- Email: [email protected]
A study of acute-phase reactant proteins in Indonesian patients with pulmonary tuberculosis
Tobercle 66 (1984)
23-39
0
Ltd.
Longman
Group
A STUDY INDONESIAN
OF ACUTE-PHASE PATIENTS WITH J.
Cardiothoracic
Institute,
University
M.
of Polmonology
Grange
of London, Fulham Road, London SW3 6HP, U.K.
T. Kardjito Departments
REACTANT PROTEINS IN PULMONARY TUBERCULOSIS
and
I. Setiabudi
and Clinical Pathology, Faculty of Medicine, Surabaya. Indonesia
Airlangga
University,
Summary The levels of 8 acute phase reactants (ul -antitrypsin, a2-macroglobulin, transferrin, al -acid glycoprotein, C-reactive protein, ceruloplasmin, haptoglobin and the third component of complement) and immunoglobulin in the IgG, IgM and IgA classes were assayed, by laser nephelometry, in sera from 107 East Javanese patients with smear-positive pulmonary tuberculosis and 144 healthy subjects. These levels were correlated with clinical, haematological and radiological features, the levels of antibody to Mycobacterium tuberculosis, and the diameters of the tuberculin skin test read at various times. Levels of all acute phase reactants increased significantly in tuberculosis except for that of transferrin which was lowered. The correlations between the various acute phase reactants in health and disease were calculated. In general, the correlations were lower in disease than in health, except for the third component of complement and a greatly increased correlation between the levels of al -antitrypsin and ceruloplasmin. There reactants and least and total IgA and
was a significant correlation between levels of some of the acute phase and those of antibodies to M. tuberculosis, mainly with IgG, less with IgA with IgM antibodies. By contrast correlations between acute phase protein immunoglobulin levels were most evident in the IgM class, less with not at all with IgG.
Although there were some associations between protein levels and age, sex and weight of controls and patients, these were not great enough to account for the differences between the two groups. There was a tendency for patients, but not controls, with intestinal helminthiasis to have higher levels of total IgM than those without evidence of parasites. In general, the levels of proteins bore very little relation to the clinical and radiological features of disease and were, with the exception of the antimycobacterial antibodies, of no diagnostic value. Likewise, protein levels were not associated with the extent of disease; better correlations were found with the ESR and leucocyte count. Transferrin levels tended to be higher in those with chronic disease and showed a correlation with the diameters of the dermal reactions to tuberculin at 24 hours, which were also significantly larger in chronic disease. Among the haematological findings, the most significant between the lymphocyte count and haptoglobin levels possible regulatory role for this protein.
was a negative correlation in disease, suggesting a
24
Grange and others Rirsum6
Par nephelometrie du laser, on a etudie le serum de 107 malades de I’est de Java atteints de tuberculose pulmonaire frottis possitif et celui de 144 sujets sains. On a mesure les taux de 8 substances de reaction de la phase aigue (ccl -antitrypcine, a2-macroglobuline, transferrine, ctl -glycoproteine acide, proteine C reagissante, ceruloplasmine, haptoglobine et le troisieme composant du complement) ainsi que les immunoglobulines des classes IgC, IgM et IgA. On a procede a des correlations entre ces taux et le elements cliniques, hematologiques et radiologiques, les taux d’anticorps vis a vis de M. tuberculosis et les diametres du test cutane a la tuberculine lus a differents moments. Les taux des substances de reaction de la phase aigue Btaient significativement augment& dans la tuberculose, a I’exception de la transferrine qui Btait diminuee. L’article discute les correlations entre les diverses substances de la phase aigue chez les malades et chez les sujets sains. D’une man&e generale, les correlations se sont montrees plus faibles en cas de tuberculose que lors de la bonne Sante, sauf pour le troisieme composant du complement et il existait en cas de tuberculose une correlation grandement augmentee entre les taux de crl -antitrypsine et ceruloplasmine. II existait une correlation significative entre les taux de certaines substances reagissantes de la phase aigue et les taux des anticorps de M. tuberculosis, essentiellement avec les anticorps des IgG, moins avec ceux de IgA et le moins avec ceux des IgM. A I’inverse, les correlations entre les taux de proteines de la phase aigue et les taux totaux d’immunoglobulines dtaient les plus marquees pour la classe de IgM, moins pour IgA et pas du tout pour les IgG. Bien qu’il existait quelques associations entre les taux de proteine et I’bge, le sexe et le poids des temoins et des malades, elles n’etaient pas assez fortes pour expliquer les differences entre les deux groupes. Les malades, mais non les temoins, porteurs d’une helminthiase intestinale avaient tendance a posseder des taux plus eleves de IgM totaux que ceux chez lesquels on n’avait pas trouve ces parasites. En general, les taux de proteine ont montre peu de rapport avec les elements cliniques et radiologiques de la maladie et sauf en ce qui concerne les anticorps antimycobacteriens, ces taux n’avaient pas de valeur diagnostique. Egalement, les taux de proteines n’btaient pas associes avec I’extension de la maladie; une meilleure correlation a et6 trouvee avec la VHS et la numerotation de leucocytes. Les taux de transferrine avaient tendance a dtre plus eleves chez les sujets porteurs d’une tuberculose chronique et montraient une correlation avec les reactions dermiques a la 248me heure, lesquelles Btaient Bgalment significativement plus grandes chez les malades chroniques. le plus significatif a et6 I’existance d’une Parmi les resultats hemotologiques, correlation negative entre les comptes de lymphocytes et les taux d’haptoglobine chez les malades, ce qui suggere un role regulateur possible de cette proteine. Resumen Se estudid el suero de 107 enfermos del este de Java con tuberculosis pulmonar con baciloscopia positiva y el de 144 sujetos sanos mediante nefelometria de rayos laser. Se midid 10s niveles de sustancias de reacci6n de la fase aguda (al -antitripsina, a2-macroglobulina, transferina, al -glucoproteina Bcida, proteina C reactiva, ceruloplasmina, haptoglobina y el tercer componente del complemento) asi coma las inmunoglobulinas de las clases IgC, IgM, e IgA. Estos niveles fueron puestos en
Acute correlation con 10s elementos anticuerposal M. tuberculosisy a diferentes momentos.
clinicos,
phase reactants
in tuberculosis
25
hematologicos y radiolbgicos, 10s niveles de de la prueba cutanea de tuberculina leida
10s ditimetros
Los niveles de todas las sustancias de reaction de la fase aguda estaban significativamente aumentados en la tuberculosis except0 el de la transferina que estaba disminuido. Se calcularon las correlaciones entre las diversas sustancias de la fase aguda en 10s enfermos y en 10s sujetos sanos. En general las correlaciones fueron. m&s bajas en 10s enfermos que en 10s sanos, salvo por el tercer componente del complemento y, en el case de tuberculosis, un gran aumento de la correlation entre 10s niveles de al -antitripsina y ceruloplasmina. Habia una correlation significativa entre algunas de las sustancias de reacion de la con 10s antifase aguda y 10s niveles de anticuerpos al M. tuberculosis, principalmente cuerpos IgG, menos con 10s IgA y afin menos con 10s IgM. Al contrario, las correlaciones entre 10s niveles de proteinas de la fase aguda y 10s niveles totales de inmunoglobulinas fueron m&s evidentes con la clase IgM, menos con la IgA y absolutamente ausentes con la IgG. A pesar de que existian algunas asociaciones entre 10s niveles de proteinas y la edad, el sexo y el peso de 10s sujetos control y de 10s enfermos, Bstas no eran suficientemente importantes coma para explicar las diferencias entre 10s dos grupos. Los enfermos, pero no 10s sujetos control, portadores de una helmintiasis intestinal tenian tendencia a presentar niveles mas elevados de IgM totales que aquellos sin evidencia de esta parasitosis. En general, 10s niveles de proteinas mostraron una relation escasa con lo elementos clinicos y radiologicos de la enfermedad y, con excepcidn de 10s anticuerpos antimicobacterianos, estos niveles no tenian valor diagnostico. Asi mismo, 10s niveles de proteinas no estaban correlacionados con la extension de la enfermedad ; se encontrd una mejor correlation con la VHS y el recuento de leucocitos. Los niveles de transferina tenian tendencia a ser mas elevados en 10s sujetos con una tuberculosis crdnica y mostraron una correlation con las reeacciones cutaneas a las 24 horas, las cuales eran tambien significativamente mas grandes en 10s enfermos crbnicos. Entre 10s resultados hematologicos, el m&s significativo fue la correlation negativa entre el recuento de linfocitos y 10s niveles de haptoglobina en 10s enfermos, lo que sugiere un posible papel regulador de esta proteina. Introduction The immunology of tuberculosis has been intensively studied by numerous workers since Robert Koch’s pioneering work in 1891 [I 1. Until recently cell mediated and humoral (i.e. antibody dependent) immune responses were often regarded as separate and independent phenomena but it is now realized that all immunological reactions are closely inter-related. It is also appreciated that a challenge by a complex set of antigens such as a mycobacterium elicits many different reactions, not just specific immunological ones, and that some of these are beneficial, some are irrelevant and some may harm the host more than the pathogen. Indeed several widespread protozoa1 and bacterial pathogens owe their success to their ability to utilize the latter type of host responses to their own advantage [2]. Accordingly the outcome of an infection such as tuberculosis is the result, not of a single protective response, but of the balance between several responses. In a previous study [3] the importance of investigating the significance of all the immune responses in tuberculosis was stressed.
26
Grange and others
It is well known that many infective, traumatic, malignant and autoimmune conditions significantly affect the levels of many serum proteins. Those proteins that rapidly increase in such conditions have been named ‘acute phase reactants’ or ‘acute phase associated proteins’. Some of these have defined functions such as the binding of metals or the neutralization of bacterial proteases but in general they have received little attention from immunologists on account of the apparent non-specificity of their effects. On the other hand, recent suggestions that these proteins may suppress or otherwise affect some specific immune responses [4, 5, 6, 71 encourage a closer look at their role in immunological reactions. The purpose of the present study is to investigate the behaviour in tuberculosis of acute phase reactants in order to determine whether their estimation might be of clinical value, and to look for correlations with other immune reactions in this disease. Materials and methods Patients
and control
subjects
A total of 107 patients with smear positive pulmonary tuberculosis and 144 control subjects were included. The patients were aged between 18 and 50, mean age 33.6 years. The control subjects were healthy volunteers with no previous history of tuberculosis and no radiological evidence of the disease and were aged from 16 to 50, mean age 32.8 years. In the patient group 68 were male and 39 were female and in the control group 57 were male and 87 were female. There was no significant difference between the ages of the males and females in either patients or controls. On admission to the study all patients and control subjects were skin tested with 5 TU of PPD (RT23) and readings were taken after 6-8 hours, 24 hours and 48 hours [8, 91. Blood was taken for the estimation of the serum proteins listed below, antibodies to Mycobacterium tuberculosis, and for haematological studies (total and differential white cell count, haemoglobin level and erythrocyte sedimentation rate). The sex, age, height, weight and BCG status were recorded and an anterior-posterior radiograph was obtained. A sample of faeces was examined microscopically for evidence of intestinal parasites. Estimation
of serum proteins
The following serum proteins were estimated by nephelometry, using the Behring laser nephelometer and commercially available antisera (Behring and Seward) : Total immunoa2-macroglobulin, transferrin, globulins in the IgG, IgM and IgA classes, ctl -antitrypsin, ceruloplasmin al -acid glycoprotein (orosomucoid), C-reactive protein, haptoglobin and the third component of complement (C3). Estimation
of antibodies
to Mycobacterium
tuberculosis
Antibodies in the IgG, IgM and IgA classes were measured by the microplate ELISA technique as described previously [lo, 11 1. The antigen was ultrasonically disrupted BCG which was partially clarified by low-speed centrifugation. The conjugates were prepared by linking antisera to the heavy chains of human IgG, IgM and IgA to horse radish peroxidase by the periodate method. The colour was developed with 2,2’-azino-di-(3-ethyl)-benzothiazoline-6sulphonic acid (ABTS) and the plates were read in a Titertek Multiskan photometer (Flow Laboratories). The results were expressed in arbitrary ELISA units i.e. optical density x 100, giving a range of 0 to 2000. Radiological
studies
The anterior-posterior
radiographs were examined by Professor J. G. Scadding and, according
Acute
phase reactants
in tuberculosis
27
to radiological appearance, the disease was classified as follows : duration (recent or chronic) ; activity (active, moderately active, indolent) ; fibrosis (none, moderate, extensive) ; cavity (none, recent origin, long-standing) ; extent of active disease (<2 lung zones, >2-~4 lung zones and >4 lung zones). One patient Statistical
was excluded
from this part of the study
as the radiograph
was missing.
methods
The significances of the differences in the protein levels between patients and controls were assessed by the Mann Whitney test. The correlations between the levels of different proteins within each group were calculated by the Spearman’s rank correlation coefficient. The partial correlation coefficients (partial r) were also calculated [12]. This procedure corrects for the influence of association with the levels of other proteins on each correlation. The differences in protein levels between patients grouped according to radiological features, age and weight were studied by variance analysis. The acute phase reactant levels were examined by stepwise discriminant analysis to determine the combinations that gave the best separation of patients from control subjects. They were also subjected to cluster analysis using the Brompton Hospital Prime computer system. The weights of the control subjects were compared according to sex, age and height with the average weights of United States citizens [13]. The average weight of the Indonesian controls was 11 .9 % below that of US citizens and this diminution was virtually constant throughout the ranges of height and age in both sexes. This correction factor was therefore used to compare the actual to the expected weights. For the purpose of analysis, both patients and controls were divided into 3 groups, of equal numbers, representing the upper, middle and lower range of weights, relative to expected weight. For the analysis of the effects of age, the patients and controls were divided into 4 groups, of almost equal numbers, aged under 24,24-30,31-40 and 41-50.
