the above-mentioned analyzer. IFN-I' and TNF-a ware measured by radioimmunoanalysis (Centocor gamma interferon radioimmunoassay; Malvern, Pa; and Medgenix TNF-a; Brussels, Belgium). Statistical Analysis
The patients were divided into two groups based on the presence or absence of RPT after treatment concluded. Clinical, biochemical, and microbiological characteristics and pleural fluid markers were compared in the two groups of patients. Parameter distributions were compared with a normal distribution using the
Kolmogorov-Smirnov test. For nonnormal distributions, the groups were compared by the Mann-Whitney test, and for normal distributions, by Student's t test. Qualitative variables were compared with the x2 method and Yates' correction. The Fisher Exact Test was used as needed.
RESULTS
Fifty-six patients with PT were included in the study: 35 were male and 21 were female, with an age range of 15 to 77 years and a mean age of 28.5 years. RPT was measured after concluding medical treatment. The interobserver kappa index was 0.9. Eleven patients (19.6%) had RPT ::::::10 mm and 24 (42.8%) had RPT >2 mm. The antituberculosis regimens given were as follows : 45 patients received isoniazid and rifampin for 6 months with pyrazinamide for the first 2 months; 5 patients received the same treatment regimen for 9 months; 1 patient received rifampin and isoniazid for 6 months combined with pyrazinamide and ethambutol the first 2 months; and 5 patients received rifampin, pyrazinamide, and ethambutol for 12 months. In nine patients, medical treatment was complemented by thoracentesis performed to reduce dyspnea; four of them developed RPT >2 mm and two patients had it ::::::10 mm. The characteristics of the pleural fluid of these patients are shown in Table l. Only one patient with a positive human immunodeficiency serology had corticosteroid treatment because of high fever. One year after concluding treatment, 13 patients who had developed RPT >2 mm were reevaluated radiologically: in 10, pleural thickening had decreased by 3 to 28 mm (mean, 7.2 mm), and in 4, RPT disappeared. The clinical, radiologic, and microbiological characteristics of the pleural fluid of patients with and without RPT are shown in Table 2. Patients with RPT >2 mm were older and male subjects predom-
Table 1-Pleural Fluid Characteristics in Patients With or Without TT*
~
RPT>2mm.
FIGURE l. Method of radiologic measurement of RPT. 1294
LDH, U/L Glucose, mgldL Proteins, giL pH Cholesterol, mgldL Amylase, U/L Cells, J.LL ADA, U/L IFN--y, U/mL Lysozyme, nglmL TNF-n, nglmL
With TT (n=9)
Without TT (n=47)
p Value
400.4 65.1 4.98 7.27 87.7 121.7 3,481 61.6 76.7 22.4 415.6
463.5 69.5 5.08 7.28 87.4 105.7 3,831 90.2 90.1 17.2 322.35
NS NS NS NS NS NS NS NS NS NS NS
*Mean concentrations of pleural fluid biochemical components and markers in patients with or without TT. NS=nonsignificant. Clinical Investigations
Table 2-General Characteristics of the Study Population*
Age, yr Male Radiologic size 1 Small Moderate Large Positive PPD Positive Lw PF Positive Lw PB TT 6-mo treatment
RPT ? 10 mm (n=ll )
RPT < 10 mm (n=45)
p V alue
RPT > 2 mm (n= 24)
RPT :S2 mm (n=32)
p Value
39 11
25 24
NS 0.01
32 19
25 16
NS NS
2 6 3 5/10
8 22 14 20/37 6/32 12/42 7/45 37/45
NS NS NS NS NS NS NS NS
3 12 8 ll/21 3/16 8/21 4/24 22/24
7 16 9 14/26 5/23 8/32 5/32 24/32
NS NS NS NS NS NS NS NS
2n
4/ll 2/ll 9/ ll
*Clinical , radiologic, and microbiological charactetistics of the study population grouped by RPT thickness. PPD = tuberculin skin test; Lw= culture in Liiwenstein-Jensen medium; PF = pleural fluid; PB = pleural biopsy. 1 One patient was excluded b ecause of encapsulated pleural fluid.
