Bull. Org. mond. Sante 1970, 42, 673-702 Bull. Wld Hithi Org.

Serology in Leprosy

J. OLIVEIRA DE ALMEIDA 1

A critical survey of the literature on serology in leprosy has shown that sera takenfrom lepromatous patients display some striking differences in comparison with sera from tuber- culoid patients. The tests most frequently employed were complement-fixation, haemag- glutination, electrophoresis, precipitation and immunofluorescence, together with a variety of antigens not only from lepromas but also from Mycobacterium tuberculosis and other actinomycetales. With the exception of the Rubino test, all these serological tests are lacking in specifi- city for leprosy since leprous sera have a broad range ofreactivity with different antigens, including those employed in the serological diagnosis of syphilis. Some features of the leprous sera could be related to a hypersensitivity state involving circulating immune complexes, low levels of complement and the presence of antibodies similar to those found in sera from patients with autoimmune diseases.

Studies on sera of leprosy patients have been made groups of lepromatous patients, before and after for the purpose of finding a reliable test for the treatment. However, a wide and clinically unex- diagnosis of leprosy, for evaluating the effect of plained fluctuation in serum titres has been observed treatment, and also for determining the humoral in the same patient. These fluctuations, resulting changes produced by the infection. from antigens released from lesions, combining Complement-fixation, since its discovery early with circulating antibodies, called attention to the this century, has been shown to be a reliable diag- dynamics of the infection. The role of the circulat- nostic test in many infections but its application to ing immune-complexes must be fully studied, but the diagnosis of leprosy was impaired by the lack their possible presence in the plasma of leprosy of a specific antigen. On the other hand, leprous patients is not incompatible with many observations sera have a polyfixing capacity with antigens not concerning the formation of autoimmune anti- related to Mycobacterium leprae. Moreover, it was bodies,-namely, the low level of complement, the found that sera from recognized cases of leprosy histamine-like contraction factors and the polyfixing failed to show any reactivity with antigens, even capacity observed with non-specific antigens. those prepared from lepromas. However, although In recent years, investigations have been carried the complement-fixation test for leprosy failed in its out with a variety of methods to study the distribu- main purpose, interest in this test has been main- tion and characterization of immunoglobulins at tained. Leprous sera were always tested in the various stages of the disease. More refined methods surveys for evaluation of the specificity of a variety for the diagnosis of other infections in leprous sera of antigens, including those for the sero-diagnosis have been used, mostly in so-called " biological of syphilis. false positive reactions" for syphilis. As far as With the advent of chemotherapy, it was expected specificity for leprosy is concerned, the Rubino test that serological tests could be employed for the is the only fully specific test known at present. evaluation of treatments, in the same way as the Although the nature of the factor is not fully under- Wassermann test for syphilis is used. This was stood, the test offers a possibility for the explora- found to be possible in the comparative serology of tion of this factor found only in leprosy patients. In many cases of leprosy, however, the test is of 1 Professor, Department of Microbiology and Immuno- low sensitivity, either of itself or from a lack of logy, Medical School of Ribeirao Pr8to, and Director, WHO Reference Laboratory for Serology in Leprosy, University the Rubino factor. of Sao Paulo, Brazil. It must be kept in mind that a serological test

2508 -673- 674 J. 0. de ALMEIDA for the early diagnosis of the disease has not yet noted that this has not so far been possible even been discovered. If cultures of Myco. leprae were with Myco. tuberculosis. obtained, specific antigens could be possibly pre- In this review, the main topics concerning the pared for the diagnosis of leprosy- but it should be serological study of leprous sera are discussed.

COMPLEMENT-FIXING REACTIONS WITH ANTIGENS PREPARED FROM ACID-FAST BACILLI AND SERA FROM LEPROSY PATIENTS

The application of complement-fixation reactions time, the methods for complement-fixation were to the diagnosis of diseases such as syphilis, plague, improved. typhoid fever and melitensis fever aroused the inter- A quantitative complement-fixation method, once est of leprologists in the development of a test to standardized for leprosy, could provide a reliable be used for the diagnosis of leprosy. procedure for the evaluation of sulfone treatment in In the same year that the complement-fixation a way similar to that in which the Wassermann test is reaction was applied to the serological diagnosis of used to evaluate treatment of syphilis. Although, syphilis, Eitner (1906) showed that an antigen pre- on the whole, complement-fixation titres correspond pared from leproma was able to fix complement to the clinical status of the patient, fluctuations have with serum from a leprosy patient. However, it was been observed in one and the same case that were soon observed that leprous sera could also react not related to the effect of the treatment. Some of with antigen preparations free from Myco. leprae this information provided by the application of and containing only a colloidal solution of lecithin complement-fixation reactions to leprosy patients (Statineanu & Damelopolu, 1909a, 1909b). under treatment might complete the knowledge of In the preparation of antigens for the complement- the pathology of the infection. fixation test for leprosy, Myco. tuberculosis, being During the period when the specificity of the easily cultivated in synthetic media, was used. When complement-fixation test was being tested, cross- cultures were obtained from lepromas, the supposed reactivity with visceral leishmaniasis was observed. leprosy bacilli were used for the preparations of The test could be applied to the serological diag- antigen and very soon it became evident that most nosis of kala-azar when the serum was diluted of the acid-fast, saprophytic mycobacteria also yield to such an extent that reactions with tuberculosis potent antigens for the complement-fixation test. and leprosy no longer occurred. A study of this Interest in complement-fixation as a specific test for cross-reactivity led to the discovery of antigens leprosy slowly decreased when its application showed present in the tubercle extract which with antibody that in a large number of sera from leprosy patients formed complexes that do not fix complement. the reactions were negative, but that, on the other hand, in cases where leprosy, and also tuberculosis, ANTIGENS USED IN COMPLEMENT-FIXATION could be excluded the reaction was positive. Although REACTIONS FOR LEPROSY purified preparations were used in tuberculosis for the serum diagnosis, the results were not encouraging Antigen preparations used in the serology of since other methods could be used, more reliable in leprosy can be presented in four main groups: clinical aspects than the serological test. (1) antigens prepared from Myco. leprae; (2) those The complement-fixation test, if not suitable for prepared from Myco. tuberculosis; (3) antigens from the detection of incipient cases of leprosy, could, other cultivable mycobacteria; (4) preparations from however, give information on the presence of circu- Myco. lepraemurium. lating antibodies, indicating conditions which were shown differently in the polar forms of leprosy. Reactions with antigens of Myco. leprae Antigens for that purpose were also prepared from Aqueous extract oflepromas reacted with sera from Myco. tuberculosis or other acid-fast mycobacteria, leprosy patients in complement-fixation tests (Eitner, and fractions could be isolated and compared with 1906), but alcoholic preparations were more suitable each other in order to have a standard antigen since they could be kept longer (Babes & Busilla, which could be taken as a reference. At the same 1909). Up to the year 1915, 185 reactions with SEROLOGY IN LEPROSY 675 leproma antigens had been achieved; the positivity culosis was standardized against immune serum was 75 Y/ for the alcoholic antigen and 84% for the from a horse and that the boiled aqueous extract aqueous preparation (Cooke, 1919). was used as antigen (Wadsworth et al., 1931). Antigen preparations from leproma were handi- The protein extract from BCG proved somewhat capped by the non-specific and anticomplementary less sensitive than the suspension itself when tested activity of tissue debris and they were replaced by with sera from lepromatous patients, but it was tubercle preparations when it was found that these more active with sera from tuberculoid cases also reacted with leprous sera. (Merclin & Pogorelov, 1964). "Crude " Myco. tuberculosis antigen preparations Purified Myco. tuberculosis antigens A suspension of Myco. tuberculosis in saline solu- The specific fraction of Myco. tuberculosis which tion reacted with 11 sera from leprosy patients, giv- reacted with tuberculous sera was insoluble in ing strong positive complement-fixation reactions acetone, ether, and chloroform, but soluble in water (Frugoni & Pisani, 1908). Not only were the bacilli and ethanol (Pinner, 1925). The presence of other used as antigen, but tuberculin also was able to fractions in an antigen preparation can affect the react with leprous sera. solubility of any other, making it very difficult to A pyridine-soluble antigen from Myco. tuber- determine the properties of the specific fractions cutlosis (WKK strain) was prepared by Witebsky (Faure, 1940). et al. (1931) to be used in the serology oftuberculosis. A boiled aqueous extract of Myco. tuberculosis, This antigen was readily available and was widely previously dried and treated with acetone, contained accepted, since the reaction was more sensitive with lipids and polysaccharides (Wadsworth et al., 1925). this antigen than with the antigen prepared by The proteins, although denatured during the prepa- Besredka (1935). The WKK antigen reacted with ration of the antigen, could adsorb the soluble sera from lepromatous patients more frequently than carbohydrates and reacted with tubercle sera (Ander- with sera from tuberculoid or indeterminate cases son, 1927). The reactivity of the antigen with (Brants, 1932). No cross-reactivity was observed tuberculous or leprous sera was related to the with syphilitic sera (Aoki & Murao, 1933), although content of polysaccharides (Meyer, 1937; Seibert, they reacted with leproma antigen (Kornel, 1933). 1941), although in some preparations the proteins When this antigen was employed in sera from could take part in the reaction (Schaefer, 1940). different forms of leprosy, it was observed that In the preparations from Myco. tuberculosis, the most of the lepromatous sera showed a high reactiv- phospholipid fractions were the antigens in the ity whereas no, or only very weak, reactions were complement-fixing complexes (Anderson, 1927; Pin- shown by tuberculoid cases (Pereira, 1936, 1938; ner, 1928; Chargaff & Schaefer, 1935; Machebceuf Rabello & Machado, 1936; Rabello et al., 1938). & Bonnefoi, 1935; Machebceuf & Faure, 1939; An aqueous extract from virulent Myco. tuber- Machebeeuf & Levy, 1935). The composition of ctulosis previously treated with acetone, has been the antigens varied considerably. The phosphatide found suitable for the standardization of the com- fraction of Anderson (1927) had 0.4% of nitrogen plement-fixation test for tuberculosis based on the but those prepared by Macheboeuf & Bonnefoi 50% end-point for haemolysis. This antigen was (1935) and Machebceuf & Levy (1935) did not. devoid of anticomplementary or thromboplastic The WKK antigen contained lipoids, probably phos- activities and when tested by the same methods pholipids and polysaccharides. The fractions iso- with leprous sera was found to be highly specific lated by Machebceuf & Faure (1939) contained (Maltaner, 1940). The fractionation of the aqueous phosphatides bound to polyalcohols. A further extract yielded two haptens whose reactivities were purification of this antigen yielded glycerophospha- not the same for human sera from tuberculosis tides containing ozides and inositol (Faure, 1940). patients and for the serum of an immunized horse No comparative study with these antigens in terms (C. Padron, personal communication). These dif- of serological activity has been made because of ferences might be related to the animal species the lack of uniform techniques and the different (Illand, 1951). Horse serum is therefore not a suitable criteria for the evaluation of their specific reactivities. reagent for the study of the aqueous extract prepara- An antigen repeatedly precipitated by acetone tions of Myco. tuberculosis. It is worth mentioning from an aqueous extract of Myco. tuberculosis was that the quantitative complement-fixation for tuber- used in 534 lepromatous sera in reactions carried 676 J. 0. de ALMEIDA out by the New York State Department of Health serum against antigen, the corresponding points of quantitative complement-fixation technique (Almeida isofixation describe an isohaemolytic curve of type I et al., 1956). The patients were distributed in three (Almeida, 1956b). When 2 doses of complement groups: the first group contained 74 untreated are used, two isohaemolytic curves can be drawn patients, the second group included 217 leprosy through the 50% end-point of haemolysis. These patients under sulfone treatment and the third lines are parallel to each other, providing conditions group, 243 lepromatous patients who became bac- for the antigen and for the antibody titrations teriologically negative after treatment. The serum (Almeida, 1958b). Other systems such as tubercu- titres determined by complement-fixation were dis- losis and Chagas' disease were found to have the tributed as follows: in the first group, 19% pre- same type of isofixation curve. In the leprosy system, sented titres less than 10, 51 % were between 10 the isofixation curve has its two branches practically and 100 and 30% higher than 100; in the second parallel to the antigen and the serum axis, meaning group, 33% of serum titres were less than 10, 52% that the immune complex formed is insoluble in an were between 10 and 100, and 15 % higher than 100; excess of antigen or in large doses of serum. in the third group, 70% of serum titres were less Paradoxical reactions. It was observed that, using than 10, 27% were between 10 and 100 and only some pyridine-soluble antigen prepared from virulent 3 % were higher than 100. These data showed that Myco, tuberculosis, the isohaemolytic curves with the level of complement-fixing antibodies in lepro- leprous serum were not parallel to each other, matous patients was related to the activity of the crossing in the zone where antigen was in excess. disease. This paradoxical reaction was observed when less The pyridine-soluble part of the antigen, precipi- fixation of complement was obtained with an tated by acetone, was found to be a mixture of increased amount of serum. It was found that the lipolysaccharides. This antigen preparation was antigen preparation might be contaminated by a titrated with lepromatous sera and with sera from second antigenic component which, when bound to tuberculosis cases. The optimal reactive dose for the leprous antibody, formed an immune complex leprosy differed from that more suitable for tuber- that failed to fix complement (Thompson & Almeida, culous sera. 1956). The isolation of this antigen from the It was found that the reactions in tuberculosis pyridine-soluble preparation was achieved by Pang- could be detected with both doses of antigen, but born (1956) with the aid of chromatography on the antigen diluted for leprosy yielded less intense silicic acid columns. The blocking antigen was a reactions. Sera from 111 lepromatous cases were polysaccharide which differed from that isolated treated with the 2 doses of antigen. Only 2 reactions by Seibert & Soto-Figueroa (1956) because it failed occurred with the dose of antigen adjusted for to fix complement, not only during incubation at tuberculosis; however, when the proper dose of 37°C but also after it had been allowed to stand antigen was used, 103 sera reacted with titres higher overnight at 4°C. than 12. When lecithin was added to this antigen, The inhibition produced by this blocking antigen its anticomplementary activity was reduced and in was more pronounced in tests on tubercuilous serum a series of complement-fixation tests it gave rather than on leprous serum. However, no inhibition encouraging results: no reaction was observed with occurred with kala-azar serum, demonstrating a lack the sera of tuberculin-negative normal children, of combining power of the contaminant antigen for whereas 69% of 177 sera from tuberculosis patients the leishmaniasis antibodies. This observation ex- showed significant reactions, and all of the 112 sera plains the high titres obtained with crude tubercle from lepromatous cases reacted with the highest antigen preparations which contain large amounts titres (Pangborn, 1955). of blocking antigen (Almeida, 1956a, 1957, 1958a). In the study of these purified tubercle antigens, This antigen, hidden in the preparations, was called there are three aspects that should also be considered; cryptoantigen and the reactions, cryptoreactions they are as follows. (Almeida, 1958a). When leprous sera reacted with Isofixation curves. When varying amounts of lep- tubercle antigen containing such an inhibitor, the rous serum react with varying doses of antigen, and reactions obtained did not express the real level of with a constant dose of complement, various com- specific antibodies, since these reacted not only binations of immune serum and antigen are able to with the fixing but also with the non-fixing antigens fix the same amount of complement. By plotting (Almeida, 1958a). Further purification of the inhibi- SEROLOGY IN LEPROSY 677 tor fraction in the tubercle antigen is in progress duce the immediate histamine-like contraction of (Pangborn, 1966). the uterine horns of normal guinea-pigs, as observed Fluctuations in the level of circulating antibodies: by Castro et al. (1963) with active lepromatous sera. circulating immune-complexes. Collier & Gehr (1957) After inactivation, this effect was no longer observed. called attention to 15 negative complement-fixation The mechanism involved in this phenomenon might results obtained among 53 bacteriologically positive be the same as that described by Kulka (1942) and leprosy patients. According to these authors " this Germuth & McKinnon (1957) as a consequence of might perhaps be explained by the fact that periodi- tissue lesions produced by a soluble antigen-antibody cally more bacteria circulate in the blood, which complex. neutralize the complement-fixing antibodies. Serial The complement-fixation method with purified examinations made in one and the same patient fractions of Myco. tuberculosis, or from other myco- might solve the question whether the negativity is bacteria including Myco. leprae, might provide infor- temporary (as we assume) or continuous ". mation on the presence of circulating free antibodies, Almeida & Bechelli (1961) applied the quantitative free antigens or formed immune-complexes. Further complement-fixation test with pyridine-soluble tuber- investigation will show if these circulating immune- cle antigen in the examination of sera collected dur- complexes play some role in the pathology of ing 1 year of observation of lepromatous patients leprosy. under suffone treatment. In 3 of 11 lepromatous patients, the complement-fixation titres decreased Antigens from other acid-fast bacteria regularly and were related to the improvement The supposed leprosy bacilli cultivated from lep- observed in these patients. In other cases, however, roma were used as a source of antigen as early as a fluctuation in the titre was observed. 1923 (Lewis & Aronson). The antigens from Clegg's, When antibody titrations were performed with Duval's or Kedrowski's cultures also were saline active and inactivated leprous sera, it became evident suspensions of such bacteria. More stable antigens that heating at 56°C for 30 minutes could reduce were prepared when the bacilli were previously the complement-fixing capacity of the sera (Almeida, treated with acetone and suspended in saline solu- 1963). Active sera might appear anticomplementary tion. With these antigens, Taylor & Malone (1924) in tests with guinea-pig complement, but this effect tested 100 leprous sera and detected 90 reactors by was lost by heating the serum. At the same time, complement-fixation; the controls (30 syphilitic and it was observed that the level of circulating comple- 14 normal sera) did not react. ment varied considerably in lepromatous patients. Dharmendra & Bose (1941) and Dharmendra When the ratios between titres with active and (1941) tried, without success, to improve the specifi- inactivated sera were plotted against the amount city of sera with absorption by mycobacterial anti- of free complement present, a correlation could be gens, including Myco. tuberculosis and Myco. phlei. found, showing a proportionality between high The absorption of leprous sera by various bacillary titres in complement-fixation tests with active serum suspensions did not make the reaction more specific and the low level of circulating complement or suitable for the serological identification of cul- (Almeida, op cit.). tures of the supposed leprosy bacilli. This behaviour of the lepromatous sera could be Antigens prepared from cultures of " Streptothrix a consequence of the presence ofcirculating immune- leproidis " by various methods, reacted in comple- complexes, formed occasionally by the combination ment-fixation tests showing a high reactivity with of antigens released from lesions with their specific sera from lepromatous patients (Gomes, 1927). antibodies. The immune-complex would fix com- This antigen was employed for the diagnosis of plement from the patient's own serum and this leprosy (Gomes & Patteo, 1928) or for detecting would explain the low levels of complementary infected contacts (Marcondes, 1929; Gomes, 1935). activity found in these cases. These circulating When lecithin was added to the antigen, its reactivity immune-complexes would be unable to bind further with leprous sera was enhanced (Freund, 1927). complement after the inactivation of the serum by If leprous sera reacted with lecithin, the additive heat and under these conditions the complement- effect would be responsible for the increased sensi- fixation titres will depend on the amount of free tivity of the reaction. antibodies in the serum. A battery of 19 antigens prepared from sapro- These circulating immune-complexes might pro- phitic acid-fast bacteria, Myco. tuberculosis and 678 J. 0. de ALMEIDA diphtheroid bacteria was tested with sera from 20lep- Cross-reactivity with kala-azar sera was also ob- rosy patients who reacted with all of them (Cooke, served with methanol extracts of Myco. tuberculosis 1919). The capacity of leprous sera to react with by Bayarri & Ahuir (1947). so many antigens was a characteristic almost diag- nostic of the disease, since it was not observed with sera from either syphilis or tuberculosis (Lewis & CONCLUSIONS Aronson, 1923). From this review, some conclusions can be drawn Antigen from Myco. lepraemurium about the application of complement-fixation in Bacillary suspensions prepared from murine lepro- leprosy. sy lesions with the addition of cephalin reacted with (1) Sera from lepromatous cases react more fre- sera from 2 among 34 contacts of leprosy. One child quently, and with higher titres, than sera from showed symptoms of leprosy 2 years later (Ichiata, tuberculoid or indeterminate cases. 1940). According to Ichiata (op. cit.), this antigen could be used for the early diagnosis of leprosy. (2) Leprous sera show a broad reactivity with Rat lepromin was used as antigen in a complement- mycobacterial antigens, which can be prepared from fixation test on 146 leprosy patients and was shown cultivated acid-fast mycobacteria as well as from to be more specific than a BCG suspension. The Myco. keprae found in leproma. Myco. lepraemurium antibodies detected in the 96 reactors correspond to is also able to react with leprous sera in complement- the antimycobacterial antigens, and the specific ones fixation tests. were those detected also by reactions performed with (3) Most of the antigens used in complement- Mitsuda lepromin as antigen (Merclin & Borisova, fixation for leprosy are prepared from Myco. tuber- 1964). culosis. The purification of the crude antigen prepa- ration yields fractions which could be tested with leprous sera; from the observation of paradoxical REACTIONS WITH SERA FROM KALA-AZAR reactions, one of these antigens was found to be an The antigen prepared from cultures of Strepto- inhibitor. thrix leproidis reacted with sera from visceral leish- (4) The non-fixing antigen is able to combine maniasis patients in complement-fixation tests with antibodies from tuberculosis and from leprosy, (Assumprao & Silveira, 1935). This cross-reactivity but not with antibodies found in the sera of visceral with sera from a disease not caused by mycobacteria leishmaniasis. The presence of this antigen in the was confirmed by Bier & Arnold (1935) with the crude preparations of Myco. tuberculosis antigen WKK antigen, or with the antigen prepared by the does not have any effect on the complement-fixing Witebsky method from cultures of streptothrix capacity of kala-azar serum. The cross-reactivity of (Bier & Planet, 1936). the tubercle antigen and kala-azar antibodies pro- When sera were diluted appropriately, the reac- vides a practical test for the diagnosis of this dis- tions for leprosy or tuberculosis decreased or disap- ease, even in countries where tuberculosis and peared, whereas the kala-azar sera still reacted. This leprosy are prevalent. cross-reactivity is the basis of the serological diag- By diluting the serum for complement-fixation nosis of kala-azar and was confirmed by parasito- tests, reactions for leprosy and tuberculosis are logical examination, even in countries where leprosy minimized, whereas that for kala-azar is maintained. is endemic (Row, 1939; Lowe & Greval, 1939; Greval et al., 1939, 1941; Lowe, 1942; Sen Gupta, (5) The reaction between leprous sera, the pyri- 1945, 1948; Dharmendra et al., 1946; Greval, 1947). dine-soluble tubercle antigen and complement pres- The WKK antigen reacted also with sera from ents the same type of isofixation curve as tubercu- cutaneous leishmaniasis, pemphigus, but not with losis. A quantitative complement-fixation test could sera from South American blastomycosis patients be standardized, based on the linear relationships (Eichbaum, 1942). When the WKK antigen was between the immune-complexes and complement prepared from cultures of Kedrowski's bacillus, its required for 50% haemolysis. specific reactivity for visceral leishmaniasis was (6) Fluctuations in complement-fixation titres 94.6% in relation to the parasitological diagnosis in could be a consequence of heat-lability of circulating a total of 1917 sera (Dharmendra et al., 1946). immune-complexes formed by the specific antibodies SEROLOGY IN LEPROSY 679 and the antigens released from lesions. These inactivation. Heating the serum was found to immune-complexes are able to fix complement in vivo decrease the titre, since it is supposed that the and, therefore, the low concentration of complement immune-complexes were unable to fix complement in lepromatous leprous sera was related to the after inactivation or produce histamine-like con- complement-fixing titre obtained with serum, before tractions in the uterine horns of normal guinea-pigs.

