RHEUMATIC FEVER

Alan L. Bisno

DEFINITION. Rheumatic fever is an inflammatory disease that occurs as a delayed, nonsuppurative sequel of upper respiratory infection with group A streptococci. The clinical manifestations include polyarthritis, carditis, subcutaneous nodules, erythema marginatum, and chorea in varying combinations. In its classic form, the disorder is acute, febrile, and largely self-limited. However, damage to heart valves may be chronic and progressive, causing cardiac disability or death many years after the initial episode.

ETIOLOGY. The development of acute rheumatic fever (ARF) requires antecedent infection with a specific organism, the group A Streptococcus, at a specific body site, the upper respiratory tract. Cutaneous streptococcal infection, a precursor of poststreptococcal acute glomerulonephritis, has never been shown to cause rheumatic fever.

Strains representing a number of the more than 80 M-protein serotypes of group A streptococci can cause ARF. There is a substantial body of evidence to indicate, however, that group A streptococci vary in their rheumatogenic potential. Strains causing clusters or epidemics usually belong to a limited number of serotypes (e.g., 3, 5, 18, 24, and others) and are often heavily encapsulated, as evidenced by their growth as mucoid colonies on blood agar plates. Streptococci epidemiologically associated with ARF outbreaks occurring during the 1980's in the United States exhibited these characteristics.

PATHOGENESIS. The mechanism by which group A streptococci elicit the connective tissue inflammatory response that constitutes ARF remains unknown. Various theories have been advanced, including (1) toxic effects of streptococcal products, particularly streptolysins S and O, both of which can initiate tissue injury; (2) inflammation mediated by antigen-antibody complexes, perhaps localized to sites of tissue injury; and (3) "autoimmune" phenomena induced by the similarity of certain streptococcal and human tissue antigens.

Efforts to discriminate among these potential pathogenetic mechanisms have been hampered by the lack of an animal model of rheumatic fever. Most authorities currently favor the theory that ARF is an "autoimmune" disorder in which tissue damage is mediated by the host's own immunologic responses to the antecedent streptococcal infection. This theory is rendered more credible by the relatively long latent period between the onset of pharyngitis and ARF and by the demonstration of numerous examples of antigenic similarity between somatic constituents of the group A Streptococcus and human tissues. The most intensively studied of these cross-reactions is that between streptococci and human heart. Many patients with ARF (as well as patients with uncomplicated streptococcal infections) have in their sera antistreptococcal antibodies that cross-react with heart tissue in a variety of test systems. Components of the streptococcal cell wall (including group A carbohydrate and M-protein) and of the cell membrane contain epitopes that share antigenic determinants with certain constituents of the human heart.

Antibodies to the cytoplasm of neurons located in the caudate and subthalamic nuclei of the brain have been identified in sera of patients with Sydenham's chorea, and such antibodies cross-react with group A streptococcal membranes. Streptococcal extracellular products appear to be present in immune complexes circulating in the blood of ARF patients. Taken together, these and other reported immunologic cross-reactions and toxic phenomena could theoretically account for most of the manifestations of ARF. As yet, however, there is no direct evidence that any of them is pathogenetically significant.

Patients with ARF have, on average, higher titers of antibodies to streptococcal extracellular and somatic antigens than do patients with uncomplicated streptococcal infections. Data relating to cellular immunity are more limited. ARF patients exhibit an exaggerated cellular reactivity to streptococcal cell membrane antigens, as demonstrated by inhibition in vitro of migration of peripheral blood lymphocytes.

Several observations suggest that development of rheumatic fever may be modulated, at least in part, by the specific genetic constitution of the host. These include (1) the tendency of rheumatic fever to affect more than one member of a given family, (2) the fact that only a small percentage of all individuals experiencing an immunologically significant streptococcal infection develop ARF, (3) the tendency of rheumatic individuals to experience recurrent attacks, (4) the propensity of rheumatic subjects to exhibit exaggerated immunologic responses to streptococcal antigens, and (5) the fact that certain class II histocompatibility antigens are encountered significantly more frequently in ARF patients than in controls. Recently, a unique antigen has been found to be strongly expressed on the B cells of virtually all ARF patients but in <20% of controls.

EPIDEMIOLOGY. The epidemiology of ARF mirrors that of streptococcal pharyngitis. The peak age of incidence is 5 to 15 years, but both primary and recurrent cases occur in adults. ARF is rare in children younger than age 4, a fact that has led some observers to speculate that repetitive streptococcal infections are necessary to "prime" the host for the disease. There is no clear-cut gender predilection, although females are more likely to develop certain manifestations such as Sydenham's chorea and mitral stenosis.

