Immunological and clinical features of smear-positive pulmonary tuberculosis in east Java

Tuber& 61 (1980) I(“; Longman Group

0041-3879/80/00420231

231-238 Ltd.

IMMUNOLOGICAL AND PULMONARY

SO2.00

CLINICAL FEATURES OF SMEAR-POSITIVE TUBERCULOSIS IN EAST JAVA T. Kardjito

Alrlangga

University,

Surabaya,

Java,

Indonesia

J. M. Grange” Cardiothoracic

Institute,

Brompton

Hospital,

Fulham

Road,

London,

S W3 6HP

Summary The immunological and clinical features of 90 Javanese patients with smear-positive pulmonary tuberculosis were investigated. Many of the patients had advanced disease at the time of diagnosis and haemoptysis was common, especially in patients with cavitating lesions. Most patients had a significant elevation of one or more nonspecific indicators of inflammation (erythrocyte sedimentation rate, third complement component, factor B and C-reactive protein). Rheumatoid factor was detected in 21% of the patients and was significantly associated with high levels of antibodies to M. tuberculosis in the IgM class. Five distinct responses were elicited by tuberculin testing ; the most marked occurred at 24 hours. The degree of reaction at 6-8 hours correlated significantly with the levels of specific antibodies in the IgG and IgA classes and the 48 hour response correlated, although less markedly, with specific antibodies in the IgG class. Neither the degree of skin test reactivity nor the level of specific antimycobacterial antibodies correlated with the extent of disease as assessed radiologically. Nine per cent of the patients were skin-test negative at 48 hours but did not differ clinically, as a group, from tuberculin positive patients. It was not possible to place the cases in a spectrum of immunological responses similar to that occurring in leprosy and it is postulated that this is due to differences in the relevance to protection of the various immunological mechanisms in the two diseases. The need to establish more rigorous criteria for assessing the immune responses in tuberculosis and for studying the interactions between the protective and non-protective reactions is stressed.

R&cum6 Les auteurs ont etudie les caracteristiques immunologiques et cliniques de 90 malades javanais atteints de tuberculose pulmonaire a frottis positif. Beaucoup d’entre eux avaient au moment du diagnostic une tuberculose avancee et I’hemoptysie etait frequente, en particulier chez les malades porteurs de lesions cavitaire. La plupart des malades avaient une augmentation significative de I’un ou de plusieurs des indicateurs non specifiques de la presence d’une inflammation (vitesse de sedimentation, fraction 3 du complement, proteine reagissant aux facteurs B et C). Le facteur rhumato’ide a 6te trouve chez 21% des malades; il etait associe de facon significative a des taux eleves d’anticorps vis-a-vis de M. tuberculosis, dans la categoric des IgM. L’epreuve a la tuberculine a mis en evidence cinq types differents de reponse dont la plus marquee est survenue a la 248me heure. L’intensite de la reaction a la 68me8Bme heure montrait une correlation significative avec les taux d’anticorps specifiques *

Requests

for reprints

should

be addressed

to Dr. J. M. Grange.

