Pediatric Annals

Rheumatic Fever: A Streptococcal-Induced Autoimmune Disease?

John B Zabriskie, MD

Abstract

While the incidence of rheumatic fever has markedly decreased in this country- as well as in most of the industrialized world- one need only visit any of the Third World countries to realize that it is still one of the world's major health problems. If the incidence of rheumatic fever is approximately one per 1000, and during epidemics one per 1 00, one can estimate 1 to 5 million new cases of rheumatic fever each year in these populations. I believe that the incidence is actually higher, since many cases are not initially diagnosed, but appear later as chronic rheumatic heart disease. Perhaps most important from a public health point of view is the fact that most patients survive the acute carditis and then go on to develop chronic valvular disease, with its attendant severe cardiologie manifestations and complications.

What is especially intriguing about this disease is that it is changing or disappearing in many parts of the world. Better medical conditions, the use of penicillin, and improved socioeconomic factors have obviously contributed to this decline; it began, however, before the widespread use of antibiotics and is occurring in parts of the world where medical and social conditions are less than optimal. Thus, there has been a subtle change in the organism's ability to cause these sequelae, in the host's susceptibility to the disease, or a combination of these factors.

It is surprising that, after 50 years of intensive investigation, we still do not have a clear picture of the pathogenesis of rheumatic fever, despite the current state of knowledge concerning its relationship to the group A Streptococcus. There are several reasons for this dilemma: 1) The latent period between the initial streptococcal infection and the onset of the clinical and pathological signs and symptoms of rheumatic fever, which severely limits prospective studies; 2) the myriad of cellular and extracellular structures and toxins associated with the streptococcal organism that may produce tissue injury, directly or indirectly (Table 1); and 3) the lack of a suitable experimental model for studying the disease process. Nevertheless, rheumatic fever needs study, not only for its public health implications, but more importantly, because knowledge of this disease might provide insights into other rheumatic diseases where even the causative agent (let alone etiology) is still not known.

Thus, it is the major purpose of this review to reexamine the theories concerning the pathogenesis of rheumatic fever, including elucidation of the relevant structural components and products of the Streptococcus, and the host's immunological response to them; theories regarding the pathogenesis of rheumatic fever, emphasizing the immune response of certain individuals to recurrent streptococcal infections; and relationship of these concepts to the clinical and laboratory findings of the disease.

STREPTOCOCCAL BACTERIOLOGY AND IMMUNOLOGY

Each streptococcal cell is surrounded by several layers (Figure 2). The M-protein, seen on the outer surface as fimbria, is the most extensively studied protein antigen. This antigen and the T-protein serve as markers for the immunological subclassification of Group A Streptococcus into specific types. The next layer consists of carbohydrate and contains the specific portion responsible for grouping of streptococci. In the case of group A Streptococcus this is N-acetylglucosamine. Beneath this is a mucopeptide layer and, finally, there is the protoplast membrane, a lipoprotein structure containing several interesting components which are cross-reactive with mammalian tissue antigens. These major components make up the cell wall of the Streptococcus.

Table

1. Manjula BN and Fischelli VA. Studies on group A streptococcal Mproieins; Purification of type 5 M-protein and comparison of its amino acid terminal sequence with two immunologically unrelated M -protein molecules J immunol 124:261-267,…

While the incidence of rheumatic fever has markedly decreased in this country- as well as in most of the industrialized world- one need only visit any of the Third World countries to realize that it is still one of the world's major health problems. If the incidence of rheumatic fever is approximately one per 1000, and during epidemics one per 1 00, one can estimate 1 to 5 million new cases of rheumatic fever each year in these populations. I believe that the incidence is actually higher, since many cases are not initially diagnosed, but appear later as chronic rheumatic heart disease. Perhaps most important from a public health point of view is the fact that most patients survive the acute carditis and then go on to develop chronic valvular disease, with its attendant severe cardiologie manifestations and complications.

What is especially intriguing about this disease is that it is changing or disappearing in many parts of the world. Better medical conditions, the use of penicillin, and improved socioeconomic factors have obviously contributed to this decline; it began, however, before the widespread use of antibiotics and is occurring in parts of the world where medical and social conditions are less than optimal. Thus, there has been a subtle change in the organism's ability to cause these sequelae, in the host's susceptibility to the disease, or a combination of these factors.

It is surprising that, after 50 years of intensive investigation, we still do not have a clear picture of the pathogenesis of rheumatic fever, despite the current state of knowledge concerning its relationship to the group A Streptococcus. There are several reasons for this dilemma: 1) The latent period between the initial streptococcal infection and the onset of the clinical and pathological signs and symptoms of rheumatic fever, which severely limits prospective studies; 2) the myriad of cellular and extracellular structures and toxins associated with the streptococcal organism that may produce tissue injury, directly or indirectly (Table 1); and 3) the lack of a suitable experimental model for studying the disease process. Nevertheless, rheumatic fever needs study, not only for its public health implications, but more importantly, because knowledge of this disease might provide insights into other rheumatic diseases where even the causative agent (let alone etiology) is still not known.