SERUM PROTEINS IN LEPROSY

The quantitative and qualitative analysis of serum not only in the total content but also in the ratio proteins in sera from leprosy patients, showed between the albumin and the globulins (Stevenson, profound departures from normal values of the 1925; Frazer & Wu, 1925; Neill & Dewar, 1927; albumin and globulin fractions. Electrophoresis, a Schossmann, 1930). more refined method of protein analysis, not only With more refined methods, electroflocculation confirmed the previous findings but also gave further and colorimetry, Ishihara (1950) observed an increase information on the identification of the protein of y-globulin in the lepromatous type of leprosy as groups, according to their mobility in an electric compared with the level found in tuberculoid and field. Electrophoresis in paper for the study of the indeterminate forms, which does not differ from leprosy serum was preferred by the majority of normal values. The albumin: globulin ratio was workers. The fractions were evaluated by densito- less than 1 in the lepromatous patients and in those metry of the fractions or by their free distribution with erythema nodosum. The changes in the serum in liquid medium. In a few studies, the preparative protein were related to the clinical status of the electrophoresis method of Tiselius was used, as well patients, since the improvement was reflected in a as separation in agar gel or in cartons. decrease of euglobulin (Uyguanco et al., 1950). The The analytical results were reported as the relative levels ofy-globulin in the lepromatous cases present a proportion of the various fractions rather than in correlation with the acceleration of blood sedimen- absolute amounts. In the evaluation of the electro- tation. Changes in the albumin: globulin ratio could phoretic pattern, it had been shown that conditions be related to the appearance of erythema nodosum other than leprosy can affect the proteinogram, (Ishihara, 1953). making any correlation between leprosy and the electrophoretic profile uncertain. On the other hand, the lack of uniformity in the methods for the evalua- ELECTROPHORETIC ANALYSIS tion of protein fractions made comparison of the With the aid of electrophoresis, a more detailed results rather difficult. However, in the majority of analysis could be made of the serum proteins in reports, the information confirmed the different leprosy. Seibert & Nelson (1943) were the first to electrophoretic profile in lepromatous leprous serum show an increase in a- and y-globulin in sera from as compared with tuberculoid leprous serum. advanced cases of leprosy, although the total pro- With the development of immunoelectrophoresis, teins were within normal limits. This finding was each one of the protein fractions has been shown to confirmed by Leon Blanco & Lavernia (1948) in be a mixture of several individual globulins. In addi- their studies on 195 sera from leprous patients; tion to this analytical work, it was possible to deter- only 2 presented hyperproteinaemia but 94 had mine the distribution of each globulin according to an increase of the globulin fractions. Hoxter et al. its electrophoretic mobility in agar gel, and by the (1951) determined the electrophoretic patterns use of pure globulin fractions it was possible to in 24 patients. In all but 1 of 13 lepromatous cases, prepare an immune serum for their further iden- a considerable increase of y-globulin was observed. tification. From the immunoelectrophoresis of lep- Values for y-globulin in the tuberculoid and inde- rous sera, the immunoglobulins were found to have terminate patients were within normal limits. The a quite different distribution in the two polar forms electrophoretic pattern described in lepromatous of leprosy. The serum analysis of proteins from leprosy was not related to the formation of antibody leprosy patients has shown some marked differences, against the Myco. leprae, being, rather, a non- 680 J. 0. de ALMEIDA

specific reaction of the reticuloendothelial system be produced also by intercurrent parasitic diseases. (Arcuri & Inzerillo, 1952). The electrophoretic patterns found in the sera The hypergammaglobulinaemia, with concomitant of lepromatous and tuberculoid leprosy patients hypoalbuminaemia, such as that found in most of could be used as criteria for the classification of the lepromatous leprosy patients, could be present the condition, according to Pozzo & Hofmann in other affections not related to leprosy, as was (1955). shown by Bechelli & Fiorillo (1953) among patients Different workers have tried to correlate electro- with schistosomiasis, collagen diseases and kala-azar. phoretic findings with the clinical status of patients Other conditions such as tuberculosis, nephritis, and many contradictory reports can be found in the nephrosis and amyloidosis, can modify the electro- literature. Miguel et al. (1954) found more pro- phoresis curves (Jardin & Beytout, 1960). nounced protein changes in lepromatous patients In tuberculoid and indeterminate (" neural " and with amyloidosis and during the lepra reaction i' macular ") leprous patients, Namba & Fujiwara while the results of Tarabini (1957a) showed no (1952) observed only a slight increase of y-globulin, differences in the globulin balance in the sera of which was markedly increased in lepromatous 100 leprosy patients. Tarabini (op. cit.) tried to (" nodular") patients. An increase of all the glo- correlate amyloidosis with the disproteinaemia in bulins was found in erythema nodosum. This obser- 50 patients and concluded that hyperproteinaemia vation was confirmed by Kono et al. (1952) who and the storage of paraproteins in the viscera are showed that when the erythema nodosum subsided relevant causes. The leprous infection could be the level of y-globulin tended to decrease. responsible for the hypergammaglobulinaemia and disproteinaema, since they are observed mainly among lepromatous patients (Bonatti & Lebron, ELECTROPHORETIC PATTERN IN THE POLAR 1958). On the other hand, the improvement of the FORMS OF LEPROSY patient is reflected in the electrophoretic profiles; this was observed by Melamed & Barcia (1958) in According to Orbaneja et al. (1953), 2 electro- patients with lepra reactions, before and after phoretic patterns could be recognized in lepro- hormone treatment. Sulfone treatment and BCG matous patients. In the first there is a hyper- vaccination might exert a favourable influence on proteinaemia, an increase lof y-globulin with minor the humoral state of leprosy patients tested by alterations in the a and fl fractions, and a reversed electrophoresis, according to Serie & Schaller (1957) albumin: globulin ratio. The other pattern pres- who observed 202 leprosy patients, each one being ents hypoproteinaemia, a normal albumin: globulin examined twice with a 6-month interval between ratio, an increase of the a-globulin, a decrease of examinations. #-globulin and no significant alteration in the A correlation was also observed between the y-globulin. In the tuberculoid or indeterminate increase of y-globulin and the bacillary contents forms, most of the cases are within normal limits, from cutaneous or nasal mucosal smears, in a but they can present electrophoretic curves of one study of 486 lepromatous leprosy patients, before or other pattern. During the lepra reaction an any treatment was given (Guinto & Mabalay, 1958). increase of the a- and the y-globulin and a decrease In a study of 102 leprosy patients and 28 contacts, of the albumin fraction was observed (Contreras a significant difference was shown in the protein et al., 1953). The fl-globulin was not greatly affected pattern according to whether the disease was during the leprous infection. lepromatous or tuberculoid. The differences in the An index used by Mauze & Arnaud (1954) to results have been found to be statistically significant express the relative proportion of electrophoretic (Lechat et al., 1961). The authors pointed out that fractions was obtained by dividing the area deter- there is one characteristic electrophoretic spectrum mined with the fraction from leprous serum by for leprosy. On the other hand, Mayama (1954) that of a normal serum. The two electrophoretic could not find any relationship between the electro- analyses were made at the same time. They found, phoresis findings and the clinical condition of whatever the clinical form of the case, that the 63 leprosy patients, although he observed in lep- album fraction diminished while the y-, a-, and romatous cases with erythema nodosum a remark- #-globulins constantly increased. In relation to the able decrease in albumin, a significant increase in index values for a- and fl-globulins, alterations can y-globulin and a rise in total serum proteins. SEROLOGY IN LEPROSY 681

LIPOPROTEINS GLYCOPROTEINS Tarabini (1957b) studied the relation between Glycoproteins were determined by Tarabini (1957) lipoproteins and blood cholesterol in the sera of in the sera of 36 leprosy patients. Among the 10 lep- 24 leprosy patients. The ratio between a- and romatous cases, a moderate decrease of glycoprotein fl-lipoproteins was normal in 4 patients. In 6 cases in the albumin fraction and an increase in the of lepra reaction or with liver affection, there was y-globulin fraction were observed. Other glyco- a relative decrease of a-lipoproteins but a normal proteins, such as a-, fi- and y-glycoproteins, were level of cholesterol. In 9 patients with nephrosis, increased in all 20 patients with lepra reactions. the cholesterol was increased and the a-lipoprotein In 4 nephrotic cases a great decrease of glycoprotein was proportionally low. In 4 lepromatous patients in the albumin fraction and a very large increase in treated with ACTH or one of the corticosteroids, the a-, ,B- and y-globulins were observed. Treatment a reduction of blood cholesterol and an increase of with corticosteroid and ACTH improved the glyco- a-lipoproteins were observed while at the same protein curve though it rarely became normal. time the proteinogram tended to be normal. Glycoprotein fractions were determined by May- Electrophoresis and analytical ultracentrifugation ama (1957) by paper-electrophoresis in 41 cases of were employed by Mayama (1958) in the analysis leprosy (25 lepromatous and 16 tuberculoid). In the of sera of from leprosy patients and 3 healthy lepromatous cases, the y-globulin was increased but persons. The lipoprotein of a " maculo-anaesthetic there was no significant difference in the albumin- case" was found not to be significantly different bound carbohydrate compared with the control. from that present in normal sera. In 5 cases of The y-globulin-bound carbohydrate, on the other tuberculoid leprosy, the molecular weights of the hand, was decreased. Furthermore, there was no lipoproteins were found to be below normal but rise in f-globulin, even though the ,B-globulin-bound the biggest difference was in the values for lipo- carbohydrate was increased. As far as the glyco- proteins isolated from the sera of 7 lepromatous and lipoproteins are concerned, there is no diag- (" nodular ") cases. nostic pattern for leprosy, although the disease alters Chekherdenian (1960), studying 50 patients with the relative proportion of the glyco- and the lipid- different forms of leprosy, found that during the bound proteins. lepra reaction, the serum lipids, as well as the a-globulin, could be increased at the same time as the y-globulin. Mayama (1960) was able to show MACROGLOBULINS that the lipoprotein isolated from a lepromatous The macroglobulin sedimentation diagram in the case differed from that from a tuberculoid case. majority of the indeterminate, tuberculoid and bor- When these lipoproteins were injected into rabbits derline cases of leprosy did not differ from the and chickens, antibodies were formed only in those normal according to Mayama (1965). However, he animals inoculated with tuberculoid lipoproteins. found a component of different sedimentation rate Antibodies could not be demonstrated by agar- in the sera of patients with lepra reactions. diffusion precipitation nor by immunoelectropho- with resis the lipoproteins isolated from lepromatous IMMUNOELECTROPHORESIS leprosy. Both of the lipoproteins, however, were able to activate the phagocytic activity of leucocytes In immunoelectrophoresis, the protein separation for the Myco. tuberculosis. is carried out on glass slides covered with a gel Studying the serum protein pattern in leprosy, (agar or starch) and an anti-human serum is used Muelling et al. (1960) considered that the determina- for detection. The precipitates can be seen and tion of lipoproteins is not particularly helpful for recognized. This method was applied to leprous diagnosis, although the mucopolysaccharide appears sera by Bergot et al. (1962), who showed that the to be related to the activity of the infection. The hypergammaglobulinaemia was due to the simulta- authors stress that more must be learned about the neous increase of IgM globulin and IgG globulins. role of lipoproteins and mucopolysaccharides in the The three forms of leprosy differ significantly in this metabolic process of the leprous infection and respect; in the lepromatous and indeterminate forms called attention to the observation of Hanks & Cray of leprosy, the increase of IgM remains moderate (1954) of the inhibitory effects of those substances while in the tuberculoid form the reaction is more on the metabolism of mycobacteria. heterogeneous and a massive increase of IgM may