Results Comparison
of protein
levels in patient
and control
groups
The levels of all the proteins, the ESR, haemoglobin levels and white cell counts (total and lymphocyte) were significantly different in the control and patient group. The range, median and mean values and the standard deviations are shown in Table I. The levels of most of the proteins were elevated in tuberculosis but those of transferrin and haemoglobin were reduced The difference in ceruloplasmin levels between patients and controls was significant (~~0.02) while the differences between all other protein levels were highly significant (ptO.OO1). The ESR and the total white cell count were elevated in disease (p
28
Grange and others
Table I. Mean, standard deviation, haematological factors.
median and range of the serum proteins, antibodies
and
Test
Control -
ul -antitrypsin u2-macroglobulin Transferrin Ceruloplasmin al -acid glycoprotein C-reactive protein Haptoglobin Complement C3 IgG antibody IgM antibody IgA antibody IgG total IgM total IgA total ESR Haemoglobin Leucocyte count Lymphocyte count *The
distribution
g. litre g. litre g. litre g. litre g. litre g. litre g. litre g, litre ELISA units ELISA units ELISA units g. litre g. litre g. litre mm. hr. g. litre cells. mm-s cells. mm-s of these
factors
Mean
Std. Dev.
Median
Range
2.56* 1.39* 2.24* 0.23* 0.78* 0.06* 0.35* 0.57* 323 445 166 17.50 1.54* 2.76* 14.9* 13.5 7183 2435
1.82 0.89 0.49 0.27 0.37 0.08 0.41 0.35 228 227 84 6.88 1.31 1.19 16.5 1.2 2041 1128
2.10 1.25 2.29 0.13 0.74 0.04 0.22 0.47 300 403 154 16.20 1.15 2.72 10 13.6 6800 2310
0.35-9.50 0.14-4.50 0.24-3.98 0.03-20.0 0.28-2.44 0.02-0.86 0.02-3.40 0.1 O-l .87 16-l 496 83-l 145 15- 539 0.70-37.50 0.12-8.75 0.18-8.44 I-103 10.7-I 6.0 840-I 2700 240-8640
showed
a marked
skewness,
mostly
Mean
Std. Dev.
Median
4.10 1.93 1.83* 0.29* 1.14 0.14* 0.42* 0.73 636* 326* 290* 21.20 2.12 3.19* 64 12.0 8409 2117
1.82 0.73 0.57 0.37 0.45 0.32 0.37 0.37 1091 166 299 7.78 1.02 1.36 35.2 1.7 2373 975
4.05 2.00 1.78 0.18 1 .I0 0.08 0.26 0.68 1104 630 328 20.50 1.95 2.92 65 12.7 8400 2000
in the form
of a strong
positive
Range
0.55-8.65 0.28-3.95 O-88-4.12 0.03-3.12 0.09-2.70 0.02-1.16 0.06-2.20 0.09-1.90 66-l 999 50-I 814 19-l 999 5.20-52.50 0.20-6.75 0.30-8.48 1-123 7.0-I 5.6 3800-l 6000 300-4420
tail.
correlation with the other proteins in disease. In two cases among the controls, namely ccl antitrypsin with ul -acid glycoprotein and haptoglobin with C3, a direct correlation became an inverse one when the partial r was calculated. On discriminant analysis of the acute phase reactants the best (although not particularly good) separation was obtained by taking al -antitrypsin, al -acid glycoprotein and transferrin in combination. By this method 39 (36.5 %) of patients and 19 (13 %) of control subjects were misclassified. This total incidence of 23.1 % misclassifications approximates closely to the theoretical incidence of 24.5 % on the null hypothesis that either form of misclassification (i.e. patient as control or vice versa) is equally significant. Further analysis of the mismatched subjects produced a discriminant function based on levels of crl -antitrypsin, transferrin and C-reactive protein. On application of this discriminant only 5 patients and 3 controls remained wrongly placed. There were no unusual features in 3 of the 5 mismatched patients but the other 2 were very ill, one with a large pleural effusion or empyaema and one with extensive miliary-like shadowing throughout both lungs. Of the 3 mismatched controls 2 were normal in all other respects but one had a very high lymphocyte count (8640 cells mm-3). Cluster analysis did not distinguish reactant protein levels.
any groups of individuals
on the basis of acute phase
Table III lists the correlations between the levels of the serum proteins (acute phase reactants and immunoglobulins) and the levels of anti-mycobacterial antibodies. In the control group, correlations between levels of acute phase reactants and antibodies were most evident with the IgG, less with the IgA and least with the IgM class. Antibody levels in all 3 classes showed a low inverse correlation with total IgA levels. These correlations disappeared in disease, with the exception of that between haptoglobin and IgA antibody.
Acute
Table II. Simple and partial correlations coefficients
phase reactants
in tuberculosis
between the acute phase reactant proteins. Test
--
Control -_
Of
With
Simple
Partial
Simple
Partial
AIA
A2M TRA CER AGP CRP HAP c3
0.259** 0.334*** 0.650*** -
0.253** -
0.771***
0.304** -
0.407*** 0.672*** 0.809*** -
0.870*** 0.344*** 0.399*** -
TRA CER AGP CRP HAP c3
0.305*** 0.353*** 0.263** -
;403*** 0.628*** 0.835*** -
1 0.464*** -
Correlation
A2M
TRA
CER
CER AGP CRP HAP c3
-0.287*** -0.390* 0.661*** 0.295*** -
* *
-0.166* -
AGP CRP HAP c3
0.188*** -
CRP
0.585*** 0.21 o* 0.229* *
CRP HAP c3
----
~___HAP c3
;383*** 0.374*** 0.206* -
0.257*** --_--_ -
-0.189* -0.324*** 0.379*** -0.203*
-0.215*
AIA=ul -antitrypsin ; ; CER=ceruloplasmin HAP= haptoglobin; *=p
0.263**
--
0.195* -
-
0.370*** 0.209*
0.247* -
0.328*** 0.442* -
0.240* ----_-_
0.655*** 0.442*** 0.203* -----~--~-------_ -0.494*** --__
c3
0.267**
0.259*
-----HAP
0.409***
-
-___ AGP
29
-0.267**
A2M=u.2-macroglobulin; AGP= al -acid glycoprotein ; CB=third component of complement. **=p
0.638*** 0.348*** 0.249*
0.395*** -
0.206* ---___--__
-
0.310**
0.230*
TRA=transferrin; CRP-= C-reactive
protein;
***~-p
Table IV show the correlations between the levels of the acute phase reactants and the total IgM levels. In all cases the correlations were closer in the patients than in the controls. No correlations were found between the acute phase reactants and total IgG while in the case of IgA there were correlations with haptoglobin (r 0.370, p
of weight and its effect on protein levels
Using the method described above the patients showed marked weight loss. Only 2 patients were above the mean control weights and 62 (58 %) were below the lower 5 % limit of the controls. There was very little relationship between weight and serum protein levels. Among
30
Grange and others
Table III. (Significant
Correlation between antimycobacterial correlations are underlined.)
antibody
Control JgM
al -antitrypsin
0.339***
-0.016
a2-macroglobulin Transferrin
0.105 0.073
0.021 -0.002
al acid glycoprotein C-reactive protein
0.344***
0.127
-0.093 0.299***
-0.112 0.012 -0.080
Haptoglobin
-m***
c3
-0.242*
Total IgM
0.113 -0.320*
Total IgA *=p
Table IV.
**=p
Correlations
Is.4
0.180* 0.121 -0.055 0.129
serum
proteins.
lgM
lg.4
0.107 0.074 0.100 0.025
-0.064 -0.032 -0.054 -0.096
0.032 -0.306**
0.077 0.060 -0.140 -0.100 -0.003
-0.295*** 0.072
-0.260** 0.224i
**
0.088 -0.207*
0.111
***=p
fgG
0.065 0.299*** -0.295***
*
0.236**
Total IgG
and
Patient -
fgG
Ceruloplasmin
levels
*
-0.049
-0.211*
0.068
-0.052 -0.005 -0.204*
0.028
0.092 0.043 0.149 -0.197*
0.148
-0x
0.056
-0.084
0.119 0.045
0.084 0.114
Assessed one-tailed.
between serum proteins and total IgM levels.
Protein
cd -antitrypsin a2-macroglobulin Transferrin Ceruloplasmin al -acid glycoprotein CRP Haptoglobin c3 NS= Not significant (p>O.O5).
Control
Patients
Correl. coeff.
p
0.290 0.391 0.019 0.237 0.421 0.355 0.094 0.151
Corn/. coeff. 0.515 0.524 0.176 0.495 0.453 0.548 -0.146 0.485
p
Assessed one-tailed.
the controls there was a tendency for ceruloplasmin and haptoglobin levels to be higher (p
Individuals in each group were placed into the 4 age groups described in the materials and methods section and were compared by variate analysis. Among the controls, levels of specific antibody in the IgM class and the leucocyte count decreased with increasing age (~~0.01) and a1 -antitrypsin levels increased (~~0.02). Haemoglobin levels were lowest in the 24-30 age group. Among the patients only C3 levels were significantly affected by age, being lower in the older patients (~~0.05).
Acute phase reactants in tuberculosis Differences
31
due to sex
Among the female controls there were higher levels of ul -acid glycoprotein (ptO.OOl), C3 (p
helminthiasis
Eleven of the 144 controls (7.6 %) and 17 of the 107 patients (16 %) had microscopical evidence of intestinal helminthiasis. The higher number among patients was not statistically significant. All cases among the controls were due to Ascarislumbricoides. Among the patients 15 cases were due to A. lumbricoides, 1 to Ancylostoma duodenale and 1 to Trichuria trichuris. The presence of worms had no effect on serum proteins among the control subjects. Among the patients, those with helminthiasis tended to have higher levels of total IgM (~~0.05). Effect of previous KG There was no significant the control subjects.
vaccination association
of protein levels to BCG status in either the patients or
Effect of skin test reactivity Among the patients there was no significant correlation between protein levels and the diameters of the skin test readings at 6-8 hours or at 48 hours. The 24 hour readings showed a negative correlation with transferrin (r -0.251, p
studies
The number of patients in the disease categories listed in the materials and methods section are shown in table V. Age, weight andsex. Chronic disease, predictably, occurred in a slightly older age group (mean 35.1, standard deviation 10.1 years) than recent disease (mean 31 .5, standard deviation 7.5 years;p<0.05). The occurrence of fibrosis was similarly age-related (p
32
Grange and others
Table V. Radiological the patients.
features and their relation to the duration
Radiological
Total in
Duration
features
each
(years)
of symptoms
category
1
2
3
4
of symptoms as reported by P
Duration on Radiology
P
5
6
>6
Recent
Chronic
2 1
-
-
Duration
Recent Chronic
48 58
30 6
IO 8
3 14
1 9
1 3
1 7
Extent of active disease
Minimal Moderate Extensive
26 46 34
4 19 13
3 10 5
7 6 4
3 3 4
0 2 2
4 2 2
5 i Not 4 significant 4
9 24 15
17 I Not 22 significant 19
Moderate Extensive Minimal
55 22 29
16 182
34
6 22
2 02
5 30
8 <0.02 50 1
20 271
35
38 41 27 106
20 133
3 122 17
1 72 10
1 21 4
1 52 8
34 140
4
Fibrosis
Cavity Total
Recent Old None
36
810 64 9 54 18
3 13
48
-
Duration of symptoms. Table V shows that the duration of disease as assessed radiologically correlated significantly with the duration of symptoms as reported by the patient. There was also a definite association between the presence of fibrosis and the duration of symptoms. Thus fibrosis was only observed in half the 36 patients with symptoms present for less than one year but in all patients reporting symptoms for 4 or more years. Cavitation also bore a relationship to duration of symptoms but the extent of disease did not. Proteins. The levels of transferrin and specific antibody in the IgG class were significantly higher in chronic disease than in recent disease (~~0.02 and <0.005 respectively). They were also both higher in patients with fibrosis than in those without (~~0.05) but IgG antibodies only were significantly higher in those with long-standing cavities (~~0.05). The levels of specific IgG antibodies were also associated with the duration of disease as reported by the patient, but at a much lower level of significance (~~0.05). Levels of haemoglobin (p
and
ctl -acid
glycoprotein
were
lower
in patients
with
cavities
Levels of specific IgA antibodies, the ESR and the leucocyte count were higher in the more extensive disease (~~0.05,
studies
The relationships between haematological values the levels of acute phase reactants and antimycobacterial antibodies are shown in table VI. Several correlations between the ESR and haemoglobin and certain proteins disappear in disease while negative correlations appear between the levels of transferrin and the ESR, haemoglobin and leucocyte count, and particularly between the lymphocyte count and haptoglobin levels. There were significant associations between the extent of disease and both the ESR and the leucocyte count (pt0.001 and co.005 respectively), both being higher in more extensive disease. In both health and disease there was a negative relationship between the ESR and haemoglobin levels. There were no significant immunoglobulin levels.