232

Kardjito and Grange des categories IgG et IgA alors que la reponse a la 488me heure etait lice, quoique de facon moins nette, aux anticorps specifiques de la categoric des IgG. Ni le degre de reactivite cutanee, ni le taux des anticorps specifiques antimycobacteriens ne montraient une association avec l’etendue de la maladie telle qu’on pouvait l’evaluer d’apres le cliche radiologique. Neuf pour cent des malades avaient des tests cutanes negatifs a la 488ms heure mais, en tant que groupe, ils ne differaient pas, au point de vue clinique, des malades tuberculino-positifs. II n’a pas ete possible de ranger les cas selon un spectre des responses immunologiques similaire a ce que I’on trouve pour la lepre. On suppose que ceci est dQ aux differences entre les deux maladies en ce qui concerne divers mecanismes immunologiques en rapport avec la protection. Les auteurs soulignent I’importance de definir des criteres plus rigoureux pour apprecier les reponses immunologiques dans la tuberculose et pour etudier les interactions entre les reactions d’ordre protecteur et celles qui ne le sont pas. Resumen Se estudian las caracteristicas inmunologicas y clinicas de 90 enfermos javaneses con tuberculosis pulmonar con frotis positivo. En el moment0 del diagnostic0 muchos de ellos presentaban una tuberculosis avanzada y la hemoptisis era frecuente sobre todo en 10s enfermos que tenian lesiones cavitarias. La mayoria de 10s enfermos tenia un aumento significativo de uno o de varios de 10s indicadores no especificos de la presencia de una inflamacion (velocidad de sedimentacibn, fraction 3 del complemento, proteina reactiva a 10s factores B y C). El factor reumatoideo se encontrd en 21% de 10s enfermos, asociado de manera significativa a tasas elevadas de anticuerpos contra el lvI. tuberculosis en la categoria IgM. La prueba de tuberculina puso en evidencia cinco tipos diferentes de respuesta, de las cuales la m$s intensa se produjo a las 24 horas. La intensidad de la reaction a las 6 ya a las 8 horas mostrd una correlation significativa con las tasas de anticuerpos especificos de las categorias IgG e IgA mientras que, la respuesta alas 48 horas estuvo ligada, aunque de manera menos neta, a 10s anticuerpos especificos de la categoria IgG. El grado de reactividad cutanea y la tasa de anticuerpos especificos antimicobacterianos no estuvieron correlacionados con la extension de la enfermedad demostrada radiolbgicamente. Nueve por ciento de 10s enfermos tuvieron tests cutaneos negativos a las 48 horas, pero, en tanto que grupo, no eran diferentes de 10s enfermos tuberculino-positivos, desde el punto de vista clinico. No fue posible ordenar 10s cases segun un espectro de respuestas inmunologicas coma sucede en el case de la lepra ; se supone que esto se debe a la diferencia entre las dos enfermedades en lo que respecta a 10s diversos mecanismos inmunoldgicos de protection. Se hate hincapie en la importancia de definir criterios m&s rigurosos para apreciar las respuestas inmunologicas en la tuberculosis y para estudiar las interacciones entre las reacciones protectoras y las no protectoras. Introduction

The immunology of tuberculosis is a subject which has attracted a growing interest in recent years but there are still many aspects which remain poorly understood. In particular the relationships of the degree of delayed hypersensitivity, and of the levels of antibodies to the causative organism, to the clinical features are far from clear. The significance of delayed hypersensitivity, as determined by the tuberculin test, to protective immunity has been hotly

Tuberculosis

in East Java

233

debated (1,2,3). The existence of more than one distinct type of dermal response to tuberculin has been demonstrated in man (4) and animals (5,6) and these may vary considerably in their relevance to protective immunity. Increase in total immunoglobulin levels and in specific antimycobacterial antibodies occur in many cases of tuberculosis (7,8) but the significance of such increases remains obscure. An attempt to divide tuberculosis into immunologically reactive, intermediate and unreactive forms has been made (9) and the existence of a spectrum of tuberculosis verysimilarto that previously described in leprosy (10) was proposed. In this study 90 cases of smear-positive pulmonary tuberculosis diagnosed in East Java have been investigated clinically, radiologically, haematologically and immunologically and compared with 50 age-matched healthy control subjects from the same region. The purpose of the study was to determine the relevance of changes in the various parameters, to what extent the changes correlated with each other and whether they enabled the cases to be arranged in a spectrum according to their clinical and immunological features. Patients and methods Ninety randomly selected patients with smear-positive pulmonary tuberculosis aged between 16 and 50 were studied. The duration of symptoms, changes in appetite and body weight and other relevant symptoms, including the occurrence of haemoptysis, were recorded.