3 682 J. 0. de ALMEIDA be observed, similar to that observed in other characteristic distribution which could be related tropical affections. The IgM globulins in the lepro- exclusively to leprosy. matous or in tuberculoid leprosy are immunologically identical. CONCLUSIONS Immunochemical studies on the albumin fraction from serum of lepromatous leprosy patients and From this review of the literature on electro- from normal blood donors were undertaken by phoresis applied to sera of leprosy patients, the Tatarinov (1962). After separation by starch- following conclusions can be drawn: electrophoresis, the albumin fractions were tested (1) There is no diagnostic electrophoretic pattern by double-diffusion in agar against the specific in leprosy. antiserum. The isolated albumin fractions gave the (2) The total proteins may be raised in lepro- same precipitation lines of identity, showing that matous leprosy or they may be within normal their antigenic properties are not changed by the limits. leprous infection. Immune sera prepared by inocu- (3) Reaction forms of leprosy may show a signifi- lating rabbits with normal serum and with lepro- cant increase in a-globulins. In erythema nodosum matous serum showed under immunoelectrophoresis the globulins also increase, with a decrease of (Tatarinov, 1964) that the qualitative composition albumin. of the antigenic components of the serum proteins no essential in cases of (4) Hypergammaglobulinaemia is present in most undergo changes leprosy. cases of lepromatous leprosy, with a corresponding The increase of IgM was also observed by Margni in The et al. (1965) in 20% of the lepromatous patients decrease albumin. albumin: globulin ratio studied; in some, an increase of IgG was observed. is lower than 1. In tuberculoid leprosy serum, the response is (5) In most tuberculoid cases, the proteinogram related to IgA rather than to IgM globulins; this is is normal, the albumin: globulin ratio being higher similar to responses observed in patients with than 1. pulmonary tuberculosis (Lim & Fusaro, 1967b). (6) Lipoproteins from tuberculoid cases of leprosy IgA was most prominent in tuberculoid and least differ from normal in their lower molecular weights. prominent in lepromatous leprosy. On the other Lipoproteins from lepromatous cases have molecu- hand, IgM globulins were present in practically all lar weights lower than those found in normal and the lepromatous patients (86) observed, whereas in tuberculoid leprosy cases. only 62% of 91 tuberculoid cases and 49 % of (7) There is a parallel relationship between the 55 indeterminate patients had this globulin (Lim & relative proportions of glycoproteins and the protein Fusaro, 1967a). A quantitative study of the glo- fractions. However, in nephrotic cases there is a bulin fractions in 216 leprosy patients (92 lepro- decrease in glyco-albumin and a very large increase matous, 83 tuberculoid and 51 indeterminate) showed in the other glycoproteins. a significant increase of IgG, IgM and IgA; IgG (8) The macroglobulin sedimentation diagram in and IgM were very elevated in young lepromatous patients (Lim & Fusaro, 1968a). The levels of IgG, leprosy does not differ from the normal. IgA and IgM in tuberculosis are comparable to (9) The balance ofthe immunoglobulins is affected those in tuberculoid leprosy but are distinguishable by tuberculoid leprosy in which IgM globulins may by the serum pattern of leprosy patients (Lim & be increased in proportion to the level found in Fusaro, 1968b). indeterminate or lepromatous forms. Although some anomalies were observed in the (10) The proteins in leprosy are not altered in immunoglobulins in leprous sera, there was no their antigenic properties.

REACTIONS WITH PREPARED ERYTHROCYTES: HAEMAGGLUTINATION AND CONDITIONED HAEMOLYSIS Two methods can be used for the reactions be- placed in close contact with erythrocytes so that tween leprous antibodies and antigens adsorbed on they may become adsorbed; various procedures erythrocytes. By the first method, antigens are can be used. When the adsorbed antigens react SEROLOGY IN LEPROSY 683 with their specific antibodies, agglutination of the Deceptive results also were obtained by Floch & erythrocytes occurs. In the second method, the Andre (1955) with sheep and group 0 human cells sensitization of the cells by the immune serum is treated with a heat-sterilized leproma suspension. the first step in the haemolysis produced by comple- One weakly positive reaction for haemagglutination ment. These two methods, haemagglutination and was obtained with 1 of 6 sera from leprosy patients. conditioned haemolysis, have been used for the Nor could ultrasonically treated lepromin sensitize detection of reactors. human erythrocytes for the haemagglutination test The surface of erythrocytes can, however, be for leprosy (Floch, 1956). modified chemically with formaldehyde. The cells so prepared are sedimented by a specific globulin Tuberculin-treated erythrocytes present in the leprous sera. The so-called " Rubino Tuberculin was used as the sensitizing agent in test" is based on the rapid sedimentation of a most of the work done on haemagglutination or formolized sheep cell suspension by sera from haemolytic tests for leprosy. When the tuberculin leprosy patients. had some lytic properties, it could be used after repeated precipitation with ethanol (Sohier et al., REACTIONS ON ERYTHROCYTES PREPARED 1950). More reproducible results could be obtained WITH MYCOBACTERIAL ANTIGENS when the human sera were previously absorbed with sheep cells in order to remove the natural anti- Antigens can be adsorbed on the surface of ery- sheep antibodies present in most human sera (Scott & throcytes which become " sensitized " to the corre- Smith, 1950). sponding antibodies. When the reaction between Levine (1925) reported the results obtained with antigen and antibody takes place on the erythrocyte tuberculin-sensitized sheep cells, in haemagglutina- an agglutination can occur, but if complement is tion tests, with sera from 256 lepromatous, 47 tuber- present the cells will haemolyse (conditioned haemo- culoid and 18 indeterminate patients. The controls lysis). The agglutination test with mycobacterial comprised 75 sera from blood donors and 109 sera antigen was described by Middlebrook & Dubos from tuberculosis patients. When the haemaggluti- (1948) and the haemolytic reaction by Muniz (1950) nation titres were computed for each of these with antigens prepared from Trypanosoma cruzi, groups, the accumulative frequency curves showed for the serological diagnosis of Chagas' disease. low titres in the control group and frequently very The haemolytic test for tuberculosis was described high titres among the leprosy patients. The curve by Fisher & Keogh (1950). for tuberculosis lay between the two. The interest raised by these very simple tests, When the frequency curves were drawn for the whose diagnostic potentialities might be applied to three forms of leprosy, it was apparent that the other mycobacterial- infections led the authors to titres tended to be high in the lepromatous cases: try several antigen preparations, with erythrocytes 21.5% were in the range 1: 256-1 : 2048, as com- of different species of animals, in a search for a pared with only 2.1 % of the tuberculoid cases. reliable test for the titration of leprous antibody. The low-titre groups (1: 8 or less) comprised only Antigens from lepromas, lepromin, Myco. lepraemu- 22.2% of the lepromatous as compared with 51.7% rium, tuberculin and extracts of Myco. tuberculosis of the tuberculoid cases and 96 % of the blood-donor were used for the sensitization of erythrocytes from group. Curves for indeterminate leprosy lay be- sheep, oxen, chickens and man. Different technical tween those for lepromatous and tuberculoid leprosy. procedures were proposed, not only for the aggluti- Haemagglutination titres computed in relation to nation but also for the haemolytic test, in order to the activity of the infection showed a quite striking make the tests more suitable for general use. contrast between the bacteriologically positive and negative groups. Thus, only 17.7 % of the positive Erythrocytes sensitized with leproma antigens individuals showed titres of 1: 8 or less, as com- Aono (1953) tried to sensitize bovine erythrocytes pared with 55.4% of those which were negative; with a leproma extract. The reaction failed in furthermore, 23.2 % of the positive group, compared 5 lepromatous and 5 tuberculoid cases. When with only 6% of the negative group, had titres of treated with tuberculin, the cells reacted better to 1: 256 or greater. the agglutination or the haemolytic tests with tuber- Levine's findings were confirmed by Gernez-Rieux culous sera than with sera from leprosy patients. et al. (1952) who used the conditioned haemolytic 684 J. 0. de ALMEIDA test and the haemagglutination reaction with sera ships among the elements of the reaction: comple- from different forms of leprosy. The titres with the ment, tuberculinized cells and immune serum. They haemolytic test were lower than those determined found that immune serum reacted on the surface by the agglutination method. Such results are to be of coated cells as a specific amboceptor, in the same expected, since the anticomplementary activity of way that an immune haemolysin combines with the leprous sera can impair the sensitivity of the reaction erythrocytes. The sensitizing agent is really the of haemolysis but not the agglutination of sensitized immune serum and when the complement is fixed erythrocytes. In this respect, the haemagglutination haemolysis occurs. In order to make the test sensi- test is more convenient than the haemolytic reac- tive enough, large doses of complement should be tion. It was also observed that the agglutination used. The titre of the immune serum is the reciprocal serum titres were lowered by sulfone treatment of the serum dilution that produces 50% haemolysis (Gernez-Rieux et al., 1953; Plissier & Secret, 1953; with 6 units of complement (50% units) (Almeida & Ross, 1954; Gernez-Rieux & Tacquet, 1956; Ker- Adura, 1954). bastard, 1957; Abe, 1963; Cichini, 1952; Madorsky Approximately 40 % of the sera from healthy chil- et al., 1953). dren residing in an area endemic for leprosy gave With sera of patients with erythema nodosum positive results when tested for antituberculopoly- leprosum, maximum titres usually occurred a few saccharides, by the haemagglutination technique but days after the temperature had fallen. In atypical the titres approached those detected in the normal cases, the maximum titre was detected between the population (Rheins et al., 1957). Haemagglutination seventh and the fourteenth day, but the titre curve and conditioned haemolysis have been applied to the was irregular (Kawaguchi, 1953). When sera of study of humoral antibodies in 200 sera from leprosy 24 leprosy patients were tested at intervals for a patients by Sindo et al. (1959). The titres in the year, it was found that the titres could be maintained haemolytic tests were higher than those found by at almost the same level during the period in which agglutination. The average serum titre for the haemo- no remarkable change occurred in their clinical lytic test was 1/85.8 and for the haemagglutination course. In some cases of lepromatous (" nodular ") 1/33.2. In the lepromatous group, the titres were leprosy, the titres showed fluctuations from time to significantly higher than in the tuberculoid group. time, although no noticeable changes were observed When the Middlebrook-Dubos test was applied in the clinical symptoms of the patients. to sera from leprosy patients, most of the conclu- Tuberculin prepared from BCG cultures was found sions indicated its usefulness, not only for the detec- suitable for the sensitization of Rh-, 0 human tion of early cases of leprosy (Roy & Banerjee, 1955; blood cells, when compared with tuberculin from Roy, 1963; Lucentini & Nazzaro, 1954), but also other sources in haemagglutination tests with sera in the study of relationships among the antigens of from leprosy patients (Andrade & Silva, 1952). Myco. leprae and other mycobacteria, including Maillard & Gagliardo (1951) described a method Myco. tuberculosis (Montestruc, 1953; Floch & for the haemolytic test with standardized reagents. Sohier, 1952). The titre of the serum was described as the highest Reaction with Myco. lepraemurium dilution in which there is a constant degree of Bovine could be sensitized with an with a constant amount of com- erythrocytes haemolysis (50%) aqueous extract of Myco. lepraemurium treated with plement (4 units). This test was applied by Maillard solvents and broken-down mechanically. These to sera from 24 from whom lipid (1952) leprosy patients sensitized cells were used in and haemo- to Reaction titres agglutination tuberculosis had be excluded. lytic tests and the results were closely interrelated were obtained in 23 of these 24 sera. This technique and coincident with those obtained with cells sen- was also employed by Ross (1954a, 1954b), who the 10 in sitized with Old Tuberculin. With leprous sera, observed significant titres higher than 44 out test with cells treated with the murine antigen of 100 sera studied and 10 tuber- the (90 lepromatous appeared more specific and sensitive (Yamada, 1958). culoid patients) and in 33 out of 56 active leproma- tous patients. The 50 % end-point technique was applied by REACTIONS ON ERYTHROCYTES TREATED Almeida & Azevedo (1952) to the conditioned WITH FORMALDEHYDE haemolysis tests with sheep cells coated with tuber- The rapid sedimentation of formolized sheep cells culin, in order to establish the quantitative relation- was observed by Rubino (1926a) in serum from a SEROLOGY IN LEPROSY 685 leprosy patient. This finding was confirmed by tive evaluation of the reaction. The results of the further investigations (Rubino, 1926b, 1929, 1931a, tests were reported in accordance with the criteria 1931b, 1931c, 1934) which showed that sheep cells proposed by Bier & Arnold (1935) as strongly posi- were better than human cells for the reaction. tive, positive, weakly positive and negative. Various adjustments were studied in relation to the Curban (1962) found a positive association be- reproducibility and sensitivity of the new test for tween the Rubino reaction and the lepromatous and leprosy, since it was observed that, although very dimorphous forms ofleprosy, and a negative associa- specific, the reaction could be negative in a large tion between the test and tuberculoid and indeter- proportion of sera from recognized cases of leprosy. minate leprosy. Statistical analysis of the results The substance responsible for the Rubino reaction obtained indicates a significant association between adheres to the formolized cells. The cells, once the Rubino reaction and the clinical condition of the sedimented by the leprous serum, could be washed patient, including the form of leprosy. The com- and resuspended in normal human sera, and again parison of the results obtained during a period of they sedimented very rapidly (Marchoux & Caro, 3 years in the same group of patients suffers from 1928). The factor responsible for the Rubino reac- the lack of a standard antigen preparation, making tion was not destroyed by heating at 56°C for it impossible to decide if different results are due 30 minutes. However, the reaction does not take to the specific Rubino factor in the serum or to place at this temperature, the optimum being 37°C different behaviour of the reagents. In one and the (Rubino, 1929). same leprosy patient, a negative Rubino reaction The agglutination of formolized sheep cells cannot exclude entirely a positive reaction in the past observed by immunohaemolysis produced in rab- or future (Curban, op. cit.). bits by the inoculation of treated cells was not of There is a consensus of opinion regarding the the same type as that caused by leprous sera absolute specificity of the Rubino reaction for (Rubino, 1931b). The technique of the reaction was leprosy. No Rubino reaction had until 1962 been improved by the previous absorption of natural detected in relation to conditions other than leprosy, heterophilic antibodies by the sheep cells (Rubino, as was shown by Curban (1962) in sera from active 1931c). pulmonary tuberculosis (12 cases), from South The mechanism of the Rubino reaction is not American blastomycosis (12 cases), from tegumen- well established. Ambrogio (1932) refers to an tary leishmaniasis (6 cases) and from syphilis alteration of the electrostatic forces on the surface (10 cases). In the experience of Curban (op. cit.), of formolized cells. On the other hand, the reaction 150 Rubino-negative tests were observed in 55 lep- occurred through the increase of globulins (Bene- romatous, 36 indeterminate, 55 tuberculoid and tazzo, 1933). Bier (1936) called the element respon- 4 dimorphous cases of leprosy. Positive Rubino sible for the Rubino reaction the "enigmatic reactions were detected only in the lepromatous (78) factor ". According to Castro et al. (1963), a haemo- and dimorphous (3) forms of leprosy. Among the globulin is involved in the reaction, together with lepromatous patients, those with nodules showed some particular electrostatic forces acting on the the highest incidence of positive Rubino reactions amino groups in the outer layer of erythrocyte with 59 out of 87 positive (Curban, 1962). membranes. The " Rubino factor" can be absorbed The failure of the Rubino reaction could be related from leprous sera by formolized sheep cells. The to the past experience with the lepra reaction in absorbed sera, however, maintains its reactivity with lepromatous patients, according to Mercau et al. tubercle (WKK) antigen, showing the existence of (1963) who studied 70 leprosy patients and 30 normal two independent globulins; one involved in the individuals. In their hands, the reaction was spe- Rubino reaction and the other the complement- cific, but had shown only a 30% sensitivity. fixing antibody (Bier & Arnold, 1935, Bier & Planet, 1936; Almeida, 1961). The factor responsible for the CONCLUSIONS Rubino reaction is associated with the y-globulin, as was shown by electrophoresis. From the review of the main aspects of reactions Work in leprosy with the Rubino test was reviewed which take place on the surface of erythrocytes, by Curban (1962), who studied sera from 231 leprosy some conclusions can be drawn: patients. In his studies, the original method of (1) The Rubino reaction has been shown to be Rubino (1931a) was used, since it gives a quantita- highly specific for leprosy. It occurs more frequently 686 J. 0. de ALMEIDA with sera of lepromatous leprosy patients although a titres were found among lepromatous than among number of such patients may be Rubino-negative. tuberculoid patients. (2) The factor responsible for the Rubino reaction (5) These serological tests are recommended for is an immunoglobulin unrelated to the complement- the investigation of humoral antibodies present in fixing antibodies detected by tubercle antigen leprous patients, and they can be used with anti- preparations. complementary sera which cannot be tested by (3) Tuberculin-prepared erythrocytes of various complement-fixation. species of animals react with leprous as well as with (6) The fractions isolated from lepromas or from tuberculous sera in the haemagglutination and the mycobacteria, once absorbed by erythrocytes-with conditioned haemolysis tests. or without previous treatment with tannic acid, (4) The average serum titre in leprosy is higher could detect the corresponding antibodies in leprous than in tuberculosis. In patients with leprosy, higher sera.