correlations
between
the haematological
features and the total
Acute
phase reactants
Table VI. Correlations of haematological features with acute bacterial antibodies. (Significant correlations are underlined.) Control
in tuberculosis
phase reactants
33
and antimyco-
Patient ---
ESR
cd -antitrypsin
Haemoglobin
-0.228*
*
-0.117
a2-macroglobulin
-0.043
Transferrin
-0.086
-0.052
Ceruloplasmin al -acid glycoprotein
-0.117
-0.005
0.107
-0.328***
C-reactive protein
-0.264
Haptoglobin
-0.001
c3
-0.067
0.282*** * *
0.042 -0.197*
Leococyte count
Lymphocyte count
-0.197
-0.055
0.074 -0.035 0.014 -0.081
ESR
-0.051
Haemoglobin 0.013
0.052
-0.036
0.077
-0.389***
-0.008 0.271**
0.085
-0.172
-0.001
0.004
-0.042
Leucocyte count -0.059 -0.247* -0.106 0.017
Lymphocyte count -0.089 0.117 -0.184 0.003 -0.158
-0.026
0.073
-0.051
-0.010
-0.083
-0.086
0.052
0.065
-0.068
-0.079
-0.112
-0.041
-0.416*** -0.049
0.041
-0.133
0.126
0.007
-0.215*
-0.075
0.048
0.099
IgM
0.050
-0.388***
-0.083
0.049
0.174
-0.118
0.081
0.078
0.058
-0.218**
-0.027
0.159
-0.211*
o-249*
0.170
antibody
IgA antibody *p
**p
0.271**
;
***p
0.232*
*
-0.032 Assessed
-0.107
0.151
0.053
IgG antibody
0.035
one-tailed.
Discussion The protective immune response in tuberculosis is generally regarded as being of the ‘cell mediated’type as opposed to one involving antibody, complement or other humoral factors [14] The effector cell in tuberculosis is the activated macrophage which, in a successful immune response, appears to engulf and destroy tubercle bacilli [15]. Although serum proteins and other soluble factors may not be the prime factors involved in the immune response to tuberculosis, there is growing evidence that they may have a substantial effect, either beneficial or detrimental, on cellular immune functions in both infectious and malignant diseases. In tuberculosis a number of specific and non-specific changes in the cellular and humoral arms of the immune system occur: these include cell-mediated reactions that manifest as a positive tuberculin test, the production of antibodies to Mycobacterium tuberculosis, a nonspecific increase in immunoglobulin levels and the appearance of autoantibodies [3]. The levels of the acute phase reactants increase in a number of physiologically stressful conditions such as infections, injuries (whether accidental or surgical), burns and myocardial infarctions. Despite the name ‘acute phase’, the levels of some of these proteins are also elevated in chronic conditions such as leprosy [I 6, 171 and intestinal helminthiasis [18]. There is no precise definition of an acute phase reactant nor is there any general consensus of opinion as to which proteins should be included. Koj [I91 has defined them as ‘traumainducible liver-produced plasma glycoproteins’ and regarded ctl -antitrypsin, ceruloplasmin, ~1 -acid glycoprotein, C-reactive protein and haptoglobin as belonging to this group. Creactive protein, however, is not a glycoprotein [7]. Sipe and Rosenstrich [7] divided acute phase reactant proteins into two groups : early and late. The early group contains endogenous pyrogen, serum amyloid A inducer and tumour necrosis factor. The late group includes Creactive protein and the glycoproteins referred to above. Opinions differ, however, as to whether x2-macroglobulin, transferrin and components of complement should be included.
34
Grange and others
In this study levels of x2-macroglobulin correlated significantly with those of some of the recognized acute phase reactants, notably ~1 -antitrypsin and C-reactive protein in both patients and control subjects. Although elevated in cases of tuberculosis, a2-macroglobulin tended to be lower in tuberculin positive than in tuberculin negative controls. There is evidence that a2-macroglobulin binds to, and inactivates, proteases liberated during tissue injury, necrosis and inflammation and that the resulting complex stimulates the biosynthesis of acute phase reactants by the liver [20]. This does not, however, explain why its levels are affected by tuberculin reactivity among the control subjects. Levels of or1-antitrypsin were found to be significantly elevated in pulmonary tuberculosis, thus confirming the findings of other workers [21, 22,231. In leprosy, on the other hand, this enzyme was not found to be significantly elevated except in those patients with erythema nodosum leprosum [24]. This is probably due to the occurrence of much less tissue destruction and inflammation in leprosy than in tuberculosis [25] except when hypersensitivity reactions occur in the former disease. Low levels of this enzyme are found in patients with chronic obstructive airway disease (COAD), and it has been postulated that this and related proteins prevent lung damage by inactivating proteases liberated by bacteria and macrophages [26,27]. In this study 3 patients had radiological evidence of emphysema but their al-antitrypsin levels were well within the normal range. No relationship between cavity formation and fibrosis and levels of al -antitrypsin was found. There is little doubt, however, that the mechanisms of lung damage and repair in tuberculosis and COAD are quite different. Although leprosy and tuberculosis differ with respect to al-antitrypsin levels they show similarities in other respects. Thus levels of total immunoglobulins are increased in leprosy although, as in tuberculosis, the classes involved appear to differ from study to study [I 61. In this study only the third of the complement components (C3) was quantitated and was found to be elevated in tuberculosis. Furthermore, its concentration correlated significantly with levels of ccl -antitrypsin in the patients but not in the controls. The fourth component (C4) is not, however, elevated in tuberculosis [3]. Likewise in leprosy increases occur in C3 but not in C4 levels [I 61. On the other hand, C4 is elevated in many acute conditions [28]. Transferrin levels were reduced in tuberculosis and they did not correlate with those of any of the other acute phase reactants in the control group with the exception of a very low inverse correlation with ceruloplasmin. This protein is, unlike other acute phase reactants, synthesised by macrophages and its levels drop in many disease states due to an excess of catabolism over synthesis [29, 301. Ceruloplasmin and haptoglobin are both involved in the transport of metal ions and are both stimulated by derivatives of arachidonic acid, which is also the precursor of prostaglandins. In rabbits with carcinoma there was a correlation between the levels of these proteins and prostaglandin E2 [31 ] but in this study there was an inverse correlation between the proteins in the controls and no significant correlation in disease. The inverse correlation in health could be due to competition during synthesis for a single precursor molecule. The significance of the correlations between the various proteins and the differences in health and disease are at present poorly understood as relatively little is known about the factors that control them, Hopefully the findings of this study will become of relevance when they can be considered in relation to in vitro measurements of cell-mediated immune reactions It is noteworthy, however, that while the levels of C3 tended to correlate more closely with other proteins in disease, the reverse was the case with those of other acute phase reactants. A remarkable exception, however, was the effect of disease on the relation between levels of ctl -antitrypsin and ceruloplasmin. The relationship of levels of acute phase reactants to those of total immunoglobulins and the
Acute phase reactants in tuberculosis
35
antimycobacterial antibodies was rather unusual. Thus, correlations between acute phase reactants and antibody levels among the control subjects was most apparent with the IgG class and least with the IgM class while the reverse was the case with the total immunoglobulins in these classes. Furthermore, the correlations with antibody levels tended to be lost in disease while those with total IgM levels increased. A further unexpected correlation was the low inverse one between mycobacterial antibody in the 3 classes, including IgA.
total
IgA and anti-
Although changes in serum protein are thought to be responsible for the increase in the ESR in disease, the acute phase reactants investigated in this study did not appear to play such a role. Indeed, when correlations appeared, they were negative ones. The negative correlation between the ESR and transferrin is almost certainly secondary to the negative correlation between the ESR and haemoglobin levels. The strong negative correlation between haptoglobin and the lymphocyte count in tuberculosis is of interest, and may be indicative of an important immunoregulatory role for this protein in this disease. The various correlations observed in this study, although statistically significant, were often not very close. This may be due to the fact that levels of serum proteins in a given individual are not constant but fluctuate significantly. Thus, in the rat, levels of IgG, a2-macroglobulin and transferrin showed distinct circadian rhythms, being higher at night. In addition IgG levels varied from day to day while levels of transferrin showed oscillations within each day [32]. It is possible, therefore, that closer correlations would be found by taking the mean of a number of protein assays on sera taken at different times and on different days, In comparing serum proteins and other factors in healthy controls and patients with tuberculosis, it was necessary to determine whether any differences were directly associated with the disease or whether they were secondary to other factors such as age, sex, weight and concurrent disease. The two groups were matched, as closely as possible, for age and socioeconomic status. There was a predominance of females among the controls and males among the patients and although some sex-related differences in serum protein levels were found in the control group these did not account for differences between controls and patients. Some protein levels were associated with age or sex among the controls but these associations were negated in disease. On the other hand, among the patients, levels of C3 and total IgA were affected by age and sex respectively. Very little information on the effect of age and sex on the levels of acute phase reactants in health is available but recently this was studied with respect to al -antitrypsin, al -acid glycoprotein, ceruloplasmin and transferrin in North American citizens [33]. No differences in the levels of these proteins in the age groups relevant to the present study were found. Levels of al -antitrypsin and ceruloplasmin, but not those of the other two proteins, were significantly higher among females than among males. This is in contrast to the present study in which, of the 4 proteins, only al -acid glycoprotein levels were sex-related, being higher in females. These differences may reflect racial or genetic variations between the subjects in the two studies. The overall levels of the acute phase reactants in the control group were, however, very similar to those of European adults, except for those of transferrin and haptoglobin which were lower in the Indonesian group. Weight loss and a low dietary intake due to loss of appetite might be expected to cause changes in serum protein levels. Indonesian children with protein-calorie malnutrition associated with acute infections have been shown to have decreased levels of transferrin and increased levels of haptoglobin [34]. In the present study, haptoglobin levels (as well as those of ceruloplasmin) were higher in the under-weight control subjects. In the patient group the association of weight to haptoglobin levels was not significant while the association to ceruloplasmin was reversed. It should, however, be stressed that weight loss was more
36
Grange and others