Thus, it is the major purpose of this review to reexamine the theories concerning the pathogenesis of rheumatic fever, including elucidation of the relevant structural components and products of the Streptococcus, and the host's immunological response to them; theories regarding the pathogenesis of rheumatic fever, emphasizing the immune response of certain individuals to recurrent streptococcal infections; and relationship of these concepts to the clinical and laboratory findings of the disease.

STREPTOCOCCAL BACTERIOLOGY AND IMMUNOLOGY

Each streptococcal cell is surrounded by several layers (Figure 2). The M-protein, seen on the outer surface as fimbria, is the most extensively studied protein antigen. This antigen and the T-protein serve as markers for the immunological subclassification of Group A Streptococcus into specific types. The next layer consists of carbohydrate and contains the specific portion responsible for grouping of streptococci. In the case of group A Streptococcus this is N-acetylglucosamine. Beneath this is a mucopeptide layer and, finally, there is the protoplast membrane, a lipoprotein structure containing several interesting components which are cross-reactive with mammalian tissue antigens. These major components make up the cell wall of the Streptococcus.

Table

TABLE 1EXTRACELLULAR PRODUCTS OF HEMOLYTIC STREPTOCOCCI

TABLE 1

EXTRACELLULAR PRODUCTS OF HEMOLYTIC STREPTOCOCCI

* M-protelns

From several points of view the M-proteins are among the most important of the surface antigens. There are over 80 serologie types of M-proteins, each of which is capable of stimulating specific protective opsonizing and precipitating antibodies. In general, each streptococcal strain possesses only one type of M-protein. M-proteins are associated with virulence, probably because they impede phagocytosis. Only a few M-proteins are associated with nephritogenicity and glomerulonephritis. However, no specific association exists between the M-type and rheumatic fever.

M-protein molecules bear a remarkable structural homology to a group of muscle proteins, which indicates that cross-reaction between these two proteins might be important in the pathogenesis of rheumatic fever.' The myriad of type-specific proteins, each with a type-specific immune response, makes the all-purpose vaccine difficult to envision, though potentially feasible.2

* Carbohydrate

Synthetic carbohydrate antigens coupled to various proteins strongly cross-react with group A streptococcal antisera.3 Certain mammalian tissues contain antigens that appear to be structurally similar to those of the streptococcal group A carbohydrate.4 Absorption studies have shown that cross-reactivity is limited specifically to the group-specific carbohydrate.5 However, these findings have not been confirmed.6

* Mucopeptlde

Current work suggests that cell-wall mucopeptide may be partly responsible for chronic, remittent, nodular lesions of connective tissue following a single injection of disrupted Group A streptococci.7 However, the histologie appearance of these lesions is unlike the hallmark lesion of rheumatic fever, Aschoffs nodule. Injection of this material in rats causes a chronic, relapsing arthritis and synovitis reminiscent of human rheumatoid arthritis.8

* Cell Membranes

The final inner layer of the streptococcal cell is a highly complex antigen lipoprotein; it contains approximately 72 percent protein, 25 percent lipid, and 2 percent carbohydrate by weight. This structure does not contain cell-waJl carbohydrates and its antigens are quite different from those in the streptococcal cell wall.

Group A streptococcal membrane structures and mammalian tissues share a number of common antigenic determinants. Recent findings may be summarized as follows: 1) Cell membranes and extracts from Group A streptococci cross react with human glomerular basement membrane antigens;9 2) rabbit antisera to these streptococcal membrane structures will bind to human muscle sarcolemmic membrane antigens (including cardiac muscle);10'1 v they also bind to the smooth muscle of blood vessel walls but not to uterine muscle; and 3) guinea pigs immunized with a number of different types of Group A streptococci developed sensitivity indistinguishable from that produced by sensitization with allogeneic tissues.12 Antigens shared by streptococcal membrane and mammalian tissue appear to include mammalian histocompatibility antigens. Evidence is derived not only from the above-mentioned experiments of graft rejection, but also from the close biochemical similarity between mammalian histocompatibility antigens, certain structural glycoproteins, and streptococcal membrane antigens.13

* Other Streptococcal Cross- React! ve Antigens

Two streptococcal mammalian tissue reactions in addition to the cross-reactive antigen reactions related to heart or valves are worthy of discussion: 1) Patients with rheumatic chorea possess an antibody that stains caudate nuclei and can be absorbed by streptococcal membrane antigens,14 indicating that streptococcal antigens share antigenic determinants with brain antigens; also, streptococcal antisera stain brain astrocytes;15 and 2) antisera bind to skin fibroblasts as well as to thymocytes,16 The latter cross-reactivity could be important in the host's immunoregulation to streptococcal antigens.