OTHER SEROLOGICAL TESTS

AUTOIMMUNE ANTIBODIES test was noted, but there appeared to be some correlation between the results of the latex-fixation The hypothesis that an autoimmune mechanism test and the clinical course of the patients. From can be present in leprosy, at least during reactions, the results of the study it was not possible to define has received the attention of many workers (Chini, the significance of rheumatoid factors found in 48 1961; Cathcart et al., 1961; Rizzi et al., 1962; out of the 101 leprosy patients, although a non- Bonomo et al., 1963; Matthews & Trautman, 1965; specific role with respect to the clinical course of Juravleva, 1965; Barua et al., 1963). Bonomo et al. the disease could be admitted. (1967) called attention to the serological overlap Matthews & Trautman (1965) pointed out the between leprosy and diseases with autoimmune clinical similarity of the reactive phenomena in phenomena, in particular those related to colla- leprosy and in collagen diseases; the similarity is genosis. A number of autoimmune-like factors has supported by the positive result of the rheumatoid been found in leprosy patients, including rheumatoid latex test and the presence of thyroglobulin anti- factors (Cathcart et al., 1961), thyroglobulin bodies, lupus cells and serum cryoproteins. Of 39 lep- antibodies (Bonomo et al., 1963), cold-precipitated rosy patients, 38 had demonstrable cryoproteins in proteins (Matthews & Trautman, 1965) and the their serum. Only 14 patients in a group of 26 who histamine-like myocontraction agent (Castro et al., were receiving corticosteroid were positive for cryo- 1963). Some other findings, such as the histamine protein. Circulating thyroglobulin was found in 25 level in the blood (Gokhale, 1958), the low level of of 65 lepromatous patients, and the rheumatoid circulating complement during the lepra reaction factor in 38 of them. Somewhat lower percentages (Azevedo & Mello, 1966) and the high prevalence of were obtained in 10 patients with dimorphous false positive reactions for syphilis in lepromatous leprosy, and all results were negative in 10 patients leprosy patients are also compatible with a state of with inactive leprosy and in 20 healthy control autoimmunization. patients. The authors believe that leprosy should Cathcart et al. (1961) undertook a study to deter- be considered as a collagen disease. mine whether or not the presence of a rheumatoid A correlation was noted between the presence of factor in the serum of patients with leprosy could thyroglobulin antibodies and positive rheumatoid- be correlated with any difference in the clinical like reactions in the serum by Bonomo et al. (1963) aspects of the disease. Latex-fixation and agglutina- in 50 cases of lepromatous leprosy. Thyroglobulin tion of sensitized sheep cells were performed on antibodies were found in 21 of them, and in 19 out 101 sera from leprosy patients, selected to represent of 41 sera tested for agglutination with tanned all stages of the disease. No relationship between erythrocytes. These authors consider that the hyper- the duration of disease and a positive latex-fixation sensitivity of the antibody-forming system may be SEROLOGY IN LEPROSY 687 responsible for positive autoimmune reactions in C-REACTIVE PROTEIN leprosy. The presence of C-reactive protein in human An investigation of lupus erythematous cells and serum indicates an inflammatory reaction, whether antinuclear factors was carried out by Bonomo bacterial or non-bacterial in origin. C-protein is et al. (1965) on sera of 55 unselected leprosy patients; never found in normal human serum. Much infor- 4 out of 10 leprosy patients whose sera were found has been accumulated about the presence to contain antinuclear factors showed also lupus mation their the of this protein in various inflammatory conditions, erythematous cells. Discussing results, but the first study relating to leprosy was published authors made some comments on the exaggerated only in 1955. Rabson (1955) studied the sera of sensitivity of the antibody-forming system which 176 leprosy patients, finding C-reactive protein in may be sustained by antigenic stimulation due to chronic infection by Myco. keprae. The adjuvant- 37 of the 79 active lepromatous cases, in 16 of the like effect of the leprous contributes to the 39 arrested lepromatous cases, and in only 7 out infection of 58 tuberculoid cases. These findings were con- establishment of hypersensitivity and to the develop- of This of view firmed by other authors who found a higher fre- ment autoimmune disease. point quency of C-reactive protein in lepromatous cases is supported by the demonstration of a higher pro- of of to in leprosy (Terencio de las Aguas, 1958; Bush, 1958; duction antibodies typhoid vaccines lepro- Ross et the matous leprosy patients as compared with the nor- al., 1959). During lepra reaction, the mal healthy population (Almeida et al., 1964), the level of C-reactive protein increases and its appear- of in ance in the sera indicates an exacerbation of the finding antinuclear factor rabbits hyperimmu- Lan- nized with killed bacteria et leprous infection (Montestruc, 1959; 1960; (Christian al., 1963) et Montestruc and by production of a rheumatoid-like factor as guillon al., 1960; Kabakov, 1961; a result of prolonged antigenic stimulus (Abruzzo & et al., 1962). & The C-reactive test was applied to sera from Christian, 1961; Williams Kunkel, 1962). 291 leprosy patients of whom 250 were lepromatous It is also worth mentioning that incomplete auto- and 41 tuberculoid. The reaction was positive in 72 anti-erythrocyte antibodies are found in a high (85 %) of 84 lepromatous cases with erythema proportion of leprosy patients, as demonstrated by nodosum. In all 27 cases with leprosy reactions, the direct Coombs test. Among 69 patients, Rizzi C-reactive protein was present. In 135 non-reactive et al. (1962) found 23 whose sera contained anti- lepromatous cases the test was positive in 2 while erythrocyte incomplete antibodies. among the 41 tuberculoid cases, only 1 serum showed These observations on the serology of leprosy are the presence of C-reactive protein. compatible with a large body of knowledge regarding C-reactive protein was absent from the sera of the immunological features of the leprosy reaction erythema nodosum patients whose symptoms were as an episode of hypersensitivity. At the same time, relatively slight but was always present in sera from the frequency of renal affections as a cause of death severe cases. Repeated tests for C-reactive protein in lepromatous patients who present a lepra reac- in erythema nodosum patients showed that this tion, reveals the coincidence of a clinical condition protein appeared in the serum within a few days of which is usually produced by autoimmune aggres- the onset of the erythema and that it disappeared sion. Data on the causes of death of 498 leprosy at the arrested stage (Abe & Hirano, 1961). patients and the influence of the lepra reaction on Wade (1955) stressed the importance of C-reac- renal disease were presented by Brusco & Masanti tive protein tests in cases showing a lepra reaction, (1963) who found a statistically significant difference and Ross (1964a, 1964b) suggested that the test not only between lepromatous and non-lepromatous should be made as a routine measure because of its patients but also between lepromatous patients, simplicity and because of the value of the informa- with and without lepra reactions. tion it provides. The reports in the literature on the serological aspects of leprosy as an immune disease show that there are some common features proper to the FLUORESCENT ANTIBODY METHOD IN LEPROSY presence of auto-antibodies. Further investigations A new approach to the detection of specific anti- are required to elucidate the immunopathology of bodies for leprosy making use of the fluorescent leprosy, mainly in relation to the clinical aspects of antibody technique was described by Morris et al. the disease. (1961). The indirect method was used to demon- 688 J. 0. de ALMEIDA strate the presence of antibodies to Myco. leprae tive fluorescent reactions were obtained with lepro- in the sera of leprosy patients. The attachment of matous sera from untreated cases in dilutions as human y-globulin with antibodies specific to Myco. high as 1/1024, whereas titres of 1/256 and 1/128 leprae in smears prepared from infected human were found in the treated cases which were bacterio- spleen was demonstrated by the addition of fluores- logically negative. In the tuberculoid cases, titres cent antihuman y-globulin sera. Morris et al. were around 1/512. The authors believed that (op. cit.) found that fluorescence could be shown titres as low as 1/8 might eliminate the diagnosis of in most lepromatous sera. Leprous sera reacted leprosy and that a limit titre up to 1/128 could also with Myco. tuberculosis but only rarely with occur without any certainty that there was infection. Binford's bacilli. Lepromatous sera absorbed with Dilution of the sera eliminates the cross-reactivity Myco. tuberculosis lost their capability to react with observed with other mycobacteria such as Myco. this organism but still reacted with Myco. leprae. lepraemurium and Myco. tuberculosis and BCG However, sera from tuberculosis patients reacted (Merklen & Cottenot, 1966). after absorption only, and partially, with smears of Myco. tuberculosis but not of Myco. leprae. With sera previously absorbed with Myco. tuberculosis, THE AGAR-GEL DOUBLE-DIFFUSION METHOD positive fluorescent tests were obtained in 19 of The cross-reactivity between leprous sera and 22 sera from lepromatous leprosy and in 5 of 17 sera antigens present in tuberculous preparations was from tuberculoid leprosy patients. In 38 control fully demonstrated by Burrell & Rheins (1957) by sera, 2 were positive and both of them were from the agar-diffusion method. Two samples of pooled individuals with histories of prolonged contact with sera from adult patients with lepromatous and leprosy patients (Morris et al., 1961). Attention tuberculoid leprosy and sera from tuberculin- was drawn to this method for the demonstration of negative children (bled before a lepromin test was specific antibodies in leprous sera by Reich (1961). made) were tested with two antigens, one being In their studies on the fluorescent antibody stain- lepromin of the Hayashi-Mitsuda type, the other, ing properties of mycobacteria, Shepard & Kirsh Old Tuberculin.' The results, obtained in plates by (1961) observed that when the organisms were the Ouchterlony method, showed that both antigens ruptured antigens were released into solution and possessed antigens in common, as well as distinctive could easily be separated from the sediment by components. Sera from children living in endemic centrifugation. Intact bacilli of the slowly growing leprosy areas reacted with lepromin with a specificity species did not stain well with fluorescent antibody, not observed with normal and tuberculous sera although the disrupted bacilli stained brightly. The obtained from places where leprosy does not occur. soluble fraction contained the greater amount of Precipitation lines were observed by Pepys et al. antigen, as judged by the ability of the solution to (1959) in reactions with an antituberculous rabbit inhibit staining of the disrupted bacilli, by the serum and antigens prepared from murine leprosy precipitation lines in agar, by diffusion tests and by and with Mitsuda tuberculin. According to Parlett its antigenicity in complement-fixation reactions. et al. (1958), the antigens prepared from Myco. Leprous antibodies could be detected by the tuberculosis gave precipitating bands in double- immunofluorescent method using preparations from diffusion with sera from tuberculous patients. The murine leprosy, as was observed with sera from 36 reaction was specific since sera from 13 patients leprosy patients (all positive) and 25 healthy persons with leprosy did not react with these antigenic (all negative). In the indeterminate and tuberculoid preparations. forms of leprosy the rate of positive fluorescent The presence of precipitating antibodies in the sera reactions is lower than that found in lepromatous of leprosy patients was demonstrated in reactions cases (Cottenot, 1964). Tuberculous sera, when with antigens prepared as filtrates of Youman's absorbed by BCG, do not show any fluorescence culture media of Myco. tuberculosis, strain H37Rv in the indirect test. The antibacillary antibodies and bovine strains (BCG 263), two atypical acid-fast detected by immunofluorescence indicate a close bacilli isolated from tuberculous sputum, and chro- relationship between Myco. leprae, Myco. lepraemu- mogenic mycobacteria. Of 86 leprous sera, 54 were rium and Myco. tuberculosis (Merklen et al., 1963). positive, 13 being positive only with H37Rv, 5 only The use of preparations with murine leprosy was recommended by Merklen & Cottenot (1965). Posi- 1 Prepared by Lederle Laboratories. SEROLOGY IN LEPROSY 689 with bovine antigen and 2 only with atypical bacilli. sera containing antibodies reacted strongly with the The other 34 sera reacted with all the antigens tested. Myco. tuberculosis culture filtrate and less so with Sera from healthy individuals did not show any Myco. keprae. All lepromatous sera contained cir- precipitation line (with the exception of one, pre- culating antibodies which reacted with tuberculous sumably non-tuberculous, patient). In a further antigen whereas the tuberculoid sera gave negative experiment, the authors observed that some sera results. One finding of interest was that 2 of the from leprosy patients and healthy individuals reacted 25 lepromatous cases possessed, in addition to with fresh-milk antigen, and that lepromatous sera precipitating antibodies, a circulating mycobacterial reacted also with PPD prepared from a human antigen which was detected by precipitation with tubercle strain (Sushida & Hirano, 1961). an antituberculous rabbit immune serum. A serum prepared in rabbits inoculated with Myco. Navalkar et al. (1965) demonstrated, by means of leprae extracted from leproma was found to react an analysis employing double-diffusion in agar-gel, with polysaccharide antigens prepared from myco- humoral antibodies against various mycobacterial bacteria (Tuma & Silva, 1962). The intensity of the antigens in 49 out of 60 sera from leprosy patients. reaction of the precipitation bands was proportional Two types of antibodies, anti-fl and anti-8, were to the polysaccharide content of the mycobacterial identified. The ,B-antigen was demonstrated in extracts. Antigens were prepared from Myco. tuber- many mycobacterial species but the 8-factor in culosis strains H37Rv and BCG, and 8 non-patho- only a few. Altogether, 12 sera formed precipitation genic mycobacteria. The leproma extract gave also bands when "Dharmendra lepromin" was used. a strong positive reaction in agar-diffusion. The The fl-precipitin is found mostly in sera from lepro- results showed a close interrelationship between the matous (13/15) and the reactional tuberculoid (8/17) carbohydrate content of the antigens and the inten- types, but in only very few sera from the tuberculoid sity of the antigen-antibody reactions and also the (2/28) type of leprosy. On the contrary, the 8-anti- presence of antigens common to Myco. keprae. body was demonstrated in the majority of sera from A group-specific polysaccharide isolated from tis- all three types (49/60). Anti-8 precipitins were sues infected with Myco. leprae gave precipitation detected in sera from bacteriologically positive cases bands in agar gel with tuberculous and leprous sera of lepromatous (15/15) and reactional tuberculoid from lepromatous cases. The leproma extract con- (13/17) leprosy and also in sera from tuberculoid- tained polysaccharides but no proteins and gave a negative cases (21/28). No precipitating antibodies precipitation band in agar, immunologically identi- were demonstrated in the sera from 30 blood cal with that produced by the group-specific antigen donors. from Nocardia brasiliensis. A common antigen was Of the various species studied, Myco. kansasii and demonstrated in Myco. lepraemurium, Myco. leprae Myco. smegmatis were found to be particularly suit- and N. brasiliensis, in reactions between the micro- able for the preparation of precipitinogenic anti- bial extracts and sera from lepromatous leprosy and gens. The authors postulated that in addition to the sera from mice inoculated with Myco. lepraemurium fl-factor, the 8-factor is a part of the antigenic mosaic (Salazar & Calderon, 1967). The leproma extract of Myco. leprae and that in leprosy there is more also reacted with sera from rabbits immunized common antibody response to this factor than to against N. brasiliensis, with serum from a patient the fl-factor. The diagnostic potentialities of this with active tuberculosis and with sera from two serological test are implied by the finding of anti-S lepromatous patients. However, extra bands could precipitins in about 80% of the leprosy cases be seen with lepromatous sera, indicating species- investigated. specific antigens since similar precipitation lines The results so far obtained by the use of the were not observed with tuberculosis sera nor with fluorescent antibody technique and of double-diffu- anti-Nocardia immune rabbit serum. Other antigens sion in agar-gel showed that these immunological were present in the crude leproma extract, as indi- methods can give information on the presence of cated by the presence of 5 precipitation bands humoral antibodies in leprosy and also provide a (Estrada-Parra et al., 1966). tool for the antigenic analysis of mycobacterial Rees et al. (1963) applied the Ouchterlony method extracts. Leproma extracts contain antigens which with sera from 25 lepromatous and 25 tuberculoid could be extracted for use in precipitin determina- leprosy patients, with antigens from mycobacteria tions. Further investigation along these lines is including Myco. tuberculosis and Myco. leprae. Most suggested. 690 J. 0. de ALMEIDA