frequent and severe in the patient group, thus the lower weight group of the controls approximated to the higher weight group of the patients. Indeed, the ceruloplasmin levels in these two groups were fairly similar but decreased in both the more overweight controls and the more underweight patients. Consequently the levels of some proteins are affected by weight loss as well as by the disease process. Patients and controls with obvious concurrent diseases or other conditions likely to affect the results (e.g. pregnancy and lactation) were excluded from the study. The only detected concurrent condition was intestinal helminthiasis. Studies on Vietnamese refugees [18] showed that children and young adults with intestinal helminthiasis tended to have elevated levels of al -antitrypsin, ceruloplasmin and al -acid glycoprotein. It was thought, however, that this was mainly due to hookworm (Ancylostoma do&ens/e) infestation. In this study, helminthiasis, which was almost entirely due to Ascaris lumbricoides, had no significant association with protein levels, with the exception of a slight elevation of total IgM levels among the patients. There have been several attempts to divide cases of tuberculosis into a ‘spectrum’ of immunological reactivity similar to that seen in leprosy, and to describe criteria for allocating cases to a point on the spectrum [35, 361. On clinical grounds such a subdivision of cases is not easy. In this study the patients were classified according to radiological appearance, but the presence or absence of fibrosis and cavities had no significant effect on the levels of most of the serum proteins. The reduction of haemoglobin in patients with cavities may be due to the more frequent occurrence of haemoptysis in such patients [3]. It seems unlikely, therefore, that an estimation of these proteins will be of use in the classification of disease. Indeed the best indicators of the extent of the disease, apart from radiology, appeared to be the ESR and leucocyte counts. Several authors have reported that levels of serum immunoglobulins increase in tuberculosis and although there appear to be some regional variations in the contribution of the various classes of immunoglobulin to this increase, most workers have found that levels of IgG and IgA increase considerably while that of IgM only increases slightly [3, 37, 38, 39, 401. These increasesarenotdue to the production of antibody to M. tuberculosis but are probably due to a polyclonal activation of B cells by mycobacterial adjuvants [3]. The reason for such an activation affecting IgG and IgA much more than IgM is not clear. In African trypanosomiasis, by contrast, the polyclonal B cell activation causes an increase principally in the IgM class [41]. It is, however, of interest to note that, of the 3 classes, the levels of IgM alone correlated with those of several acute phase reactants in health and, more closely, in disease. It is possible, therefore, that this class of immunoglobulin is controlled by a mechanism that also governs the synthesis of acute phase reactants, and that this mechanism is refractory to the effect of mycobacterial adjuvants. Indeed IgM has been included among the acute phase reactants [42]. It is noteworthy that, among the patients, those with intestinal helminthiasis tended to have higher total IgM levels than those without, but that this trend did not occur among the control subjects. Whether or not this represents some form of synergy between adjuvants from helminth and mycobacteria remains to be determined but the phenomenon might account for the apparent geographical variations in the effect of tuberculosis on total IgM levels. It is, however, doubtful whether intestinal helminthiasis alone is a cause of polyclonal B cell activation [41]. In view of the immunoregulatory properties reported for various acute phase reactants, their relationship to the diameter of the skin test reactions was studied. The negative correlation between the 24 hour reaction and transferrin levels in both health and disease is of interest. The 24 hour reaction in man probably corresponds with the Listeria type reaction in mice [43] which is associated with macrophage activation. As mentioned previously, transferrin differs from the other acute phase reactants in being synthesised by the macrophage. Thus the size
Acute phase reactants in tuberculosis
37
of the 24 hour reaction and the transferrin levels may both reflect variations in macrophage function. It has been shown on several occasions since the pioneering work of Pickert and Lowenstein in 1908 [44] that serum from patients with tuberculosis may enhance or suppress the 48 hour dermal reaction to tuberculin. The responsible factors were termed ‘procutins’ and ‘anticutins’. None of the proteins examined in this study appeared to have such properties. It has been shown previously that an immediate reaction to tuberculin, presumably caused by IgE or short-term sensitizing IgG antibodies, tends to reduce the diameter of the 48 hour response [8, 451 but otherwise very little is known about the humoral factors that affect the tuberculin test. Cluster analysis of the acute phase reactant levels did not reveal any distinct groups of patients or control subjects and this statistical method did not separate the former from the latter. Stepwise discriminant analysis, on the other hand, showed that certain combinations of the levels of these proteins could be used to achieve a fairly good separation of patients from the controls. This is unlikely, however, to be of diagnostic usefulness owing to the nonspecificity of changes affecting the levels of these proteins. It is possible though that such an analysis might prove useful in the early detection of relapse after therapy or reactivation in individuals with evidence of previous infection. In conclusion, tuberculosis has been shown to be associated with significant alterations in the levels of many serum proteins and, although these changes are of no diagnostic value, they may prove to be of considerable reievance to the pathogenesis of the disease and the efficacy of the immune reponses. Thus, although humoral factors are not of prime importance in the destruction of the causative agent of tuberculosis, they should not be ignored or forgotten when considering the host-pathogen interactions in this important and prevalent disease. Acknowledgements
We are most grateful to Professor J. G. Scadding and Dr M. Oesman for reading the radiographs, Dr S. Darmadi for assistance with the laser nephelometry, Dr I. Handoyo for assistance with the ELISA studies and Dr H. Notodipuro for the haematological examinations. We also thank Dr D. lman Santoso and S. Karim and their colleagues for their enthusiastic help in the clinical studies. We also received valuable technical assistance from Miss R. Anggraini and Miss U. Haryanti. We also thank Professor John Swanson Beck and Dr Harald Mauch for their advice and encouragement; Miss Margaret Rehahn for statistical advice and for performing the stepwise discriminant analysis; Mr Stephen Saunders for writing the program for cluster analysis on the Brompton Hospital Prime computer system and the British Council for financial support of one of us (T.K.). References 1 Koch, Ft. (1891). Weitere Mitteilungen uber ein Heilmittel gegen Tuberkulose. Deursche Mediziniscbe Wochenschrift, 17,101. 2 Bloom, B. R., Tanowitz, H., Et Wittner, M. (1979). Games parasites play. Nature, 279, 21. 3 Kardjito, T., Et Grange, J. M. (1980). Immunological and clinical features of smear-positive pulmonary tuberculosis in East Java. Tubercle, 61, 231. 4 Tormey, D. C., Er Mueller, G. C. (1972). Biological affects of transferrin on human lymphocytes. Experimental Cell Research. 74. 220. 5 Mortensen, R. F., Braun, D., 8 Gewurz, H. (1977). Effects of C-reactive proteins on lymphocytefunction. III. Inhibition of antigen-induced lymphocyte stimulation and lymphokine production. Cellularlmmunology, 28, 59. 6 Bullock, W. E. (1979). Mechanisms of anergy in infectious diseases. In lmmono/ogicalAspects of infectious Diseases. Edited by G. W. A. Dick: p. 269, MTP Press, England.
38
Grange and others
7 Sipe, J. D., Et Rosenstrich, D. L. (1981). munity and fnffammations. Edited Arnold, London.
Serum factors associated with inflammation. In Cellular Functions In lmby J. J. Oppenheim, D. L. Rosenstrich Et M. Potter: p. 411, Edward
8 Kardjito. T., Donosepoetro, M., 8 Grange, J. M. (1981). The Mantoux test in tuberculosis: diameters of the dermal responses and the serum protein levels. Tubercle, 62, 31.
correlations
between
fhe
9 Kardjito, T., 8 Grange, J. M. (1982). Diagnosis of active tuberculosis by immunological methods. 2. Qualitative differences in the dermal response to tuberculin in patients with active pulmonary disease and healthy tuberculinpositive individuals. Tubercle, 63, 275. 10
11
12 13 14 15 16
17 18
Grange, J. M., Gibson, J., Nassau, E., Et Kardjito, T. (1980). Enzyme-linked immunosorbent assay (ELISA) : A study of antibodies to Mycobacterium tuberculosis in the IgG, IgA, and IgM classes in tuberculosis, sarcoidosis and Crohn’s disease. Tubercle, 61, 145. Kardjito,T., Handoyo, I., & Grange, J. M. (1982). Diagnosis of activetuberculosis by immunological methods. 1. Effect of tuberculin reactivity and previous BCG vaccination on the antibody levels determined by ELISA. Tubercle, 63, 269. Stuart, A., & Kendall, M. (1965). TheAdvanced Theory of Statistics. Vol. 2. p. 317, Charles Griffin, London. Documenta Geigy (1970). Scientific Tables, 7th edition. Edited by K. Diem & C. Lentner: p. 711, Geigy, Basle. Mackaness, G. B. (1968). The immunology of antituberculous immunity. American Review of Respiratory Disease. 97, 337. Lowrie, D. 8. (1981). Survival of intracellular bacteria inside macrophages. Anna/es d’lmmunologie (/nstitutPasteur), 132D, 151. Srivastava, L. M., Agarwal, 0. P., & Goedde, H. W. (1975). Biochemical, immunological and genetic studies in leprosy. II. Profile of immunoglobulins, complement components and C-reactive protein in sera of leprosy patients and healthy controls. Tropenmedizin und Parasitologic, 26, 212. Celikoglu, S. I., Fahir, M. G., & Turkan, 8. (1976). A preliminary report of a study on serum alpha-l -antitrypsin and immunoglobulin levels in lepromatous leprosy. Anna/s of the New York Academy of Science, 278,125. Blom, M., Prag, J. B., Et Nmrredam, K. (1979). ct2-acid glycoprotein, crl -antitrypsin and ceruloplasmin in human intestinal helminthiasis. American Journal of Tropical Medicine and Hygiene, 28, 76.
19
Koj, A. (1974). Acute phase reactants. Their synthesis, turnover and biological significance. In Structure and Function of Plasma Proteins. Vol. 1, Ch. 4, p. 73. Edited by A. C. Allison, Plenum Press, London.
20
Metcalfe, J., 8 Tavill, A. S. (1975). A proposed role for alpha-l -acute phase globulin synthesis by the perfused rat liver. British Journal of Experimental Pathology, 56, 570. Geisler, L. S., Bachmann, G. W., Lauman, F., Nolte, D., Wenzel, H., Et Rost, H. D. (1972). Alpha-l -antitrypsin and Deutsche Mediziimmunoglobulins in chronic non-specific pulmonary disease and pulmonary tuberculosis, nische Wochenschrift, 97,329. Poh, S. C., Et Seet, A. M. (1975). Alpha-I-antitrypsin levels in chronic obstructive lung disease and pulmonary tuberculosis in Singapore. Singapore Medical Journal, 16, 89.
21
22 23
Gatner,
E. M. S., Et Anderson, R. (1980). An in-vitro assessment tuberculosis: correction of defective neutrophil motility Clinical and Experimental Immunology, 40, 327.
of cellular and humoral immune function in pulmonary by ascorbate, levamisole, metoprolol and propanolol.
24
Agarwal, S. K., Et Sahah, K. (1978). Serum alpha-l -antitrypsin in various forms of leprosy. lndian Journal of Medical Research, 68, 136. 25 Skinsness, 0. K. (1968). Comparative pathogenesis of the mycobacterioses. Annals of the New York Academy of Sciences, 154, 19. 26 Lieberman, J. (1973). Alpha-l -antitrypsin deficiency. Medical Clinics of North America, 57, 691. 27 Norse, J. O., Lebowitz, M. D., Knudson, R. J., Et Burrows, B. (1977). Relation of protease inhibitor phenotypes to obstructive lung disease in a community. New England Journal of Medicine, 296, 1190. as a diagnostic aid. New England Journalof Medicine, 291,287. 28 Alper, C. A. (1974). Plasma protein measurements 29 30
Haurani, F. I., Meyer, A., 8 endothelial Society. Gordon, A. H. (1976). The 8 A. S. McFarlane,
O’Brien, R. (1973). Production of transferrin by the macrophage. Journal of the Reticufo‘I 4. 309. acute phase plasma proteins. In Plasma Protein Turnover. Edited by R. Bianchi, G. Mariani, p. 381, University Park Press, Baltimore.
31 Voelkel,
32
33 34
E. F., Levine, L., Apler, C. A., Et Tshijan, A. H. (1978). Acute phase reactants ceruloplasmin and haptoglobin and their relationship to plasma prostaglandins in rabbits bearing the VX2 carcinoma. Journalof Experimental Medicine, 147. 1078. Schade, R., Rex, H., Kastner, D., Friedrich, A., Et Sprang, C. (1980). Biorhythms of four serum proteins of the laboratory rat: IgG, transferrin, a2-acute phase protein and a VLD lipoprotein. Acta Biologica et Medica Germanica, 88, 1183. Denko, C. W., Et Gabriel, P. (1981). Age and sex related levels of albumin, ceruloplasmin, al -antitrypsin, ctl -acid glycoprotein, and transferrin. Annals of Clinical and Laboratory Science, 11, 63. Schelp,
F. P., Migasena, P., Pongpaew, P., Schreurs, W. H. P., Et Supawan, V. (1977). Serum proteinase and other serum proteins in protein-energy malnutrition. British Journal of Nutrition, 38, 31.
inhibitors
Acute 35 36 37 38 39 40 41 42 43
44 45
Lenzini,
L., Rottoli,
P., 8 Rottoli,
L. (1977).
The spectrum
phase reactants
of human tuberculosis.
Clinicaland
in tuberculosis
39
Experimenfa//mmuno/ogy
27. 230. Daniel, T. M., Oxtoby, M. J., Pinto, E. M., Et Moreno, E. S. (1981). The immune spectrum in patients with pulmonary American Review of Respiratory Disease, 123, 556. tuberculosis. Faulkner, J. B., Carpenter, R. L., B Patnode, R. A. (1967). Serum protein and immunoglobulin levels in tuberculosis. American Journal of Clinical Pathology, 48. 556. Malomo, I., McFarlane, H., & Idowu, J. (1970). Serum immunoglobulins in pulmonary tuberculosis in Ibadan, Nigeria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 64,427. Rowinska, E. (1972). lmmunoglobulin levels in pulmonary tuberculosis. Polish Medical Journal, 11, 524. Jha, V. K., Eajpai, B. K., Et Gupta, Ft. M. (1974). Levels of serum immunoglobulins in pulmonary tuberculosis patients. Indian Journal of Chest Diseases and Allied Sciences, 16, 361. Terry, R. S., Hudson, J. M., 8 Faghihi-Shrazi, M. (1980). Polyclonal activation by parasites. In The Host-Invader /nterp/ay. Edited by H. Van den Bossche: p. 259, Elsevier/North Holland Biomedical Press, Amsterdam. Coffey, R. L., Zile, M. R., Et Luskin, A. T. (1981). Immunologic tests of value in diagnosis. 1. Acute phase reactants and Postgraduate Medicine, 70, 163. autoantibodies. Rook, G. A. W., Et Stanford, J. L. (1979). The relevance to protection of three forms of delayed skin test response evoked by M. leprae and other mycobacteria in mice. Correlation with the classical work in the guinea pig. Parasite Immunology, I, 111. Deutsche Medizinische Pickert, M., Et Lowenstein, E. (1908). Eine neue Methodezur Priifung derTuberkulin-immunitit. Wochenschrift, 34. 2262. Pepys, J. (1955). The relationship Tuberculosis, 71, 49.
of nonspecific
and specific
factors
in the tuberculin
reaction.