Patients

A posteroanterior chest x-ray was obtained and the presence or absence of cavitation noted. The extent of disease was assessed as follows :

was

The total extent of disease did not involve more than an area of lung equivalent I Minimal. to that above the second chondrosternal junction and the spine of the fourth thoracic vertebra. I I Moderate/y advanced. Dense and confluent lesions did not exceed an area of one-third of the volume of one lung; non-confluent lesions of slight or moderate density did not exceed the area of one lung. III Far advanced, More extensive than II. Mantoux testing with 5 T.U. of PPD (RT23) was performed and the diameter of the response was read immediately, after 30 minutes, 6-8 hours, 24 hours and 48 hours. Laboratory studies The number of acid-fast bacilli on direct smear examination were graded from 1 to 10 on the Gaffkey scale (11). Haemoglobin was estimated spectrophotometrically by the method of Drabkin and Austin (12). Total leucocyte and lymphocyte counts and differential white cell counts were performed. A commercial test kit (Merck) was used to measure serum albumin levels. The erythrocyte sedimentation rate (Westergren) was read at 1 and 2 hours.

Total serum levels of immunoglobulins in the IgG, IgM and IgA classes were assayed by laser antigens were quantitated nephelometry and the specific antibodies binding to M. tuberculosis by enzyme-linked immunosorbent assay (8). Serum levels of C-reactive protein and the complement components C3, C4, and factor B were assayed by radial immunodiffusion (13) using commercially available antisera (Seward). Rheumatoid factor and antinuclear factor were detected by latex agglutination (Wellcome Laboratories) and immune complexes were detected by Cl q binding (14). Data were stored in, and statistically computer.

analysed

in, a Co

modore PET 2001-I

6N micro-

Results of Disease Six patients were classified radiologically as having minimal disease (grade I). 33 had moderately advanced disease (grade II) and 51 had far advanced disease (grade Ill).

Extent

234

Kardjito and Grange

For analytical purposes patients of grade I and II were considered together in view of the small number of grade I cases. Cavitation was observed in 48 patients (53.3 %). Of the 90 patients, 89 presented with a productive cough and 66 (73 %) had a recent or past history of haemoptysis which was subjectively classified as mild in 10 patients, moderate in 53 patients and severe in 3 patients. A history of haemoptysis was commoner in patients with advanced disease (x2=06.42, p-0.01) and there was a highly significant association between haemoptysis and cavitation (x2=1 9.3, p- >O.OOOl ). Patients with advanced (grade III) disease excreted significantly more acid-fast bacilli in their sputum (p=O.O2), they had higher serum levels of total IgG and IgA (p==O.O5) and a higher ESR (~~0.02) but they did not differ from less advanced cases with respect to specific antibody levels, rheumatoid factor, skin test readings, white cell count, duration of symptoms and serum levels of C3, C4, factor B and C-reactive protein. and haematological findings: The mean values and standard deviations of the ESR, total IgG, IgA and IgM levels, antibodies against M. tuberculosis in these three classes, levels of C3, C4 and factor B, total leucocyte and lymphocyte counts, haemoglobin and serum albumin in the patients and control subjects are shown in table I. Serological

Table I. Comparisons control subjects.

of seriological

and haematological

Patients

Controls

fnvestigation

parameters in patients and Significance

_

Total IgG Total IgA Total IgM *Specific antibody *Specific antibody *Specific antibody Complement C3 Complement C4 Complement factor ESR 1 hr. ESR 2 hr. Leucocyte count Lymphocyte count Haemoglobin Albumin

mgldl

mg/dl w/d1 IgG IgA IgM

B

mg/dl mg/dl mg/dl mm mm mme3 mmT3 g/d1 g/d1

Mean

Std. Dev.

Mean

Std. Oev.

1839 413 155 63.3 7.06 16.3 145 32.7 26.5 64.2 88.5 8062 2397 12.9 4.42

(810) (1204) 108) (35.4) (2.07) (15.8) (40.9) 144) (5.88) (48.0) (41 .O) (2560) (1283)

1475

(769) (162) (98) (16.8) (1.55) (8.73) (24.7) (10.3) (4.57) (20.8) (25.9) (1942) (1394) (1.24) (0.79)

(2.29) (1.11)

261 176 19.5 4.23 17.9 105 33.4 19.9 15.4 27.6 6764 2813 13.6 4.93

(P) 0.001

: 0.001 )

0.1t > 0.001 :> 0.001 0.q . .O.OOl ~:.0.5-f ; 0.001 .-z-0.001 .‘O.OOl -;0.001 0.05 0.02 0.001

*arbitrary ELISA units. tnot significant.

