Immunopathology of nephritis in Africa*

emoranda Immunopathology of nephritis in Africa * Clinicians have long suspected a relationship between malaria and nephritis in Africa. The results of tests made several years ago suggested that the relationship might be an immunological one. This memorandum discusses clinical, epidemiological, morphological, and immunopathological aspects of malaria-associated nephropathy, will special emphasis on immunological investigations. Immunofluorescence studies on renal biopsies from patients with the nephrotic syndrome and Plasmodium malariae parasitaemia have shown the presence of immunoglobulin (Ig) deposits with certain complement components on glomerular basement membranes. IgG with anti-P. malariae specificity has been found in eluates of kidney tissue from such patients and P. malariae antigen was identified in the glomerular basement membrane by immunofluorescence studies. These observations support the view that the nephropathy associated with P. malariae infections is a form of immune complex nephritis initiated by circulating P. malariae antigens and anti-P. malariae antibodies. Additional support is obtained from electron microscope studies, which show that electron-dense material is associated with the glomerular basement membrane in certain diseases of the kidney in which immune complexes have been detected in glomeruli by immunofluorescence methods. The view that malarial nephritis is aform ofimmune complex disease should be useful in stimulating new approaches to the study of the pathogenesis of both the initiating and the perpetuating immunopathological lesion. Earlier reports of an association between Plasmo- dium malariae infections and renal disease led to studies being made in West and East Africa in which modem immunological techniques were employed to study the pathogenesis of the nephrotic syndrome associated with malaria in these areas. The demon- stration that antigen-antibody reactions may cause the renal lesions of the nephrotic syndrome in Africa would be important evidence for the hypothesis that similar immunological mechanisms may underlie the nephrotic syndrome in other parts of the world. It would also provide guidance and justification for fur- ther therapeutic efforts in a disease in which therapy has hitherto failed to halt the progression of the renal lesion. This memorandum reviews the evidence that immunological mechanisms might produce the renal lesion responsible for the nephrotic syndrome in East and West Africa, and makes several suggestions for further research. Pending definitive international agreement on nomenclature, the terms " nephrotic syndrome associated with Plasmodium malariae ", " malarial nephritis ", and " malaria-associated renal disease" are used. IMMUNOPATHOLOGICAL MECHANISMS IN EXPERIMENTAL GLOMERULONEPHRITIS AND PARALLELS IN HUMAN DISEASE Glomerulonephritis is produced by at least two pathogenic mechanisms that are mediated by anti- body (Dixon, 1971). The first of these depends on the production of antibody that reacts with the glomerular basement membrane (GBM). The result- * This memorandum was drafted by the signatories listed on page 394. Requests for reprints should be addressed to the Chief Medical Officer, Immunology, World Health Organization, 1211 Geneva, Switzerland. The memorandum will be reprinted in Clinical and Experimental Immunology. ing disease is characterized by linear immunofluores- cent staining of GBM, and deposits are not observed by electron microscopy. The second mechanism is based on the production of antibody reacting in the circulation with non-glomerular endogenous or exo- genous antigens to form immune complexes. These complexes are trapped in the glomerular capillary walls and in the mesangium by a non-immunological mechanism. The usual immunofluorescent pattern is granular and corresponds to the scattered sub- 2814 - 387 MEMORANDA endothelial and subepithelial deposits seen by elec- tron microscopy. Where possible, the most frequent forms of experimental and human glomerulonephri- tis are discussed in terms of these two major immuno- pathogenic mechanisms. Delayed hypersensitivity does not seem to be as important as the humoral response, and mononuclear cells are usually not observed in the areas of glomerular lesions. GLOMERULONEPHRITIS PRODUCED BY ANTIBODIES TO GBM ANTIGENS Nephrotoxic nephritis When anti-GBM antibodies raised in a foreign species are injected intravenously into an experi- mental animal, the GBM is the principal site of binding. During the first phase only minimal glo- merular lesions are visible by electron microscopy. After a few days the animal produces antibodies against the heterologous globulin implanted in the GBM (second phase). The resulting inflammatory reaction is characterized by severe proliferative and exudative changes (Unanue & Dixon, 1967). In man. A comparable mechanism has produced glomerulonephritis in man in exceptional cases. When anti-human-cancer horse serum was injected into three patients, both the horse globulin and human antibody to horse gamma-globulin bound in a linear pattern to the GBM and