Current Concepts of Rheumatic Fever
G Stollerman
Citation
G Stollerman. Current Concepts of Rheumatic Fever. The Internet Journal of Cardiology. 2002 Volume 2 Number 1.
Abstract
A half-century ago, students of medicine in the United States learned that rheumatic fever (RF) killed more school-age children than all other diseases combined (1). Within the next several decades, acute RF and rheumatic heart disease (RHD) became remarkably rare in the most developed countries of the world (2). Yet, in these same countries pharyngitis due to group A streptococci (GAS) continued to account for approximately 20% of sore throats . What explains these remarkable changes in the rheumatogenicity of GAS pharyngitis? Evidence for the marked decline in the virulence of sporadic GAS pharyngitis in developed countries has been accumulating for decades, but the critical importance of GAS strain virulence in the pathogenesis of RF is still not widely appreciated (3). Recent studies of the genetic control of the expression of various virulence factors of group A streptococci (GAS) (3, 4) are beginning to explain the wide spectrum of group A streptococcal diseases and their striking epidemiological variation . Experimental manipulation of the genome to produce mutants with and without specific virulence factors may soon provide greater insight into their respective roles in the pathogenesis of the various streptococcal diseases.
Pathogenesis of GAS infection
In the log phase of growth, streptococci divide approximately every 20 minutes. Only when they do so are they rapidly killed by penicillin. When phagocyted, they are also readily killed because they are highly susceptible to the antibacterial action of oxygen radicals and other antibacterial substances within phagosomes of white blood cells. Thus,
GAS toxins may contribute to morbidity in several ways. Weight for weight, the cell-surface bound hemolysin, streptolysin S, is one of the most toxic proteins known (10). It causes rapid destruction of cell membranes and is very cardiotoxic. Experimentally, the oxygen labile hemolysin, Streptolysin O, is also a powerful cardiac toxin. Streptokinase and desoxyribonuclease liquefy fibrin and nucleic proteins, respectively, accounting for the thin pus of GAS infections. Streptococcal hyaluronidase promotes rapid spread of the organisms through tissues (e.g., cellulitis and lymphangitis). The erythrogenic toxins are responsible for the rash of scarlet fever and are considered an important factor in toxic shock (11). Many of these secreted toxins have the properties of superantigens, nonspecifically and powerfully stimulating the host's immune response. M protein also contains moieties that similarly behave as such superantigens, thus further boosting immune responses to virulent strains (12, 13) (see “vaccines” below). The way in which these substances interact to produce the various complications of GAS infection is currently a subject of intense investigation.
During convalescence from GAS pharyngitis, virulent GAS strains strains, rich in M protein and heavily encapsulated, progressively lose these virulence factors (1,3). Attenuated strains of group A streptococci, however, may be transmitted and carried stubbornly for weeks or months (28, 29). Throat carriage is often more difficult to eradicate with penicillin (see “treatment”). The rapid dissociation of virulent strains in artificial media also makes maintenance of the virulent phase of the organisms difficult, requiring frequent mouse passage or passage though fresh human blood whereby only phagocyte-resistant clones survive. These are best preserved for storage or transport by prompt freeze-drying. Strains sent to reference laboratories without careful preservation often attenuate by the time they are studied in the laboratory. Indeed, very few published clinical studies record encapsulation of the strains of GAS isolated, although they may be readily recognized by their mucoid colonies on blood agar (30).
Primary Prevention Of Rheumatic Fever
Throat cultures, however, have the advantage of revealing the presence of mucoid colonies on blood agar to alert clinical laboratories. Early detection of clusters of large mucoid GAS colonies in throat cultures signal danger (3). In association with RF, AGN, or invasive disease, such strains should be sent to research or reference laboratories for detailed study if we are to learn more about them. For example, in the 1950's, in throat cultures from naval recruits with epidemic pharyngitis at the Great Lakes Naval Training Center the sudden appearance of a highly encapsulated single M type regularly predicted the onset of an outbreak of ARF (37). In contrast, at the same base ARF was not nearly as common among the naval personnel in housed in separate, non-recruit training units. The strains recovered from these epidemics have been an important source of studies of GAS vaccines and other research. More recently, outbreaks of RF in the U.S. have been associated with a single clone belonging to a single M serotype of GAS (23). (see below, “RF pathogenesis”).