The frequency with which ARF develops following untreated group A streptococcal upper respiratory infection differs with the prevalence of highly rheumatogenic strains in the population and the epidemiologic circumstances. In the years following World War II, careful prospective studies were conducted among personnel in military camps suffering from exudative tonsillitis or pharyngitis caused by M-typable group A streptococci. Under such circumstances, in which cases of streptococcal pharyngitis tend to be clinically severe and to appear in epidemics, approximately 3% of untreated patients developed ARF. Studies of endemically occurring streptococcal infection among open populations of children are complicated by the difficulties of differentiating cases of streptococcal pharyngitis from viral pharyngitis occurring in streptococcal carriers; nevertheless, the ARF attack rate in such circumstances is clearly lower than in the military experience, with an overall attack rate of <1%.

Certain features of the antecedent streptococcal infection are associated with an increased risk of ARF. Among these are the magnitude of the antistreptolysin O (ASO) titer rise and the persistence of the infecting organism in the pharynx. Although ARF is more likely to occur following clinically severe exudative pharyngitis than following mild nonexudative illness, one third or more of cases occur after streptococcal infections that are asymptomatic or so mild as to have been forgotten by the patient.

Patients with a history of ARF are at greatly increased risk of recurrent disease following an immunologically significant streptococcal infection. In one long-term prospective study of rheumatic subjects at a rheumatic fever sanitarium, one of every five documented streptococcal infections gave rise to a recurrence of ARF. The risk of recurrence is greater in patients with pre-existing rheumatic heart disease and in those experiencing symptomatic throat infections; the risk declines with advancing age and with increasing interval since the most recent rheumatic attack. Nevertheless, rheumatic patients remain at increased risk well into adult life, perhaps indefinitely.

Rheumatic fever occurs in all parts of the world; there is no known racial predisposition. In temperate climates, ARF peaks in the cooler months of the year, in the winter and early spring or shortly after schools open in the fall. The major environmental factor favoring occurrence appears to be crowding, as in military barracks or similar closed institutions and large households. Crowding favors interpersonal spread of group A streptococci and perhaps enhances streptococcal virulence by frequent human passage.

ARF remains rampant in developing areas such as the Middle East, the Indian subcontinent, and many nations of Africa and South America. It has been estimated that rheumatic heart disease causes 25 to 40% of all cardiovascular disease in the Third World. In striking contrast, the incidence of ARF and the prevalence of rheumatic heart disease have declined both in North America and in western Europe during the course of the twentieth century. Rates of fewer than 2 per 100,000 school children have been reported from several areas of the United States. The disease has become extremely uncommon in the affluent suburbs of many U.S. cities, while persisting among lower socioeconomic groups, particularly in the densely populated core areas of major urban centers. The higher incidence rates reported for African-Americans than Caucasians thus appears to be due to socioeconomic rather than genetic factors.

The mid-1980's, however, witnessed some startling developments in the epidemiology of ARF in the United States. Outbreaks of the disease were reported in Salt Lake City, Utah, Columbus and Akron, Ohio, Pittsburgh, Pennsylvania, Nashville and Memphis, Tennessee, and a number of other communities. The largest outbreak was in Salt Lake City and its environs, where approximately 200 cases occurred between 1985 and 1989. Equally surprising was the fact that, in many of these outbreaks, the victims were predominantly white, middle-class children dwelling in the suburbs. Moreover, ARF epidemics occurred in military training bases in Missouri and California, a phenomenon that had not been observed for two decades. Group A streptococci recovered from ARF patients, their families, and community and training camp surveys were generally highly mucoid and belonged to well-established rheumatogenic serotypes (e.g., serotypes 3 and 18). There is as yet no evidence that these events presage a major national resurgence of ARF.

PATHOLOGY. ARF is characterized by exudative and proliferative inflammatory lesions in the connective tissues, especially those of heart, joints, and subcutaneous tissues. The early lesions consist of edema of the ground substance, fragmentation of collagen fibers, cellular infiltration, and fibrinoid degeneration. In the heart, diffuse degeneration and even necrosis of muscle cells may be observed. At a slightly later stage, focal perivascular inflammatory lesions develop. These so-called Aschoff nodules (Fig. 277-1), considered virtually pathognomonic of rheumatic fever, consist of a central area of fibrinoid surrounded by lymphocytes, plasma cells, and large basophilic cells, some of them multinucleate. Many of these cells have elongated nuclei with a distinctive chromatin pattern, sometimes called "caterpillar" or "owl-eye" nuclei, depending on their orientation in microscopic cross section. Cells containing these nuclei are called "Anitschkow myocytes," despite the fact that most authorities believe them to be of mesenchymal origin.