SEROLOGICAL REACTIONS FOR SYPHILIS IN LEPROSY

The first observation of a positive complement- et al., 1954). Syphilitic sera, however, did not suffer fixation reaction for syphilis in leprosy was made from this treatment. When several reactions were by Eitner (1908) who examined the serum of one carried out in the same sera, it was observed that of his leprosy patients, employing antigen from the Meinicke test gave fewer false positive reactions guinea-pig heart (i.e., Wassermann antigen prepared (Kvittingen et al., 1952, Ruge, 1955). Although it according to Landsteiner). This observation of would not be justifiable to say that the Meinicke cross-reactivity between the Wassermann and the test never produces false positive reactions in leprosy Eitner tests, was confirmed by Photinos & Michae- patients, the clinical data indicate that such reac- lides (1912) who obtained the same results when tions are comparatively rare, and that positive they used the leproma preparation, or the original reactions occur mainly in cases showing evidence of Wassermann antigen, with 204 sera from leprosy syphilis (Kvittingen et al., 1952). patients. The main differences between Meinicke antigen Cooke (1919) reviewed the literature on the sero- and those used for complement-fixation or floccula- logy of leprosy up to 1916 and found that of 1397 tion, is that the Meinicke antigen contains Tolu cases described 50.4 % reacted with the Wassermann balsam instead of cholesterol; it was thought that test, indicating clearly that false positive reactions the increase in blood cholesterol in leprosy patients for syphilis occur frequently with leprous sera. might account for false positive reactions when On the basis of a comparison of results obtained antigens containing cholesterol are used (Sagher, with sera from leprosy patients and from healthy 1953; Gollerkeri et al., 1952). persons, it could be assumed that positive Wasser- In two surveys for syphilis made in 1934 and mann tests may occur in leprosy even though 1952 in the present Republic of Viet-Nam, Montel syphilis is absent (Hasseltine, 1924). (1952) found in 1000 leprosy patients that the Leprous sera not only fix complement with Wasser- proportion of positive results was identical with that mann antigen, but also give positive reactions with in non-leprous populations, although he pointed lipoid antigens used for flocculation or precipitation out that sera of advanced lepromatous leprosy have tests for syphilis (Badger, 1931). In his own experi- a polyfixing property for many antigens. False ence, Badger (op. cit.) found an incidence of 20.2% positive reactions for syphilis are noted less often for the Wassermann test and 27.5% for the Kahn now than they were before sulfones came into use test among 207 leprosy patients. However, when the (Mertslin & Loginov, 1963). These authors com- euglobulins were precipitated with hydrochloric acid pared the results of the Wassermann test with those (Sachs method) the false positive reaction in leprous of Sachs-Witebsky reactions in 202 leprosy patients; sera became weaker or disappeared; the opposite all but 1 of the 11 reactions they obtained were occurred with syphilitic sera (Boncinelli, 1937). specific. In leprosy, degradation products from Myco. The evaluation of serological procedures in terms Ieprae and infected tissues may act as complete of human specimens is commonly plagued by uncer- antigens, resulting in circulating antibodies (specific tainties regarding the true status of the patients, as and heterophile). Using a lipid extract prepared can be seen when false positive reactions must be from lepromas for a micro-flocculating test, Castro distinguished from genuine syphilitic reactions. & Bonatti (1951) examined 162 lepromatous sera, Applying the Kolmer complement-fixation test, and resulting in 73.5% of positive tests; with 62 sera the precipitation test of Kline and the VDRL test from tuberculoid leprosy, 4.8 % reacted whereas to 126 healthy Sudanese, Lauret & Kerbastard only 0.2% of 1108 healthy persons reacted to the (1955) found 79 negative, 4 dissociated and 43 test. The reactivity with syphilitic sera (140) was (34%) positive reactions. In 239 leprosy patients, very low (1.8%) but among leprosy contacts, 3.7% negative reactions were obtained in 163, 3 were reacted. The antibody responsible for this test was dissociated and 73 (31 %) were positive. All these absorbed by formolized sheep cells. patients had been receiving sulfone treatment for When leprous sera were heated for 5 minutes at a long time. In 142 untreated cases (56 lepromatous, 65°C, the reactions with lipoid antigens were less 72 tuberculoid and 14 indeterminate), the authors intense or negative (Narula & Gupta, 1953; Gupta found 77 negative, 4 dissociated and 61 (42%) SEROLOGY IN LEPROSY 691 positive reactions. Of the 30 patients with positive CARDIOLIPIN WITH CEPHALIN AND CHOLESTEROL serology, 6 gave histories of syphilitic chancres and 13 of yaws, in 11 no definite history of trepone- Complement-fixation reactions using an antigen of matosis was found. In 40 cases out of 54 with cardiolipin, cephalin and cholesterol, produced a positive serology for syphilis, a drop was seen in positive reaction with 57 % of leprous sera but serological titres after penicillin treatment. From with only 10% of syphilitic sera (Honda & Yoshino, their experience, these authors believe that the 1951). A comparison of this method with that of classical notion of false syphilitic reactions in the Ogata favoured the antigen with cephalin. In 62 lep- sera of leprosy patients should be abandoned. romatous cases, the complement-fixation test was The reactivity of the antigens is related to their positive in 55 and the agglutination test in 44; content in lecithin, as shown by Kent et al. (1957) in 29 neural cases, the former was positive in 16 with sera of 34 autochthonous Cubans with leprosy; and the latter in 9; in 5 macular cases the comple- none presented clinical or anamnestic evidence of ment-fixation test was positive in 3 whereas aggluti- treponematosis and all were negative to Treponema nation was not observed (Honda et al., 1956). pallidum immobilization (TPI) tests. However, with With an antigen prepared with cardiolipin (1), lipoidal antigens used in VDRL, Kline, Rein- cholesterol (5) and cephalin (10), Tanimura et al. Bossak and Mazzini tests, the frequency of positive (1956) obtained 87.3% positive reactions in 110 reactions varied from 6 to 21 and from 7 to 17 in leprous sera, which had been submitted to comple- tests with cardiolipin. The specificity of cardiolipin ment-fixation. Only 3 out of 38 syphilitic cases was correlated with the content of lecithin (Kent were positive and none of the non-leprosy sera, et al., 1957). of which 19 were from tuberculosis cases, 6 from cancer patients and 40 from healthy persons. How- ever, this antigen was not found to be of diagnostic THE CARDIOLIPIN-LECITHIN ANTIGENS value in leprosy, since reactions were obtained with FOR LEPROSY sera from syphilis, kala-azar and tuberculosis (Ghosh et al., 1964). Antigens prepared from lecithin and cephalin Ogata & Abe (1958) presented a new serological of and extracted from human and murine lepromas did test based on the proportions cardiolipin any improvement to the specificity and lecithin. It was found that the solutions containing not make of and when absorbed sensitivity of reactions complement-fixation, as far equal parts cardiolipin lecithin, as the testing of human leprous sera is concerned with kaolin particles, give the highest agglutinating 1949). end-point titres with leprosy sera, in strong contrast (Yoshinaga, to syphilis sera, which give the highest titre with 1 part of cardiolipin and 10 parts of lecithin. When CEPHALIN-CHOLESTEROL FLOCCULATION TEST leprosy and syphilis are present in the same patient, the agglutination end-titre curve presents a different A preparation of cephalin and cholesterol without shape. The antigen of Ogata & Abe (1958) can cardiolipin was used by Ross (1951) with sera from show some weak reactions with normal human sera. 425 cases of leprosy (398 of the lepromatous and These non-specific reactions could be completely 27 of the tuberculoid type). Serum was diluted eliminated by the use of lysolecithin in the antigen for reaction with the cephalin-cholesterol emulsion. preparation. It was found that phosphatidyletha- Blood specimens of 30 healthy persons were simi- nolamine intensifies the agglutination of cardiolipin- larly tested. The patterns obtained were suggestive lecithin antigen with leprous sera (Ogata & Hara, of liver dysfunction. The use of the test at intervals 1960). The Ogata (1953) reaction permits the detec- may be useful as a guide to therapy and prognosis tion of syphilitic sera with " syphilitic antigen " in the treatment of leprosy, or of concomitant but it seems that the same reaction with " leprous hepatic dysfunction (Ross, op. cit.). antigen " cannot be regarded as specific for leprosy (Floch & Andre, 1956). CARDIOLIPIN-CHOLESTEROL ANTIGEN The addition of cholesterol to the Ogata antigen, in varying proportions and using the slide-aggluti- A mixture of cardiolipin and cholesterol was nation technique, did not improve,the specificity tested with leprous sera in complement-fixation tests, and sensitivity of the reaction. in comparison with a cardiolipin-lecithin-cholesteroI 692 J. 0. de ALMEIDA antigen. The lecithin-free antigen, called " cardchol " (Collier & Mesquita, 1958), since mass serological (Schmidt, 1959), showed an especially pronounced examination of the non-leprous population revealed reactivity in sera from false positive reactors. Sera a small proportion (1 %-2%) of positive syphilis from leprosy patients were found to be highly tests and so the high percentage in leprosy patients reactive with cardchol but not reactive with the cannot be explained by a concomitant infection. complete cardiolipin antigen. In a study of 300 sera These authors, analysing the results obtained in from leprous patients, Schmidt (1961) confirmed the lepromatous and in tuberculoid leprosy, concluded high reactivity of cardchol for these sera, as com- that a positive syphilis test and a positive leprosy pared with syphilitic, non-leprous sera. The anti- complement-fixation test are independent of each lipoidal antibodies in leprosy were probably different other. In the active form of leprosy, ubiquitous from those found in treponematoses. Among syphi- lipids are released and antibodies form against litic leprosy patients, the percentage reactivity was them, giving positive Wassermann, Kahn and significantly higher with cardchol than with cardio- Meinicke reactions. The increase in titre of anti- lipin-lecithin-cholesterol antigen. Although card- bodies reacting in the leprosy complement-fixation chol is by no means a specific antigen for the demon- test runs parallel with the increase in titre of anti- stration of leprosy, it has a fixing capacity for anti- bodies reacting with the syphilis antigen. Both kinds bodies of leprous sera, which do not react with the of antibodies are completely independent of each ordinary cardiolipin antigen (Schmidt, op. cit.). other, but those reacting to syphilis antigen depend Reactions of cardchol with antigens prepared from on the activity of the pathological process (Collier & Myco. tuberculosis were found to be related to the Mesquita, 1958). concentration of antibodies demonstrated by com- These conclusions could not be confirmed by plement-fixation. Positive reactions with cardchol Almeida (1964) who found in 509 leprous sera were found more frequently among high-titre sera, only 3.3 % of complement-fixation reactors with showing a strong correlation between the degree of cardiolipin and in 389 sera, 5.9% whose titres for reactivity with tubercle antigen and the incidence leprosy were higher than 2000. The former group of complement-fixation reactions with lecithin-free comprised sera with titres of less than 20. Although antigen (Almeida, 1962). it is possible that in the high-titre sera some cross- The antibodies found in leprous sera retained reactivity may occur, the incidence of these false their reactivity with cardchol after prolonged heat- positive reactors was less than 5 % in a total of ing at 560C. Altogether, 24 sera from cases of lep- 2324 lepromatous sera (Almeida, 1964). In a lep- romatous leprosy with negative TPI tests were romatous group of 234 cases, Collier & Mesquita tested and 16 reacted with cardchol antigen and (1957) found 41 (17.5%) of Wassermann-positive 3 with the cardiolipin complete preparation after sera; in 112 tuberculoid cases, 11 were Wassermann being heated for 30 minutes. Prolonged heating reactors (10%). These differences may be due to resulted in complete loss of reactivity for cardiolipin- the method employed in the quantitative comple- lecithin-cholesterol antigen, but 11 sera still showed ment-fixation test for syphilis. reactivity for cardchol. This heat-stability is a property of the immune globulins themselves (Holst, 1964). THE KAHN UNIVERSAL SEROLOGICAL REACTION FOR LEPROSY

REACTING ANTIBODIES IN LEPROSY The precipitation test described by Kahn (1951) as a " universal reaction " is based on the effect of When the results of Wassermann, Kahn and different concentrations of sodium chloride on the Meinicke tests were compared with the titres ob- mixture of antigen and serum. Reactions of different tained by quantitative complement-fixation of tuber- types are differentiated by a characteristic pattern, cle antigen with sera from leprosy patients, it was according to the degree and arrangement of precipi- observed that sera from lepromatous cases tended tation observed in three reacting zones of the uni- to give more positive syphilis reactions than those versal serological technique. A distinctive pattern from tuberculoid cases. Also, active lepromatous was described for lepromatous, but not for tuber- cases gave more false positive reactions than quies- culoid, leprosy. Pinto & Zeo (1951), however, could cent or arrested cases. In this study, the possibility not find these patterns in leprosy. The test has of an existing syphilitic infection could be excluded been found to be of doubtful practical value (Floch SEROLOGY IN LEPROSY 693

& Sohier, 1953) and the results obtained from and the New York State Department of Health 110 lepromatous cases, 20 tuberculoid cases and complement-fixation tests) were completely non- 20 healthy controls showed that there is no distinc- reactive to all specimens. The reactivity rates among tive serological pattern in lepromatous leprosy (Ross other tests vary from 6.9% for the Army comple- & Gemar, 1953). However, Romanci (1952) reported ment-fixation and the New York State Department variations in precipitation curves during a period of of Health microprecipitation tests to 72.4% for the 1 year, which may be attributed to the clinical Rein-Bossak test (United States Public Health Ser- course of the disease; later, Olitzki & Sagher (1956) vice, 1954). In the experience of Zarco & Chan observed the results of the universal Kahn reaction (1958) who examined 108 sera from leprosy patients in 27 leprosy patients under sulfone treatment for employing the New York State Department ofHealth a period of 3 years and concluded that changes in complement-fixation method and other standard the curves occurred with clinical improvement. serological tests, the results obtained by the former were in close agreement with those of the TPI test, thus showing its high specificity for syphilis; this THE NEW YORK STATE DEPARTMENT was in contrast to the high proportion of false OF HEALTH QUANTITATIVE COMPLEMENT-FIXATION positive reactions obtained with the Kahn and TEST WITH CARDIOLIPIN VDRL tests. The quantitative complement-fixation method for The results obtained so far indicate that the syphilis, as developed by Wadsworth, Maltaner & "quantitative technique" provides a reliable crite- Maltaner (1938a, 1938b) based on the linear relation- rion of syphilitic infection, even in leprosy patients. ship between serum and the complement required While it seems unnecessary to point out that specifi- for 50% haemolysis, was applied in the examination city depends on the principles and procedure of a of 47 leprous sera using a beef-heart extract as serological test as well as on proper and well- antigen. It was found that 7 of the 8 sera which adjusted doses of the reagent, it may be noted that reacted were from individuals with histories of if the data presented are representative, the diagnosis " venereal exposure or promiscuity " (Maltaner, of syphilis by laboratory procedures is feasible 1940). This method yielded equally discriminative among leprosy and other patients and that the New information on 60 sera from leprosy patients which York State Department of Health quantitative were examined in the Interstate Evaluation Study of complement-fixation test has a unique value under Serologic Methods held in Washington, D.C., USA, such circumstances. in 1942 (Parran et al., 1942). With cardiolipin antigens adjusted to give maximal reactions with syphilitic THE LEVEL OF ANTIBODIES AGAINST ANTIGENS sera, the quantitative complement-fixation technique PREPARED FROM ACID-FAST BACILLI AND was employed with 467 lepromatous sera and only THEIR REACTIVITY WITH LIPID ANTIGEN, 28 of them (6%) reacted. A total of 118 non- IN LEPROUS SERA reacting specimens were also tested by other techniques for the diagnosis of syphilis. Positive Almeida (1961) described the behaviour of 40 lep- reactions were also tested by other techniques for the romatous sera in a battery of tests, including diagnosis of syphilis. Positive reactions were complement-fixation and flocculation reactions. frequently encountered, in conformity with the Complement-fixation was performed with antigens results of Shively & Kuhns (1950), Grasso (1951), prepared from virulent Myco. tuberculosis in accor- Trespalacios & Otero (1951), Portella & Almeida dance with methods based on isofixation curves, (1952), Kvittingen et al. (1952) and Portnoy et al. previously described by Almeida (1958b), to deter- (1952). The Kolmer complement-fixation technique mine the level of specific antibodies in the sera yielded 16.8 % of reactions, the Kahn standard of leprosy patients. The same sera were tested 65 % and the VDRL test, 36% (Almeida et al., 1955). with cardiolipin 72 and cardchol (Schmidt, 1959), with In a serological evaluation carried out in Wash- formolized sheep cells for the Rubino test, and with ington, in 1956-67 with different tests used for the T. cruzi antigen (Freitas & Almeida, 1949). In this diagnosis of syphilis on a sample of 29 lepromatous preliminary study, 10% of lepromatous sera reacted sera, the non-treponemal tests were shown to have with cardiolipin, 72.40% with cardchol, 25% with the greatest reactivity, and the TPI test the least. Rubino antigen and 15 % with T. cruzi antigen. Only two of the non-treponemal tests (the Hinton All 40 sera reacted with tubercle antigen in the 694 J. 0. de ALMEIDA