produced lethal glomerulonephritis (De la Pava et al., 1962). In animals. Antisera produced in animals by immunization with non-renal epithelial cell cultures show mild nephrotic properties because species- specific antigens of nucleated cells are present in GBM. When such antisera are injected they bind to the GBM of the host in a pattern resembling nephrotoxic nephritis (Kano & Milgrom, 1971). Likewise, in patients treated with anti-human-lym- phocyte horse globulin the horse globulin and human antibody to horse globulin may bind to the GBM (Wilson et al., 1971). Experimental glomerulonephritis Another type of glomerulonephritis is produced by antibody cross-reacting with the host's GBM. When sheep are immunized with human GBM, the resulting antibody localizes in a linear fashion in the GBM of the sheep, which develops a fulminating proliferative glomerulonephritis (Steblay, 1962). In man, Goodpasture's syndrome and certain forms of rapidly progressive proliferative glomerulonephritis may be produced by a similar pathogenic mechanism. Anti-GBM antibodies are eluted from the glomeruli and may be demonstrable in the serum after bilateral nephrectomy (Lerner et al., 1967). The antigen that provokes the antibody response may be an endo- genous GBM antigen normally present in the serum and in the urine, or an exogenous antigen cross- reacting with human GBM. In vitro cross-reaction between antigens of streptococci group A, type 12, and human GBM has been shown by Markowitz et al. (1960). GLOMERULONEPHRITIS PRODUCED BY CIRCULATING ANTIGEN-ANTIBODY COMPLEXES Acute serum sickness In rabbits. A massive intravenous injection of a foreign protein into rabbits is followed by the development of acute proliferative and exudative glomerulitis, which spontaneously regresses (Feld- man, 1958). Focal precipitates containing antigen- antibody and complement, conceivably in the form of immune complexes, can be demonstrated on the epithelial side of the GBM by immunofluorescence and electron microscopy. The antigen, however, can be detected only in a small number of animals and for a limited period. In man. Acute post-streptococcal glomerulonephri- tis is probably produced in man by a comparable mechanism (Andres et al., 1966). Chronic serum sickness Both homologous and heterologous antigens may produce chronic serum-sickness nephritis. Rabbits. The daily inoculation of small amounts of heterologous protein in rabbits over a long period leads to chronic progressive glomerulonephritis, which covers the whole morphological spectrum of human proliferative and membranous glomerulo- nephritides (Dixon et al., 1961). The quantity and quality of antibody response, complement poly- morphonuclear leucocytes, and other mediators may account for the different morphological lesions that are observed. With electron microscopy, immune precipitates are seen to be irregularly distributed on the endothelial and epithelial sides of the GBM. Immunofluorescent staining is granular. The antigen may not be readily detectable. 388 IMMUNOPATHOLOGY OF NEPHRITIS IN AFRICA Mice. Chronic serum sickness nephritis develops in mice neonatally inoculated with lymphocytic choriomeningitis virus (Oldstone & Dixon, 1967). An example of autologous non-glomerular antigen- antibody complex glomerulonephritis is the disease produced in rats by immunization with rat kidney homogenate mixed with Freund's adjuvant (Hey- mann et al., 1962); the antigen is localized in the brush border of proximal convoluted tubules and is present in the circulation. Therefore, the immune response promotes the formation of circulating com- plexes that damage the glomeruli (Dixon, 1971). Man. Malarial glomerulonephritis is the most important example in man of a chronic disease that is exogenous in nature (Houba et al., 1970). Auto- logous antigen might be responsible for membranous nephropathy but the evidence is still lacking. Lupus nephritis The spontaneous systemic disease of NZB mice is similar to human systemic lupus erythematosus (SLE). The NZB glomerulonephritis is proliferative and membranous. Three types of immune complex have been shown in glomerular lesions-namely, DNA-anti-DNA (Seegal et al., 1969), RNA-anti- RNA (Lambert & Dixon, 1968), gross antigen-a soluble antigen common to all murine leukaemia viruses (Steinberg et al., 1969)-and anti-gross anti- gen (Mellors et al., 1968). A viral infection in new- born mice, especially if the causative agent is an RNA virus, sharply hastens the onset and increases the severity of the nephritis (Tonietti et al., 1970). Genetic predisposition, immune hyper-reactivity, and viral infections are all important pathogenic factors. Human SLE glomerulonephritis is also characterized by polymorphic lesions and a granular deposition of immune complexes in glomerular structures. An antigen, DNA, has been demonstrated in glomerular lesions, together with specific antibodies and com- plement (Koffler et al., 1967; Andres et al., 1970). The occurrence of increased titres of serum antibody to measles (Phillips & Christian, 1970) and the obser- vation of endothelial structures similar to nucleo- protein strands of