American
Review
of
23-39
0
Ltd.
Longman
Group
A STUDY INDONESIAN
OF ACUTE-PHASE PATIENTS WITH J.
Cardiothoracic
Institute,
University
M.
of Polmonology
Grange
of London, Fulham Road, London SW3 6HP, U.K.
T. Kardjito Departments
REACTANT PROTEINS IN PULMONARY TUBERCULOSIS
and
I. Setiabudi
and Clinical Pathology, Faculty of Medicine, Surabaya. Indonesia
Airlangga
University,
Summary The levels of 8 acute phase reactants (ul -antitrypsin, a2-macroglobulin, transferrin, al -acid glycoprotein, C-reactive protein, ceruloplasmin, haptoglobin and the third component of complement) and immunoglobulin in the IgG, IgM and IgA classes were assayed, by laser nephelometry, in sera from 107 East Javanese patients with smear-positive pulmonary tuberculosis and 144 healthy subjects. These levels were correlated with clinical, haematological and radiological features, the levels of antibody to Mycobacterium tuberculosis, and the diameters of the tuberculin skin test read at various times. Levels of all acute phase reactants increased significantly in tuberculosis except for that of transferrin which was lowered. The correlations between the various acute phase reactants in health and disease were calculated. In general, the correlations were lower in disease than in health, except for the third component of complement and a greatly increased correlation between the levels of al -antitrypsin and ceruloplasmin. There reactants and least and total IgA and
was a significant correlation between levels of some of the acute phase and those of antibodies to M. tuberculosis, mainly with IgG, less with IgA with IgM antibodies. By contrast correlations between acute phase protein immunoglobulin levels were most evident in the IgM class, less with not at all with IgG.
Although there were some associations between protein levels and age, sex and weight of controls and patients, these were not great enough to account for the differences between the two groups. There was a tendency for patients, but not controls, with intestinal helminthiasis to have higher levels of total IgM than those without evidence of parasites. In general, the levels of proteins bore very little relation to the clinical and radiological features of disease and were, with the exception of the antimycobacterial antibodies, of no diagnostic value. Likewise, protein levels were not associated with the extent of disease; better correlations were found with the ESR and leucocyte count. Transferrin levels tended to be higher in those with chronic disease and showed a correlation with the diameters of the dermal reactions to tuberculin at 24 hours, which were also significantly larger in chronic disease. Among the haematological findings, the most significant between the lymphocyte count and haptoglobin levels possible regulatory role for this protein.
was a negative correlation in disease, suggesting a
24
Grange and others Rirsum6
Par nephelometrie du laser, on a etudie le serum de 107 malades de I’est de Java atteints de tuberculose pulmonaire frottis possitif et celui de 144 sujets sains. On a mesure les taux de 8 substances de reaction de la phase aigue (ccl -antitrypcine, a2-macroglobuline, transferrine, ctl -glycoproteine acide, proteine C reagissante, ceruloplasmine, haptoglobine et le troisieme composant du complement) ainsi que les immunoglobulines des classes IgC, IgM et IgA. On a procede a des correlations entre ces taux et le elements cliniques, hematologiques et radiologiques, les taux d’anticorps vis a vis de M. tuberculosis et les diametres du test cutane a la tuberculine lus a differents moments. Les taux des substances de reaction de la phase aigue Btaient significativement augment& dans la tuberculose, a I’exception de la transferrine qui Btait diminuee. L’article discute les correlations entre les diverses substances de la phase aigue chez les malades et chez les sujets sains. D’une man&e generale, les correlations se sont montrees plus faibles en cas de tuberculose que lors de la bonne Sante, sauf pour le troisieme composant du complement et il existait en cas de tuberculose une correlation grandement augmentee entre les taux de crl -antitrypsine et ceruloplasmine. II existait une correlation significative entre les taux de certaines substances reagissantes de la phase aigue et les taux des anticorps de M. tuberculosis, essentiellement avec les anticorps des IgG, moins avec ceux de IgA et le moins avec ceux des IgM. A I’inverse, les correlations entre les taux de proteines de la phase aigue et les taux totaux d’immunoglobulines dtaient les plus marquees pour la classe de IgM, moins pour IgA et pas du tout pour les IgG. Bien qu’il existait quelques associations entre les taux de proteine et I’bge, le sexe et le poids des temoins et des malades, elles n’etaient pas assez fortes pour expliquer les differences entre les deux groupes. Les malades, mais non les temoins, porteurs d’une helminthiase intestinale avaient tendance a posseder des taux plus eleves de IgM totaux que ceux chez lesquels on n’avait pas trouve ces parasites. En general, les taux de proteine ont montre peu de rapport avec les elements cliniques et radiologiques de la maladie et sauf en ce qui concerne les anticorps antimycobacteriens, ces taux n’avaient pas de valeur diagnostique. Egalement, les taux de proteines n’btaient pas associes avec I’extension de la maladie; une meilleure correlation a et6 trouvee avec la VHS et la numerotation de leucocytes. Les taux de transferrine avaient tendance a dtre plus eleves chez les sujets porteurs d’une tuberculose chronique et montraient une correlation avec les reactions dermiques a la 248me heure, lesquelles Btaient Bgalment significativement plus grandes chez les malades chroniques. le plus significatif a et6 I’existance d’une Parmi les resultats hemotologiques, correlation negative entre les comptes de lymphocytes et les taux d’haptoglobine chez les malades, ce qui suggere un role regulateur possible de cette proteine. Resumen Se estudid el suero de 107 enfermos del este de Java con tuberculosis pulmonar con baciloscopia positiva y el de 144 sujetos sanos mediante nefelometria de rayos laser. Se midid 10s niveles de sustancias de reacci6n de la fase aguda (al -antitripsina, a2-macroglobulina, transferina, al -glucoproteina Bcida, proteina C reactiva, ceruloplasmina, haptoglobina y el tercer componente del complemento) asi coma las inmunoglobulinas de las clases IgC, IgM, e IgA. Estos niveles fueron puestos en
Acute correlation con 10s elementos anticuerposal M. tuberculosisy a diferentes momentos.
clinicos,
phase reactants
in tuberculosis
25
hematologicos y radiolbgicos, 10s niveles de de la prueba cutanea de tuberculina leida
10s ditimetros
Los niveles de todas las sustancias de reaction de la fase aguda estaban significativamente aumentados en la tuberculosis except0 el de la transferina que estaba disminuido. Se calcularon las correlaciones entre las diversas sustancias de la fase aguda en 10s enfermos y en 10s sujetos sanos. En general las correlaciones fueron. m&s bajas en 10s enfermos que en 10s sanos, salvo por el tercer componente del complemento y, en el case de tuberculosis, un gran aumento de la correlation entre 10s niveles de al -antitripsina y ceruloplasmina. Habia una correlation significativa entre algunas de las sustancias de reacion de la con 10s antifase aguda y 10s niveles de anticuerpos al M. tuberculosis, principalmente cuerpos IgG, menos con 10s IgA y afin menos con 10s IgM. Al contrario, las correlaciones entre 10s niveles de proteinas de la fase aguda y 10s niveles totales de inmunoglobulinas fueron m&s evidentes con la clase IgM, menos con la IgA y absolutamente ausentes con la IgG. A pesar de que existian algunas asociaciones entre 10s niveles de proteinas y la edad, el sexo y el peso de 10s sujetos control y de 10s enfermos, Bstas no eran suficientemente importantes coma para explicar las diferencias entre 10s dos grupos. Los enfermos, pero no 10s sujetos control, portadores de una helmintiasis intestinal tenian tendencia a presentar niveles mas elevados de IgM totales que aquellos sin evidencia de esta parasitosis. En general, 10s niveles de proteinas mostraron una relation escasa con lo elementos clinicos y radiologicos de la enfermedad y, con excepcidn de 10s anticuerpos antimicobacterianos, estos niveles no tenian valor diagnostico. Asi mismo, 10s niveles de proteinas no estaban correlacionados con la extension de la enfermedad ; se encontrd una mejor correlation con la VHS y el recuento de leucocitos. Los niveles de transferina tenian tendencia a ser mas elevados en 10s sujetos con una tuberculosis crdnica y mostraron una correlation con las reeacciones cutaneas a las 24 horas, las cuales eran tambien significativamente mas grandes en 10s enfermos crbnicos. Entre 10s resultados hematologicos, el m&s significativo fue la correlation negativa entre el recuento de linfocitos y 10s niveles de haptoglobina en 10s enfermos, lo que sugiere un posible papel regulador de esta proteina. Introduction The immunology of tuberculosis has been intensively studied by numerous workers since Robert Koch’s pioneering work in 1891 [I 1. Until recently cell mediated and humoral (i.e. antibody dependent) immune responses were often regarded as separate and independent phenomena but it is now realized that all immunological reactions are closely inter-related. It is also appreciated that a challenge by a complex set of antigens such as a mycobacterium elicits many different reactions, not just specific immunological ones, and that some of these are beneficial, some are irrelevant and some may harm the host more than the pathogen. Indeed several widespread protozoa1 and bacterial pathogens owe their success to their ability to utilize the latter type of host responses to their own advantage [2]. Accordingly the outcome of an infection such as tuberculosis is the result, not of a single protective response, but of the balance between several responses. In a previous study [3] the importance of investigating the significance of all the immune responses in tuberculosis was stressed.
26
Grange and others
It is well known that many infective, traumatic, malignant and autoimmune conditions significantly affect the levels of many serum proteins. Those proteins that rapidly increase in such conditions have been named ‘acute phase reactants’ or ‘acute phase associated proteins’. Some of these have defined functions such as the binding of metals or the neutralization of bacterial proteases but in general they have received little attention from immunologists on account of the apparent non-specificity of their effects. On the other hand, recent suggestions that these proteins may suppress or otherwise affect some specific immune responses [4, 5, 6, 71 encourage a closer look at their role in immunological reactions. The purpose of the present study is to investigate the behaviour in tuberculosis of acute phase reactants in order to determine whether their estimation might be of clinical value, and to look for correlations with other immune reactions in this disease. Materials and methods Patients
and control
subjects
A total of 107 patients with smear positive pulmonary tuberculosis and 144 control subjects were included. The patients were aged between 18 and 50, mean age 33.6 years. The control subjects were healthy volunteers with no previous history of tuberculosis and no radiological evidence of the disease and were aged from 16 to 50, mean age 32.8 years. In the patient group 68 were male and 39 were female and in the control group 57 were male and 87 were female. There was no significant difference between the ages of the males and females in either patients or controls. On admission to the study all patients and control subjects were skin tested with 5 TU of PPD (RT23) and readings were taken after 6-8 hours, 24 hours and 48 hours [8, 91. Blood was taken for the estimation of the serum proteins listed below, antibodies to Mycobacterium tuberculosis, and for haematological studies (total and differential white cell count, haemoglobin level and erythrocyte sedimentation rate). The sex, age, height, weight and BCG status were recorded and an anterior-posterior radiograph was obtained. A sample of faeces was examined microscopically for evidence of intestinal parasites. Estimation
of serum proteins
The following serum proteins were estimated by nephelometry, using the Behring laser nephelometer and commercially available antisera (Behring and Seward) : Total immunoa2-macroglobulin, transferrin, globulins in the IgG, IgM and IgA classes, ctl -antitrypsin, ceruloplasmin al -acid glycoprotein (orosomucoid), C-reactive protein, haptoglobin and the third component of complement (C3). Estimation
of antibodies
to Mycobacterium
tuberculosis
Antibodies in the IgG, IgM and IgA classes were measured by the microplate ELISA technique as described previously [lo, 11 1. The antigen was ultrasonically disrupted BCG which was partially clarified by low-speed centrifugation. The conjugates were prepared by linking antisera to the heavy chains of human IgG, IgM and IgA to horse radish peroxidase by the periodate method. The colour was developed with 2,2’-azino-di-(3-ethyl)-benzothiazoline-6sulphonic acid (ABTS) and the plates were read in a Titertek Multiskan photometer (Flow Laboratories). The results were expressed in arbitrary ELISA units i.e. optical density x 100, giving a range of 0 to 2000. Radiological
studies
The anterior-posterior
radiographs were examined by Professor J. G. Scadding and, according
Acute
phase reactants
in tuberculosis
27
to radiological appearance, the disease was classified as follows : duration (recent or chronic) ; activity (active, moderately active, indolent) ; fibrosis (none, moderate, extensive) ; cavity (none, recent origin, long-standing) ; extent of active disease (<2 lung zones, >2-~4 lung zones and >4 lung zones). One patient Statistical
was excluded
from this part of the study
as the radiograph
was missing.
methods
The significances of the differences in the protein levels between patients and controls were assessed by the Mann Whitney test. The correlations between the levels of different proteins within each group were calculated by the Spearman’s rank correlation coefficient. The partial correlation coefficients (partial r) were also calculated [12]. This procedure corrects for the influence of association with the levels of other proteins on each correlation. The differences in protein levels between patients grouped according to radiological features, age and weight were studied by variance analysis. The acute phase reactant levels were examined by stepwise discriminant analysis to determine the combinations that gave the best separation of patients from control subjects. They were also subjected to cluster analysis using the Brompton Hospital Prime computer system. The weights of the control subjects were compared according to sex, age and height with the average weights of United States citizens [13]. The average weight of the Indonesian controls was 11 .9 % below that of US citizens and this diminution was virtually constant throughout the ranges of height and age in both sexes. This correction factor was therefore used to compare the actual to the expected weights. For the purpose of analysis, both patients and controls were divided into 3 groups, of equal numbers, representing the upper, middle and lower range of weights, relative to expected weight. For the analysis of the effects of age, the patients and controls were divided into 4 groups, of almost equal numbers, aged under 24,24-30,31-40 and 41-50.