complement-fixation test. A correlation between lipin and 30 (5.9%Y.) with T. cruzi antigen; in 912 sera the antileprous antibody content and the reactivity in the second group (titres from 21 to 200), 39 (4.3 %) with cardchol antigen was found in a study of reacted for syphilis and 89 (9.7%Y.) for Chagas' dis- 431 lepromatous patients (Almeida, 1962). When ease; in the third group of 623 sera (titres from 201 syphilis was present in a lepromatous case, it was to 2000), 30 sera (4.8%) reacted with cardiolipin possible to follow the effect of the specific treatment and 84 (13.5°/) with T. cruzi antigen; in the last of the two infections with sulfones and penicillin, group 389 sera (with titres higher than 2000), by drawing two serological curves, one for leprosy 23 (5.9%) reacted with cardiolipin and 62 (15.9%) and one for syphilis. When cardiolipin was used with trypanosoma antigen. A correlation between and the test was performed by the New York State the level of antileprous antibodies and the rate of Department of Health microflocculation technique, reactors with cardiolipin and with T. cruzi antigens 43 of the 431 sera from lepromatous patients reacted was observed, but not between reactions for syphilis whereas among 428 sera from blood donors, only and for Chagas' disease. However, 4 sera reacted 25 showed a reaction. There was a significant with Brucella abortus antigen, in complement- difference in the distribution of positive results, fixation, performed with 265 chagasic sera. This confirming the higher prevalence of reactors among incidence was significantly higher than that observed lepromatous leprosy patients, but much lower than among non-reactors with T. cruzi antigen, namely, that observed with the technique of Rein & Bossak 7 reactions among 2168 sera, although it is true that (1946) by Portella & Almeida (1952). When com- the same living conditions favour both infections. plement-fixation with cardiolipin 72 was employed in the quantitative method, only 16 reactors were detected among 283 lepromatous sera and 13 in the LEPROMATOUS LEPROSY AS A CONDITION FOR control group of 428 sera from healthy subjects. HIGH LEVEL OF ANTIBODIES Although no significance was found in this difference If this correlation found in lepromatous patients in the frequency of positive results in the two is a result of higher content of antibodies formed groups, there seems to have been a trend towards a against other antigens, much of the cross-reactivity higher frequency of positive reactions in lepromatous found in lepromatous sera could be understood. sera. The adjuvant action of mycobacteria potentiates Since reactions with cardchol were frequent among the effect of antigens when Freund's (1951) adjuvant leprosy patients (Schmidt, 1961), a total of 275 lep- is added to the inoculum. Guinea-pigs were good rous sera were titrated by the quantitative comple- producers of antibody, as shown by Lewis & Loomis ment-fixation with tubercle antigen and it was (1924) in tuberculous animals inoculated with sheep found (Almeida, 1962) that positive reactions with blood cells. Lepromatous lesions could also have cardchol were more frequent among the high-titre an adjuvant effect on antigens unrelated to acid- sera, showing a correlation between the degree of fast bacilli, as found by Almeida et al. (1964) in reactivity with tubercle antigen and that with the lepromatous patients inoculated with typhoid vac- lecithin-free cardiolipin antigen. cines. A group of 60 lepromatous patients showed In order to verify the possible influence of anti- an agglutinin response significantly higher than the leprous antibodies found in the lepromatous cases controls (60 healthy soldiers). The higher capacity on the serum reactivity with cardiolipin and with for antibody formation found in lepromatous lep- T. cruzi antigens, which could be masked when rosy favours the view of the formation of circulating total results are computed, a large number of leprosy antigen-antibody complexes, which may fix comple- patients were examined (Almeida, 1964). Leprous ment in vivo, depressing the amount of free comple- antibodies were titrated by the Wadsworth, Maltaner ment and giving to the sera, before inactivation, & Maltaner (1931) quantitative complement-fixation some features of anticomplementary sera (Almeida, method, using a lipo-polysaccharide antigen prepared 1963). from Myco. tuberculosis strain 37 RV (Almeida, TPI IN LEPROSY 1958b) in 2433 sera. At the same time, reactions were performed with cardiolipin 72 and with T. cruzi Nelson & Mayer (1949) described a complement- antigen by essentially the same serological proce- fixation test with a living and active motile suspen- dure. The results showed that in 509 sera with sion of Treponema pallidum. These treponemas titres of less than 20, 17 (3.3 Y.) reacted with cardio- remained motile after 18 hours' incubation in SEROLOGY, IN LEPROSY 695 anabiotic conditions in the presence of non-syphilitic rosy sera showed a diminution of sera reactivity. serum and guinea-pig complement.- When syphilitic However, false negative results were obtained with antibodies were present the treponemas were immo- the VDRL test in 9 and 5 TPI reactors, respectively, bilized and killed. This test was applied to 75 leprous when the reactions were carried out with saline or sera from Carville, La., USA, and the immobilization with choline chloride. of treponemas occurred in 11 (19%), whereas In a group of 197 leprosy patients, Rollier et al. standard serological tests for 57 syphilitic sera were (1955) found 10% had reacting sera for TPI, while positive. The incidence of syphilis in this population for other serological tests the proportions were lies between 15 % and 20% and the TPI test was found comparable with those in the non-leprous population; to be highly specific for syphilis, even with sera from 26.5% had dissociated results related to the albumin- leprosy patients (Nelson, 1952). globulin rates. TPI antibodies were found to be Comparing the tests of Kline and Debains with present in 2 sera from 20 leprosy patients who had the TPI test, Thivolet et al. (1953) found 12 positive positive Wassermann reactions with Budet antigen TPI reactions in a group of 50 reacting sera from and had also positive Meinicke, Kahn and cytochol a total of 90 leprous sera. At the same time, Ranque reactions, either partly or totally. Employing a et al. (1953), testing 62 leprous sera from French standardized 50% end-point quantitative technique Guiana and 40 from Dakar, found 22 positive for the TPI test, Portella (1956) examined 164 lepro- TPI reactions whereas 47 reacted with the standard matous sera. The test was positive in 6% of the serological tests. Gat6 et al. (1953) examined sera group under investigation and these can be accepted from 18 lepromatous, 8 neural and 2 indeterminate as representing the residual syphilitic cases. cases of leprosy and found all of them negative In patients with tuberculoid or indeterminate to the TPI, but positive to the Wassermann, Mei- leprosy, the serology for syphilis is comparable to nicke II, VDRL and Kline-cardiolipin tests. that of healthy persons. A dissociation of results is The incidence of 15% positive TPI reactions often present in lepromatous leprosy, occurring as among 77 leprous sera was similar to the results an abnormally high incidence of positive reactions obtained by Nelson (1953) and Floch et al. (1953). "except with the Nelson test " (Rollier & Plan- The specificity of the TPI test for syphilis on leprous chet, 1957). sera was pointed out by Vilanova & Catasus (1953) In a comparison between TPI and other serological and Rollier & Pelbois (1953) in their comparative reactions for syphilis, Rossetti et al. (1957) showed studies with Wassermann tests performed with that the TPI was highly specific for syphilis since in Kolmer's technique. the 4 positive cases among 300 leprosy patients, In a study of 224 leprosy patients, 120 tuberculoid it was possible "for the anamnestic and clinical and 204 lepromatous, all the clinical information results to say, without doubt, that they were also being available, the reactivity rates of the various affected with syphilis ". In these groups, the inci- tests employed were as follows: Kolmer C, 63.4%; dence of positive reactions varied from 25 % for the Kahn standard, 52.7%; Rein-Bossak flocculation, citochol test to 59% for the Kahn test. Only 26% 46.9%; VDRL, 46.9%; TPI, 11.2%. Altogether, of the sera were negative to the Wassermann, Kahn, 17 (68%.) of the 25 TPI reactive patients were found Meinicke II and cytochol tests, and 10% reacted to be reactive in all the other tests used, although 66 with all four tests (Rossetti et al., 1957). (33.2%) of the 199 TPI-negative patients also had The TPI test confirmed the 4 positive results positive reactions to all the other tests. It was obtained in the examination of sera from 108 leprous concluded that the TPI results were much more patients by the New York State Department of consistent with clinical opinions regarding the occur- Health quantitative complement-fixation test. In this rence of syphilis in these patients than were the group, 57 reacted with Kahn antigen and 29 with results obtained with the tests using lipid antigens, VDRL. The low incidence of positive results of by complement-fixation or by precipitation tests this complement-fixation method with cardiolipin (Edmundson et al., 1954). and their close agreement with TPI tests showed The VDRL test can show more reactivity with the high specificity of this serological test for syphilis TPI-negative leprous sera when the antigen is among leprosy patients, in contrast to the high per- diluted with buffered saline solution than when it is centages of false positive reactions obtained with prepared with a solution ofcholine chloride (Portnoy Kahn and VDRL tests (Zarco & Chan, 1958). & Edmundson, 1954); the non-syphilitic reactive lep- An incidence as high as 21 % of TPI-reacting sera 696 J. 0. de ALMEIDA was observed among 515 treated leprosy cases in as 100, or as a secondary reference standard such Casablanca, where the serological index of syphilis as the replica ALBI made to be virtually equivalent " infection is 20 % (Planchet, 1958). Clinical data were (Portella & Thompson, 1959, 1965). The potency not available to determine the specificity of the should be 1 or less for all sera evaluated as negative. Nelson test among these leprosy patients. Those previously rated as " positive " appear almost A comparison of serological tests for syphilis in always to have a potency greater than 3. When the leprous patients showed that TPI and complement- potency of the serum falls between 1 and 3 further fixation tests with Reiter-protein as antigen (RPCF) tests should be made, preferably with first specimens have almost the same specificity. The sera of 248, (Portella & Thompson, 1965). presumably non-syphilitic, leprosy patients after 5 The high specificity and sensitivity of RPCF, different serological tests, resulted in the following respectively 99.0% and 83.2%, and the lowest rate reactivities: RPCF, 3.2 %; TPI, 4.8 %; TPCF, 15.3 %; (2.9%) of biological false positive reactions when VDRL, 27.4%; Kahn, 35.9%. The RPCF and TPI compared with all other tests for syphilis, including most closely approximate to the expected serological the TPI and the TPCF reactions, were shown in a activity in relation to clinical and anamnestic infor- serological evaluation carried out by several labora- mation. Thus, the reliability of the RPCF test with tories under the sponsorship of the National Com- serum from patients with leprosy is of the same municable Disease Center, Atlanta, Ga., USA order of magnitude as the TPI test and more reliable (United States Public Health Service, 1959). In this than TPCF (with T. pallidum antigen), VDRL and survey with a small sample of 29 leprous sera, the Kahn tests (Cannefax et al., 1959). In only 3 in- greatest reactivity occurred in the non-treponemal stances was there a positive reaction to the RPCF tests and the least in the T. pallidum tests. Only two test, and reactions to other tests were negative. of the non-treponemal tests (the Hinton and the The RPCF test was reactive in 5 sera in which New York State Department of Health complement- results of the other tests were also reactive. This fixation test) were completely non-reactive to all study shows that the RPCF test is as reliable as the specimens. The reactivity rates among the others TPI test but has the advantage of being a much varied from 6.9 % for the Army complement-fixation simpler procedure (Ross, 1964a, 1964b). and New York State Department of Health micro- This low degree of non-specific reactivity for precipitation tests to 72.4% for the Rein-Bossak syphilis among leprosy patients was confirmed by test. Only one of the Reiter treponema tests was non- Ruge (1960) who observed 1 RPCF reacting serum, reactive to all specimens. In the others, reactivity among 54 false positive reactions from 327 leprous was limited to 1-3 specimens. sera. The higher proportion of TPI reactions (31 Some discrepancies observed in the TPI test are against 13 RPCF and 14 with standard methods) to be expected. This was shown by Rollier (1963) could be related to the presence of anti-treponema who submitted the same sera from 250 leprosy antibodies, since yaws may affect 1.5% or 2% of patients to two laboratories. Altogether, 42 TPI the population of the Philippine Islands where the reacting sera were detected by one laboratory and sera came from. 33 by the other; presumably false positive reactions In the experience of Ruge (1960), the complement- were found in 48 sera with negative TPI tests; fixation test with Reiter protein antigen gave only 23.3 % reacted with VDRL and 14.8 % with Kolmer 1 non-specific reaction, and even that was doubtful. and Kline tests. This result, obtained with 327 leprous sera, confirmed With leprous sera, the problem of false positive the high specificity of the RPCF test for syphilis. reactors for syphilis can be solved on practical The TPI test has many sources of error since grounds by the RPCF test. This test can be per- treponemas may be immobilized by some factors formed in those areas where leprosy is endemic; in the serum which are not of syphilitic origin. the equipment required for the test can be found Only a quantitative method can provide reliable in any laboratory for clinical pathology. The anti- results (Fribourg-Blanc, 1957a, 1957b) for the TPI gen is available from various supply houses and it tests. When a weakly reactive serum is tested, the can be kept under ordinary laboratory conditions. determination of its relative potency should be made On the other hand, TPI tests require more skill as a bio-assay with a reference standard serum. and the technical procedures and evaluation can be " Potencies are expressed relative to that of TPI carried out only in special, well-equipped serological control serum in whole serum form (WHO 2 W) centres. Any problem concerning serum should, SEROLOGY IN LEPROSY 697 when necessary, be sent to a specialized laboratory, CONCLUSIONS as has been done until now, so that a check may be kept on the specificity and sensitivity of the tests. Some conclusions on the use of serological tests As far as the reproducibility of the TPI test is for the diagnosis of syphilis in places where leprosy concerned, this is by no means unquestionable. is endemic can be drawn. As pointed out Ruge (1960), some marked discrep- ancies in the results of TPI tests were observed when (1) A preliminary screen test should be performed sera from the same leprosy patients were re-examined with the New York State Department of Health after sulfonamide treatment. Like all biological quantitative complement-fixation test, in order to tests, the TPI test is not free from false positive exclude non-reacting sera with cardiolipin antigen. reactions. Ruge (op. cit.) stated that the higher (2) Sera reacting to the complement-fixation test proportion of positive TPI reactions in his experi- should be tested again by the same method but ments (31 as compared with 13 RPCF and 14 STS) using the Reiter protein antigen. could be attributed to causes other than syphilis, e.g., to yaws. It is of interest that Ross et al. (1959) (3) Fluorescent-antibody technique, such as observed among 250 leprosy patients in Carville, FTA-200, can be employed and TPI is recom- where yaws does not exist, only 10 (4 %) with positive mended for further investigations on the specificity TPI reactions and the same number with a positive of the test. For that purpose, a bio-assay should RPCF test. be performed with the WHO reference serum.

ACKNOWLEDGEMENTS

The author is gratefulto the World Health Organization Departamento de Profilaxia da Lepra, Sao Paulo, Brazil, forsupport andadvicegiveninthis investigation. Gratitude for kindly providing the facilities for bibliographical is also expressed to Mrs L. Keffer, Librarian of the research.