Guidelines by expert committees of the American Academy of Pediatrics (34), the Infectious Disease Society of America (33), and the American Heart Association (38) favor greater precision in diagnosis by the use of throat cultures. In the interest of reducing excessive antibiotic usage that promotes emergence of resistant organisms, the American College of Physicians' published its own guidelines for diagnosis of GAS in adults (35,36). These guidelines eschew throat cultures in favor of RADTs, and suggest that even the latter may be unnecessary in the presence of clearly expressed clinical manifestations of GAS pharyngitis. It is apparent that the threat of rheumatic fever, or other severe complications, and the economic resources available to a given population will influence the practicality of the use of laboratory tests for support of the diagnosis of GAS pharyngitis. Moreover, annual and seasonal variation in the severity of GAS disease may be another factor influencing the local perceived need for precision in the diagnosis of GAS pharyngitis.
Issues In The Treatment Of GAS Pharyngitis
Variation in the treatment of GAS pharyngitis should now be considered in relation to the varying prevalence of RF, invasive GAS diseases and AGN in different geographical and social settings. Despite more than a half century of intense clinical use, penicillin resistant GAS strains have not emerged. Group A streptococci are uniformly highly sensitive to the action of penicillin. For rapidly multiplying organisms, penicillin G is bactericidal in a concentration of 0.01 to 0.04 units/mL in a standard broth culture. Thus, sustained low bactericidal blood levels eradicate proliferating group A streptococci as well as high penicillin blood levels.
Since World War II, the treatment of GAS pharyngitis has been strongly directed toward the primary prevention of rheumatic fever and suppurative complications. Where rheumatic fever persists in the world, and particularly in undeveloped countries, primary rheumatic fever prevention is still a major consideration in the treatment of group A streptococcal pharyngitis. Such treat.ment should ensure effective penicillin levels for at least 10 days.(39). Because this result can be achieved by a single intramuscular injection of 1.2 million units of benzathine penicillin G, or 600,000 units for children who weigh less than 27 kg or 60 pounds (40), this regimen is a favored one (38). Intramuscular injections of repository penicillins, however, produce some local pain and discomfort, and must be administered by physicians or nurses. Injectable .benzathine penicillin G for pharyngitis, therefore, has declined in popularity in developed countries in which ARFrheumatic fever is no longer feared. In such venues, uncertain as compliance with the full 10-day-regimen may be, oral penicillin, usually penicillin V, is currently most popular..
Penicillin V may .be given twice daily in 1.0 g doses and has been shown to be at least as effective as 0.5 g administered four times daily. Greater compliance has been seen with a twice-daily regimen. Oral cephalosporins are also highly effective in the treatment of streptococcal pharyngitis, and some reports show a slightly higher rate of eradication of convalescent carriage than that achieved with penicillin therapy (41).
If penicillin allergy is suspected or known to exist, erythromycin may be used in divided doses not exceeding 1 g/d), also for a period of 10 days. Although erythromycin resistance of GAS is not a serious problem in most regions of the United States, this drug has caused GAS resistance with striking frequency when it has been used extensively as the first line drug for treatment of sore throat. Newer macrolides, azithromycin, clarithromycin are also as effective but are much more expensive. Treating streptococcal pharyngitis with bacteriostatic agents like sulfonamides does not prevent rheumatic fever. Sulfonamides are quite effective, however, as preventatives of GAS infection and are therefore used quite effectively as secondary prophylactic agents for rheumatic fever recurrences (see later). Tetracycline-resistant group A streptococci are prevalent in many areas, and therefore this drug is not recommended.