myxoviruses (Gyorkey et al., 1969) have led to the formulation of the hypothesis that in human SLE also a viral infection may influence the development of the disease. EPIDEMIOLOGICAL CONSIDERATIONS IN THE ASSOCIATION BETWEEN P. MALARIAE AND THE NEPHROTIC SYNDROME There is convincing evidence of an association between P. malariae infections and the nephrotic syn- drome in childhood in Africa (Gilles & Hendrickse, 1963; Kibukamusoke et al., 1967; Macfie & Ingram, 1917). It is generally believed that P. malariae infec- tions are causally related to the nephrotic syndrome, but the alternative, that children with the nephrotic syndrome have an increased susceptibility to P. mala- riae infection, has not been completely disproved. Evidence in support of the concept of a causal relationship includes the following: (1) data on the incidence of the nephrotic syn- drome, which indicate that its prevalence is much higher in malarious areas than in non-malarious areas (Hendrickse & Gilles, 1963; Kibukamusoke, 1966); (2) case history studies; and (3) time-trend studies on the incidence of malaria and renal diseaseincommunities (Giglioli, 1930,1962). Further studies are needed for the following reasons: (1) incidence studies were hospital-based and cannot be applied directly to the general popu- lation; (2) it is necessary to define more completely the relationship between P. malariae infections and the nephrotic syndrome in children; and (3) a recent publication (Mandle, 1970) shows that in the 30-year period preceding the eradication of malaria in Guyana there was a decline in deaths from all causes, including renal disease and malaria. Following the eradication of malaria there was a notable acceleration in the decline of deaths from renal disease. A similar decline in tuberculosis was also observed. Epidemiological evidence in support of a causal relationship between P. malariae infec- tions and the nephrotic syndrome needs to be strengthened. 389 MEMORANDA CLINICAL FINDINGS In children, the nephrotic syndrome is associated with P. malariae parasitaemia but parasite prevalence in adult nephrotic patients is similar to that in the general population. Adults show a much higher incidence of hypertension and azotaemia early in the course of the disease than do children, and also show marked proliferative glomerular lesions more frequently. CHILDREN Observations on the disease in Africa Observations on children in East and West Africa show that, irrespective of treatment, the disease may take the following courses: (1) stable remission; (2) transient remission; (3) loss of oedema but persistent symptomless proteinuria; (4) slowly deteriorating renal function and hyper- tension; or (5) rapidly progressive renal failure. The disease in the vast majority of patients follows courses (3) or (4). Stable remission is seldom spon- taneous, but in a minority of patients it can be associated with therapy. Rapidly progressive renal failure (5) is rarely found. Steroid-sensitive patients Differential protein clearances permit the steroid- sensitive minority of patients to be identified (Soothill & Hendrickse, 1967), but it has been noted that steroid response is poor when highly selective pro- teinuria is associated with definite histological changes. Steroid therapy Steroids are ineffective and are associated with severe toxicity in patients with moderately selective proteinuria (Hendrickse, 1966). Among patients with highly selective proteinuria, steroid therapy is usually not associated with toxic effects and stable remissions can be induced in about 50% of the cases. Incidence of P. malariae parasitaemia The incidence of parasitaemia has been reported to be significantly lower in patients with highly selective proteinuria than in those with poorly selec- tive proteinuria (Adeniyi, unpublished data). Treatment Observations in Ibadan, Nigeria, and Kampala, Uganda, indicate that cyclophosphamide and aza- thioprine may induce remission in some cases. How- ever, a controlled clinical trial in Ibadan of cyclo- phosphamide (3 mg per kg of body weight per day) and azathioprine (2.5-3 mg per kg of body weight per day) for 3 months in patients with poorly selec- tive proteinuria did not demonstrate any significant difference in the course of the disease in the two groups (Adeniyi et al., 1970). The observation in Kampala that remission of symptoms may occur in untreated patients (Kibukamusoke, 1968) indi- cates that improvements noted in patients given cyclophosphamide and azathioprine may not be an effect of treatment. Toxicity of azathioprine at the dose mentioned above was considerable but the doses of 1.5-2.5 mg per kg of body weight per day that were used in Kampala caused less toxicity. Cyclophosphamide appeared to cause less toxicity than azathioprine. Immunoglobulin deposits Patients who were found to have immunoglobulin deposits on the GBM in the first renal biopsy, and who subsequently went into remission, showed no evidence of such deposits in subsequent renal biopsies (Adeniyi et al., 1970). Immunofluorescence The different patterns of renal immunofluores- cence described elsewhere in this report