RIUSUMI8 LA SIROLOGIE DE LA LLPRE

On a eu recours a toute une s6rie de m6thodes pour de fortes variations des titres d'anticorps, une faible etudier les aspects serologiques de la lepre: fixation du teneur du serum en complement, qui pourrait etre due compl6ment, 6lectrophorese, immunoelectrophorese, a une fixation de la substance in vivo par des immun- h6magglutination et h6molyse d'6rythrocytes sensibilises, complexes circulants. precipitation, immunodiffusion et immunofluorescence. Bien qu'il n'existe pas de tableau electrophoretique Le present article passe en revue ces divers proc6des, leurs caract6ristique de la lepre, les formes lepromateuses applications et les resultats obtenus. s'accompagnent dans la plupart des cas d'hypergamma- La reaction de fixation du complement a ete la plus globulinemie, la teneur en a-globulines etant augment6e frequemment utilisee, mettant en aeuvre diff6rents anti- pendant les episodes reactionnels. Dans la lepre tubercu- genes: Mycobacterium leprae, Myco. tuberculosis, Myco. loide, le rapport albumine-globuline est generalement lepraemurium, autres mycobact6ries cultivables. Les 6gal 'a l'unite; il y a un desequilibre des immunoglobulines, serums des malades atteints de lepre offrent malheureuse- le taux des immunoglobulines M etant plus eleve que dans ment un tres large spectre de r6activite qui nuit a la les formes l6promateuses ou indeterminees. specificite de l'epreuve. Celle-ci peut cependant servir au L'6preuve d'hemagglutination des hematies de mouton titrage des anticorps en pr6sence d'antigenes mycobacte- trait6es par le formol (test de Rubino) est remarquable- riens dans les s6rums des malades lepromateux ou ment specifique de la lepre. Sa positivite est la plus nette tuberculoides. Dans les formes l6promateuses, les anti- dans les formes lepromateuses, quoique certains serums de corps sont deceles plus fr6quemment et en plus grande malades donnent une reaction negative. L'hemagglutina- quantite que dans les formes tuberculoides ou indeter- tion d'erythrocytes sensibilises par des antigenes myco- min6es. Au cours de la reaction l6preuse, on note, outre bacteriens permet de titrer les anticorps antilepreux. Le

4 698 J. 0. de ALMEIDA titres sont les plus eleves dans la lepre lepromateuse, et fractions de Myco. leprae et des anticorps spdcifiques de ils augmentent dans une mesure significative dans les groupe r6agissant avec d'autres antigenes mycobacteriens. etats reactionnels. Les reactions s6rologiques utilisees dans la syphilis, On a d6couvert dans le serum de malades atteints de et notamment les reactions de Bordet-Wassermann, de lepre une s6rie de facteurs analogues it ceux que l'on Kahn et de Meinicke, sont frequemment positives dans la rencontre dans les maladies auto-immunes: anticorps 1epre, mais certaines epreuves de fixation du complement antithyr6oglobuline, facteurs antinucleaires, facteur rhu- font preuve d'une specificite suffisante permettant le matoide, proteine C-reactive, auto-anticorps antierythro- diagnostic de syphilis chez les malades atteints de lepre. cytaires. L'infection l6preuse accroit la sensibilite du II en est de meme du test d'immobilisation des treponemes systeme formateur des anticorps a l'egard des antigenes, et de la reaction de fixation du complement utilisant com- meme non apparentes a Myco. leprae, avec apparition de me antigene la proteine de Reiter. symptomes cliniques rappelant ceux des maladies auto- La m6thode des anticorps fluorescents pratiqu6e sur immunes, particuli6rement dans les etats reactionnels. des serums absorbes par Myco. tuberculosis montre la La methode de double diffusion en gel de gelose permet pr6sence d'anticorps specifiques chez les malades atteints de deceler des anticorps specifiques diriges contre diverses de 1epre.

REFERENCES

Abe, M. (1963) La Lepro, 32, 87-88 Almeida, J. O., Lima, L. S. & Carvalho, R. P. S. (1955) Abe, M. & Hirano, T. (1961) La Lepro, 30, 186-189 Amer. J. trop. Med. Hyg., 4,41-46 Abruzzo, J. L. & Christian, C. L. (1961) J. exp. Med., Ambrogio, A. (1932) Pathologica, 24, 721-804 114, 791 Anderson, R. J. (1927) J. bio. Chem., 74, 537-551 Almeida, J. 0. (1956a) J. Immunol., 76, 259-263 Andrade, L. & Silva, S. (1952) Hospital (Rio de J.), 41, Almeida, J. 0. (1956b) Complement-fixation reactions 831-868 with antigens isolatedfrom tubercle bacilli. In: Annual Aoki, Y. & Murao, K. (1933) Z. Immun.-Forsch., 79, Report of the Division of Laboratories and Research, 365-371 New York State Department of Health, Albany, Aono, G. (1953) Tohoku J. exp. Med., 57, 311-315 pp. 12-13 Arcuri, F. & Inzerillo, R. (1952) Acta med. ital. Med. Almeida, J. 0. (1957) Rev. bras. Leprol., 25, 296-297 trop., 7, 39-43 Almeida, J. 0. (1958a) Cienc. Cult., 10, 71-76 Assump,ao, L. & Silveira, G. F. (1935) Rev. Leprol. Almeida, J. 0. (1958b) Rev. bras. Leprol., 26, 181-271 S. Paulo, 2 13-36 Almeida, J. 0. (1961) Reactivity ofleprous sera in comple- Azevedo, M. P. & Homem de Mello, P. (1966) Int. J. ment fixation reactions using lipids from tubercule Leprosy, 34, 34-38 bacilli and cardchol as antigens. In: Primera Conferencia Babes, V. & Busilla, V. (1909) C. R. Soc. Biol. (Paris), Interamericana de Leprologla Experimental, pp. 126-130 67, 817-819 Almeida, J. 0. (1962) Bull. Wld Hlth Org., 26, 233-240 Badger, L. F. (1931) Publ. Hlth Rep. (Wash.), 46, 957- Almeida, J. 0. (1963) The meaning of antibody titre in 970 leprosy, determined by complement fixation with anti- Barua, D., Ghosh, S., Mukherjee, N., Banerjee, R. & gens prepared from tubercle bacilli. In: Anais do Bose, R. (1963) Bull. Calcutta, Sch. trop. Med., 11, VIII Congresso Internacional de Leprologia, Rio de 61-63 Janeiro, vol. 3, pp. 263-275 Bayarri, S. V. & Ahuir, M. R. (1947) Rev. clin. esp., 25, Almeida, J. 0. (1964) Rev. bras. Leprol., 32, 15-22 103-107 Almeida, J. 0. & Adura, M. (1954) Clin. tisiol., 9, 371- Bechelli, L. M. & Fiorillo, A. M. (1953) Lepromino- 384 reafao e alteracoes eletroforeticas em individuos nao- Almeida, L. 0. & Azevedo, M. P. (1952) Rev. bras. Biol., doentes de lepra eportadores de afecVoes sistematicas. In: 12, 129-150 doentes de lepra e portadores de afecpoes sistematicas. Almeida, J. 0. & Bechelli, L. M. (1961) Avaliafjo da In: Memoria del VI Congreso Internacional de Lepro- terapeutica antileprosa em lepromatosos pela rea_vdo logia, Madrid, pp. 496-500 quantitativa de fixapao de complemento. In: Primera Benetazzo, G. (1933) Dermosifilografo, 7, 241-256 Conferencia Interamericana de Leprologia Experimental, Bergot, J., Nicoli, J., Ziegler, P. & Demarchi, J. (1962) pp. 131-140 Bull. Soc. Path. exot., 55, 776-782 Almeida, J. O., Branddo, H., Lima, E. G. & Lippelt, A. Besredka, A. (1935) C. R. Soc. Biol. (Paris) 119, 252-253 (1964) Int. J. Leprosy, 32, 292-296 Bier, 0. G. (1936) Rev. bras. Leprol., 4, 211-222 Almeida, J. O., Carvalho, R. P. S. & Padron, C. (1956) Bier, 0. G. & Arnold, K. (1935) Folia clin. biol. (S. Paulo), Rev. bras. Leprol., 24, 17-26 7, 9-11 SEROLOGY IN LEPROSY 699

Bier, 0. G. & Planet, N. (1936) Folia clin. biol. (S. Paulo), Estrada-Parra, S., Calderon M., S., Salazar M., M. & 8, 72-75 Amezana, M. A. (1966) Int. J. Leprosy, 34, 294-297 Bonatti, A. A. & Lebron, E. J. (1958) Rev. Fac. Med. Faure, M. (1940) Les haptenes de fixation de l'alexine, Tucumdn, 1, 143-149 Lille, Douriez-Bataille Boncinelli, U. (1937) G. ital. Derm. Sif., 78, 1175- Fisher, S. & Keogh, E. V. (1950) Nature (Lond.), 165-248 1201 Floch, H. & Andrd, J. (1955) Arch. Inst. Pasteur Guyane Bonomo, L., Dammacco, F., Pinto, L. & Barbieri, G. fran_., 375 (1963) Lancet, 2, 807-809 Floch, H. & Andrd, J. (1956) Arch. Inst. Pasteur Guyane Bcnomo, L., Dammacco, F., Turso, A. & Barbieri, C. fran9., 393 (1967) Int. J. Leprosy, 35, 65-71 Floch, H. & Sohier, R. (1952) Arch. Inst. Pasteur Guyane Bonomo, L., Tursi, A., Trimigliozzi, G. & Dammacco, F. franc., 272 (1965) Brit. med. J., 2, 689-690 Floch, H. & Sohier, R. (1953) Bull. Soc. Path. exot., 46, Brants, J. (1932) Derm. Wschr., 47, 1688-1891 992-1001 Brusco, C. M. & Masanti, J. G. (1963) Int. J. Leprosy, 31, Floch, H., Sohier, R., Thivolet, J. & Rolland, M. (1953) 14-25 Etude comparee des resultats des reactions serologiques Burrell, R. G. & Rheins, M. S. (1957) Int. J. Leprosy, 25, classiques de la syphilis et du test d'immobilisation des 223-229 treponemes chez les lepreux. In: Memoria del VI Con- Bush, 0. B., Jr (1958) Int. J. Leprosy, 26, 123-126 gresso Internacional de Leprologia, Madrid, pp. 811-812 Cannefax, G. R., Ross, H. & Bancroft, H. (1959) Publ. Frazier, C. N. & Wu, H. (1925) Amer. J. trop. Med., 5, Hlth Rep. (Wash.), 74, 45-48 297-305 Castro, I., Almeida, S. M. & Castro, N. (1963) Prova de Fretias, J. L. P. & Almeida, J. 0. (1949) Hospital (Rio miocontraplo sensivel aos soros de leprosos. In: Anais de J.), 35, 787-800 do VIII Congresso Internacional de Leprologia, Rio de Freund, J. (1927) J. Immunol., 13, 161-169 Janeiro, vol. 9, pp. 332-346 Freund, J. (1951) Amer. J. clin. Path., 21, 645-656 Castro, N. 0. & Bonatti, A. A. (1951) Int. J. Leprosy, 19, Fribourg-Blanc, A. (1957a) Ann. Derm. Syph. (Paris), 309-321 84, 286-304 Cathcart, E. S., Williams, R. C., Ross, H. & Calkins, E. Fribourg-Blanc, A. (1957b) Ann. Derm. Syph. (Paris), (1961) Amer. J. Med., 31, 758-765 84, 410-420 Chargaff, E. & Schaefer, W. (1935) Ann. Inst. Pasteur, 54, Frugoni, C. & Pisani, E. (1909) Berl. klin. Wschr., 46, 708-714 1530 Chekherdenian, M. (1960) Leprologia, 5, 69-79 Gat6, J., Rousset, J. & Coudert, J. (1953) Le test de Chini, V. (1961) Policlinico Sez. prat., 68, 461-482 Nelson chez les lepreux. Comparaison avec les reactions Christian, C. L., de Simone, A. R. & Abruzzo, J. L. classiques et celles de la cardiolipine. In: Memoria del (1963) Arthr. and Rheum., 6, 766 VI Congreso Internacional de Leprologia, Madrid, Cichini, T. (1952) Arch. Ital. Sci. med. trop., 33, 151-160 pp. 809-810 Collier, W. A. & Gehr, E. (1957) Docum. Med. geogr. Gernez-Rieux, C., Montestruc, E. & Tacquet, A. (1952) trop. (Amst.), 9, 165-168 Arch. Inst. Pasteur Martinique, 5, 3-14 Collier, W. A., Mesquita, S. J. B. & Van Zanten, E. Gernez- Rieux, C., Montestruc, E. & Tacquet, A. (1953) (1957) Docum. Med. geogr. trop. (Amst.), 9, 345-356 Int. J. Leprosy, 21, 21-25 Collier, W. A., Mesquita, S. J. B. & Van Zanten, B. Gernex-Rieux, C. & Tacquet, A. (1956) Sem. H6p. Paris, (1958) Docum. Med. geogr. trop. (Amst.), 10, 171-174 32, 1129-1136 Contreras, F., Miguel, S., Roldan, A., Guillen, J., Germuth, F. G., Jr, & McKinnon, G. E. (1957) Bull. Terencia, J. & Tarabini, J. (1953) Fontilles (Alicante), Johns Hopk. Hosp., 101, 13-42 12, 467-478 Ghosh, S., Mukherjee, N. & Bose, R. (1964) Indian J. Cooke, J. V. (1919) J. infect. Dis., 25, 474-492 med. Res., 52, 458-461 Cottenot, F. (1964) C. R. Soc. Biol. (Paris), 158, 1004- Gokhale, B. B. (1958) Leprosy Rev. 29, 155-157 1005 Golierkeri, P. B., Gokhale, B. B. & Ranade, S. N. (1952) Curban, G. V. (1962) Rev. bras. Leprol., 30, 179-216 Indian J. med. Sci., 6, 357-363 Dharmendra (1941) Indian J. med. Res., 29, 523-525 Gomes, J. M. (1927) Rev. Biol. & Hyg., 1, 17-38 Dharmendra & Bose, R. (1941) Indian J. med. Res., 29, Gomes, J. M. (1935) Int. J. Leprosy, 3, 283-290 7-22 Gomes, J. M. & Patteo, J. D. (1928) Rev. Biol. Hyg., 1, Dharmendra, Bose, R. & Sen Gupta, P. C. (1946) Indian 89-104 J. med. Res., 34, 1-2 Grasso, P. (1951) Ann. ital. Derm. Sif., 6, 72-77 Eichbaum, F. W. (1942) Rev. bras. Biol., 2, 285-300 Greval, S. D. S. (1947) Indian med. Gaz., 82, 668-671 Eitner, E. (1906) Wien. klin. Wschr., 19, 1555 Greval, S. D. S., Chandra, S. N. & Das, B. C. (1941) Eitner, E. (1908) Wien. klin. Wschr., 21, 729 Indian med. Gaz., 76, 474-475 Edmundson, W. F., Wolcott, R. R., Olansky, S. & Ross, Greval, S. D. S., Lowe, J. & Bose, R. (1939) Indian J. med. H. (1954) Int. J. Leprosy, 22, 440-449 Res., 26, 843-849 700 J. 0. de ALMEIDA

Guinto, R. S. & Mabalay, M. (1958) Paper electrophoresis Lim, S. D. & Fusaro, R. M. (1968a) Int. J. Leprosy, 36, in the study ofserum proteins in leprosy. In: Review of 144-153 the second conference on progress and potentials in Lim, S. D. & Fusaro, R. M. (1968b) Int. J. Leprosy, 36, leprosy investigation, Carville, La., US Department of 154-161 Health, Education, and Welfare (Public Health Publi- Lowe, J. (1942) Int. J. Leprosy, 10, 156 cation No. 641), pp. 21-32 Lowe, J. & Greval, S. D. S. (1939) Indian J. med. Res., Gupta, N. P., Mehrotra, T. N., Mangalik, V. S. & 26, 833-841 Chandy, P. J. (1954), Indian J. med. Sci., 7, 23-32 Lucentini, L. & Nazzaro, P. (1954) Ann. ital. Derm. Sif., Hanks, J. H. & Gray, C. T. (1954) Int. J. Leprosy, 22, 9, 97-100 147-161 Macheboeuf, M. & Bonnefoi, A. (1935) Bull. Soc. Chim. Hasseltine, H. E. (1924) Publ. Hlth Bull. (Wash.), 141, biol. (Paris), 17, 1201-1209 37-54 Macheboeuf, M. & Faure, M. (1939) C. R. Acad. Sci. Holst, E. (1964) Acta path. microbiol. scand., 62, 367-376 (Paris), 209, 700-702 Honda, H., Oshima, T., Morita, Y., Shivai, S., Sakurai, Machebeeuf, M. & Levy, G. (1935) Bull. Soc. Chim. biol. H. & Iguchi, K. (1956) La Lepro, 25, 45-49 (Paris), 17, 1210-1234 Honda, H. & Yoshino, K. (1951) La Lepro, 20, 128-132 Madorsky, M., Bachrach, U., Gurevitch, J. & Sagher, F. Hoxter, G., Batista, L. & Vellini, L. L. (1951) Rev. bras. (1953) Bull. Res. Coun. Israel, Series E, 2, 449 leprol, 19, 27-40 Maillard, E. R. (1952) Amer. J. publ. Hlth, 42, 175-181 Ichiata, T. (1940) La Lepro, 11, Suppl., abstract 8386 Maillard, E. R. & Gagliardo, F. J. (1951) Amer. Rev. Illand, C. N. (1951) J. Path. Bact., 63, 735-741 Tuberc., 64, 675-681 Ishihara, S. (1950) La Lepro, 19, 5-11, 12-18 Maltaner, E. (1940) Amer. J. trop. Med., 20, 843-848 Ishihara, S. (1953) Int. J. Leprosy, 21, 187-199 Marchoux, E. & Caro, J. (1928) Ann. Inst. Pasteur, 42, Jardin, C. & Beytout, D. (1960) Mid. trop., 20, 81-101 542-552 Juravleva, G. F. (1965) Voprosi Leprologii i. Dermato- Marcondes, J. R. (1929) A reacVao de Gomes para o diag- logii, 19, 89-92 nostico precoce da lepra e sen valor prophylactico. Tese Kabakov, E. N. (1961) J. Microbiol. Epidemiol. Epi- da Faculdade de Medicina, SAo Paulo demiol. Immunobiol., 32, 448-455 Margini, R. M. Houvilier, 0. A., Acerbo, E. O., Gutierrez, Kahn, R. I. (1951) Ann. N. Y. Acad. Sci., 54, 40-47 O., Bobbi, C., Cordal de Bobbi, M. E. & Casavola, H. Kawaguchi, Y. (1953) La Lepro, 22, 59-62 (1965) Rev. argent. Leprol., 2, 19-22 Kent, J. F., Otero, A. G. & Harrigan, R. E. (1957) Matthews, L. J. & Trautman, J. R. (1965) Lancet, 2, Amer. J. clin. Path., 27, 539-545 915-918 Kerbastard, P. (1957) Med. trop., 17, 251-263 Mauze, J. & Arnaud, G. (1954) Int. J. Leprosy, 22, 55-60 Kono, M., Tsugami, H. & Sakurau, H. (1952) La Lepro, Mayama, A. (1954) Rep. Res. Inst. Tuberc. Leprosy 21, 98-108 Tohoku, 5, 273-282 Kornel, G. (1933) Z. Immun.-Forsch., 78, 207-216 Mayama, A. (1957) La Lepro, 26, 160-165 Kulka, A. M. (1942) J. Immunol., 43, 273-288 Mayama, A. (1958) La Lepro, 27, 233-241 Kvittingen, J., Cutler, J. C., Amador Guevara, J., Mayama, A. (1960) La Lepro, 29, 100-103 McCullough, J. C., Rose, E., Ford, V., Tampi, R. B,. Mayama, A. (1965) Sci. Rep. Res. Inst. Tohoku Univ., Sen, S., Lakshwir, Khan, J. & Dass, J. (1952) Bull. Series C, 12, 267-274 Wld Hlth Org., 5, 481-504 Melamed, A. J. & Barcia, A. (1958) Leprologia, 3, 57-67 Languillon, J., Bourrell, P., Boissan, R. H. & Picard, P. Mercau, A. R., DePaoli, E. A., Culasso, R., Canoirana, (1960) Med. trop., 20, 365-367 J. M., Gonzales, A., Podesta, D. & Martinez Prieto, Lauret, L. & Kerbastard, P. (1955) Bull. Soc. Path. exot., P. R., (1963) Leprologia, 8, 226-227 48, 789-795 Merclin, G. V. & BorisQva, S. S. (1964) Itogovaja nau6naja Lechat, M. F., Close, J. & Robyns, E. (1961) Ann. Soc. Konferencija dermato-venerologov, Zdanov, pp. 61-63 belge Med. trop., 41, 535-566 Merclin, G. V. & Pogorelov, V. N. (1964) U6enye zapiski Leon Blanco, L. & Lavernia, F. (1948) Las proteinas en Instituta dlja izu6enija lepry, 4, 113-115 el suero de los leprosos. In: Memoria del V Congreso Merklen, F. P., Cottenot, F. & Gallistin, P. (1963) C. R. Internacional de la Lepra, Havana, p. 573 Acad. Sci. (Paris), 257, 2212-2213 Levine, M. (1952) Int. J. Leprosy, 20, 201-202 Merklen, F. P. & Cottenot, F. (1965) Bull. Soc. Path. Lewis, P. A. & Aronson, J. D. (1923) J. Exp. Med., 38, exot., 58, 332-335 219-232 Merklen, F. P. & Cottenot, F. (1966) Rev. Path. comp., Lewis, P. A. & Loomis, D. (1924) J. Exp. Med., 40, 503- 66, 333-335 515 Mertslin, G. B. & Loginov, V. K. (1963) Leprosy Rev., Lim. S. D. & Fusaro, R. M. (1967a) Int. J. Leprosy, 35, 34, 46 355-360 Meyer, K (1937) Ann. Inst. Pasteur, 59, 477-491 Lim, S. D. & Fusaro, R. M. (1967b) Int. J. Leprosy, 35, Middlebrook, G. & Dubos, R. J. (1948) J. exp. Med., 361-365 88, 521-528 SEROLOGY IN LEPROSY 701