Insistence by some authors that antibiotic regimens produce
“Mass” Primary Prophylaxis Of RF In Epidemics
When rheumatic sequelae are associated with a focal epidemic of streptococcal pharyngitis, prophylactic treatment of an entire populationcohort may be required
Oral prophylactic regimens are also effective but are less reliable. They are recommended when the risk of rheumatic recurrences is relatively low. Penicillin V orally is recommended in doses of 250 mg bid. Sulfadiazine is also about as effective for secondary prevention and is inexpensive. The recommended dose of oral sulfadiazine is 0.5 g once daily for patients who weigh less than 27 kg (60 pounds) and 1 g daily for heavier persons. For the rare patient who is sensitive to both penicillin and sulfonamides, erythromycin may be substituted in a dose of 250 mg twice daily.
Figure 1
1 Taranta A, et al, Rheumatic fever in children and adolescents. Along-term epidemiological study of subsequent prophylaxis, streptococcal infections, and clinical sequelae IV. Relation of the rheumatic fever recurrence rate per streptococcal infectrion to the titers of streptococcal antibodies. Ann Intern Med 1964 60(Suppl 5) 5;47.
Problems With The Diagnosis Of RF
A clear diagnosis of RF, and particularly rheumatic carditis, is important since it commits an individual to prolonged prophylactic antibiotic therapy. As RF becomes rare in developed countries, its familiarity to younger physicians also wanes. Moreover, the diagnosis of RF may be particularly difficult when it presents as an isolated major manifestation. Unfortunately, RF remains a clinical syndrome without a single pathognomic feature. In the 1940s, T. Duckett Jones adopted the constellation of major manifestations of RF that were first recognized as a single disease at the end of the 19 th century by William Cheadle (53). The Jones criteria (54) became particularly useful in clinical investigation to ensure admission to clinical studies of a uniform cohort of clear-cut cases of ARF. Thus, these guidelines avoid overdiagnosis but do not always capture the more subtle manifestations of the disease. The major manifestations are
In the 1960s, when antistreptolysin O and other GAS antibody titers generally became available to clinical laboratories a committee of the American Heart Association revised the Jones criteria suggesting that, particularly those of polyarthritis, could be strengthened by including evidence of antecedent GAS infection (55). Some limitations were emphasized; circumstances in which a diagnosis of ARF may be made without strict adherence to the Jones criteria (56). For example, in contrast to arthritis, chorea, the latest-appearing of the major manifestations following the antecedent infection, may present without any other major or minor features of ARF - so-called “pure chorea” (see below). Also, isolated acute carditis may first come to medical attention several months into or after the rheumatic attack. By then, antibody titers may have declined to normal levels and the minor manifestations of systemic inflammation (fever, ESR, C-reactive protein, etc) may have abated.
Most patients with recurrent ARF also fulfill the Jones criteria, but in some the diagnosis of a recurrence is less obvious. For example, when rheumatic valvular disease preexists, clear recognition of a new bout of carditis requires evidence of fresh cardiac injury such as pericarditis, acute cardiac enlargement and/or congestive heart failure, or a newly detected murmur from a valve not previously affected. The Jones Criteria, therefore, apply more readily to initial attacks, and more diagnostic latitude is sometimes needed to interpret recurrent carditis in patients with pre-existing rheumatic heart disease. The steps in the evolution of the modification of the Jones Criteria have been reviewed recently in detail (57).
Chronicity of the arthritis, and particularly its recurrence in the absence of a new GAS pharyngeal infection, the appearance of joint deformity, or the presence of rheumatoid factor or DNA antibodies may eventually reveal a different disease, (e.g. rheumatoid arthritis, systemic lupus, polyarteritis, etc.). Although typically migratory, many authorities have observed patients with the polyarthritis of ARF that was not initially “migratory”, but rather, “additive”, persisting in many joints at once, and furthermore stubbornly “rebounding” once or twice after six week courses of anti-rheumatic therapy (more often with corticosteroids than NSAIDs ( 1, 58). In prospective studies of acute rheumatic attacks that occur in the absence of a new GAS infection, relapses of ARF have been noted as late as 5 weeks after completion of six weeks of antirheumatic therapy (1). And some patients do not respond brilliantly to salicylates, requiring supplemental corticosteroids. Nonetheless, these cases finally heal without deformity. In a few patients, rheumatic heart disease has been noted years later (3).