do not cor- relate precisely with protein selectivity or outcome, but the diffuse pattern was usually associated with poorly selective proteinuria and a bad outcome. ADULTS The incidence of hypertension (60-70%) and raised serum creatinine (about 30%.) found in Afri- 390 IMMUNOPATHOLOGY OF NEPHRITIS IN AFRICA 391 can adults at the initial examination is higher than in those from non-tropical areas. The incidence of the AS haemoglobin genotype is similar to that of the population in general. The incidence of P. malariae parasitaemia in African adults is similar to that in the control population. In eight patients without parasitaemia, P. malariae antigen was demonstrated in the kid- ney. The five courses that the disease may take in children may occur also in adults. There is an indication that azathioprine may increase the remission rate (Kibukamusoke, 1968); cyclophosphamide, however, has not been used. MORPHOLOGY With regard to the morphological changes in renal biopsies of children with the nephrotic syndrome, the frequency of the different types differed from that described in non-malarious areas. Kidneys showing minimal changes, and proliferative and membranous nephritis were not common. In an extensive " blind " study of biopsy specimens from nephrotic children in Ibadan,' by far the most common lesion seen was one in which the essential component was a localized or diffuse thickening in the capillary wall of the tuft. The lesion was associated with periodic-acid-Schiff's reagent (PAS)-positive segmental sclerosis of peri- pheral capillary loops and with mesangial cell in- crease. Mesangial cell proliferation (segmental) has been considered by some workers to be an inconstant feature. These lesions appeared to progress to total glomerular sclerosis. In a number of repeated biop- sies the lesions observed were gross thickenings of the basement membrane with segmental sclerosis of capillary loops without proliferative changes. Electron microscope studies revealed the segmental fusion of foot processes of the epithelial cells with thickenings of the lamina densa and aggregations of similar material of varying densities on the sub- endothelial surface of the basement membrane. These thickenings extended into the lumen and pro- duced marked irregularities of the endothelial sur- face. Within the thickened basement membrane, there were also small lacunae that often contained electron-dense material. Whether these deposits represent antigen-antibody complexes remains to be determined by immunological electron microscope studies. In adults the histopathological pattern of the renal biopsies of patients suffering from the nephrotic syndrome differed from that seen in children. The most common lesion was a proliferative glomerulo- nephritis characterized by a swelling of the glo- merular tuft, proliferation of the endothelial cells, and occasional lobulation. A few patients in the adolescent age group showed severe sclerozing lesions. IMMUNOLOGICAL STUDIES IMMUNOFLUOESCENCE MICROSCOPY A total of 82 renal biopsies from nephrotic patients in Ibadan (50 children and 32 adults) and 30 from patients in Kampala were examined by fluorescence microscopy. The vast majority of patients showed deposits of immunoglobulin and complement along the capillary walls in the glomeruli. Differences in pat- tern, size, and distribution of deposits were reported. In most of the patients the fluorescence was typically 1 Carried out by D. E. F. Glasgow and R. H. R. White, The Childrens' Hospital, Birmingham, England. granular (coarse and medium deposits), in others there was apparently diffuse fluorescence (?fine depos- its). A typical linear pattern of fluorescence was very rare, 2 cases only being seen (Houba et al., 1971). The immunoglobulins involved were of both clas- ses G and M; IgA was found only exceptionally, while IgD and IgE were not tested (Houba et al., 1971). There was no special prevalence of any indi- vidual IgG subclass in the granular pattern of stain- ing, and the deposits were mostly complement- positive; the sublass IgG2 was found to be pre- 8 392 MEMORANDA dominant in the diffuse pattern of staining and these deposits were mostly complement-negative. The nature of the antigens P. malariae antigens were demonstrated in 30% of children and adults by immunofluorescence stud- ies of tissue sections; P. falciparum and streptoly- sine 0 were each detected in only one case (Houba et al., 1970, 1971; Ward & Kibukamusoke, 1969). More than 40% of patients showed positive stain- ing for immunoglobulins within tubular cells but only about half of them were also positive for complement and P. malariae antigen. Two different patterns of fluorescence were noticed within the cytoplasm of tubular cells-namely, coarse granular fluorescence, which may represent a reabsorption phenomenon, and diffuse fluorescence, which may be an autoimmune reaction to tubular antigens (Houba et al., 1971). Repeated biopsies Biopsies were obtained from 18 patients, 10-15 months being allowed to elapse between biopsies. Marked clinical improvement was closely associated with the disappearance ofimmunoglobulins and