The ESR, total IgG and IgA levels, specific antibodies in these two classes, C3 and factor I3 levels were all significantly elevated in disease. The total white count was significantly higher in patients than in control subjects but the lymphocyte count was lower. The differential count showed that the elevation of the white cell count in disease was due to a significant increase in both neutrophils and monocytes. Haemoglobin and serum albumin levels were lower in the patients than in the controls but the mean levels of C4, total IgM and specific antibodies in this class did not differ significantly between the two groups. Rheumatoid factor was detected in 19 (21.1 %) patients and in 1 (2 %) of the control subjects but the antinuclear factor test was negative in all patients and controls. An increased

Tuberculosis

in

East Java

235

level of C-reactive protein (>6 mg/di) was present in 54 (60 %) patients but not in any control subject. Significant levels of Clq binding were detected in 20 patients and there was a very close association with the presence of rheumatoid factor (L’ -11.9, p :~0.0001). There was a significant correlation between high total IgG levels and a high ESR, C-reactive protein and C3 levels (p=O.OOl, 0.001 and 0.05 respectively) but not with factor B levels (p *:0.5). Significant elevations (above the mean -tl standard deviation of the control levels) of the ESR, C3 and factor B levels were found in 75,67 and 60 (83.3,74.4 and 66.6 %) patients respectively. Of the 90 patients, 84 had a significant elevation of one or more nonspecific indicators of inflammation. Of the 6 patients without such significant elevations, 5 had non-significant levels of antimycobacterial antibody and 4 were classified radiologically as cases of quiescent or inactive disease. Antimycobacterial antibodies The levels of specific antibodies in correlated significantly with the diameter of the skin test response at 0.05 respectively). The IgG antibody levels also correlated, but to diameter of the skin test at 48 hours (p-0.05). Levels of antibody correlate with any of the skin test peaks.

the IgG and IgA classes 6-8 hours (p =: 0.001 and a lesser extent, with the in the IgM class did not

The levels of specific antibody did not correlate significantly with the extent of disease, the presence of cavities, the bacterial content of the sputum, the duration of disease, the total immunoglobulin levels or the levels of C3, C4, C-reactive protein or factor B. Patients with rheumatoid factor had significantly higher levels of specific antibody in the IgM class (mean 25.0 arbitrary units, standard deviation =19.9) than those without rheumatoid factor (mean 14.7 arbitrary units, standard deviation ==I 2.4) (p-0.02). Skin test The response to tuberculin consisted of 5 distinct peaks occurring immediately after injection and after 30 minutes, 6-8 hours, 24 hours and 48 hours. The mean diameters of the skin test responses and the number of patients with distinct peaks at the various times are shown in table II. The most marked responses and the greatest number of distinct peaks occurred at 24 hours.

Table II.

Size of skin test reactions

Time after skin test

Diameter

and number

of reaction

(mm)

of distinct Patients

peaks at different

times.

with distinct peaks -

Mean

Std. deviation

No.

Per cent

15 10 12 44 24

16.7 1 1.1 13.3 48.9 26.7

__I__-Immediate 30 minutes 6-8 hours 24 hours 48 hours

-__ 9.01 11.3 13.2 31.5 18.3

(2.28) (6.58) (9.92) (7.89) (12.5)

Four patients (4.5 per cent) were skin-test negative throughout the 48 hour period and a further four patients were positive at 24 hours but negative at 48 hours. The serum albumin levels were significantly lower in those patients with a skin-test of 10 mm or less at 48 hours (mean 4.33 g/dl) than in those with more marked skin-test responses (mean 4.59 g/dl, p-:0.05). There was an association between a history of weight loss and poor appetite and a low serum albumin level (p=O.O5). None of the skin-test peaks correlated with the extent of disease nor with the presence of cavitation.