Cardiac findings may include pericarditis, myocarditis, and endocarditis. Foci of coronary arteritis also may be observed. A thickened and roughened area ("MacCallum's patch") is frequently present in the left atrium above the posterior leaflet of the mitral valve. Valvular lesions appear early as small verrucae along the line of closure. Later, as healing occurs, the valves may become thickened and deformed, the chordae shortened, and the commissures fused. These changes result in valvular stenosis or insufficiency. The mitral valve is involved most commonly, followed by the aortic, the tricuspid, and rarely, the pulmonic valves.

Pathologically, the arthritis of ARF is characterized by a fibrinous exudate and sterile effusion without erosion of the joint surfaces or pannus formation. Subcutaneous nodules have many histologic features in common with the Aschoff nodules. These consist of central zones of fibrinoid necrosis surrounded by histiocytes, fibroblasts, occasional lymphocytes, and rare polymorphonuclear cells. Inflammation of the smaller arteries and arterioles may occur throughout the body. Despite pathologic evidence of diffuse vasculitis, aneurysms and thrombosis are not typical features of ARF.

CLINICAL MANIFESTATIONS. Rheumatic fever may involve a number of different organ systems, most notably the heart, joints, skin, subcutaneous tissues, and central nervous system. The clinical picture of the disease may thus be quite variable (Table 277-1), depending on which systems are attacked, whether they are involved singly or in combination, and the severity of the involvement. Five clinical features of the disease are so characteristic that they are recognized as "major manifestations" according to the revised Jones criteria (see below) for the diagnosis of ARF: carditis, polyarthritis, chorea, subcutaneous nodules, and erythema marginatum. Certain other findings, frequently present but nonspecific, have been designated "minor manifestations." These include arthralgia, fever, and certain laboratory findings (see below).

In cases in which it can be determined, the latent period between the antecedent streptococcal infection and the onset of symptoms of ARF ranges between 1 and 5 weeks. The average latent period is 19 days for both primary and recurrent attacks. When acute polyarthritis is the presenting complaint, the onset is often rather abrupt and may be marked by high fever and toxicity. If isolated carditis is the initial manifestation, the onset may be insidious or even subclinical. Between these two extremes, diverse gradations exist in the initial presentation of ARF (see Table 277-1). In most attacks, fever and joint involvement are the earliest clinical manifestations, although occasionally they may be preceded by abdominal pain localized to the periumbilical or infraumbilical areas. At times, the location and severity of the pain, as well as fleeting signs of peritoneal inflammation, may lead to a misdiagnosis of acute appendicitis. Carditis, if it is to appear, usually does so within the first 3 weeks of the illness. In contrast, chorea tends to occur later in the course of the disease, sometimes after all other manifestations have subsided. Fortunately, chorea and polyarthritis almost never occur simultaneously. Epistaxis may be a feature of ARF, occurring both at the onset and throughout the acute phase of the illness; it may be quite severe.

The incidence of major manifestations varies in reported series. Overall, however, arthritis occurs in approximately 75% of first attacks of ARF, carditis in 40 to 50%, chorea in 15%, and subcutaneous nodules and erythema marginatum in <10%. The frequency of individual manifestations varies with age. Carditis is more frequent in the youngest age groups and is relatively uncommon in first attacks occurring in adults. Chorea occurs primarily in persons between age 5 and puberty. It is seen more frequently in females and virtually never occurs in adult males. Thus the majority of ARF attacks occurring in adults are manifested primarily by arthritis.

Arthritis. Joint involvement ranges from arthralgia alone to acute, disabling arthritis, characterized by swelling, warmth, erythema, severe limitation of motion, and exquisite tenderness to pressure. The larger joints of the extremities are usually involved-most frequently the knees and ankles but also the wrists and elbows. The hips and small joints of the hands and feet are affected occasionally. Involvement of shoulders and lumbosacral, cervical, sternoclavicular, and temporomandibular joints occurs in a relatively small percentage of cases. The synovial fluid contains thousands of white blood cells, with a marked preponderance of polymorphonuclear leukocytes; bacterial cultures are sterile. Characteristically, the articular involvement in ARF assumes a pattern of migratory polyarthritis. This does not mean that inflammation in one joint disappears before the next is attacked. Rather, a number of joints are affected in succession, and the periods of involvement overlap. Inflammation in one joint may subside while another is becoming symptomatic so that the process seems to migrate from joint to joint. In untreated cases, as many as 16 joints may be affected, and about half the patients develop arthritis in more than 6 joints. When effective anti-inflammatory therapy is administered early in the course of the disease, the involvement not infrequently remains monarticular or pauciarticular.