In spite of the obvious interest created by these numerous cross-reactions, a word of caution should be introduced: Group A streptococci, as do tissue antigens, have Ig molecule Fc receptors on their surface. ' 7 Many of these cross-reactions are real, but each report must be carefully studied with proper controls.

* Extracellular Products

A number of substances with biologic importance are produced by hemolytic streptococci - several with potent toxic properties. A list of the known extracellular products of group A streptococci is presented in Table I, but there are undoubtedly other as yet unidentified substances. (For instance, at least 20 extracellular antigens in group A streptococcal culture filtrates have been demonstrated by immunoelectrophoresis.18

Many extracellular products are directly cytotoxic to mammalian cells. For example streptolysin "O" causes a loss in myocardial contractility when injected into experimental animals19 and both streptolysin "S" and "O" cause hemolysis. Erythrogenic toxins can cause (scarlet) fever and have important biological effects on the immune system.20 Proteinase is also cardiotoxic in experimental animals.21 Since all group A streptococci can produce these toxins, and all infected humans develop an immune response to them, it appears likely that toxins are primarily involved in the disease process, although they may initiate or augment an immunologie process that leads to rheumatic fever.

Figure 1. Some known side effects of Streptococcus antigens on human tissues.

Figure 1. Some known side effects of Streptococcus antigens on human tissues.

Figure 2. Diagrammatic representation of the subcellular components of hemolytic streptococci and some of their known cross reactions with human tissue components.

Figure 2. Diagrammatic representation of the subcellular components of hemolytic streptococci and some of their known cross reactions with human tissue components.

PATHOGENESIS OF RHEUMATIC FEVER

* Theoretical Considerations

Confronted with this imposing array of cellular and extracellular products of the group A Streptococcus, it is surprising that knowledge of the exact pathological mechanisms leading to rheumatic fever has advanced so slowly. The accumulated evidence can be conveniently grouped into three theories of pathogenesis (Table 2): 1) An unusual persistence of streptococcal infection; 2) a direct toxic reaction by sireptococcal products; and 3) an abnormal immune response to the initial streptococcal infection. Each of these theories will be discussed.

Persistent Infection

From a pathologic point of view, the cardiac lesions of rheumatic fever do not have the appearance of an infectious process, nor are streptococci visible on histologie examination. Scattered reports of the isolation of group A Streptococcus from heart valves of patients dying with acute rheumatic fever have not been confirmed." Support for this view of streptococcal persistence comes from studies that prevention of rheumatic fever was successful only when sufficient penicillin was administered to eradicate streptococcus from the pharynx at the time of the pharyngitis.21 In contrast, treatment of streptococcal pharyngitis with sulfonamides did not reduce subsequent attacks of rheumatic fever, presumably because these drugs do not effectively eradicate the organism. As a result of these studies, intensive penicillin therapy for a sixweek period, similar to that employed for subacute bacterial endocarditis, has been utilized for the treatment of rheumatic fever in an effort to eliminate any possible latent infection and to terminate the rheumatic process. Extensive clinical trial provided no evidence that this treatment resulted in clinical improvement of the acute disease or reduced the development of permanent valvulities.24

Toxic Reaction

Particular efforts were undertaken to determine the possible importance of the extracellular enzymes and products mentioned in the preceding section on the pathogenesis of rheumatic fever.25 Many of the toxins have cardiotoxic properties, but there are several observations which make it difficult to evaluate their roles; with the exception of proteinase, few have been obtained in pure form. Thus, an observed biologic or toxic property may represent the actual property of the toxin or could be the property of some other, unidentified, streptococcal product. Recent discovery that the mitogenic properties of streptolysin "S" are not a property of the toxin itself, but of another streptococcal product in streptolysin "S" preparations, illustrates this point.26

Moreover, the histologie damage produced by these toxins does not mimic the chronic granuloma lesions seen in rheumatic fever. A variety of cellular and extracellular structures may cause necrosis and vasculitis nodules when injected into experimental animals; yet these lesions are not analogous to the ones seen in rheumatic fever. It is possible that these toxins do not cause rheumatic fever directly but may play a role in initiation of the pathologic events in the disease.

An Abnormal Immune Response

Most investigators concerned with the pathogenesis of rheumatic fever now favor the concept that it is the result of abnorma"! host humoral and /or cellular immune responses to the streptococcal infection.

For instance, rheumatic fever patients have antibodies to streptococcal products such as streptolysin "O" (ASO) higher than non-rheumatics.27 In addition, the sera of patients with acute rheumatic fever and rheumatic heart disease have been shown to contain antibodies that react with constituents of human heart tissue;""" antibodies that are present in very low titers or absent in uncomplicated streptococcal infections. Elevated antibody titers to streptococcal antigens may reflect a greater antigenic challenge, though an asymptomatic streptococcal infection, so mild that the patient cannot recall the symptoms of pharyngitis, can precipitate an attack of rheumatic fever.