236

Kardjito and Grange Discussion

In Java sputum microscopy is the major tuberculosis case-finding technique: facilities for culture are not available in most rural areas. This study has shown that many cases have reached an advanced stage at the time of detection of the disease. The incidence of a history of haemoptysis was high, especially in patients with cavitating disease. It has been shown in previous studies, using enzyme-linked immunosorbent assay (ELISA) (8,15), that a high proportion of patients with tuberculosis have significant levels of antibody to NI. tuberculosis. The antibody levels did not correlate with the extent of the disease nor with the other parameters in this study except for the skin test response at 6-8 hours. This response is an Arthus-type and is due to the formation of antigen-antibody complexes and is thus the most dependant upon circulating antibody levels (4). The 48 hour response also correlated to some extent with specific IgG antibody levels even though delayed hypersensitivity is not antibodydependant. The existence of 5 distinct peaks of reactivity elicited by tuberculin was observed. The nature of the 24 hour peak which was the most prominent one is unknown. Lenzini and his colleagues (9) observed a similar peak in smear-positive cases but not in smear-negative cases with minimal disease. Wilhelm and Romer (18) found that a ribonucleoprotein isolated from M. tuberculosis elicited a 24 hour response only and that this differentiated patients with active tuberculosis from tuberculin-positive individuals without active disease. The degree of tuberculin positivity at 48 hours was affected by the nutritional status as assessed by the serum albumin levels and a history of weight loss but did not correlate with other clinical features. The adverse effect of weight loss and a restricted protein intake on cell-mediated delayed hypersensitivity has been described by Abbassy and his colleagues (17). Caplin has also observed that the degree of tuberculin hypersensitivity bears no relationship to the amount of active disease (18). The non-specific indicators of inflammation, although of little diagnostic usefulness, may be of considerable value as indicators of reactivation of disease. In all except 6 patients at least one indicator was significantly elevated and of these 6 exceptions 4 were regarded by the radiologist as being cases of inactive or quiescent disease, despite the presence of acid-fast bacilli in the sputum. The reason for the high incidence of rheumatoid factor and its association with a high level of specific antibodies in the IgM class is unknown. Rheumatoid factor has been detected in 50 per cent of patients with lepromatous leprosy, a disease in which there is a polyclonal elevation of immunoglobulin levels (19). In this study, however, there was no correlation between the presence of rheumatoid factor and the total of IgG, IgM or IgA levels. The close association between rheumatoid factor and Cl q binding strongly suggests that the latter test was detecting rheumatoid factor in most cases rather than complexes containing specific antibody. Antinuclear factor was not detected in any patient although this factor was present in 30 per cent of patients with lepromatous leprosy (19). The increase in total IgG and IgA but not in IgM in tuberculosis has been reported previously (7,8) and a correlation between IgA levels and extent of disease has also been observed (7). In leprosy a negative skin test occurs at the anergic (lepromatous) pole of the spectrum of disease. In tuberculosis the relationship between tuberculin negativity and immunological anergy is not so straightforward. Although in some regions of the world tuberculin nonreactivity in tuberculosis is rare and is usually associated with serious disease, this is not the case in other regions. In Burma, for example (20), 13 per cent of patients did not react to PPD and it was suggested that excessive sensitization by environmental mycobacteria was able to cause a suppression of delayed hypersensitivity without necessarily compromising the protective immune response. In the present study 9 per cent of cases were tuberculin negative