com- plement bound in tissue. Patients with poor response or none to therapy with anti-malarial drugs, steroids, and cyclophosphamide showed no significant change in positivity and/or pattern of glomerular fluores- cence. ELUATES FROM KIDNEY SPECIMENS The examination of 24 eluates from nephrotic kid- ney specimens (4 post-mortem samples, 20 open biopsies) confirmed the presence of immunoglobu- lins G and M by precipitation and the presence of antibodies to P. malariae by indirect fluorescence tests, using blood smears from malaria-infected mon- keys of the genus Aotus as antigen. In two cases, antibodies to cytoplasmatic antigens of tubular cells were demonstrated by indirect fluorescence tests on normal kidney tissue sections. Antibodies to glo- merular basement membrane antigen were not found. ELECTRON-MICROSCOPY In order to determine the differences between pat- terns of glomerular immunofluorescence, selected biopsy specimens from children and adults at Ibadan were examined for the presence of deposits by electron microscopy (Allison et al., 1969; Houba et al., 1970).1 In 4 out of 6 specimens (2 from children, 2 from adults) with granular fluorescence, the deposits of electron-dense material were found beneath the epi- thelium or within the basement membrane, or in both sites. In 8 biopsy specimens (5 from children, 3 from adults) with diffuse or mixed patterns of glomerular fluorescence the deposits of electron-dense material were found most frequently at the endothelial surface of the thickened basement membrane (sometimes totally occluding the capillary lumen) or within the basement membrane (see also Morphology, p. 391). SEROLOGY A wide range of immunoglobulin levels was found in serum; IgM levels were generally elevated, espe- cially in patients with an active phase of the disease. IgA levels were within the normal range or slightly elevated. A marked decrease of IgG reported in one series of adult patients was not observed in a second series. Levels of the C3 component of complement were normal or occasionally raised. Some children with disease of recent onset showed low values. Electrophoretically altered beta-l-C, pointing to in vivo activation, was detected in 8 of 11 patients (Soothill & Hendrickse, 1967). The indirect fluorescence antibody technique was applied to estimate plasmodial antibodies in neph- rotic sera and controls. One study (Kibukamusoke et al., 1967) indicated significantly higher titres to P. malariae among adult cases when compared with carefully matched controls. In another study, how- ever, no such difference has yet been found. 1 In collaboration with A. C. Allison and S. de Petris, Clinical Research Centre, Northwick Park Hospital, Harrow, England, and J. R. Goodman, Veterans Administration Hospital, and C. Piel, Department of Pediatrics, University of California Medical Center, San Francisco, Calif., USA. IMMUNOPATHOLOGY OF NEPHRITIS IN AFRICA THE POSSIBLE ROLE OF INFECTIONS OTHER THAN MALARIA IN THE PATHOGENESIS OF NEPHROTIC SYNDROME IN TROPICAL AREAS The high incidence of poorly selective proteinuria and of poor steroid response of children in India with nephrotic syndrome where there is no P. mala- riae (Chandra et al., 1970) suggests that other infec- tions prevalent in that area might be contributing to the incidence. There have also been reports from Brazil of an association between Schistosoma mansoni infections and histological evidence of nephropathy compatible with soluble-complex disease (Andrade et al., 1971; Silva et al., 1970; Brito et al., 1969). Eight young adult patients with the nephrotic syndrome were studied in Cairo (Bassily, personal communication); the patients had both chronic Salmonella infections and Schistosoma mansoni infections of the gastrointestinal tract (but no S. haematobium infection). It is not known whether either of the infections led to the nephro- pathy but the patients showed a marked improve- ment in renal function with diminution or dis- appearance of proteinuria when they were treated with ampicillin and niridazole. The possible role of other infections in the pathogenesis of nephro- tic syndrome associated with malaria must also be considered. PATHOGENESIS OF MALARIAL NEPHRITIS Immunofluorescence and electron microscope studies on malarial nephritis have given results con- sistent with the hypothesis that this disease is elicited by the deposition of immune complexes in the glo- meruli. The following immunological findings and considerations appear relevant to the pathogenesis of this disease. PLASMODIAL ANTIGEN Data available at present indicate that P. malariae, but not P. falciparum, is involved in the pathogene- sis of malarial nephritis. This observation shows some similarity to post-streptococcal nephritis, in which some types only of streptococci (nephritogenic strains) are involved. Evidence for the participation of a plasmodial antigen in the formation of immune complexes was obtained from immunofluorescence studies, which demonstrated such antigen in deposits on the GBM (Houba et al., 1970; Ward & Kibukamusoke, 1969). Significantly, the staining pattern for P. malariae antigen or