Several observations support the idea that cellmediated immunity is also involved in the rheumatic fever process: 1) The rare occurrence of the disease before four years of age suggests that several streptococcal infections are needed to sensitize;30 2) streptococcal antigens can be shown to induce delayed hypersensitivity in both animals and man;31 and 3) examination of human hearts from acute rheumatics reveals lymphocyte infiltrates, both perivascular and between muscle bundles.32

Table

TABLE 2PATHOGENESIS OF RHEUMATIC FEVER: THEORETICAL CONSIDERATION

TABLE 2

PATHOGENESIS OF RHEUMATIC FEVER: THEORETICAL CONSIDERATION

Table

TABLE 3CLINICAL MANIFESTATIONS OF RHEUMATIC FEVER AND THE IMMUNE RESPONSE

TABLE 3

CLINICAL MANIFESTATIONS OF RHEUMATIC FEVER AND THE IMMUNE RESPONSE

The incidence of rheumatic fever, in one to three percent of streptococcal pharyngitis cases, appears to be constant regardless of the area studied. This appears to hold true even in areas such as Trinidad, where the rate of both skin and throat streptococcal infection in the population is extremely high and where periodic epidemics of glomerulonephritis occur. But the yearly attack rate of rheumatic fever remains quite constant, suggesting that there are only a limited number of individuals who are uniquely susceptible to this disease. (The concept of a predisposing genetic factor will be discussed below.)

CORRELATION WITH CLINICAL MANIFESTATIONS

What evidence is there for this preferred concept of abnormal immune response, and how well do these observations fit with the observed signs and symptoms of the disease (Table 3)?

* Humoral Immunity

Carditis

The finding that a component of the group A Streptococcus had the capacity to elicit an antibody which bound to cardiac tissue naturally led to a search for similar antibodies in the sera of patients with recent streptococcal infections and their sequelae. When the serum from a patient with acute rheumatic fever is layered over a cardiac section and stained with a fluor esceinconjugated anti-IgG antiserum, the staining pattern bears a striking resemblance to the pattern observed with rabbit antisera to group A streptococci (Figure 2). In general, the localization of the staining, the limitation to muscle tissue and the lack of species specificity were exhibited by this antibody.

Table

TABLE 4A COMPARISON OF THE AVERAGE TITERS OF CROSS-REACTIVE ANTIBODY IN PATIENTS WITH UNCOMPLICATED SCARLET FEVER, SCARLET FEVER WITH RHEUMATIC FEVER AND OTHER DISEASES

TABLE 4

A COMPARISON OF THE AVERAGE TITERS OF CROSS-REACTIVE ANTIBODY IN PATIENTS WITH UNCOMPLICATED SCARLET FEVER, SCARLET FEVER WITH RHEUMATIC FEVER AND OTHER DISEASES

Examination of a large number of sera (Table 4) from patients with recent streptococcal infections and their sequelae; revealed heart staining antibody in most of them. The amount of antibody detected in rheumatic fever individuals at the onset of their diseases was strikingly different from that in patients with uncomplicated streptococcal infections. Two to three weeks after uncomplicated streptococcal infection, sera from the latter had little or no heart-reactive antibody, while sera from patients with acute rheumatic fever had antibodies detectable at a ten-fold dilution. The presence of these high titers of heart-reactive antibodies has been an important additional diagnostic tool in cases of suspected rheumatic fever. Bright fluorescent staining of cardiac tissue at dilutions of serum greater than l:10is considered indicative of acute rheumatic fever. Furthermore, the absence of any significant levels of heart-reactive antibody in patients with unrelated arthritic and immunological diorders, coupled with the characteristic nuclear staining of sera from lupus erythematosus patients has been helpful in the differential diagnosis of other clinical disorders.29

Serial studies of the sera obtained from patients with acute rheumatic fever reveal that the antibody declines rapidly during the first three to six months after the initial attack; then more gradually over the next two to three years. At the end of five years, the vast majority of patients have little or no detectable antibody present in their serum unless there is an intercurrent streptococcal infection or a recurrence of rheumatic fever.

When viewed in the light of clinical observations that most rheumatic recurrences occur within the first two years after the initial attack, and are rare after five years,33 the presence of a heart-reactive antibody appears to have more than a casual relationship to the disease process.

Observations of long intervals between attacks, and appearance of heart-reactive antibodies in serum prior to recurrence makes it tempting to speculate that repeated streptococcal infections (perhaps with subclinical symptoms of disease) may be necessary to stimulate the production of heart-reactive antibodies. Only when titers are sufficiently elevated does full-blown disease complex appear. Examination of a small number of rheumatic fever recurrences has indicated that, in each case, elevated heart-reactive antibody titers were present at the time of, or just prior to, the second attack.

If this rise in heart-reactive antibody in the serum is related to rheumatic heart disease, then one should be able to demonstrate gamma-globulin in rheumatic hearts. Gamma-globulin has, in fact, been demonstrated as bound to the sarcolemma of hearts from rheumatic fever patients at the site of histological damage. Knowledge concerning the nature and specificity of this gammaglobulin will be crucial to an understanding of the role these antibodies play in the pathologic process.