Tuberculosis

in East Java

237

at 48 hours but did not, as a group, have more serious disease than the tuberculin-reactive patients. The spectrum of immune responses in leprosy has been well described(10) : at the reactive or tuberculoid pole there is a brisk granulomatous response, very few organisms in the lesions, low levels of specific antibodies and a positive skin test. The anergic or lepromatous cases differ in all these respects. Lenzini and his colleagues (9) have postulated the existence of a similar spectrum in tuberculosis. Our findings, however, indicate that there may be fundamental differences in the spectra of immune responses in the two diseases. Neither the levels of specific antibodies nor the degree of skin test reactivity correlated with the extent of disease, indicators of inflammation or bacillary load in the sputum. Antibody levels were not higher in the tuberculin non-reactors and, in contrast to leprosy, there was a significant positive correlation between the levels of specific antibodies in the IgG class and the size of the 48 hour skin test responses. The relevance of the 48 hour skin test response to protective immunity may be quite different in leprosy and tuberculosis. Evidence from experimental studies in animals (21) and natural infections in humans (22) indicate that a marked degree of tuberculin hypersensitivity may have an adverse effect on protective immunity. The principal protective immune response in tuberculosis is macrophage activation (23) which is distinct from the necrotic Koch phenomenon (6). The latter phenomenon effectively eliminates viable organisms in the skin but has leprae differs from n/r. tuberculosis in no effect on systemic disease (24). Mycobacterium that it only appears to replicate within living cells. In this case a necrotic response which kills the infected cells may be advantageous (6) and the spectrum in leprosy may therefore be linked to this phenomenon. It is highly unlikely that NI. tuberculosis evolved from an obligate intracellular parasite like NI. leprae. it is much more likely that its precursor was, like almost all other mycobacteria, an environmental saprophyte. The clinical forerunners of tuberculosis may therefore have been dermal infections following the contamination of scratches by such organisms. Under these circumstances a Koch-type reaction would have been of undoubted value. If pulmonary tuberculosis was of much later onset in mammalian history the evolutionary acquisition of a more appropriate balance of immune responses may not yet be complete. In this context it is noteworthy that strains of mice in whom mycobacteria elicit a marked Koch phenomenon are the most resistant to dermal infections with M. tuberculosis yet the most susceptible to intravenous challenges with this organism (6). Although anergic cases of tuberculosis occur with disseminated lesions showing numerous bacilli but a minimal cellular response (25), this type of disease is usually rapidly fatal owing to the toxicity of IV. tuberculosis. Such cases are uncommon. In the normal course of events it is probable that the balance between the degree of macrophage activation and the inappropriate Koch-type reaction is a major determinant of the nature and outcome of a tuberculous infection. It is likely, therefore, that the observed immunological features of tuberculosis in man are the result of a balance between a number of different effector and suppressor T-cell subsets. This study has illustrated the complexity of the clinical and immunological features of tuberculosis and has emphasised the need for more rigorous criteria for assessing the exact nature of the host-parasite relationships. In particular techniques are required to measure all the various immune responses in this disease and to assess their synergistic and antagonistic interactions. Acknowledgements

We are very grateful to our clinical colleagues, Drs. Santoso, Soetomo, Karim and Habel, for their assistance and to Professors Asimo and Marsetio for performing the radiological and haematological investigations, We are also most grateful to Miss Julia Gibson, Miss Antonia