inated. In patients with RPT :::::10 mm, there was a significant difference in sex: all the patients >vith RPT were male. There were no differences between the patients with and without RPT in the radiologic size of the pleural effusion or in the microbiology of pleural fluid or tissue. One patient with encapsulated pleural fluid was excluded from analysis of the radiologic size of the pleural effusion. The pleural-fluid biochemical values, cell count, and immunologic markers in patients with and without RPT are shown in Tables 3 and 4 We found that patients with RPT :::::10 mm had significantly lower glucose (p
DISCUSSION
The incidence of RPT after antituberculosis treatment varies from one study to another. Lee et al4
Table 3-Pleural Fluid Characteristics (RPT 10 mm)*
LDH, U!L Glucose, mgldL Proteins, giL pH Cholesterol , mgldL Amylase, U!L Cells, JJ.L ADA, U/L IFN--y, U/mL Lysozyme, nglmL TNF -a, nglmL
reported 10%, Soler et al 13 reported 72%, and Barbas et al 14 and Moudgil et al 5 reported intermediate values of 52% and 66%, respectively; we found an incidence of 42.8% . These variations can be attributed to the lack of a uniform concept of RPT; some define a pleural thickness of >2 mm as abnormal and others require a thickness of :::::10 mm. Still others do not clarify the definition of RPT. For this reason, we used two RPT values, :::::10 mm, which was considered more important because functional disturbances occur with this thickness, and >2 mm, which is radiologically abnormal. Studies of RPT in PT have examined the epidemiologic characteristics and biochemical and microbiological parameters in pleural fluid because of the following: RPT is a consequence of an inflammatory mechanism;l3 patients with RPT have higher LDH levels or a hypersensitivity reaction; 14 and there is a higher incidence of positive tuberculin test and fewer positive Lowenstein cultures of pleural fluid. However, these findings have not been confirmed in every series .15 Table 4-Pleural Fluid Characteristics (RPT >2 mm)*
RPT Present (n=ll )
RPT Absent (n=45 )
p V alue
627.7 51 5.01 7.20 82.7 96.2 4,467 74.1 91.7 26.4 577.2
410.8 73.2 5.07 7.30 88.5 111.6 3,606 87.3 87 16.7 275.5
NS 0.004 NS 0.02 NS NS NS NS NS 0.01 0.003
*Mean concentrations of pleural fluid biochemical components and markers in patients separated b ypresence or absence ofRPT ? 10 mm.
LDH, U/L Glucose, mgldL Proteins, giL pH Cholesterol, mgldL Amylase, U/L Cells, JJ.L ADA, U/L IFN --y, U/mL Lysozyme, nglmL TNF -a , nglmL
RPT Present (n=24 )
RPT Absent (n= 32)
p Value
490.9 63.5 5.08 7.25 89 104.3 3,961 117.3 75.2 20.1 424.2
425.3 72.8 5.04 7.30 86.4 111.8 3,635 62.8 97.8 16.7 267
NS NS NS NS NS NS NS NS NS NS NS
*Mean concentrations of pleural fluid biochemical components and markers in patients separated b ypresence or absence of RPT > 2 mm. CHEST I 112 I 5 I NOVEMBER, 1997
1295
144 i
I
I
7,5 !
1241 I
--"'0
!
I
'j
' 7,3 1
i
!
1041
I
.sw 841 0>
(j)
0
0
::J _J
<.9
!
..... or;"•
I 64 /
i
!
I
6,91
.!.
l
441 1
''I 1
RPT~10mm
RPT<10mm
65 ' -
,
RPT~10mm
RPT<10mm
~--~--------------~
I
1.545 1
...!....
I
I
45 L__------------~------~ RPT~10mm
RPT<10mm
FIGURE 2. Glucose concentration, pH, and TN F-a level in pleural fluid of patients with and without
RPT ~10 mm. Statistically significant differences were found in all of them (p=0.004, p=0.02, and p=0.003, respectively).
We found that neither age, treatment regimen, nor TT influenced the later development of RPT. Male subjects predominated among patients with RPT, as reported by Soler et al,l3 but in our study, this finding was significant only for RPT ~10 mm, whereas Soler et al 13 obtained significant sex differences for RPT >2 mm. 1296
'Ve found no relation between the development of RPT and either tuberculin test results or microbiological findings in pleural fluid or tissue, so our findings do not support the theory of increased hypersensitivity as the origin of RPT. Our biochemical results concur with the findings of Barbas et al 14 in that no difference in LDH Clinical Investigations