In most instances, inflammation in any one joint begins to subside spontaneously within a week, and the total duration of involvement is no more than 2 or 3 weeks. The entire bout of polyarthritis rarely lasts more than 4 weeks and resolves completely, leaving no residual joint damage. Some authors have described the rare occurrence of Jaccoud's arthritis, so-called chronic post-rheumatic fever arthropathy of the metacarpophalangeal joints, following repetitive bouts of rheumatic polyarthritis. This entity is not a true arthritis but a form of periarticular fibrosis; its relationship to rheumatic fever remains unresolved.

Carditis. Rheumatic fever may involve the endocardium, myocardium, and pericardium (Table 277-2), and thus the disease is capable of inducing a true pancarditis. Carditis is the most important manifestation of ARF because it is the only one that can cause significant permanent organ damage or death. Although the clinical picture may at times be fulminant, it is more frequently mild or even asymptomatic and may escape notice in the absence of more obvious associated findings, such as arthritis or chorea. The diagnosis of carditis requires the presence of one of the following four manifestations: (1) organic cardiac murmurs not previously present, (2) cardiomegaly, (3) pericarditis, or (4) congestive heart failure. In practice, the characteristic murmurs of ARF are almost always present in cases of rheumatic carditis, unless the ability to hear them is obscured (e.g., loud pericardial friction rub, large pericardial effusion, low cardiac output, severe tachycardia). The diagnosis of carditis should be made with caution in the absence of one of the following three murmurs: apical systolic, apical mid-diastolic, and basal diastolic. Such murmurs, if they are destined to develop, do so usually within the first week and almost always within the first 3 weeks of illness. (An exception to this rule may occur in the patient with "pure" chorea; see later discussion.) The apical systolic murmur of relative or actual mitral regurgitation encompasses most of systole. It is blowing, relatively high pitched, and heard best at the apex; it radiates to the axilla and at times to the base of the heart or the back. It must be distinguished carefully by quality, location, and radiation from a variety of functional precordial systolic murmurs heard in normal individuals, especially in children. The apical mid-diastolic (Carey-Coombs) murmur is a low-pitched sound replacing or immediately following the third heart sound and ending distinctly before the first heart sound. It may be heard in a variety of conditions associated with increased flow across the mitral valve and is thus not pathognomonic of ARF. It may be differentiated from the diastolic rumble of mitral stenosis by the absence of an opening snap, presystolic accentuation, or accentuated first sound at the mitral area. The high-pitched, decrescendo basal diastolic murmur of aortic regurgitation is best heard along the upper left sternal border or over the aortic area. It may be brief and faint, best heard after expiration with the patient leaning forward. The prognostic significance, if any, of echocardiographically recorded valvular regurgitation in the absence of audible murmurs remains to be determined.

Other prominent auscultatory findings in patients with active rheumatic carditis include tachycardia, which persists during sleep; protodiastolic, presystolic, or summation gallops; an indistinct or "mushy" quality to the first heart sound (resulting in some cases from first-degree heart block); pericardial friction rub; or muffling of heart tones caused by pericardial effusion. In the early stages of congestive heart failure, rapid distention of the hepatic capsule may lead to right upper quadrant aching and tenderness over the liver. All the usual clinical findings of pericarditis or congestive failure may be observed.

A number of different rhythm disturbances may occur during the course of ARF. By far the most common is first-degree atrioventricular block. Second- and third-degree heart block, nodal rhythm, and premature contractions also may be observed; atrial fibrillation, on the other hand, is usually a feature of chronic rather than acute rheumatic involvement. Conduction disturbances do not in themselves indicate acute carditis, and their presence or absence is unrelated to the subsequent development of rheumatic heart disease.

In cases of ARF with severe carditis, areas of patchy pneumonitis are sometimes seen. Many observers feel that these pulmonary infiltrates represent a specific rheumatic pneumonia. The case is difficult to prove, however, because of the confusion induced by such confounding clinical entities as pulmonary edema, pulmonary embolization, superimposed bacterial pneumonia, and the acute respiratory distress syndrome in these severely ill and toxic patients.

Sydenham's Chorea (Chorea Minor, "St. Vitus' Dance"). This neurologic syndrome occurs after a latent period that is variable but on average longer than that associated with the other manifestations of ARF. It frequently occurs in "pure" form, either unaccompanied by other major manifestations or, after a latent period of several months, at a time when all other evidence of acute rheumatic activity has subsided. Chorea is characterized by rapid, purposeless, involuntary movements, most noticeable in the extremities and face. The arms and legs flail about in erratic, jerky, uncoordinated movements that may sometimes be unilateral (hemichorea). Facial tics, grimaces, grins, and contortions are evident. The speech is usually slurred or jerky. The tongue, when protruded, retracts involuntarily, while asynchronous contractions of lingual muscles produce a "bag of worms" appearance. The involuntary motions disappear during sleep and may be partially suppressed by rest, sedation, or volition.