With respect to valvulitis, the sera of a number of patients with rheumatic fever and rheumatic heart disease have been shown to contain antibodies which bind to Nacetyl glucosamine the carbohydrate specific to group A Streptococcus. Patients with valvular disease have higher titers of this antibody than patients without valvular disease, and these antibodies persist for years after the initial attack. In contrast, in patients without valvular disease, the titers decline rapidly after the initial rheumatic insult. The persistence of high titers in patients with rheumatic valvular disease may be related to the slow and sustained release of valvular cross- react i ve glycoproteins, thus perpetuating the valvular damage. It remains to be seen whether the presence of these antibodies is specific to rheumatic patients or whether they also occur in patients with advancing arteriose lero tic heart disease and other forms of non-rheumatic heart disease.

Arthritis

Swelling of the major joints has always been a prime feature of rheumatic fever. Yet even this manifestation appears to vary with geographic area. It has been our own experience in Trinidad, as well as that of other investigators, that incipient carditis without arthritis is often the major clinical manifestation of the disease. Analysis of the synovial fluid in well-documented rheumatic fever patients with arthritis35 generally reveals a decrease of complement components Ciq, Ci and C4, indicating their consumption by immune complexes in the joint fluid.

More recently, we have noted the presence of high levels of circulating immune complexes in the sera of well-documented acute rheumatic fever patients,36 which persist for several months after the acute attack. Dr. Friedman in our laboratory has isolated these complexes and raised antibodies to them in rabbits. The anticomplex antibody gives a reaction against extracellular products of a rheumatogenic strain distinctly different from that in antisera to nephritis complexes.

Another mechanism for the observed synovitis in rheumatic fever could be antibodies to streptococcal antigens that bind to the smooth muscle of blood vessel walls (see the discussion of other streptococcal crossreactive antigens). Thus, one might envision a vasculitis secondary to streptococcal cross-reactive antibodies, perhaps augmented by the immune complexes circulating in the sera of these patients. Immunoglobulins have been noted in the blood vessel walls of these patients and make this concept attractive, but still unproven.

Chorea

This clinical syndrome has always been enigmatic as a frequent late manifestation of rheumatic fever. The sera of patients with Sydenham's chorea contain an antibody that binds to the cytoplasm of caudate nuclei in the thalamic and subthalamic areas of the brain.'4 This antibody correlates quite well with the clinical course of the disease. The "streptococcal connection" was firmly established by the observation that Group A streptococcal membranes completely abolished immunofloure scent staining, while other streptococcal antigens did not. Unpublished data by our group reveal the presence of this antibody in the spinal fluid in five well-documented cases of rheumatic chorea.

* Cell-mediated Immunity

The question of whether delayed hypersensitivity to hemolytic streptococci and their products may play a role in the pathogenesis of the nonsuppurative sequelae has been the subject of investigation for many years. Using skin tests to indicate delayed sensitivity to streptococcal products, several investigators agreed that hypersensitivity to streptococci and their products was a common occurrence in man, and increased in intensity with the age of the individual tested.37 These reactions were also more intense in rheumatic subjects than in nonrheumatic controls; the greatest number of positive reactions were obtained with autogenous streptococci, suggesting typespecificity to the reaction. Repeated streptococcal infections and exposure to their products may be vital to the disease process, in view of the rarity of rheumatic fever before age three to four years.29

A number of authors have reported studies on cellular reactivity to streptococcal antigens in both normal, healthy subjects31'38 and rheumatic individuals.39140 Albeit with different techniques, most authors noted a generalized cell-mediated response to streptococcal antigens in most individuals tested. The lymphocyte response to streptococcal cellular antigens appeared specific. Chord blood lymphocytes did not respond to these antigens according to one investigator; another investigator found the antigens produced a definite response in these lymphocytes, and he felt that the cellular reactivity, "nonspecific", was akin to that produced by phytohemagglutinin and other mitogens.

During the past five years our laboratory has been reexploring this question of cellular reactivity to streptococcal antigens in patients with the known sequelae of streptococcal infection, rheumatic fever and post-streptococcal glomerulonephritis. This is a unique opportunity for us, since in Trinidad both diseases occur simultaneously in the same age group and there is no seasonal variation in their occurrence.

It was demonstrated in vitro that patients with acute rheumatic fever have an increased cellular response compared with nephritics, primarily to membranes of streptococcal strains commonly associated with rheumatic fever in Trinidad. This reaction persists for at least two years after the initial attack. Since this reactivity was not seen in patients with acute post-streptococcal glomerulonephritis, a specific reactivity to these antigens in rheumatic fever patients is suggested. Moreover, there was no increased reactivity to those streptococcal antigens isolated from strains commonly associated with glomerulonephritis. The nature of the antigen(s) responsible for the observed reactivity is presently unknown.