238

Kardjito and Grange

Batty and Mr Philip Townsend for technical assistance. We thank the Dean of the Faculty of Medicine, Airlangga University and the head of the district health service of Probolinggo and Kediri for their help and encouragement. The British Council supported one of us (T.K.) and the Clinical Research Committee of the National Heart and Chest Hospitals provided a grant for materials for the serological studies. References 1 Youmans, G. P. (1975). Relation between delayed hypersensitivity and immunity in tuberculosis. American Review of Respiratory Disease, 111, 109. 2 Lefford, M. J. (1975). Delayed hyerpsensitivity and immunity in tuberculosis. American Review of Respiratory Disease, 111. 243. between delayed hypersensitivity and cell-mediated immunity. 3 Salvin, S. B. Et Neta, R. (1975). A possible relationship American Review of Respiratory Disease, 111, 373. 4 Pepys, J. (1976). Allergic manifestations in tuberculosis. In : Textbook of lmmunopathology. Edited by P. A. Miecsher and H. J. Muller-Eberhard, p. 453. Grune and Stratton, New York. 5 Rook, G. A. W. (1978). Three forms of delayed skin-test response evoked by mycobacteria. Nature, 271, 64. 6 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 lmmonology, 1, Ill, 2. (1979). lmmunoprofile studies in patients with pulmonary tuberculosis. 7 Skvor, J., Trnka, L, & Kugukovova, Scandinavian Journal of Respiratory Diseases, 60. 148. 8 Grange, J. M.. Gibson, J., Nassau, E. Et Kardjito, T. (1980). Enzymy-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. 9 Lenzini, L.. Rottoli. P. & Rottoli, L. (1977). The spectrum of human tuberculosis. Clinicaland Experimental lmmunology, 27, 230. 10 Ridley. D. S. Et Jopling, W. H. (1966). Classification of leprosy according to immunity. A five group system. lnternational Journal of Leprosy, 34, 255. 11 Baldwin, E. R.. Petroff, S. A. & Gardner, L. U. (1927). Tuberculosis: Bacteriology, Pathology and Laboratory Diagnosis. p. 245. Bailliere Tindall and Cox, London. 12 Drabkin, D. L. Et Austin, J. H. (1932). Spectrophotometric studies: spectrophotometric constants for common 96,719. hemoglobin derivatives in human, dog and rabbit blood. Journalof Biologicalchemistry, 13 Mancini, G., Carbonara, A. 0. Et Heremans, J. F. (1965). lmmunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry, 2, 235. 14 Zubler, R. H., Lange, G., Lambert, P. H. Et Miescher, P. A. (1976) Detection of immune complexes in unheated sera by a modified I-Clq binding test. Journalof Immunoiogy, 116, 232. 15 Nassau, E., Parsons, E. R. Et Johnson, G. D. (1976). The detection of antibodies to Mycobacterium tuberculosis by microplate enzyme-linked immunosorbent assay (ELISA). Tobercle, 67, 67. of a newly isolated oligopeptide from Mycobacterium tuberculosis 16 Wilhelm, G. Et Riimer, C. (1977). An investigation with antigenic properties specific for the diagnosis of tuberculosis. Zentralblatt fiirBakterio/ogie, Parasitenkunde, lnfektionskrankheiten und Hygiene, I. Originale, A., 239, 379. 17 Abbassy, A. S., Badr El-Din, M. K., Hassan, A. I., Aref, G. H., Hammad, S. A., El-Arady. I. I., et a/. (1974). Studies of cell-mediated immunity and allergy in protein energy malnutrition. 1. Cell-mediated delayed hypersensitivity. Journal of Tropical Medicine and Hygiene, 77. 13. The Tuberculin Test in Clinical Practice, p. 59. Bailliere Tindall, London. 18 Caplin, M. (1980). 19 Wager, 0. (1969). Immunological aspects of leprosy with special reference to autoimmune disease. Bulletin of the World Health Organisation, 41. 793. 20 Shield. M. J.. Stanford, J. L., Paul, R. C. Et Carswell, J. W. (1977). Multiple skin testing of tuberculosis patients using a wide range of new tuberculins and a comparison with leprosy and Mycobacterium ulcerans infection. Journal of Hygiene, 76, 331. 21 Wilson, G. S., Schwabacher, H. Et Maier, I. (1940). The effect of the desensitization of tuberculous guinea-pigs. Journal of Pathology and Bacteriology, 60, 89. 22 Turner, H. M. (1953). Correlation of quantitative Mantoux reactions with clinical progress in pulmonary tuberculosis. Tubercle, 34,155. of antituberculous immunity. American Review ofRespiratory Disease, 97, 23 Mackaness, G. B. (1967). The immunology 337. 24 Koch. R. (1891). Fortsetzung der Mittheilungen iiber ein Heilmittel gegen Tuberculose. Deotsche Medizinische Wochenschrift, 17. 101. 25 Proudfoot. A. T. (1971). Cryptic disseminated tuberculosis. British Journalof Hospital Medicine, 6,773.