Patients with chorea display generalized muscle weakness and an inability to maintain a tetanic muscle contraction. Thus, when the patient is asked to squeeze the examiner's fingers, a squeezing and relaxing motion occurs that has been described as "milkmaid's grip." The knee jerk may have a pendular quality. There is no cranial nerve or pyramidal involvement, and sensory modalities are unaffected. The electroencephalogram may display abnormal slow wave activity.

Emotional lability is characteristic of Sydenham's chorea and often may precede other neurologic manifestations, leaving teachers and parents puzzled over apparently inexplicable personality changes.

Subcutaneous Nodules. These are firm, painless subcutaneous lesions that vary in size from a few millimeters to approximately 2 cm. The skin overlying them is freely movable and is not inflamed. The lesions tend to occur in crops over bony surfaces or prominences and over tendons. Sites of predilection include the extensor surfaces of the elbows, knees, and wrists, the occiput, and spinous processes of the thoracic and lumbar vertebrae (Fig. 277-2). Nodules are virtually never the sole major manifestation of ARF; they almost always appear in association with carditis, and the cardiac involvement in such cases tends to be clinically severe. Nodules ordinarily do not appear until at least 3 weeks after the onset of an attack, usually lasting 1 to 2 weeks. They may appear in repeated crops in patients with protracted carditis. Similar nodules may be seen in systemic lupus erythematosus (SLE) and in rheumatoid arthritis. Subcutaneous nodules in the latter disease are larger and more persistent than those in rheumatic fever.

Erythema Marginatum. The rash begins as an erythematous macule or papule which then extends outward while the skin in the center returns to normal. Adjacent lesions coalesce, forming circinate or serpiginous patterns. The lesions may be raised or flat, are neither pruritic nor indurated, and they blanch on pressure. They vary greatly in size and appear mostly on the trunk and proximal extremities, sparing the face. The lesions are evanescent, migrating from place to place, at times changing before the observer's eyes, and leaving no residual scarring. The erythema may be brought out by applying heat. Individual lesions may come and go in minutes to hours, but the process may go on intermittently for weeks to months uninfluenced by anti-inflammatory therapy; its persistence is not necessarily an adverse prognostic sign. In the great majority of cases, erythema marginatum is accompanied by carditis; it also tends to be associated with subcutaneous nodules.

LABORATORY FINDINGS. No specific laboratory test is diagnostic of ARF. Usually there is a leukocytosis with an increase in the proportion of polymorphonuclear leukocytes. A mild to moderate normocytic, normochromic anemia is the rule. In some patients, the serum aspartate aminotransferase (AST) level is elevated. Evidence of acute inflammation is prominent, including readily detectable quantities of C-reactive protein in the blood and elevation of the erythrocyte sedimentation rate. An exception is "pure" chorea, which may appear long after indices of inflammation have returned to normal.

The urine may contain protein, white cells, and red cells. Biopsy studies have revealed a variety of renal abnormalities, but the classic proliferative glomerular abnormalities that characterize poststreptococcal acute glomerulonephritis occur quite rarely in ARF. Electrocardiographic and radiographic studies may reveal evidence of rhythm disturbances, pericarditis, or congestive heart failure. Echocardiography may document myocardial and valvular dysfunction and pericardial effusion.

The major laboratory contribution to the diagnosis of ARF is the documentation of recent group A streptococcal infection. Throat culture should always be performed but is positive in only a minority of cases. The low rate of culture positivity remains unexplained, although it may be due in part to the time lapse of several weeks between the onset of the pharyngeal infection and the throat culture. The serum titer of ASO is elevated in 80% or more of ARF patients. If two streptococcal antibody tests, e.g., ASO plus either anti-DNase B or antihyaluronidase, are performed, an elevated titer of at least one will be found in 90% of ARF patients. A battery of three tests establishes the presence of recent, immunologically significant streptococcal infection in >95% of individuals experiencing an acute rheumatic attack. The definition of an "elevated" titer varies, depending on the test used, the patient's age, and geographic locale. ASO titers of at least 240 Todd units per milliliter in adults and 320 Todd units in children are generally considered elevated. At times, serial sampling may detect a rising titer of streptococcal antibodies in patients seen early in the course of a rheumatic attack.