Although a different etiology of polyarthritis may be inadvertently included, we prefer to retain so-called PSRA patients within the framework of the diagnosis of RF and administer antibiotic prophylaxis to them, but perhaps for a modified duration, the exact time dependent on other variables, particularly the prevalence of ARF in the community. Moreover, some PSRA patients apparently have developed rheumatic valvular disease after several years of follow-up, indeed, reported in some children to be as high as 7% of PSRA (60). Although the numbers of the reported cases of so-called PSRA are still rather few, and not always similarly defined, they deserve further study (61,62). Whether or not PSRA is part of RF, it is generally agreed that secondary prophylaxis to prevent recurrences and possible heart disease is prudent.
Isolated Chorea And Post-Infectious Autoimmune Neurological Diseases (PANDAs)
Sydenham's chorea may also occur as an isolated manifestion, and frequently recurs following new streptococcal pharyngitis (1). After puberty, Sydenham's chorea is almost entirely limited to women . Like polyarthritis, it is most often evanescent, over in a few weeks, but occasionally it may be stubborn, persisting for many months. The pathogenesis of chorea, (similar to that of the synovitis of polyarthritis), seems to be associated with immune complex disease produced by non-destructive antoantibodies localized to the basal ganglia and striatal system of the brain (63,64,65,66). Severe chorea seems to respond sometimes to treatment with intravenous IgG (67) . It also seems to be closely related in pathogenesis to the so-called PANDAs (post-infectious autoimmune neurological diseases (64, 65, 68). These include tics, Tourettes syndrome, and obsessive-compulsive behavior, all of which have been observed in some patients during or after an attack of rheumatic chorea. PANDA cases that did not express choreiform movement and were not previously referred to rheumatic fever centers were more often referred to pediatric neurologists. Pandas are often associated with antecedent GAS infection (69) and some cases have clearly been shown to be associated with deposition of streptococcal antigens in the basal ganglia (65).
If studies in progress reveal at least some of these neurological manifestations to be preventable by antistreptococcal prophylaxis, they might well be included, like PSRA, as variable features of the syndrome of ARF. This is notwithstanding the fact that, as in Sydenham's chorea, other autoimmune disease (e.g.,systemic lupus erythematosus ) may occasionally cause them, just as SLE may also cause endocarditis. A possible association of PANDAS with rheumatic heart disease, such as seen in long-term follow-up of patients with chorea, should also be carefully studied. Currently, the role of GAS pharyngitis as a cause of recurrent episodes of obsessive-compulsive disorders in children without chorea or other PANDA manifestations is being evaluated.(69, [[[69a]]]). The relation of recurrences of all of the manifestions of the PANDAs could be assessed by continuous antistreptococcal prophylaxis, preferably with monthly benzathine penicillin G to insure compliance.
Eventually, the immunologic and host factors deciding the development of the major manifestations of ARF, their severity, and their chronicity may warrant separate etiological classifications other than their coexistence, as noted by T. Duckett Jones , “with a frequency exceeding chance”. Meanwhile, in my opinion they still warrant the same prophylactic management currently advised for RF patients.
Although EC, particularly accompanied by Doppler studies, offers greater sensitivity and specificity for the assessment of valvular regurgitation it need not be considered essential for the diagnosis of RF by experienced primary care physicians in settings where the disease is common and medical resources limited (56). Nonetheless, cardiologists proficient in echo-Doppler technology now use this method routinely to detect abnormal valve structure and function more sensitively and accurately than can be achieved by auscultation alone. Despite the relatively good prognosis of “silent” rheumatic mitral regurgitation, EC can, indeed, provide a more accurate assessment of the presence and severity of valvulitis, especially in an era when cardiac auscultation has been taught less extensively and is used with less confidence by young clinicians. In any case, it is doubtful that such a powerful diagnostic tool as EC will be neglected in the assessment of valvular disease wherever the instrument is available, and certainly where its expense may not be too great a concern as in developed countries. It is now important to extend recent long term studies (75), to establish more precisely the natural history of the subauscultatory valvular rheumatic valvular regurgitation as diagnosed by echocardiography. Such information will further influence the choice and duration of secondary prophylaxis.