concentration was found in pleural fluid, in contrast with the findings of Soler et al. 13 To our knowledge, there have been no repmts of differences in glucose concentration in pleural fluid and only the study of Soler et aJl 3 analyzed pH, although no significant differences were observed. In contrast, we found significantly lower glucose concentrations and pH in the pleural fluid of patients with RPT ::::::10 mm. In patients with RPT ::::::10 mm, TNF-a and lysozyme levels were significantly higher than in those with lower levels of RPT. These findings and the biochemical findings described suggest enhanced inflammatory activity in the pleural space of patients who eventually developed intense RPT. Cytokines participate in the regulation of inflammation and fibrosis by activating or inhibiting fibroblast proliferation. The effects on fibroblasts of different cytokines, TNF-a, IFN-')', and interleukins vary depending on whether they act individually or synergistically.16 The participation of TNF -a in the development of pleural and pulmonary fibrosis has been demonstrated.17- 19 The fibrosis is characterized by a deposit of type I collagen as a result of fibroblast proliferation and stimulation. In conjunction \'Vith the cytokines, TNF -a stimulates fibroblast proliferation so that increased TNF -a in the pleural space leads to more intense localized fibrosis. In inflammatory processes, IFN -')' is one of the main macrophage activators,20 but it has been d emonstrated in vitro that it reduces collagen production by fibroblasts, thus impeding the development of fibrosis. 2 1 Therefore, the presence of a higher TNF -a concentration in pleural fluid may increase fibroblast proliferation and stimulation, thus enhancing the deposit of collagen in pleura and originating more pleural fibrosis. To conclude, in our region of central Spain, 42.8% of our patients vvith PT after antituberculosis treatment developed RPT >2 mm. Glucose, pH, TNF-a, and lysozyme values in pleural fluid differed significantly benveen patients who developed RPT ::::::10 mm and those who did not. REFERENCES Ame1ican Thoracic Society. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1994; 149:1359-74 2 Grupo de trabajo sobre tuberculosis. Consenso nacional para el control de Ia tuberculosis e n Espana. Med Clin (Bare) 1992; 98:24-31
3 Chan CHS, Arnold M, Chan CY, et al. Clinical and pathological features of tuberculous pleural effusion and its longterm consequences. Respiration 1991; 58:171-75 4 Lee CH , Wang WJ, Lan RS , et al. Corticosteroids in the treatment of tuberculous pleurisy: a double-blind, placebocontrolled, randomized study. Chest 1988; 94:1256-59 5 Moudgil H, Sridhar G, Leitch AG. Reactivation disease: the commonest form of tuberculous pleural effusion in Edinburgh, 1980-1991. Respir Med 1994; 88:301-04 6 Soler JJ, Gil Suay V, Cordero P, et al. Corticoides en Ia tuberculosis pleural i R educen el engrosamiento pleural residual [abstract]? Arch Bronconeumol1994; 30(suppl1):43 7 Galarza I, Caiiete C, Granados A, et la. Randomised trial of corti costeroids in the treatment of tuberculous pleurisy. Thorax 1995; 50:1305-07 8 Morrone N, Lombardi MC, Machado 0 . Prevention of pleural thickening through pleural aspiration in patients with tuberculous effusion. J Pneu mol (Sao Paulo) 1989; 15:180-84 9 Valdes L, San Jose E ,Alvarez D, e t al. Diagnosis of tuberculous pleurisy using the biologic parameters adenosine deaminase, lysozyme, and interfe ron gamma. Chest 1993; 103: 458-65 10 Sorribas A, Blanco F, Gomez JA, eta!. Adenosina desaminasa: caracterfsti cas bioqufmicas y significaci6n clfnica de una enzima clave para la inmunidad celular. Med Clin (Bare) 1988; 90:548-52 11 Barnes P, Fong S, Brennan P, et al. Local production of tumor necrosis factor and IFN-gam ma i n tuberculous pleuritis. J Immunol 1990; 145:149-54 12 Light RW. Pleural disease. 3rd ed. Baltimore: Williams & Wilkins, 1990 13 Soler JJ, Gil Suay V, Cordero P, et al. Engrosamiento pleural en el d errame pleural tuberculoso: factores asociados . Arch Bronconeumol 1995; 31:157-61 14 Barbas CS, Cukier A, Varvalho CR, et al. The relationship between pleural fluid findings and the development of pleural thickening in patients with pleural tuberculosis. Chest 1991; 100:1264-67 15 Flores J, Perez-Rod1iguez E, Gaud6 J, et al. Valor de La presencia de micobacterias en Ia tuberculosis pleural. Arch Bronconeurnol 1994; 30(suppl 1):38-39 16 Elias JA, Freundlich B , Kern JA, e t la. Cytokine networks in the regulation of inflammation and fibrosis in the lung. Chest 1990; 97:1439-45 17 Owens MW, Grimes SR. Pleural mesothelial cell response to inflammation: tumor necrosis factor-induced mitogenesis and collagen synthesis. Am J Physiol 1993; 265(pt 1):L382-88 18 Law K, Weiden M, Harkin T, et al. Increased r elease of interleukin-1 b eta, interleukin-6, and tumor necrosis factoralpha by bronchoalveolar cells al vaged from involved sites in pulmonary tuberculosis. Am J Respir Crit Care Med 1996; 153:799-804 19 Piguet P, Collart M, Grau G, et la. Tumor necrosis factor/ cachectin plays a key role in bleomycin-induced pneumopathy and fibrosis. J Exp Med 1989; 170:655-63 20 Rook GA, Attiyah R. Cytokines and the Koch phenomenon. Tubercle 1991; 72:13-20 21 Goldstein RH, Fine A. Potential th erapeutic initiatives for fibrogenic lung diseases. Chest 1995; 108:848-55
CHEST I 112 I 5 I NOVEMBER, 1997
1297