antigens was found to be similar to the pattern for immunoglobulin and complement. The evidence for the presence of plasmodial antigen in deposits should be strengthened by further experi- ments; specifically, the possibility that there is non- specific binding of the conjugate should be excluded as convincingly as possible. ANTI-PLASMODIAL ANTIBODIES Immunofluorescence studies have shown the pres- ence of IgG and/or IgM in the deposits on GBM. Evidence has been presented that at least some of these immunoglobulins are actually antibodies to P. malariae. It has been demonstrated that low-pH eluates of renal tissue from cases of malarial nephri- tis contain immunoglobulins combining specifically with antigens of P. malariae (Houba et al., 1970). Further elution studies should strengthen this evi- dence. Specifically, the possibility should be excluded that antibody in the eluates is serum antibody trap- ped in the tissue along with other serum components. COMPLEMENT Immunofluorescence studies have shown the pres- ence of complement components in the deposits on GBM in many cases of malarial nephritis. Comple- ment probably plays a role in pathogenesis. On the other hand, staining for complement components was negative in several instances, and this was cor- related with the presence of IgG2 immunoglobulin in the deposits. Therefore, the possibility of non- complement-mediated injury should also be con- sidered. NATURAL HISTORY It appears very likely that malarial nephritis is initi- ated by the formation of immune complexes com- posed of some soluble plasmodial antigen and its corresponding antibody. It is interesting to note that evidence for the presence of such complexes in the circulation has come from cases of malaria caused by P. falciparum, but not so far in cases caused by P. malariae. The complexes are deposited on the 393 394 MEMORANDA GBM and this is followed in most instances by the binding of complement. With the progress of the disease, immune complexes remain on the GBM and possibly become larger. The disease has a chronic character in which it resembles lupus nephritis. The factors responsible for the chronic character of malarial nephritis are poorly understood but the following possibilities may be considered. (1) The chronicity is the result of constant avail- ability of the antigen, similar to the availability of DNA in lupus nephritis, and this allows complexes to be formed continuously. The majority of malarial nephritis cases are not affected by antimalarial therapy; therefore, the chronicity of the disease is not likely to depend on the constant availability of plasmodial antigen. Speculations could be made on the possible role of autologous antigen or antigens in perpetuating the formation and deposition of immune complexes. Formation of autoantibodies could be stimulated by an autologous tissue antigen released as a result of plasmodium-inflicted damage. Otherwise, anti- bodies formed in response to a plasmodial antigen may be autoantibodies in that they may cross-react with some autologous antigens. Autoantibody formation seems to occur in the course of malaria, as indicated by demonstrations of Wassermann antibodies and antibodies to an antigen located within the erythrocyte membrane (but not on the erythrocyte surface) (Kano et al., 1968). It should be stressed, however, that no evi- dence for an autologous immune complex in malarial nephritis is available at present. (2) The persistence of immune complexes could be due to low-affinity antibodies participating in their formation (Christian, 1970; Soothill & Steward, 1971). Low-affinity antibodies would be very ineffi- cient in eliminating immune complexes through the reticuloendothelial system. They would tend to form complexes that would remain in the circulation for a long time. If this hypothesis is true, patients who respond with low-affinity antibodies to P. malariae would develop malarial nephritis, whereas those responding with the formation of high-affinity anti- bodies would not develop nephritis. Further experi- ments will be required to provide the evidence to support this hypothesis. (3) The formation of immune complexes may elicit a self-perpetuating immune process, a " vicious cycle ". Antibody molecules undergo molecular transformation in the course of the reaction with the corresponding antigen. This exposes hidden antigenic sites which may stimulate the formation of either IgG or IgM antiglobulin antibodies or of both classes. Complement molecules altered in their reaction with immune complexes may stimulate the forma- tion of immunoconglutinin. Once this process has started immune complexes may persist, and even increase, in the absence of plasmodial antigen. It is consistent with the hypothesis that as the disease progresses it becomes more difficult to demonstrate plasmodial antigen by immunofluorescence methods. There is, however, no experimental evidence for the hypothesis. OTHER FACTORS Environmental factors (including malnutrition and/or other infections) as well as genetic factors may play some role in predisposing a patient to malarial nephritis. Some evidence for a