Right Ventricular Endomyocardial Biopsy
When the characteristic murmurs of rheumatic carditis are detected early in the course of a rheumatic attack and are associated with other major and minor manifestations of ARF, such as arthritis and fever, the yield of useful additional clinical information from endomyocardial biopsy (EB) has been low. Its diagnostic sensitivity in one relatively large study was only 27% (77). EB has, however, confirmed the presence of underlying carditis in unexplained congestive heart failure of acute onset in some patients with preexisting rheumatic heart disease and elevated antistreptolysin titers, suggesting a rheumatic recurrence. In patients with chronic rheumatic heart disease, however, EB does not appear to provide additional diagnostic information. In my opinion, in patients with rheumatic carditis endomyocardial biopsy should be limited to clinical investigation.
Troponin 1 levels in rheumatic carditis are low (except when underlying a severe pericarditis, arguing against extensive myocardial muscle necrosis as a significant feature of rheumatic myocarditis (78,79). Earlier studies found that serum glutamic-oxalacetic transaminase (SGOT) levels are not elevated in acute carditis in the absence of toxic doses of salicylates (80, 81), supporting our view that rheumatic myocarditis is principally interstitial rather than severe cardiac muscle damage (1). The heart failure associated with acute rheumatic carditis is considered to be due primarily to severe valvular insufficiency, although severe interstitial myocardial inflammation causing myoicardial dysfunction has not been entirely excluded as a contributing factor.
Treatment of RF
In a recent well-controlled trial of 59 patients with acute rheumatic fever, of whom 39 had carditis, intravenous gamma globulin did not affect the course of the illness. No detectable differences in the clinical, laboratory or echocardiographic parameters of the disease process were found between treated and control patients during the subsequent 12 months (82).
Treatment remains supportive. There is no longer significant doubt that corticosteroids, however symptomatically beneficial, do not prevent valvular damage (83, 84). For mild rheumatic carditis, lingering doubt in the minds of some investigators about the possible long-term benefit of corticosteroids over salicylates was based upon the results of but a few studies suggesting favorable outcomes of such treatment of minimal rheumatic mitral regurgitation (84). These minimal mitral murmurs are most difficult to standardize (perhaps EC will help) but spontaneous healing occurs in 80% or more of patients with minimal mitral regurgitation from acute RF (
Current Research Of RF Pathogenesis
As noted above, the entire genome of three M serotypes, M1, M3 and M18 have already been determined, (20,21,22,23). Such analysis may help to establish the clonal nature of the strains that cause either RF, AGN or invasive streptococcal infections as well. Bacteriophage infection of the GAS genome is common and such genomic transformation by this agent as well as other causes of mutations may help account for the sudden appearance and focal outbreaks of RF and invasive disease. The tendency of such virulent strains to spread to contacts has been well demonstrated in nosocomial invasive GAS infections (88). What remains at issue is whether a rheumatogenic strain requires some unique toxin or antigen, or whether a number of highly virulent GAS strains can initiate the various manifestations of the rheumatic process in predisposed hosts.
Immune complexes containing host antigens identical to some streptococcal epitopes have been identified in synovia, heart and brain (63,65, 66, 90). These may cause the non-destructive, reversible rheumatic inflammation seen in joints, skin and brain (polyarthritis, nodules/ erythema, chorea and PANDAs .Cytotoxic autoimmunity, on the other hand, has been hypothesized to cause destruction of heart valves (91,92,93). But how is such putative autoimmune intolerance be initiated? Do large amounts of group A streptococcal antigens swallowed and absorbed in the course of GAS pharyngitis result in an intense immunologic stress that breaks immune tolerance to certain antigens in susceptible hosts? Knowledge of the immunophysiology of the immune system of the gut is still evolving, particularly with regard to its role in immune tolerance and autoimmunity.