Results Comparison
of protein
levels in patient
and control
groups
The levels of all the proteins, the ESR, haemoglobin levels and white cell counts (total and lymphocyte) were significantly different in the control and patient group. The range, median and mean values and the standard deviations are shown in Table I. The levels of most of the proteins were elevated in tuberculosis but those of transferrin and haemoglobin were reduced The difference in ceruloplasmin levels between patients and controls was significant (~~0.02) while the differences between all other protein levels were highly significant (ptO.OO1). The ESR and the total white cell count were elevated in disease (p
28
Grange and others
Table I. Mean, standard deviation, haematological factors.
median and range of the serum proteins, antibodies
and
Test
Control -
ul -antitrypsin u2-macroglobulin Transferrin Ceruloplasmin al -acid glycoprotein C-reactive protein Haptoglobin Complement C3 IgG antibody IgM antibody IgA antibody IgG total IgM total IgA total ESR Haemoglobin Leucocyte count Lymphocyte count *The
distribution
g. litre g. litre g. litre g. litre g. litre g. litre g. litre g, litre ELISA units ELISA units ELISA units g. litre g. litre g. litre mm. hr. g. litre cells. mm-s cells. mm-s of these
factors
Mean
Std. Dev.
Median
Range
2.56* 1.39* 2.24* 0.23* 0.78* 0.06* 0.35* 0.57* 323 445 166 17.50 1.54* 2.76* 14.9* 13.5 7183 2435
1.82 0.89 0.49 0.27 0.37 0.08 0.41 0.35 228 227 84 6.88 1.31 1.19 16.5 1.2 2041 1128
2.10 1.25 2.29 0.13 0.74 0.04 0.22 0.47 300 403 154 16.20 1.15 2.72 10 13.6 6800 2310
0.35-9.50 0.14-4.50 0.24-3.98 0.03-20.0 0.28-2.44 0.02-0.86 0.02-3.40 0.1 O-l .87 16-l 496 83-l 145 15- 539 0.70-37.50 0.12-8.75 0.18-8.44 I-103 10.7-I 6.0 840-I 2700 240-8640
showed
a marked
skewness,
mostly
Mean
Std. Dev.
Median
4.10 1.93 1.83* 0.29* 1.14 0.14* 0.42* 0.73 636* 326* 290* 21.20 2.12 3.19* 64 12.0 8409 2117
1.82 0.73 0.57 0.37 0.45 0.32 0.37 0.37 1091 166 299 7.78 1.02 1.36 35.2 1.7 2373 975
4.05 2.00 1.78 0.18 1 .I0 0.08 0.26 0.68 1104 630 328 20.50 1.95 2.92 65 12.7 8400 2000
in the form
of a strong
positive
Range
0.55-8.65 0.28-3.95 O-88-4.12 0.03-3.12 0.09-2.70 0.02-1.16 0.06-2.20 0.09-1.90 66-l 999 50-I 814 19-l 999 5.20-52.50 0.20-6.75 0.30-8.48 1-123 7.0-I 5.6 3800-l 6000 300-4420
tail.
correlation with the other proteins in disease. In two cases among the controls, namely ccl antitrypsin with ul -acid glycoprotein and haptoglobin with C3, a direct correlation became an inverse one when the partial r was calculated. On discriminant analysis of the acute phase reactants the best (although not particularly good) separation was obtained by taking al -antitrypsin, al -acid glycoprotein and transferrin in combination. By this method 39 (36.5 %) of patients and 19 (13 %) of control subjects were misclassified. This total incidence of 23.1 % misclassifications approximates closely to the theoretical incidence of 24.5 % on the null hypothesis that either form of misclassification (i.e. patient as control or vice versa) is equally significant. Further analysis of the mismatched subjects produced a discriminant function based on levels of crl -antitrypsin, transferrin and C-reactive protein. On application of this discriminant only 5 patients and 3 controls remained wrongly placed. There were no unusual features in 3 of the 5 mismatched patients but the other 2 were very ill, one with a large pleural effusion or empyaema and one with extensive miliary-like shadowing throughout both lungs. Of the 3 mismatched controls 2 were normal in all other respects but one had a very high lymphocyte count (8640 cells mm-3). Cluster analysis did not distinguish reactant protein levels.
any groups of individuals
on the basis of acute phase
Table III lists the correlations between the levels of the serum proteins (acute phase reactants and immunoglobulins) and the levels of anti-mycobacterial antibodies. In the control group, correlations between levels of acute phase reactants and antibodies were most evident with the IgG, less with the IgA and least with the IgM class. Antibody levels in all 3 classes showed a low inverse correlation with total IgA levels. These correlations disappeared in disease, with the exception of that between haptoglobin and IgA antibody.
Acute
Table II. Simple and partial correlations coefficients
phase reactants
in tuberculosis
between the acute phase reactant proteins. Test
--
Control -_
Of
With
Simple
Partial
Simple
Partial
AIA
A2M TRA CER AGP CRP HAP c3
0.259** 0.334*** 0.650*** -
0.253** -
0.771***
0.304** -
0.407*** 0.672*** 0.809*** -
0.870*** 0.344*** 0.399*** -
TRA CER AGP CRP HAP c3
0.305*** 0.353*** 0.263** -
;403*** 0.628*** 0.835*** -
1 0.464*** -
Correlation
A2M
TRA
CER
CER AGP CRP HAP c3
-0.287*** -0.390* 0.661*** 0.295*** -
* *
-0.166* -
AGP CRP HAP c3
0.188*** -
CRP
0.585*** 0.21 o* 0.229* *
CRP HAP c3
----
~___HAP c3
;383*** 0.374*** 0.206* -
0.257*** --_--_ -
-0.189* -0.324*** 0.379*** -0.203*
-0.215*
AIA=ul -antitrypsin ; ; CER=ceruloplasmin HAP= haptoglobin; *=p
0.263**
--
0.195* -
-
0.370*** 0.209*
0.247* -
0.328*** 0.442* -
0.240* ----_-_
0.655*** 0.442*** 0.203* -----~--~-------_ -0.494*** --__
c3
0.267**
0.259*
-----HAP
0.409***
-
-___ AGP
29
-0.267**
A2M=u.2-macroglobulin; AGP= al -acid glycoprotein ; CB=third component of complement. **=p
0.638*** 0.348*** 0.249*
0.395*** -
0.206* ---___--__
-
0.310**
0.230*
TRA=transferrin; CRP-= C-reactive
protein;
***~-p
Table IV show the correlations between the levels of the acute phase reactants and the total IgM levels. In all cases the correlations were closer in the patients than in the controls. No correlations were found between the acute phase reactants and total IgG while in the case of IgA there were correlations with haptoglobin (r 0.370, p
of weight and its effect on protein levels
Using the method described above the patients showed marked weight loss. Only 2 patients were above the mean control weights and 62 (58 %) were below the lower 5 % limit of the controls. There was very little relationship between weight and serum protein levels. Among
30
Grange and others
Table III. (Significant
Correlation between antimycobacterial correlations are underlined.)
antibody
Control JgM
al -antitrypsin
0.339***
-0.016
a2-macroglobulin Transferrin
0.105 0.073
0.021 -0.002
al acid glycoprotein C-reactive protein
0.344***
0.127
-0.093 0.299***
-0.112 0.012 -0.080
Haptoglobin
-m***
c3
-0.242*
Total IgM
0.113 -0.320*
Total IgA *=p
Table IV.
**=p
Correlations
Is.4
0.180* 0.121 -0.055 0.129
serum
proteins.
lgM
lg.4
0.107 0.074 0.100 0.025
-0.064 -0.032 -0.054 -0.096
0.032 -0.306**
0.077 0.060 -0.140 -0.100 -0.003
-0.295*** 0.072
-0.260** 0.224i
**
0.088 -0.207*
0.111
***=p
fgG
0.065 0.299*** -0.295***
*
0.236**
Total IgG
and
Patient -
fgG
Ceruloplasmin
levels
*
-0.049
-0.211*
0.068
-0.052 -0.005 -0.204*
0.028
0.092 0.043 0.149 -0.197*
0.148
-0x
0.056
-0.084
0.119 0.045
0.084 0.114
Assessed one-tailed.
between serum proteins and total IgM levels.
Protein
cd -antitrypsin a2-macroglobulin Transferrin Ceruloplasmin al -acid glycoprotein CRP Haptoglobin c3 NS= Not significant (p>O.O5).
Control
Patients
Correl. coeff.
p
0.290 0.391 0.019 0.237 0.421 0.355 0.094 0.151
Corn/. coeff. 0.515 0.524 0.176 0.495 0.453 0.548 -0.146 0.485
p
Assessed one-tailed.
the controls there was a tendency for ceruloplasmin and haptoglobin levels to be higher (p
Individuals in each group were placed into the 4 age groups described in the materials and methods section and were compared by variate analysis. Among the controls, levels of specific antibody in the IgM class and the leucocyte count decreased with increasing age (~~0.01) and a1 -antitrypsin levels increased (~~0.02). Haemoglobin levels were lowest in the 24-30 age group. Among the patients only C3 levels were significantly affected by age, being lower in the older patients (~~0.05).
Acute phase reactants in tuberculosis Differences
31
due to sex
Among the female controls there were higher levels of ul -acid glycoprotein (ptO.OOl), C3 (p
helminthiasis
Eleven of the 144 controls (7.6 %) and 17 of the 107 patients (16 %) had microscopical evidence of intestinal helminthiasis. The higher number among patients was not statistically significant. All cases among the controls were due to Ascarislumbricoides. Among the patients 15 cases were due to A. lumbricoides, 1 to Ancylostoma duodenale and 1 to Trichuria trichuris. The presence of worms had no effect on serum proteins among the control subjects. Among the patients, those with helminthiasis tended to have higher levels of total IgM (~~0.05). Effect of previous KG There was no significant the control subjects.
vaccination association
of protein levels to BCG status in either the patients or
Effect of skin test reactivity Among the patients there was no significant correlation between protein levels and the diameters of the skin test readings at 6-8 hours or at 48 hours. The 24 hour readings showed a negative correlation with transferrin (r -0.251, p
studies
The number of patients in the disease categories listed in the materials and methods section are shown in table V. Age, weight andsex. Chronic disease, predictably, occurred in a slightly older age group (mean 35.1, standard deviation 10.1 years) than recent disease (mean 31 .5, standard deviation 7.5 years;p<0.05). The occurrence of fibrosis was similarly age-related (p
32
Grange and others
Table V. Radiological the patients.
features and their relation to the duration
Radiological
Total in
Duration
features
each
(years)
of symptoms
category
1
2
3
4
of symptoms as reported by P
Duration on Radiology
P
5
6
>6
Recent
Chronic
2 1
-
-
Duration
Recent Chronic
48 58
30 6
IO 8
3 14
1 9
1 3
1 7
Extent of active disease
Minimal Moderate Extensive
26 46 34
4 19 13
3 10 5
7 6 4
3 3 4
0 2 2
4 2 2
5 i Not 4 significant 4
9 24 15
17 I Not 22 significant 19
Moderate Extensive Minimal
55 22 29
16 182
34
6 22
2 02
5 30
8 <0.02 50 1
20 271
35
38 41 27 106
20 133
3 122 17
1 72 10
1 21 4
1 52 8
34 140
4
Fibrosis
Cavity Total
Recent Old None
36
810 64 9 54 18
3 13
48
-
Duration of symptoms. Table V shows that the duration of disease as assessed radiologically correlated significantly with the duration of symptoms as reported by the patient. There was also a definite association between the presence of fibrosis and the duration of symptoms. Thus fibrosis was only observed in half the 36 patients with symptoms present for less than one year but in all patients reporting symptoms for 4 or more years. Cavitation also bore a relationship to duration of symptoms but the extent of disease did not. Proteins. The levels of transferrin and specific antibody in the IgG class were significantly higher in chronic disease than in recent disease (~~0.02 and <0.005 respectively). They were also both higher in patients with fibrosis than in those without (~~0.05) but IgG antibodies only were significantly higher in those with long-standing cavities (~~0.05). The levels of specific IgG antibodies were also associated with the duration of disease as reported by the patient, but at a much lower level of significance (~~0.05). Levels of haemoglobin (p
and
ctl -acid
glycoprotein
were
lower
in patients
with
cavities
Levels of specific IgA antibodies, the ESR and the leucocyte count were higher in the more extensive disease (~~0.05,
studies
The relationships between haematological values the levels of acute phase reactants and antimycobacterial antibodies are shown in table VI. Several correlations between the ESR and haemoglobin and certain proteins disappear in disease while negative correlations appear between the levels of transferrin and the ESR, haemoglobin and leucocyte count, and particularly between the lymphocyte count and haptoglobin levels. There were significant associations between the extent of disease and both the ESR and the leucocyte count (pt0.001 and co.005 respectively), both being higher in more extensive disease. In both health and disease there was a negative relationship between the ESR and haemoglobin levels. There were no significant immunoglobulin levels.