COURSE AND PROGNOSIS. The average duration of an untreated attack of ARF is approximately 3 months. The duration tends to be longer, up to 6 months, in patients with severe carditis. Fewer than 5% of patients have continuing rheumatic activity for longer than 6 months. In a few of these the disease is limited to chorea and is otherwise benign. Other patients exhibit evidence of persistent inflammatory activity, including arthritis, carditis, and subcutaneous nodules. "Chronic rheumatic fever" occurs more frequently in patients who have had one or more previous attacks; cardiac involvement in chronic rheumatic fever tends to be frequent and severe.

Death from intractable myocarditis during the acute phase of ARF is now very rare. Once the acute attack has subsided, the only long-term sequel is that of rheumatic heart disease, manifested primarily by scarring and/or calcification (Fig. 277-3) of the mitral and aortic valves (see Ch. 42) and leading to insufficiency and/or stenosis. The prognosis from a cardiac standpoint very much depends on the clinical findings when the patient is first seen. In one large study, for example, 347 patients were examined during an acute rheumatic attack and again 10 years later. Among patients who had been free of carditis during their acute attack, only 6% had residual heart disease on follow-up. Patients with no pre-existing heart disease and with mild carditis during their acute attack (i.e., apical systolic murmur without pericarditis or heart failure) had a relatively good prognosis in that only approximately 30% had heart murmurs 10 years later. About 40% of subjects with apical or basal diastolic murmurs and 70% of subjects with failure and/or pericarditis during their acute attacks had residual rheumatic heart disease. The prognosis was worse in patients with pre-existing heart disease and in those who had experienced recurrent attacks of ARF in the 10-year interval.

These data indicate that patients who do not develop carditis during an acute attack and are protected from ARF recurrences are most unlikely to suffer from rheumatic heart disease. The patient with "pure" chorea represents an exception to this rule. A significant proportion of such patients who have no evidence of carditis when first examined may develop rheumatic valvular disease on prolonged follow-up. Although the explanation for this phenomenon is unknown, it is conceivable that in view of the long latent period associated with chorea, signs of carditis might have been present earlier but subsided by the time the neurologic abnormality became evident.

DIAGNOSIS. Although ARF is readily recognized in the individual who presents with multiple major manifestations or in epidemic circumstances, at other times the disease may be extraordinarily difficult to diagnose with confidence.

The diagnostic criteria of T. Duckett Jones, initially proposed in 1944 and subsequently modified by committees of the American Heart Association, attempt to minimize overdiagnosis and underdiagnosis (Table 277-3). The most recent (1992) revision specifies that the guidelines are designed to assist in the diagnosis of the initial ARF attack. Although most patients with recurrent ARF fulfill the criteria, in some cases the diagnosis of a recurrence may be less apparent.

Two major manifestations, or one major and two minor manifestations, indicate a high probability of ARF, provided that there is supporting evidence of recent streptococcal infection. Although a positive throat culture or rapid antigen test for group A streptococci technically satisfies this requirement, streptococcal carriage rates of 15% are not uncommon among school-aged children during the fall and winter. Elevated titers of antibodies to streptococcal extracellular products, although not diagnostic of ARF, do indicate a recent, immunologically significant streptococcal infection. Conversely, if a battery of streptococcal antibody tests fails to reveal any evidence of recent infection, the diagnosis of ARF must be considered unlikely.

The modified Jones criteria are, of course, only guidelines. They are most difficult to apply confidently when polyarthritis is the single major manifestation. Under such circumstances, the diagnosis of ARF should be made only after excluding other causes of polyarthritis, such as rheumatoid arthritis, Still's disease, Lyme disease, viral arthritides (e.g., rubella, hepatitis B), the early prepurpuric phase of Henoch-Schönlein purpura, and septic arthritis, including gonococcal arthritis. The latter diagnosis cannot be excluded unequivocally by negative cultures of blood and synovial fluid. Therefore, if the clinical and epidemiologic picture is compatible with disseminated gonococcal infection, a trial of antigonococcal therapy should precede initiation of treatment with anti-inflammatory drugs.

Some patients have been described as manifesting polyarthritis that is atypical in time of onset and duration, does not respond dramatically to salicylate therapy, and is unassociated with other clinical features of ARF. Such individuals have on occasion been categorized as having "post-streptococcal reactive arthritis." The existence of this entity as a distinct syndrome, however, and its relationship to rheumatic fever remain uncertain. Pending further clarification, such individuals should be considered to have ARF if they fulfill the Jones criteria and alternative diagnoses have been excluded.