In contrast to the lack of a definitive association with specific HLA DR antigens, a strong relationship has been detected with a non HLA B cell antigen originally designated 883 and detected in widely distributed populations from New York to Bogota, Colombia, and New Mexico to India (97, 99). Studies with a series of monoclonal antibodies directed against B cells from rheumatic fever patients have identified another B cell alloantigen labeled D8/17.(99). It is present in a relatively large percentage of the total B cells of rheumatic fever probands: 33.5% compared with 14.6% and 13%, respectively, of the B cells of unaffected siblings and parents. Two sets of identical twins were included in these studies. The proband with rheumatic fever had 43% positive B cells, whereas the unaffected twin had only 15%. In the other set of unaffected twins, 20% and 10%, respectively, had D8/17 B cells. Thus, this B cell alloantigen is not unique to rheumatic hosts but is expressed more vigorously in those who have had rheumatic fever. Thus, it may be an acquired feature of the GAS exaggerated immune response. At any rate, perhaps a predisposing host factor, present to some degree in all persons, is more expressible in rheumatic hosts who are stimulated by antigens, and perhaps, as recently shown (100), by the varied host responses to superantigens contained in virulent pharyngeal strains of GAS (see below).
Prospects For A Vaccine Against Rheumatic Fever
Because immunity to GAS is type specific and dependent on antibodies to M protein, attempts at vaccine production have focused primarily on M protein purification. Since the extraction of M protein by Rebecca Lancefield (101), its further purification has led to its molecular definition (102,103). A clean separation of the type-specific N-acetyl terminal peptide (the type specific antigenic determinant) from the proximal part of the M molecule has been achieved freeing it from the more proximal region of the M molecule that contains the epitopes cross-reactive with heart, brain, skin and synovial tissue antigens . The terminal type-specific M epitope is antigenic in humans without raising host-tissue cross reactions and it is non-toxic in human skin (104). An effective vaccine against rheumatic fever may not require the inclusion of all known M protein serotypes, but rather those identified most clearly as containing rheumatogenic strains. In fact, a recombinant, multivalent vaccine containing the type-specific epitopes of some 26 M serotypes associated the great majority of serious GAS infections is currently under field trial (105). Newly identified M types containing dangerous strains could be added subsequently as necessary. The potential for the production of IgA antibodies to M proteins by employing such preparations for oral human immunization is suggested by recent experimental studies (106). The protective role of mucosal Iga and its production by oral streptococcal vaccines is under vigorous current investigation.
Summary and Conclusions
Rheumatogenic strains of GAS still infest the majority of the world's population. Jet travel may at any time spread such strains from high to low-risk regions of the globe. Dangerous GAS strains may emerge in any location by mutation, transduction, or by other as yet unknown environmental factors. Adverse social conditions causing crowding affect their spread. Therefore, in my opinion the appearance of but one case of RF in any community should be greeted with alarm and all infected contacts among cohorts identified and properly treated in the same fashion as the contacts of the recently reported nosocomial septic infections caused by clones of invasive strains (107). Vaccines are under trial that may afford protection against the most virulent and dangerous of known GAS strains. If we learn how to identify and immunize safely susceptible hosts, prevention will be greatly simplified. Meanwhile, even those privileged to live in affluent surroundings must continue to diagnose and treat GAS pharyngitis effectively, remain alert to the threat of ARF , and continue to strive toward the eradication of RF from the human race.
Correspondence to
Gene H. Stollerman, M.D. Professor of Medicine & Public Health Boston University #21 The Courtyard Hanover, New Hampshire 03755 Email: gstollerman@valley.net Tel/Fax 603-643-3336