correlations
between
the haematological
features and the total
Acute
phase reactants
Table VI. Correlations of haematological features with acute bacterial antibodies. (Significant correlations are underlined.) Control
in tuberculosis
phase reactants
33
and antimyco-
Patient ---
ESR
cd -antitrypsin
Haemoglobin
-0.228*
*
-0.117
a2-macroglobulin
-0.043
Transferrin
-0.086
-0.052
Ceruloplasmin al -acid glycoprotein
-0.117
-0.005
0.107
-0.328***
C-reactive protein
-0.264
Haptoglobin
-0.001
c3
-0.067
0.282*** * *
0.042 -0.197*
Leococyte count
Lymphocyte count
-0.197
-0.055
0.074 -0.035 0.014 -0.081
ESR
-0.051
Haemoglobin 0.013
0.052
-0.036
0.077
-0.389***
-0.008 0.271**
0.085
-0.172
-0.001
0.004
-0.042
Leucocyte count -0.059 -0.247* -0.106 0.017
Lymphocyte count -0.089 0.117 -0.184 0.003 -0.158
-0.026
0.073
-0.051
-0.010
-0.083
-0.086
0.052
0.065
-0.068
-0.079
-0.112
-0.041
-0.416*** -0.049
0.041
-0.133
0.126
0.007
-0.215*
-0.075
0.048
0.099
IgM
0.050
-0.388***
-0.083
0.049
0.174
-0.118
0.081
0.078
0.058
-0.218**
-0.027
0.159
-0.211*
o-249*
0.170
antibody
IgA antibody *p
**p
0.271**
;
***p
0.232*
*
-0.032 Assessed
-0.107
0.151
0.053
IgG antibody
0.035
one-tailed.
Discussion The protective immune response in tuberculosis is generally regarded as being of the ‘cell mediated’type as opposed to one involving antibody, complement or other humoral factors [14] The effector cell in tuberculosis is the activated macrophage which, in a successful immune response, appears to engulf and destroy tubercle bacilli [15]. Although serum proteins and other soluble factors may not be the prime factors involved in the immune response to tuberculosis, there is growing evidence that they may have a substantial effect, either beneficial or detrimental, on cellular immune functions in both infectious and malignant diseases. In tuberculosis a number of specific and non-specific changes in the cellular and humoral arms of the immune system occur: these include cell-mediated reactions that manifest as a positive tuberculin test, the production of antibodies to Mycobacterium tuberculosis, a nonspecific increase in immunoglobulin levels and the appearance of autoantibodies [3]. The levels of the acute phase reactants increase in a number of physiologically stressful conditions such as infections, injuries (whether accidental or surgical), burns and myocardial infarctions. Despite the name ‘acute phase’, the levels of some of these proteins are also elevated in chronic conditions such as leprosy [I 6, 171 and intestinal helminthiasis [18]. There is no precise definition of an acute phase reactant nor is there any general consensus of opinion as to which proteins should be included. Koj [I91 has defined them as ‘traumainducible liver-produced plasma glycoproteins’ and regarded ctl -antitrypsin, ceruloplasmin, ~1 -acid glycoprotein, C-reactive protein and haptoglobin as belonging to this group. Creactive protein, however, is not a glycoprotein [7]. Sipe and Rosenstrich [7] divided acute phase reactant proteins into two groups : early and late. The early group contains endogenous pyrogen, serum amyloid A inducer and tumour necrosis factor. The late group includes Creactive protein and the glycoproteins referred to above. Opinions differ, however, as to whether x2-macroglobulin, transferrin and components of complement should be included.
34
Grange and others
In this study levels of x2-macroglobulin correlated significantly with those of some of the recognized acute phase reactants, notably ~1 -antitrypsin and C-reactive protein in both patients and control subjects. Although elevated in cases of tuberculosis, a2-macroglobulin tended to be lower in tuberculin positive than in tuberculin negative controls. There is evidence that a2-macroglobulin binds to, and inactivates, proteases liberated during tissue injury, necrosis and inflammation and that the resulting complex stimulates the biosynthesis of acute phase reactants by the liver [20]. This does not, however, explain why its levels are affected by tuberculin reactivity among the control subjects. Levels of or1-antitrypsin were found to be significantly elevated in pulmonary tuberculosis, thus confirming the findings of other workers [21, 22,231. In leprosy, on the other hand, this enzyme was not found to be significantly elevated except in those patients with erythema nodosum leprosum [24]. This is probably due to the occurrence of much less tissue destruction and inflammation in leprosy than in tuberculosis [25] except when hypersensitivity reactions occur in the former disease. Low levels of this enzyme are found in patients with chronic obstructive airway disease (COAD), and it has been postulated that this and related proteins prevent lung damage by inactivating proteases liberated by bacteria and macrophages [26,27]. In this study 3 patients had radiological evidence of emphysema but their al-antitrypsin levels were well within the normal range. No relationship between cavity formation and fibrosis and levels of al -antitrypsin was found. There is little doubt, however, that the mechanisms of lung damage and repair in tuberculosis and COAD are quite different. Although leprosy and tuberculosis differ with respect to al-antitrypsin levels they show similarities in other respects. Thus levels of total immunoglobulins are increased in leprosy although, as in tuberculosis, the classes involved appear to differ from study to study [I 61. In this study only the third of the complement components (C3) was quantitated and was found to be elevated in tuberculosis. Furthermore, its concentration correlated significantly with levels of ccl -antitrypsin in the patients but not in the controls. The fourth component (C4) is not, however, elevated in tuberculosis [3]. Likewise in leprosy increases occur in C3 but not in C4 levels [I 61. On the other hand, C4 is elevated in many acute conditions [28]. Transferrin levels were reduced in tuberculosis and they did not correlate with those of any of the other acute phase reactants in the control group with the exception of a very low inverse correlation with ceruloplasmin. This protein is, unlike other acute phase reactants, synthesised by macrophages and its levels drop in many disease states due to an excess of catabolism over synthesis [29, 301. Ceruloplasmin and haptoglobin are both involved in the transport of metal ions and are both stimulated by derivatives of arachidonic acid, which is also the precursor of prostaglandins. In rabbits with carcinoma there was a correlation between the levels of these proteins and prostaglandin E2 [31 ] but in this study there was an inverse correlation between the proteins in the controls and no significant correlation in disease. The inverse correlation in health could be due to competition during synthesis for a single precursor molecule. The significance of the correlations between the various proteins and the differences in health and disease are at present poorly understood as relatively little is known about the factors that control them, Hopefully the findings of this study will become of relevance when they can be considered in relation to in vitro measurements of cell-mediated immune reactions It is noteworthy, however, that while the levels of C3 tended to correlate more closely with other proteins in disease, the reverse was the case with those of other acute phase reactants. A remarkable exception, however, was the effect of disease on the relation between levels of ctl -antitrypsin and ceruloplasmin. The relationship of levels of acute phase reactants to those of total immunoglobulins and the
Acute phase reactants in tuberculosis
35
antimycobacterial antibodies was rather unusual. Thus, correlations between acute phase reactants and antibody levels among the control subjects was most apparent with the IgG class and least with the IgM class while the reverse was the case with the total immunoglobulins in these classes. Furthermore, the correlations with antibody levels tended to be lost in disease while those with total IgM levels increased. A further unexpected correlation was the low inverse one between mycobacterial antibody in the 3 classes, including IgA.
total
IgA and anti-
Although changes in serum protein are thought to be responsible for the increase in the ESR in disease, the acute phase reactants investigated in this study did not appear to play such a role. Indeed, when correlations appeared, they were negative ones. The negative correlation between the ESR and transferrin is almost certainly secondary to the negative correlation between the ESR and haemoglobin levels. The strong negative correlation between haptoglobin and the lymphocyte count in tuberculosis is of interest, and may be indicative of an important immunoregulatory role for this protein in this disease. The various correlations observed in this study, although statistically significant, were often not very close. This may be due to the fact that levels of serum proteins in a given individual are not constant but fluctuate significantly. Thus, in the rat, levels of IgG, a2-macroglobulin and transferrin showed distinct circadian rhythms, being higher at night. In addition IgG levels varied from day to day while levels of transferrin showed oscillations within each day [32]. It is possible, therefore, that closer correlations would be found by taking the mean of a number of protein assays on sera taken at different times and on different days, In comparing serum proteins and other factors in healthy controls and patients with tuberculosis, it was necessary to determine whether any differences were directly associated with the disease or whether they were secondary to other factors such as age, sex, weight and concurrent disease. The two groups were matched, as closely as possible, for age and socioeconomic status. There was a predominance of females among the controls and males among the patients and although some sex-related differences in serum protein levels were found in the control group these did not account for differences between controls and patients. Some protein levels were associated with age or sex among the controls but these associations were negated in disease. On the other hand, among the patients, levels of C3 and total IgA were affected by age and sex respectively. Very little information on the effect of age and sex on the levels of acute phase reactants in health is available but recently this was studied with respect to al -antitrypsin, al -acid glycoprotein, ceruloplasmin and transferrin in North American citizens [33]. No differences in the levels of these proteins in the age groups relevant to the present study were found. Levels of al -antitrypsin and ceruloplasmin, but not those of the other two proteins, were significantly higher among females than among males. This is in contrast to the present study in which, of the 4 proteins, only al -acid glycoprotein levels were sex-related, being higher in females. These differences may reflect racial or genetic variations between the subjects in the two studies. The overall levels of the acute phase reactants in the control group were, however, very similar to those of European adults, except for those of transferrin and haptoglobin which were lower in the Indonesian group. Weight loss and a low dietary intake due to loss of appetite might be expected to cause changes in serum protein levels. Indonesian children with protein-calorie malnutrition associated with acute infections have been shown to have decreased levels of transferrin and increased levels of haptoglobin [34]. In the present study, haptoglobin levels (as well as those of ceruloplasmin) were higher in the under-weight control subjects. In the patient group the association of weight to haptoglobin levels was not significant while the association to ceruloplasmin was reversed. It should, however, be stressed that weight loss was more
36
Grange and others