higher inci- dence of this disease in the lower socioeconomic strata of the population has been presented. Forma- tion of low-affinity antibody, if such an antibody is indeed involved in the pathogenesis of malarial nephritis, may be genetically conditioned. Injury to the reticuloendothelial system occurring in the course of malaria may result in altered immunological reac- tivity, possibly predisposing the patient to the devel- opment of the renal lesion. TUBULAR LESION In many cases of malarial nephritis a tubular lesion was demonstrated by positive immunofluo- rescent staining for immunoglobulin of the cyto- plasm of proximal tubular cells (Houba, personal communication). This lesion was apparently induced by an antibody to the antigen characteristic for cells of proximal convoluted tubules. No information is available about the relation between glomerular and tubular lesions; it is possible that they are com- pletely independent pathological processes. Alter- natively, the participation of tubular antigen in the formation of glomerular deposits should be con- sidered. * * A. Adeniyi, Institute of Child Health, University College Hospital, Ibadan, Nigeria 0. 0. Akingkugbe, Faculty of Medicine, University of Ibadan, Ibadan, Nigeria IMMUNOPATHOLOGY OF NEPHRITIS IN AFRICA 395 G. A. Andres, Department of Pathology and Micro- biology, State University of New York at Buffalo, Buffalo, N.Y., USA S. Bassily, US Naval Medical Research Unit No. 3, Cairo, Arab Republic of Egypt T. Brunner, Institut Suisse de Recherches experimentales sur le Cancer, Lausanne, Switzerland G. M. Edington, Department of Morbid Anatomy, Uni- versity of Ibadan, Ibadan, Nigeria W. P. Faulk, Medical Officer, Immunology, World Health Organization, Geneva, Switzerland H. C. Goodman, Chief Medical Officer, Immunology, World Health Organization, Geneva, Switzerland R. G. Hendrickse, Liverpool School of Tropical Medi- cine, Liverpool, England V. Houba, WHO Immunology Research and Training Centre, Ibadan, Nigeria J. W. Kibukamusoke, Department of Medicine, Makerere University, Kampala, Uganda F. Milgrom, Department of Microbiology, State Uni- versity of New York at Buffalo, Buffalo, N.Y., USA T. 0. Ogunlesi, Department of Medicine, University of Ibadan, Ibadan, Nigeria A. B. 0. 0. Oyediran, Department of Preventive and Social Medicine, University of Ibadan, Ibadan, Nigeria J. F. Soothill, Department of Immunology, Institute of Child Health, London, England M. A. 0. Soyannwo, Department of Medicine, Univer- sity College Hospital, Ibadan, Nigeria RESUME IMMUNOPATHOLOGIE DE LA NEPHRITE EN AFRIQUE Le present document expose les resultats obtenus a l'aide des techniques immunologiques modernes dans l'etude de la pathogenie du syndrome nephrotique associe au paludisme en Afrique orientale et occidentale. La demonstration du r6le joue par les r6actions anti- gene-anticorps dans I'apparition de cette forme d'affec- tion renale si frequente en Afrique pourrait etayer l'hypothese selon laquelle des mecanismes immunolo- giques similaires seraient a l'origine du syndrome dans d'autres regions du monde. On examine d'abord deux des principaux mecanismes immunopathologiques intervenant dans la glomerulo- nephrite experimentale et les processus similaires respon- sables de la maladie chez l'homme. Dans certains cas, la maladie resulte d'une production d'anticorps qui reagissent avec la couche sous-epitheliale du glomerule. Dans d'autres, il y a production d'anticorps qui reagissent dans le sang circulant avec des antigenes endogenes ou exogenes non glomerulaires; les immuncomplexes ainsi form6s sont retenus dans le rein. Les alterations morpho- logiques du tissu renal dans les syndromes nephrotiques associes au paludisme, telles qu'elles apparaissent a l'examen au microscope optique ou electronique, peuvent etre reproduites experimentalement. Elles consistent notamment en un epaississement de la couche sous-epi- theliale et en un dep6t de materiel opaque aux electrons dans les glomerules. On a obtenu des preuves indiscutables d'un lien entre les infections a Plasmodium malariae et le syndrome nephrotique chez les enfants en Afrique. On admet gene- ralement l'existence d'une relation de cause a effet entre le paludisme a P. malariae et la nephropathie, mais une autre hypothese, attribuant aux enfants atteints du syn- drome une receptivite accrue au parasite n'a pas ete refutee. Chez les enfants, le syndrome nephrotique est associe a la parasitemie a P. malariae; chez les adultes, la prevalence de la parasitemie est similaire chez les sujets presentant le syndrome et dans 1'ensemble de la population. Du point de vue clinique, l'atteinte renale se manifeste plus frequemment chez les adultes que chez les enfants par de 1'hypertension et une azotemie pre- coces et par des 1lsions glomerulaires proliferatives. Le traitement par les corticosteroldes se montre efficace dans 50% des cas environ chez les enfants presentant une proteinurie tres selective. La cyclophosphamide et 1'aza- thioprine (plus toxique) peuvent entrainer dans certains cas une regression des sympt6mes, mais de telles remis- sions sont aussi observees en l'absence de