Serum sickness is frequently a serious consideration, particularly if the patient has received penicillin or other antibiotics for a preceding respiratory infection. SLE, sickle cell hemoglobinopathies, and infective endocarditis may involve the joints and the heart. Other differential diagnostic considerations include congenital heart lesions, viral and idiopathic forms of myocarditis and pericarditis, and functional heart murmurs. Nonfamilial forms of chorea have been described in SLE and rarely in association with the use of birth control pills. It remains uncertain how often episodes of chorea occurring during pregnancy ("chorea gravidarum") represent attacks of rheumatic fever. Other disorders that may at times be confused with ARF are gout, sarcoidosis, Hodgkin's disease, and acute leukemia.

There are certain circumstances in which ARF can be diagnosed even when the guidelines set forth in Table 277-3 have not been met. Patients whose only rheumatic manifestation is Sydenham's chorea may not fulfill the Jones criteria. Because of the long latent period between the antecedent streptococcal infection and the appearance of the neurologic abnormalities, evidence of inflammation encompassed in the minor manifestations may no longer be present, and previously elevated antibody titers may have declined to normal. A similar situation occasionally occurs in patients with indolent carditis, who may not come to medical attention until months after the onset of rheumatic fever. In patients with established rheumatic heart disease, it may be difficult to distinguish new from pre-existing cardiac involvement unless the patient has been under careful prospective follow-up, a previously undamaged valve is involved, or pericarditis is evident. Thus the diagnosis of recurrent ARF must be strongly entertained in the presence of suggestive clinical findings, provided there is evidence of recent group A streptococcal infection.

TREATMENT. Antibiotics neither modify the course of a rheumatic attack nor influence the subsequent development of carditis. Nevertheless, it is conventional to give a course of antibiotics designed to eradicate any rheumatogenic group A streptococci remaining in the tonsils and pharynx, in order to prevent spread of the organism to close contacts. The recommended regimens are those conventionally used for treatment of acute streptococcal pharyngitis (see Ch. 276). Benzathine penicillin G is preferred in non-penicillin-allergic patients. Following completion of this therapy, continuous antistreptococcal prophylaxis should commence (see below).

Treatment with anti-inflammatory agents is effective in suppressing many of the signs and symptoms of ARF. These agents do not "cure" the disease, nor do they prevent the subsequent evolution of rheumatic heart disease. They should be avoided in very mild or equivocal cases because, by suppressing the clinical manifestations, they may obscure the diagnosis. The two drugs most widely used are aspirin and corticosteroids. The former is used in patients with acute polyarthritis, provided that carditis is either absent or mild and there is no evidence of congestive heart failure. Aspirin is very effective in decreasing fever, toxicity, and joint inflammation. It should be given in a dosage of 90 to 100 mg per kilogram per day in children and 6 to 8 grams per day in adults. This is administered in equally divided doses, every 4 hours for the first 24 to 36 hours; thereafter it may be given in four doses during waking hours. A salicylate level of 25 mg per deciliter is usually satisfactory. The incidence of nausea and vomiting may be minimized by starting somewhat below the optimal dosage level and gradually increasing over a few days. The patient should be observed for evidence of significant gastrointestinal bleeding and for signs and symptoms of salicylism (see Ch. 19). After 2 weeks, the dosage is reduced to 60 to 70 mg per kilogram per day for an additional 6 weeks. These dosage schedules represent general guidelines only. The precise aspirin dose must be determined by the patient's clinical response, blood salicylate levels, and tolerance of the drug.

Corticosteroids are generally reserved for patients who have severe carditis manifested by congestive heart failure, who are unable to tolerate large doses of salicylates, or whose signs and symptoms are inadequately suppressed by aspirin. As with aspirin, the dosage must be individualized. Prednisone, 40 to 60 mg per day in divided doses, may be used initially. After 2 to 3 weeks it should be withdrawn slowly over an additional 3-week period. In cases of fulminating carditis with profound heart failure, intravenous corticosteroids may be used. As is the case for other patients receiving corticosteroids, the physician should be alert to problems such as gastrointestinal bleeding, sodium and water retention, and impaired glucose tolerance. Suppression of the pituitary-adrenal axis or of the host immune system is a potential problem but not ordinarily a major one during this relatively short course of treatment. The role of nonsteroidal anti-inflammatory agents in managing ARF remains to be defined.

After cessation of anti-inflammatory therapy, clinical or laboratory evidence of ARF may reappear. Such therapeutic "rebounds" occur more frequently after corticosteroid therapy than after treatment with aspirin. They may be minimized by prolonging salicylate therapy for 9 to 12 weeks and, when corticosteroids have been required, by continuing aspirin for a month after corticosteroids have been discontinued. Congestive heart failure is managed by conventional measures. If digitalis is used, the potential risk of drug-induced arrhythmias in the patient with active myocarditis must be kept in mind. Patients with Sydenham's chorea require a quiet environment, and sedatives such as phenobarbital may be helpful.