While these results strongly suggest that there is a heightened response to streptococcal antigens in rheumatic individuals, the exact role played in the disease process by these sensitized cells remains unknown. The present finding that there is abnormal cell-mediated response to membrane antigens, coupled with previous reports of an abnormal humoral response to the streptococcal membrane, argues strongly for a crucial role of this cell structure in the pathogenesis of rheumatic fever. The cross-reactive properties of these antigens might result in autosensitization to tissue antigens with cytotoxic effects in host tissues. This concept agrees with the histologie findings of a large number of lymphocytic cells in and near the pathologic heart lesions of rheumatic fever. There are only two reports that suggest that lymphocytes are in fact cytotoxic for target organs such as the heart."1 42,43

* Genetics

No discussion of rheumatic fever would be complete without a reference to the possibility that rheumatic fever individuals may be genetically programmed to abnormally respond to a streptococcal infection. A little more than 80 years ago, it was pointed out that rheumatic fever frequently occurred in more than one member of an affected family.44 Since then, numerous investigators45 have postulated that there is an inherited susceptibility to rheumatic fever, but neither the mode of inheritance nor the methods of its expression have been satisfactorily delineated. Observations in 56 rheumatic fever patients and their respective co-twins suggest that genetic factors, if operative, have limited penetrance (only three of 16 pairs of monozygotic twins were concordant for rheumatic fever).46 Other studies have concentrated on the presence or absence of selected blood groups and the secretor status of rheumatic subjects in an effort to identify a susceptible genotype.45

While our efforts to find an association with either the HLA-A or B locus were unsuccessful,*7 recent work46 lends strong support to the existence of a genetic marker in rheumatic fever patients. Using a serum containing a Bcell alloantigen (883+), a positive reaction was found in approximately 72 percent of alfrheumatic fever patients from both New York and Bogota, Colombia; this also suggests that the antigen is worldwide. Since 25 percent of the well-documented rheumatics did not react positively, it appears that there is more than one allele for this genetic marker. Screening of this serum against a panel of well-defined, homozygous D-locus cell lines indicates that the antigen is not one of the established B-cell alloantigens such as Dw2 or Dw4. Whether this antigen is present on other cells or acts as a streptococcal antigen receptor is not known at present, nor are we sure how this antigen fits into the pathological picture of rheumatic fever. However, the presence of a genetic marker that indicates a high risk factor for contracting the disease, and that can be identified at birth, has broad preventive medicine and public health implications.

CONCLUSIONS

Although current evidence strongly implicates an immunological mechanism in the pathophysiology of rheumatic fever, the details of the manner in which the disease process develops are by no means clear. There is an abnormal immune response in rheumatics to streptococcal antigens, particularly humoral and cell-mediated responses against cell membrane antigens of the group A Streptococcus - antigens cross-reactive with heart and other muscle tissue antigens. The relationship between a B-cell alloantigen and rheumatic fever susceptibility suggests that abnormal reactivity to streptococcal antigens may be genetically determined on the basis of immune response genes, linked to histocompatibility genes.

Our working concept of the pathogenic mechanism operative in rheumatic fever is as follows. Exposure in a susceptible individual to an infection with group A Streptococcus gives rise to an exaggerated humoral and cell-mediated immune response to those streptococcal antigens cross-reactive with human tissues (especially muscle antigens). This reaction could be augmented or abetted by the presentation of buried cross-reactive determinants in the host, resulting from direct toxic effect on the tissues during the early stages of the infection. At a critical level of the exaggerated response, tolerance to the host's "self antigens is broken. There is a direct attack on host tissues by cells sensitized to streptococcal antigens that are cross-reactive with heart, brain or smooth muscle antigens. The immunological arm (cellular or humoral) predominantly responsible for the actual damage is still open to question. Circulating complexes of streptococcal antigens and antibodies with particular affinity for synovial receptors could account for the arthritic manifestations of the disease.

It is clear that further work needs to be done to clearly delineate the relative roles of cell-mediated and humoral immunity, as well as genetic predisposition in rheumatic fever. Since there is both an abnormal cellular and humoral response to streptococcal antigens in these patients, and since only group A streptococci contain the cross-reactive antigens, the concept of a streptococcalinduced "autoimmune" process in rheumatic fever is suggested.

To return finally to why the disease pattern and incidence is changing, a deviation in any of several main areas of this host-microbial circle could affect the rheumatic fever cycle. A decrease in the organism's virulence or potential to carry cross-reactive antigens would obviously affect the incidence of disease. One can speculate that smaller families, decreased spread of a virulent strain and /or better medical care could change the type of streptococcus present in the community. Moreover, if fewer genetically susceptible individuals carrying the 883+ marker are present in the human gene pool, this might have a strong impact on the relative number of new cases in the community. H is my belief that a subtle change in socioeconomic environment, possibly related to ingestion of greater amounts of animal muscle protein, might be the key factor in this change. For example, in those countries where meat or meat products (possibly capable of eliciting cross-reactive antibodies) are consumed, the incidence of rheumatic fever will be low secondary to the protective blocking effects of these antibodies. The converse will occur in those countries whose low economic status would not generally include the consumption of meat by its inhabitants. Although this is pure speculation, it may prove to be a more important factor than originally considered in the decrease of rheumatic fever.