traitement. L'etude par immunofluorescence de biopsies renales prelevees chez des sujets presentant un syndrome nephro- tique et une parasitemie a P. malariae revele la presence de dep6ts d'immunoglobulines et de certains composants du complement dans la couche sous-epitheliale du glomerule. L'examen d'eluats de tissu renal recueilli a l'autopsie ou par biopsie confirme la presence d'immuno- globulines G et M specifiques de P. malariae, et l'im- munofluorescence pratiquee a l'aide de serums humains anti-P. malariae decele l'antigene plasmodique dans la couche sous-epitheliale du glomerule. Ces observations etayent l'hypothese selon laquelle la nephropathie asso- ciee a l'infection a P. malariae est due au dep6t dans les glomerules d'immuncomplexes formes dans le sang circu- lant par la reaction entre l'antigene P. malariae et l'anticorps correspondant. Un autre argument en faveur de cette pathogenie est fourni par la microscopie elec- tronique qui indique la presence de materiel opaque aux electrons dans la couche sous-epitheliale dans certaines affections renales oui l'immunofluorescence, pour sa part, montre le dep6t d'immuncomplexes dans les glomerules. Le fait de reconnaitre a la nephropathie paludeenne le caract&e d'une maladie liee a la formation d'immun- complexes devrait permettre d'aborder sur de nouvelles bases 1'etude des mecanismes declenchant et entretenant les troubles immunopathologiques et augmenter les chances de mettre au point une therapeutique plus efficace. 396 MEMORANDA REFERENCES Adeniyi, A., Hendrickse, R. G. & Houba, V. (1970) Lancet, 1, 644 Allison, A. C., Houba, V., Hendrickse, R. G., Petris, S. de, Edington, G. M. & Adeniyi, A. (1969) Lancet, 1, 1232 Andrade, Z. A., Andrade, S. G. & Sadigursky, M. (1971) Amer. J. trop. Med. Hyg., 20, 1, 77 Andres, G. A., Accinni, L., Beiser, C. L., Ginotti, G. A., Erlanger, B. G., Hsu, K. C. & Seegal, B. C. (1970) J. clin. Invest., 49, 2106 Andres, G. A., Accinni, L., Hsu, K. C., Zabriskie, J. N. & Seegal, B. C. (1966) J. exp. Med., 2, 399 Brito, T. de, Boni, D. de, Lopes, J. D. & Silva, L. C. da (1969) Rev. Inst. Med. trop. S. Paulo, 11, 62 Chandra, R. K., Manchanda, S. S., Srivastava, R. N. & Soothill, J. F. (1970) Arch. Dis. Childh., 45, 491 Christian, C. L. (1970) Immunology, 18, 457 De la Pava, S., Nigogosyan, G. & Pickren, J. W. (1962) Arch. intern. Med., 109, 391 Dixon, F. J. (1971) In: R. A. Good & D. W. Fisher, ed., Immunobiology, Stamford, Conn., Sinauer Associa- tion, p. 167 Dixon, F. J., Feldman, J. D. & Vazquez, J. J. (1961) J. exp. Med., 113, 899 Feldman, J. D. (1958) J. exp. Med., 108, 957 Giglioli, G. (1930) Malarial nephritis, London, Churchill Giglioli, G. (1962) Ann. trop. Med. Parasit., 56, 225 Gilles, H. M. & Hendrickse, R. G. (1963) Brit. med. J., 2, 27 Gyorkey, F., Min, K. W., Sincorics, J. G. & Gy6rkey, P. (1969) New Engl. J. Med., 280, 333 Hendrickse, R. G. (1966) Clin. Pediat. (Phila), 5, 315 Hendrickse, R. G. & Gilles, H. M. (1963) E. Afr. med. J., 40, 186 Heymann, W., Hackel, D. B., Harwood, S., Wilson, S. G. & Hunter, J. L. P. (1962) Proc. Soc. exp. Biol. (N.Y.), 111, 568 Houba, V., Allison, A. C., Adeniyi, A. & Houba, J. (1971) Clin. exp. Immunol., 8, 761 Houba, V., Allison, A. C., Hendrickse, R. C., de Petris, S., Edington, G. M. & Adeniyi, A. (1970) In: J. Bonomo & J. L. Turk, ed., Immune complex diseases, Milan, Carlo Erba Foundation, p. 23 Kano, K., McGregor, I. A. & Milgrom, F. (1968) Proc. Soc. exp. Biol. Med., 129, 849 Kano, K. & Milgrom, F. (1971) Int. Arch. Allergy, 40,471 Kibukamusoke, J. W. (1966) W. Afr. med. J., 15, 213 Kibukamusoke, J. W. (1968) Arch. Dis. Childh., 43, 598 Kibukamusoke, J. W., Hutt, M. S. & Wilks, N. E. (1967), Quart. J. Med., 36, 393 Koffler, D., Schur, P. H. & Kunkel, H. G. (1967) J. exp. Med., 126, 607 Lambert, P. H. & Dixon, F. J. (1968) J. exp. Med., 127, 507 Lerner, R. A., Glassock, F. J. & Dixon, F. J. (1967) J. exp. Med., 126, 989 Mandle, J. R. (1970) Demography, 7, 301 Markowitz, S., Armstrong, S. H. & Kushner, D. S. (1960) Nature (Lond.), 187, 1095 Macfie, J. W. S. & Ingram, A. (1917) Ann. trop. Med. Parasit., 11, 1 Mellors, R. C., Aoki, T. & Heubner, R. J. (1968) J. exp. Med., 129, 1045 Oldstone, M. B. A. & Dixon, F. J. (1967) Science, 158, 1193 Phillips, P. E. & Christian, C. L. (1970) Science, 168, 982 Seegal, B. C., Accini, L., Andres, G. A., Beiser, S. M., Christian, C. L., Erlanger, B. G. & Hsu, K. C. (1969) J. exp. Med., 130, 203 Silva, L. C. de, Brito, T. de, Boni, D. de, Camargo, M. E., Lopes, J. D. & Gunji, J. (1970) Bull. Wld Hlth Org., 42, 907-910 Soothill, J. F. & Hendrickse, R. G. (1967) Lancet, 1, 629 Soothill, J. F. & Steward, M. W. (1971) Clin. exp. Immunol., 9, 193 Steblay, R. W. (1962) J. exp. Med., 116, 253 Steinberg, A. D., Baron, S. H. & Talal, N. (1969) Proc. nat. Acad. Sci. (Wash.), 63, 1102 Thunold, S. et al. (1970) J. Immunol., 104, 685 Tonietti, G., Oldstone, M. B. A. & Dixon, F. J. (1970) J. exp. Med., 132, 89 Unanue, E. R. & Dixon, F. J. (1967) Advanc. Immunol., 6, 1 Ward, P. A. & Kibukamusoke, J. W. (1969) Lancet, 1, 283 Wilson, C. B., Dicon, F. J., Fortner, J. G. & Cerilli, G. J. (1971) J. clin. Invest., 50, 1525