frequent and severe in the patient group, thus the lower weight group of the controls approximated to the higher weight group of the patients. Indeed, the ceruloplasmin levels in these two groups were fairly similar but decreased in both the more overweight controls and the more underweight patients. Consequently the levels of some proteins are affected by weight loss as well as by the disease process. Patients and controls with obvious concurrent diseases or other conditions likely to affect the results (e.g. pregnancy and lactation) were excluded from the study. The only detected concurrent condition was intestinal helminthiasis. Studies on Vietnamese refugees [18] showed that children and young adults with intestinal helminthiasis tended to have elevated levels of al -antitrypsin, ceruloplasmin and al -acid glycoprotein. It was thought, however, that this was mainly due to hookworm (Ancylostoma do&ens/e) infestation. In this study, helminthiasis, which was almost entirely due to Ascaris lumbricoides, had no significant association with protein levels, with the exception of a slight elevation of total IgM levels among the patients. There have been several attempts to divide cases of tuberculosis into a ‘spectrum’ of immunological reactivity similar to that seen in leprosy, and to describe criteria for allocating cases to a point on the spectrum [35, 361. On clinical grounds such a subdivision of cases is not easy. In this study the patients were classified according to radiological appearance, but the presence or absence of fibrosis and cavities had no significant effect on the levels of most of the serum proteins. The reduction of haemoglobin in patients with cavities may be due to the more frequent occurrence of haemoptysis in such patients [3]. It seems unlikely, therefore, that an estimation of these proteins will be of use in the classification of disease. Indeed the best indicators of the extent of the disease, apart from radiology, appeared to be the ESR and leucocyte counts. Several authors have reported that levels of serum immunoglobulins increase in tuberculosis and although there appear to be some regional variations in the contribution of the various classes of immunoglobulin to this increase, most workers have found that levels of IgG and IgA increase considerably while that of IgM only increases slightly [3, 37, 38, 39, 401. These increasesarenotdue to the production of antibody to M. tuberculosis but are probably due to a polyclonal activation of B cells by mycobacterial adjuvants [3]. The reason for such an activation affecting IgG and IgA much more than IgM is not clear. In African trypanosomiasis, by contrast, the polyclonal B cell activation causes an increase principally in the IgM class [41]. It is, however, of interest to note that, of the 3 classes, the levels of IgM alone correlated with those of several acute phase reactants in health and, more closely, in disease. It is possible, therefore, that this class of immunoglobulin is controlled by a mechanism that also governs the synthesis of acute phase reactants, and that this mechanism is refractory to the effect of mycobacterial adjuvants. Indeed IgM has been included among the acute phase reactants [42]. It is noteworthy that, among the patients, those with intestinal helminthiasis tended to have higher total IgM levels than those without, but that this trend did not occur among the control subjects. Whether or not this represents some form of synergy between adjuvants from helminth and mycobacteria remains to be determined but the phenomenon might account for the apparent geographical variations in the effect of tuberculosis on total IgM levels. It is, however, doubtful whether intestinal helminthiasis alone is a cause of polyclonal B cell activation [41]. In view of the immunoregulatory properties reported for various acute phase reactants, their relationship to the diameter of the skin test reactions was studied. The negative correlation between the 24 hour reaction and transferrin levels in both health and disease is of interest. The 24 hour reaction in man probably corresponds with the Listeria type reaction in mice [43] which is associated with macrophage activation. As mentioned previously, transferrin differs from the other acute phase reactants in being synthesised by the macrophage. Thus the size
Acute phase reactants in tuberculosis
37
of the 24 hour reaction and the transferrin levels may both reflect variations in macrophage function. It has been shown on several occasions since the pioneering work of Pickert and Lowenstein in 1908 [44] that serum from patients with tuberculosis may enhance or suppress the 48 hour dermal reaction to tuberculin. The responsible factors were termed ‘procutins’ and ‘anticutins’. None of the proteins examined in this study appeared to have such properties. It has been shown previously that an immediate reaction to tuberculin, presumably caused by IgE or short-term sensitizing IgG antibodies, tends to reduce the diameter of the 48 hour response [8, 451 but otherwise very little is known about the humoral factors that affect the tuberculin test. Cluster analysis of the acute phase reactant levels did not reveal any distinct groups of patients or control subjects and this statistical method did not separate the former from the latter. Stepwise discriminant analysis, on the other hand, showed that certain combinations of the levels of these proteins could be used to achieve a fairly good separation of patients from the controls. This is unlikely, however, to be of diagnostic usefulness owing to the nonspecificity of changes affecting the levels of these proteins. It is possible though that such an analysis might prove useful in the early detection of relapse after therapy or reactivation in individuals with evidence of previous infection. In conclusion, tuberculosis has been shown to be associated with significant alterations in the levels of many serum proteins and, although these changes are of no diagnostic value, they may prove to be of considerable reievance to the pathogenesis of the disease and the efficacy of the immune reponses. Thus, although humoral factors are not of prime importance in the destruction of the causative agent of tuberculosis, they should not be ignored or forgotten when considering the host-pathogen interactions in this important and prevalent disease. Acknowledgements
We are most grateful to Professor J. G. Scadding and Dr M. Oesman for reading the radiographs, Dr S. Darmadi for assistance with the laser nephelometry, Dr I. Handoyo for assistance with the ELISA studies and Dr H. Notodipuro for the haematological examinations. We also thank Dr D. lman Santoso and S. Karim and their colleagues for their enthusiastic help in the clinical studies. We also received valuable technical assistance from Miss R. Anggraini and Miss U. Haryanti. We also thank Professor John Swanson Beck and Dr Harald Mauch for their advice and encouragement; Miss Margaret Rehahn for statistical advice and for performing the stepwise discriminant analysis; Mr Stephen Saunders for writing the program for cluster analysis on the Brompton Hospital Prime computer system and the British Council for financial support of one of us (T.K.). References 1 Koch, Ft. (1891). Weitere Mitteilungen uber ein Heilmittel gegen Tuberkulose. Deursche Mediziniscbe Wochenschrift, 17,101. 2 Bloom, B. R., Tanowitz, H., Et Wittner, M. (1979). Games parasites play. Nature, 279, 21. 3 Kardjito, T., Et Grange, J. M. (1980). Immunological and clinical features of smear-positive pulmonary tuberculosis in East Java. Tubercle, 61, 231. 4 Tormey, D. C., Er Mueller, G. C. (1972). Biological affects of transferrin on human lymphocytes. Experimental Cell Research. 74. 220. 5 Mortensen, R. F., Braun, D., 8 Gewurz, H. (1977). Effects of C-reactive proteins on lymphocytefunction. III. Inhibition of antigen-induced lymphocyte stimulation and lymphokine production. Cellularlmmunology, 28, 59. 6 Bullock, W. E. (1979). Mechanisms of anergy in infectious diseases. In lmmono/ogicalAspects of infectious Diseases. Edited by G. W. A. Dick: p. 269, MTP Press, England.
38
Grange and others
7 Sipe, J. D., Et Rosenstrich, D. L. (1981). munity and fnffammations. Edited Arnold, London.
Serum factors associated with inflammation. In Cellular Functions In lmby J. J. Oppenheim, D. L. Rosenstrich Et M. Potter: p. 411, Edward
8 Kardjito. T., Donosepoetro, M., 8 Grange, J. M. (1981). The Mantoux test in tuberculosis: diameters of the dermal responses and the serum protein levels. Tubercle, 62, 31.
correlations
between
fhe
9 Kardjito, T., 8 Grange, J. M. (1982). Diagnosis of active tuberculosis by immunological methods. 2. Qualitative differences in the dermal response to tuberculin in patients with active pulmonary disease and healthy tuberculinpositive individuals. Tubercle, 63, 275. 10
11
12 13 14 15 16
17 18
Grange, J. M., Gibson, J., Nassau, E., Et Kardjito, T. (1980). Enzyme-linked immunosorbent assay (ELISA) : A study of antibodies to Mycobacterium tuberculosis in the IgG, IgA, and IgM classes in tuberculosis, sarcoidosis and Crohn’s disease. Tubercle, 61, 145. Kardjito,T., Handoyo, I., & Grange, J. M. (1982). Diagnosis of activetuberculosis by immunological methods. 1. Effect of tuberculin reactivity and previous BCG vaccination on the antibody levels determined by ELISA. Tubercle, 63, 269. Stuart, A., & Kendall, M. (1965). TheAdvanced Theory of Statistics. Vol. 2. p. 317, Charles Griffin, London. Documenta Geigy (1970). Scientific Tables, 7th edition. Edited by K. Diem & C. Lentner: p. 711, Geigy, Basle. Mackaness, G. B. (1968). The immunology of antituberculous immunity. American Review of Respiratory Disease. 97, 337. Lowrie, D. 8. (1981). Survival of intracellular bacteria inside macrophages. Anna/es d’lmmunologie (/nstitutPasteur), 132D, 151. Srivastava, L. M., Agarwal, 0. P., & Goedde, H. W. (1975). Biochemical, immunological and genetic studies in leprosy. II. Profile of immunoglobulins, complement components and C-reactive protein in sera of leprosy patients and healthy controls. Tropenmedizin und Parasitologic, 26, 212. Celikoglu, S. I., Fahir, M. G., & Turkan, 8. (1976). A preliminary report of a study on serum alpha-l -antitrypsin and immunoglobulin levels in lepromatous leprosy. Anna/s of the New York Academy of Science, 278,125. Blom, M., Prag, J. B., Et Nmrredam, K. (1979). ct2-acid glycoprotein, crl -antitrypsin and ceruloplasmin in human intestinal helminthiasis. American Journal of Tropical Medicine and Hygiene, 28, 76.
19
Koj, A. (1974). Acute phase reactants. Their synthesis, turnover and biological significance. In Structure and Function of Plasma Proteins. Vol. 1, Ch. 4, p. 73. Edited by A. C. Allison, Plenum Press, London.
20
Metcalfe, J., 8 Tavill, A. S. (1975). A proposed role for alpha-l -acute phase globulin synthesis by the perfused rat liver. British Journal of Experimental Pathology, 56, 570. Geisler, L. S., Bachmann, G. W., Lauman, F., Nolte, D., Wenzel, H., Et Rost, H. D. (1972). Alpha-l -antitrypsin and Deutsche Mediziimmunoglobulins in chronic non-specific pulmonary disease and pulmonary tuberculosis, nische Wochenschrift, 97,329. Poh, S. C., Et Seet, A. M. (1975). Alpha-I-antitrypsin levels in chronic obstructive lung disease and pulmonary tuberculosis in Singapore. Singapore Medical Journal, 16, 89.
21
22 23
Gatner,
E. M. S., Et Anderson, R. (1980). An in-vitro assessment tuberculosis: correction of defective neutrophil motility Clinical and Experimental Immunology, 40, 327.
of cellular and humoral immune function in pulmonary by ascorbate, levamisole, metoprolol and propanolol.
24
Agarwal, S. K., Et Sahah, K. (1978). Serum alpha-l -antitrypsin in various forms of leprosy. lndian Journal of Medical Research, 68, 136. 25 Skinsness, 0. K. (1968). Comparative pathogenesis of the mycobacterioses. Annals of the New York Academy of Sciences, 154, 19. 26 Lieberman, J. (1973). Alpha-l -antitrypsin deficiency. Medical Clinics of North America, 57, 691. 27 Norse, J. O., Lebowitz, M. D., Knudson, R. J., Et Burrows, B. (1977). Relation of protease inhibitor phenotypes to obstructive lung disease in a community. New England Journal of Medicine, 296, 1190. as a diagnostic aid. New England Journalof Medicine, 291,287. 28 Alper, C. A. (1974). Plasma protein measurements 29 30
Haurani, F. I., Meyer, A., 8 endothelial Society. Gordon, A. H. (1976). The 8 A. S. McFarlane,
O’Brien, R. (1973). Production of transferrin by the macrophage. Journal of the Reticufo‘I 4. 309. acute phase plasma proteins. In Plasma Protein Turnover. Edited by R. Bianchi, G. Mariani, p. 381, University Park Press, Baltimore.
31 Voelkel,
32
33 34
E. F., Levine, L., Apler, C. A., Et Tshijan, A. H. (1978). Acute phase reactants ceruloplasmin and haptoglobin and their relationship to plasma prostaglandins in rabbits bearing the VX2 carcinoma. Journalof Experimental Medicine, 147. 1078. Schade, R., Rex, H., Kastner, D., Friedrich, A., Et Sprang, C. (1980). Biorhythms of four serum proteins of the laboratory rat: IgG, transferrin, a2-acute phase protein and a VLD lipoprotein. Acta Biologica et Medica Germanica, 88, 1183. Denko, C. W., Et Gabriel, P. (1981). Age and sex related levels of albumin, ceruloplasmin, al -antitrypsin, ctl -acid glycoprotein, and transferrin. Annals of Clinical and Laboratory Science, 11, 63. Schelp,
F. P., Migasena, P., Pongpaew, P., Schreurs, W. H. P., Et Supawan, V. (1977). Serum proteinase and other serum proteins in protein-energy malnutrition. British Journal of Nutrition, 38, 31.
inhibitors
Acute 35 36 37 38 39 40 41 42 43
44 45
Lenzini,
L., Rottoli,
P., 8 Rottoli,
L. (1977).
The spectrum
phase reactants
of human tuberculosis.
Clinicaland
in tuberculosis
39
Experimenfa//mmuno/ogy
27. 230. Daniel, T. M., Oxtoby, M. J., Pinto, E. M., Et Moreno, E. S. (1981). The immune spectrum in patients with pulmonary American Review of Respiratory Disease, 123, 556. tuberculosis. Faulkner, J. B., Carpenter, R. L., B Patnode, R. A. (1967). Serum protein and immunoglobulin levels in tuberculosis. American Journal of Clinical Pathology, 48. 556. Malomo, I., McFarlane, H., & Idowu, J. (1970). Serum immunoglobulins in pulmonary tuberculosis in Ibadan, Nigeria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 64,427. Rowinska, E. (1972). lmmunoglobulin levels in pulmonary tuberculosis. Polish Medical Journal, 11, 524. Jha, V. K., Eajpai, B. K., Et Gupta, Ft. M. (1974). Levels of serum immunoglobulins in pulmonary tuberculosis patients. Indian Journal of Chest Diseases and Allied Sciences, 16, 361. Terry, R. S., Hudson, J. M., 8 Faghihi-Shrazi, M. (1980). Polyclonal activation by parasites. In The Host-Invader /nterp/ay. Edited by H. Van den Bossche: p. 259, Elsevier/North Holland Biomedical Press, Amsterdam. Coffey, R. L., Zile, M. R., Et Luskin, A. T. (1981). Immunologic tests of value in diagnosis. 1. Acute phase reactants and Postgraduate Medicine, 70, 163. autoantibodies. Rook, G. A. W., Et Stanford, J. L. (1979). The relevance to protection of three forms of delayed skin test response evoked by M. leprae and other mycobacteria in mice. Correlation with the classical work in the guinea pig. Parasite Immunology, I, 111. Deutsche Medizinische Pickert, M., Et Lowenstein, E. (1908). Eine neue Methodezur Priifung derTuberkulin-immunitit. Wochenschrift, 34. 2262. Pepys, J. (1955). The relationship Tuberculosis, 71, 49.
of nonspecific
and specific
factors
in the tuberculin
reaction.
American
Review
of