REFERENCES

1. Manjula BN and Fischelli VA. Studies on group A streptococcal Mproieins; Purification of type 5 M-protein and comparison of its amino acid terminal sequence with two immunologically unrelated M -protein molecules J immunol 124:261-267, 1980.

2. Whiinack E and Bisno AL. In: Clinical Immunology, Vol. 2. Parker CW (ed). Chapter 29, p. 894, 1980.

3. McCarty M. Missing links in the sireptococcal chain leading lo rheumatic fever. Circulation 29:488, 1964.

4. Goldslein I. et al. Isolation from heart valves of glycoproteins which share immunological properties with Streptococcus hemolyiicus group A poly saccha rides. Nature 219:866-868, 1968.

5. Halpern B, parlebas J, Goldstein I. Isolement a parlin du cytoplasme du streptocogue A d'une glycoproteine qui parente immunologique aree les gly e opro te i nés des valvules cardiques comptes. Rendus des Seances d L'Academie des Sciences. 273:995, 1971.

6. Kasp-Grochowska E, Kingston D, Glynn LE. Cross reaction with the Cpotysaccharide of Sireptocut -cus , pyugenes. Ann Rheum Dis 31:82, 1972.

7. Cromartie WJ. Schwab JH, Craddock JG. The effect of a toxic cellular component of Group A streptococci on connective tissue. Am J Palhol 37:79, I960.

8. Cromartie WJ, et al. Arthritis in rats after systemic injection of streptococcal cells or cell walls. J Exp Med 146:1585, 1977.

9. Ra pa port FT, et al. Induction of renal disease with a mísera to group A streptococcal membranes. Transpl Proc 1:981-984, 1969.

10. Kaplan M H and Frengley JD. Autoimmunity to the heart in cardiac disease. Current concepts of the relation of autoimmunity to rheumatic fever, postcard io to m y and post infarction syndromes and cardio myopathies. Amer J Cardio! 24:459. 1969.

11. Zabriskie JB and Freimer EH. An immunological relationship between the group A Streptococcus and mammalian muscle. J Exp Med 124:661, 1966.

12. R a pa pon FT and Chase RM Jr. Homograft sensitivity induction by group A streptococci. Science 145:407, 1964.

13. Robert L, et al. Homology of amino acid composition of structural gjycoproteins. transplantation antigens, cell wall glycoproteJns, and Streptococcus A cell membrane. Transpl Proc 4:415, 1972.

14. Husby G, et al. Antibodies reacting with cytoplasm of subthalamic and caudate nuclei neurons in chorea and acute rheumatic fever. J Exp Med 144:1094-1110, 1976.

15. Kingston D and Glynn LE. Antistrepiococcal antibodies reacting with brain tissue. I. lmmunoftuorescent studies. Br J Exp Pathol 57:1 14, 1976.

16. Lyampert IM, et al. C ros s- react i ve antigens of group A streptococci: Their role in the autoimmune process. In: Pathogenic Streptococci. Parker TM (cd). Surrey, Eng: Reedbooks, 1979, p. 105.

17. Holm SE, Christensen P, Shalen C. Alternate interpretation of cross reactions between streptococci and human tissue. In: Pathogenic Strepioeocci. Parker TM (ed). Surrey, Eng: Reedbooks, 1979, p. 71.

18. Haibert SP and Keatings SL. The analysis of streptococcal infections. Vl. lmmunoelcctrophoretic observations on extracellular antigens detectable with human antibodies. J Exp Med 113:1013. 1961.

19. Haibert SP, Bircher R, Dahle E. The analysis of streptococcal infections. V. Cardiotoxicity of streptolysin O from rabbits in vivo. J Exp Med 1 13:759, 1961.

20. Watson DW. Host-parasite factors in group A streptococcal infections; pyogenic and other effects of immunologie distinct endoloxins related to scarlet fever toxins. J Exp Med 1 1 1 :255. I960.

21. Kellner A, et al. Loss of myocardial contractility induced in isolated mammalian hearts by streptolysin O. J Exp Med 104:361. 1956.

22. Watson RF, Hirsh GK, La nee field RC. Bacteriological studies on cardiac tissue obtained at autopsy from Il patients dying with rheumatic fever. Arthritis Rheum 4:74. 1961.

23. Dcnny FW Jr, et al. Prevention of rheumatic fever; treatment of the preceding infection. JAMA 143:151, 1950.

24. Vaisman S, et al. The failure of penicillin to alter acute rheumatic valv ulitis. JAMA 194:1284, 1965.

25. Ginsburg I. Mechanisms of cell and tissue injury induced by group A streptococci. Relation to post strepiococca! sequelae. J Infect Dis 126:294, 1972.