Miguel, S., Roldan, A., Guillen, J. Terencio, J. & Pon- Pereira, P. C. R. (1936) Int. J. Leprosy, 4, 207-214 ciani, J. (1954), Int. J. Leprosy, 22, 47-54 Pereira, P. C. R. (1938) Rev. mid. Sul de Minas, 5, 23-31 Montel, L. R. (1952) Bull. Soc. franc. Derm. Syph., 59, Photinos, A. & Micha6lides, N. (1912) Lepra, Bibliotheca 157-158 Internationalis, 12, 207-209 Montestruc, E. (1953) Rev. colon. Med. Chir., 25, 172-174 Pinner, M. (1925) Amer. Rev. Tuberc., 12, 142-155 Montestruc, E. (1959) Bull. Soc. Path. exot., 52, 561-564 Pinner, M. (1928) Amer. Rev. Tuberc., 18, 497-501 Montestruc, E. (1960) Int. J. Leprosy, 28, 140-143 Pinto, T. J. & Zeo, A. (1951) Ann. N. Y. Acad. Sci., 54, Montestruc, E., Papa, F. & Hermantin, G. (1962) Bull. 48-52 Soc. Path. exot., 55, 16-19 Planchet, L. (1958), Maroc. med., 37, 1113-1129 Morris, J. A., Aulisio, C. G., Boseman, F. M. & Guinto, Pfissier, M. & Secret, E. (1953) Maroc. med., 32, Special R. S. (1961) Leprosy Briefs, 12, 2 Number, pp. 779-783 Muelling, R. J., Goetz, C. G., Ross, H. (1960) Int. J. Portella, 0. B. (1956) Treponema immobilization test. In: Leprosy, 28, 144-154 Annual report of the Division of Laboratories and Muniz, J. (1950) On the value of " conditioned hemolysis" Research, New York State Department of Health, for the diagnosis of American trypanosomiasis. A com- Albany, pp. 84-87 parative study with complement fixation and sensitized Portella, 0. B. & Almeida, J. 0. (1952) Rev. bras. Leprol., erythrocytes agglutination tests. In: Transactions of the 20, 32-40 5th International Congress for microbiology, Rio de Portella, 0. B. & Thompson, W. R. (1959) Immunology, Janeiro, vol. 2, pp. 326-332 2, 295-308 Namba, M. & Fujiwara, H. (1952) La Lepro, 21, 5-10 Portella, 0. B. & Thompson, W. R. (1965) Bull. Wld Hlth Narula, R. M. & Gupta, V. P. (1953) Bull. WId Hlth Org., 33, 705-720 Org., 9, 877-880 Portnoy, J., Galvez, R. & Cutter, J. C. (1952) Amer. J. Navalkar, R. G., Norlin, M. & Ouchterlony, 0. (1965) Syph., 26, 566-570 Int. Arch. Allergy., 28, 250-260 Portnoy, J. & Edmundson, W. F. (1954) Int. J. Leprosy, Neill, M. H. & Dewar, M. M. (1927) Publ. Hlth Bull. 22, 181-184 (Wash.), 168, 1-19 Pozzo, G. & Hofmann, F. M. (1955) G. ital. Derm. Sif., Nelson, R. A. (1952) Brit. J. vener. Dis., 23, 160-168 96, 403-429 Nelson, R. A. (1953) Le test d'immobilization des trepo- Rabello, E. & Machado, J. C. (1936) Rev. bras. Leprol., nemes dans la syphilis. In: Charpy, J., ed., Le T.P.I. 4, 307-318 test de Nelson-Mayer et les nouveaux aspects immuno- Rabello, E., Machado, J. C. & Pinto, J. T. (1938) Derm. logiques de la syphilis, Paris, Masson, pp. 87-100. Wschr., 107, 1214 Nelson, R. A., Jr, & Mayer, M. M. (1949) J. exp. Med., Rabson, A. S. (1955) Int. J. Leprosy, 23, 155-161 89, 369-393 Ranque, J., Gallais, P., Depieds, R. & Moignoux, J. B. Ogata, T. (1953) Some chemical andbiochemicalproperties (1953) La Lepre, 199-205 ofcardiolipin and lecithin in serological testsfor syphilis. Rees, R. J. W., Chatterjee, K. R., Peys, J., Tee, R. & In: Transactions of the 6th International Congress for Waters, M. F. R. (1963) Serological studies in Leprosy. Microbiology, Rome, vol. 2, pp. 181-182 In: Anais VIII Congresso Internacional de Leprologia, Ogata, T. & Abe, M. (1958) Leproagglutination-sero- Rio de Janeiro, Vol. 3, pp. 439-444 logical agglutination of leprous sera using cardiolipin- Reich, C. V. (1961) Leprosy Briefs, 12, 1-2 lecithin antigen. In: Transactions of the VII Inter- Rein, C. R. & Bossak, H. N. (1946) Amer. J. Syph., 30, national Congress of Leprology, Tokyo, pp. 215-219 40-46 Ogata, T. & Hara, I. (1960) La Lepro, 26-28, 49-57 Rheins, M. S. Birkeland, J. M., Burrell, R. G. & Guinto, Orbaneja, J. G., Perez, G. A. & Arjona, E. (1953) Pro- R. S. (1957), Int. J. Leprosy, 25, 104-109 teinas plasmaticas en las distintas formas clinicas de la Rizzi, D., Daniele, F. & Marano, R. (1962) G. Mal. lepra. In: Memoria del VI Congreso Internacional de infett., 14, 70-86 Leprologia, Madrid, pp. 777-780 Rollier, M. R. (1963) Bull. Soc. franC. Derm. Syph., 70, Olitzki, A. L. & Sagher, F. (1956) J. invest. Derm., 27, 551-553 35-42 Rollier, R. & Pelbois, F. (1953) Maroc. mid., 32, 777-778 Pangbom, M. C. (1955) Antigens of the tubercle bacillus. Rollier, R., Pelbois, F. & Chraibi, L. (1955) Bull. Soc. In: Annual Report of the Division of Laboratories and fran9. Derm. Syph. 62, 326-329 Research, New York State Department of Health, Rollier, R. & Planchet, H. (1957) Maroc med., 36, 1206- Albany, p. 18 1212 Pangbom, M. C. (1956) Tuberculology, 16, 134-137 Romanci, C. A. (1952) Igiene mod., 45, 163-171 Parlett, R. C., Youmans, G. P., Reher, C. & Lester, W. Ross, H. (1951) Int. J. Leprosy, 19, 51-57 (1958) Amer. Rev. Tuberc., 77, 462-472 Ross, H. S. (1954a) Int. J. Leprosy, 22, 174-180 Parran, T. et al. (1942) J. Vener. Dis. Inform., 23, 355-359 Ross, H. S. (1954b) Int. J. Leprosy, 22, 328-330 Pepys, J., Augustin, R. & Patterson, A. B. (1959) Tubercle Ross, H. S. (1964a) Leprosy in India, 36, 93-107 (Edinb.), 20, 163-173 Ross, H. S. (1965b) Leprosy in India, 36, 193-211 702 J. 0. de ALMEIDA

Ross, H. S., Butler, C. F. & Laukaitis, R. B. (1959) Int. Stevenson, W. D. H. (1925) Indian J. med. Res., 12 J. Leprosy, 27, 129-133 583-600 Ross, H. & Gemar, F. (1953) Int. J. Leprosy, 21, 27-33 Sushida, K. & Hirano, N. (1961) La Lepro, 30, 81-88; Rossetti, C., Doglioni, L., Guillen, J. & Tarabini, C. 89-95 (1957), Fontilles (Alicante), 4, 213-226 Tanimura, T., Honda, H. & Oshima, T. (1956) Int. J. Roy, A. N. (1963) Ann. Biochem., 23, 1-4 Leprosy, 24, 434-446 Roy, A. T. & Banerjee, G. (1955) Ann. Biochem., 15, 55-58 Tarabini, C. G. (1957a) Fontilles (Alicante), 4, 227-234 Row, R. (1939) Trans. roy. Soc. trop. Med. Hyg., 32, Tarabini, C. G. (1957b) Fontilles (Alicante), 4, 235-245 497-504 Tarabini, C. G. (1957c) Fontilles (Alicante), 4, 247-254 Rubino, M. C. (1926a) Rev. med. Uruguay, 29, 143-155 Tatarinov, Y. S. (1962) Trudy Instituta dlja izuxnija lepry, Rubino, M. C. (1926b) Rev. Soc. argent. Biol., 2, 407-416 8 (3), 52-55 Rubino, M. C. (1929) Rev. mid. Uruguay, 32, 85-116 Tatarinov, Y. S. (1964) Leprosy, 4, 106-112 Rubino, M. C. (1931a) Bull. Soc. Path. exot., 24, 519 Taylor, J. & Malone, R. H. (1924) Indian J. med. Res., Rubino, M. C. (1931b) Ann. Inst. Pasteur, 47, 147-172 22, 127-137 Rubino, M. C. (1931c) Zbl. Bakt., I. Abt. Orig., 120, Terencio de Las Aguas, J. (1958) Fontilles (Alicante), 4, 378-384 445-453 Rubino, M. C. (1934) Arch. urug. Med., 5, 414-431 Thivolet, J., Floch, H., Rolland, M. & Sohier, R. (1953) Ruge, H. (1955) Bull. Wid Hlth Org., 13, 861-886 Etude des T.P.L chez les lepreux. In: Charpy, J., ed., Ruge, H. G. S. (1960) Bull. Wld Hlth Org., 23, 793-802 Le T.P.L-test de Nelson-Mayer et les nouveaux Sagher, F. (1953) Int. J. Dermatol., 107, 182-224 aspects immunologiques de la syphilis, Paris, Masson, Salazar M., M. & Calderon M., S. (1967) Rev. Inst. Salud. pp. 221-223 publ. (Mex.), 27, 125-128 Thompson, R. W. & Almeida, J. 0. (1956) Use of model Schaefer, W. (1940) Ann. Inst. Pasteur, 64, 301-315 systems to indicate presence of a second agent in tuber- Schmidt, H. (1959) Bull. Wld Hlth Org., 20, 1175-1191 culosis antigen preparations. In: Annual Report of the Schmidt, H. (1961) Bull. Wld Hlth Org., 25, 189-195 Division of Laboratories and Research, New York Schossmann, K. (1930) Z. Immun.-Forsch., 68, 154 State Department of Health, Albany, pp. 33-34 Scott, N. B. & Smith, D. T. (1950) J. Lab. clin. Med., 35, Trespalacios, F. & Garcia-Otero, A. (1951) Bol. Soc. 303-307 cuba Derm. Sif., 8, 19-29 Seibert, F. (1941) Bact. Rev., 5, 69-95 Tuma, M. & Silva, C. (1962) Int. J. Leprosy, 30, 71-76 Seibert, F. B. & Nelson, J. W. (1943) Amer. Rev. Tuberc., United States Public Health Service (1958) Serology 47, 66-77 evaluation and research assembly, 1956-57, Washing- Seibert, F. B. & Soto-Figueroa, E. (1956) Proc. Soc. exp. ton, D.C., US Government Printing Office (Public Biol. (N.Y.), 92, 811-814 Health Service Publication No. 650) Sen Gupta, P. C. (1945) Indian med. Gaz., 80, 396-398 Uyguanco, M. L., Rodriguez, J. N. & Concepcion, I. Sen Gupta, P. C. (1948) Researches on kala-azar in India, (1950) J. Philipp. med. Ass., 26, 65-73 1938-48. In: Proceedings of the Fourth International Vilanova, X. & Catasus, J. M. (1953) Int. J. Leprosy, 21, Congresses on Tropical Medicine and Malaria, Washing- 453-457 ton, D.C., US Government Printing Office, vol. 2, pp. Wade, W. H. (1955) Int. J. Leprosy, 23, 315 1135-1142 Wadsworth, A. B., Maltaner, F. & Maltaner, E. (1925) S,fid, C. & Schaller, K. F. (1957) Bull. Soc. Path. exot., J. Immunol., 10, 241-431 50, 17-20 Wadsworth, A. B., Maltaner, F. & Maltaner, E. (1931) Shepard, C. C. & Kirsch, D. (1961) Proc. Soc. exp. Biol. J. Immunol., 21, 313-340 (N. Y.), 106, 685-691 Wadsworth, A. B., Maltaner, F. & Maltaner, E. (1938a) Shively, J. A. & Kuhns, D. M. M. (1950) Int. J. Leprosy, J. Immunol., 35, 105-115 New Orleans, 18 (2), 169-175 Wadsworth, A. B., Maltaner, F. & Maltaner, E. (1938b) Sindo, T., Nakanishi, S., Yamaguchi, I., Itoi, M., J. Immunol., 35, 217-234 Hiyama, A., Haga, K., Utahashi, T., Narita, M. & Williams, R. C. & Kunkel, H. G. (1962) J. clin. Invest., Okada, S. (1959) La Lepro, 28, 6-12 41, 666 Sohier, R., Juillard, J. & Trimberger, D. (1950) Ann. Inst. Witebsky, E., Klingenstein, R. & Kuhn, H. (1931) Klin. Pasteur, 78, 283-285 Wschr., 10, 1068-1071 Statineanu, A. & Damelopolu, D. (1909a) Zbl. Bakt., 4, Yamada, N. (1958) La Lepro, 27, 126-135 480-483 Yoshinaga, T. (1949 La Lepro, 18, 57-58; 82-85 Statineanu, A. & Damelopolu D. (1909b) C. R. Soc. Biol. Zarco, R. M. & Chan, V. (1958) J. Philipp. med. Ass., (Paris), 66, 332-334 34, 205-212