Once the acute attack has subsided completely, the patient's subsequent level of physical activity depends on cardiac status. Patients without residual heart disease may resume full and unrestricted activity. It is important that patients not be subjected to unwarranted invalidism, because of either their own inaccurate perceptions of the nature of the rheumatic process or those of parents, teachers, or employers.

PREVENTION. "Primary prevention" of ARF consists of accurate diagnosis and appropriate treatment of streptococcal sore throat (see Ch. 276). Although straightforward in theory, primary prevention is often frustratingly difficult to achieve. In many of the densely populated indigent communities in which the risk of ARF is greatest, children with self-limited illnesses such as sore throats may never come to medical attention, and throat culture services are usually unavailable to aid in diagnosis. Moreover, in one third or more of cases, ARF may arise after a clinically inapparent streptococcal infection.

Perhaps the most effective strategy for avoiding the mortality and chronic cardiac disability associated with ARF is that of "secondary prevention." This strategy focuses on the group of persons who have already suffered a rheumatic attack and who are inordinately susceptible to a recurrence following an immunologically significant streptococcal upper respiratory infection. Recurrent attacks tend to be mimetic in nature, so patients who have suffered carditis with their previous attack are likely to have repetitive cardiac involvement and progressive cardiac damage. Because even patients who experienced only arthritis or chorea may develop carditis with recurrent attacks of ARF, all patients who have experienced a documented attack of ARF should receive continuous antimicrobial prophylaxis to prevent either symptomatic or asymptomatic streptococcal infections. The specific regimens to be used are indicated in Ch. 42 and 276. By far the most effective of these is intramuscular benzathine penicillin G every 4 weeks. Rheumatic recurrences are very unusual in patients faithfully adhering to this regimen. In areas of the world where the incidence of ARF and the risk of recurrence are extremely high, however, injections every 3 weeks may provide more complete protection.

The total duration of intramuscular or oral rheumatic fever prophylaxis remains unresolved. The risk of rheumatic recurrence is known to diminish with increasing age and increasing interval since the most recent rheumatic attack. Patients who escape carditis during their initial attack are less likely to experience rheumatic recurrences and less likely to develop carditis if a recurrence does ensue. These facts suggest that prophylaxis need not be perpetual for all rheumatic subjects. Continuous prophylaxis should be maintained indefinitely for those with clinically significant rheumatic heart disease. Other rheumatic subjects should be protected until reaching adulthood, for at least 5 years after their most recent attack, and if they are in an epidemiologic circumstance that places them at high risk of streptococcal acquisition (e.g., parents of small children, school teachers, military recruits, nurses, pediatricians, or residents of areas with a high incidence of ARF). The decision to remove a rheumatic subject from continuous prophylaxis should be an individualized one, based on the physician's assessment of the risk and likely consequences of recurrence, and taken with the patient's informed consent. Patients taken off prophylaxis must be instructed to return immediately for medical follow-up whenever symptoms of pharyngitis occur.

Patients with rheumatic valvular heart disease must receive prophylaxis designed to avoid bacterial endocarditis whenever they undergo dental or surgical procedures likely to evoke bacteremia. This is not necessary in the rheumatic subject who is free of residual heart disease. The regimens to prevent endocarditis (see Ch. 278) are different from those prescribed for preventing ARF, and the fact that a patient is receiving rheumatic fever prophylaxis does not exempt him/her from endocarditis prophylaxis. This is a frequent point of confusion not only among patients but among physicians and dentists as well.

REFERENCES

Berrios X, Del Campo E, Guzman B, Bisno AL: Discontinuance of rheumatic fever prophylasis in selected adolescents and young adults. Ann Intern Med 118:401, 1993. Presents new data, along with a review of previous studies, on the circumstances under which rheumatic fever prophylaxis might be discontinued.

Bisno AL: Group A streptococcal infections and acute rheumatic fever. N Engl J Med 325:783, 1991. Review of the biology of the group A streptococcus as it relates to civilian and military outbreaks of ARF in the 1980's and the more recent resurgence of life-threatening streptococcal infections.

Stollerman GH: Rheumatic Fever and Streptococcal Infection. New York, Grune & Stratton, 1975. A comprehensive, extremely readable summary of rheumatic fever. Detailed descriptions of clinical manifestations will be particularly valuable to physicians unfamiliar with the disease.

Veasy LG, Tani LY, Hill HR: Persistence of acute rheumatic fever in the intermountain area of the United States. J Pediatr 124:9, 1994. A summary of the demographic and clinical data on 274 cases of ARF hospitalized in Salt Lake City between 1985 and 1992.

Texto de Domínio Público, retirado do Cecil TextBook of Medicine - 20th Edition