26. Taranta A, Cuppari C, Quagliata F. Dissociation of hemolytic and lymphocyte transforming activities of streptolysin S preparations. J Exp Med 129:605, 1969.

27. St o Her man GH. Rheumatic Fever and Sireptococcal Infection. New York: Grune, Chapter 8. 1975, p. 181.

28. Kaplan MH, Meyeresian M. Kushner L Immunologie studies of heart tissue. IV. Serologie reactions with human heart tissue as revealed by immunofluorescent methods: Isoimmune, Wassermann and autoimmune reactions. J Exp Med 113:17, 1961.

29. Zabriskie JB, Hsu KC, Seegal BC. Heart-reactive antibody associated with rheumatic fever; Characterization and diagnostic significance. Clin Exp lmmunol 7:147. 1970.

30. Rantz LA, Maroney M, DiCaprio JC. Hemolytic streptococcal infection in childhood. Pediat 12:498. 1953.

31. Francis TC, Oppenheim JJ, Barile NJ. Lymphocyte transformation by streptococcal antigens in guinea pigs and man. In: Proceedings 3rd Annual Leukocyte Culture Conference. Ricke WD (ed). New York: ACC, 1967. p. 501.

32. Murphy GE. Personal communication.

33. Stollerman GH. In: Rheumatic Fever and Streptococcal Infection. New York: Grune, Chapter 9. 1975. p. 218.

34. Dudding BA and Ayoub EM, Persistence of sireptococcal group A antibody in patients with rheumatic valvular disease. J Exp Med 128:1081. 1968.

35. Svartman M, et al. Immunoglobins and complement components in synovial fluid of patients with acute rheumatic fever. J Clin Invest 56: 111-117. 1975.

36. van de RiJn I, et al. Serial studies in circulating immune complexes in post streptococcal sequelae. Clin Exp lmmunol 34:318-325, 1978.

37. Read SE and Zabriskie JB. Immunological concepts in rheumatic fever pathogenesis. In: Textbook of lmmunopathology. Miescher PA and Muller-Eberhard HJ (eds). New York: Grune. 1976, p. 471.

38. Plate JM and Amos DB. Studies on a lymphocyte circulating factor from Group A streptococci. In: ProcetdinKs of Third Annual leukocyte Culture Conference. Rieke WD (ed). New York: ACC, 1967. p. 485.

39. Hirschhorn K. et al. The action of strepiolysin S on peripheral lymphocytes of normal subjects and patients with acute rheumatic fever. Proc Nat Acad Science 52:1151. 1964.

40. Keiser H, Kusher I, Kaplan M H. "Non-specific" stimulation of lymphocyte transformation by cellular fractions and acid extracts of group A streptococci. J lmmunol 106:1593, 1971.

41. Read SE, et al. Cellular reactivity studies to slreplococcal antigens in patients with streptococcal infections and their sequelae. J Clin Invest 54:439-450. 1974.

42. Gowrishankar R and Agarwal SC. Leukocyte migration inhibition with human heart valve glycoproteins and group A streptococcal ribonucleic acid proteins in rheumatic heart disease and post streptococcal glomerulonephritis. Clin Exp lmmunol 39:519-525. 1980.

43. Yang LC. et al. Streptococcal induced cell mediated immune destruction of cardiac fibers in vitro. J Exp Med 146:344. 1977.

44. Cheadle WB. Harveian lectures on the various manifestations of the rheumatic state as exemplified in childhood and early life. Lancet 821, 871, 921. 1889.

45. Glynn LE and H al borrow EJ. Relationships between blood groups. secretion status and susceptibility to rheumatic fever. Arthritis Rheum 4:203. 1961.

46. Taranta A. et al. Rheumatic fever in monozygotic and dizygotic twins. Circulation 20:778. 1959.

47. Falk JA, et al. A study of HLA antigen pnenotype in rheumatic fever and rheumatic heart disease patients. Tissue Antigens 3:173-178, 1973.

48. Patarroyo ME, et al. Association of B-cell alloantigen with susceptibility to rheumatic fever. Nature 278:173-174. 1979.

TABLE 1

EXTRACELLULAR PRODUCTS OF HEMOLYTIC STREPTOCOCCI

TABLE 2

PATHOGENESIS OF RHEUMATIC FEVER: THEORETICAL CONSIDERATION

TABLE 3

CLINICAL MANIFESTATIONS OF RHEUMATIC FEVER AND THE IMMUNE RESPONSE

TABLE 4

A COMPARISON OF THE AVERAGE TITERS OF CROSS-REACTIVE ANTIBODY IN PATIENTS WITH UNCOMPLICATED SCARLET FEVER, SCARLET FEVER WITH RHEUMATIC FEVER AND OTHER DISEASES

10.3928/0090-4481-19820401-09

Sign up to receive

Journal E-contents