Update on acute rheumatic fever

Literature review

Two distinct schools of literature exist for ARF: studies and commentaries from North America and Europe that view rheumatic fever as a rare disease and discuss the limited efficacy of screening for streptococcal sore throat, ARF’s presumed precursor; and literature from the developing world and international aboriginal literature that documents a robust discussion of the presentations, epidemiology, and control of rheumatic fever and its sequela, rheumatic heart disease (RHD), which is of ongoing relevance.

In a meta-analysis of antibiotics for the primary prevention of ARF in patients with documented pharyngitis (N = 7665), Robertson et al found a protective effect (relative risk 0.32, 95% confidence interval 0.21 to 0.48) for a reduction in risk of almost 70% (level I evidence).1 Interestingly, inclusion of randomized controlled studies done in the 1950s by the US Army might make such overviews less informative, as they were done in an era of endemic group A streptococcus (GAS) and ARF.1

A Cochrane review by Del Mar et al looked at the benefit of antibiotics for sore throats (N = 2835) and found that they reduced ARF by more than two-thirds (relative risk 0.22, 95% confidence interval 0.02 to 2.08), although in developed societies most patients would derive no benefit given the low incidence of ARF (level I evidence).2

Natural history

Repeated GAS infections are thought to occur and prime the immune response before the first episode of ARF.3 Symptoms of arthritis, carditis, erythema marginatum, subcutaneous nodules, or chorea usually present 1 to 3 weeks after GAS pharyngitis (level II evidence).4 In recent outbreaks in the United States, affected patients reported only mild pharyngitis, for which only a few sought medical attention (level II evidence).4,5 The subsequent outcome, RHD, might be the first presentation: 46% of RHD patients in the Northern Territory of Australia, for example, had no known prior diagnosis of ARF (level II evidence).6

Risk factors are poorly understood but likely include host factors such as susceptibility to the immune response to GAS (limited to 3% to 6% of the population); housing and overcrowding remain important considerations (level III evidence).2,7 The Jones criteria, established in 1944, were revised in 1965 and 1984, then updated in 1992 and 2002 to provide diagnostic guidelines for clinical diagnosis.8–10 They require 2 major or 1 major and 2 minor criteria and evidence of prior streptococcal infection (throat culture positive for the bacteria, positive rapid antigen detection test results, or elevated antistreptolysin O titre [ASOT]). Major and minor criteria are outlined in Box 1. Echocardiogram is not part of the criteria but is often part of the cardiac workup.

Disease frequency

In Canada ARF is not a reportable disease. Available data place Canadian, American, and Western European incidences at 0.1 to 2 cases per 100 000 persons (level II evidence).11,12 Isolated cases can occur anywhere from time to time. A series of 3 cases in Nova Scotia was reported in 1998 (level II evidence).13,14 Another 3 cases were reported in a UK teaching centre during a 6-month period in 2000 (level II evidence).15

Acute rheumatic fever is now generally seen as a disease of “emerging economies,” indigenous communities, and tropical regions, with incidences in these settings of 10 to 20 cases per 100 000 persons (level II evidence).12 Hot spots, such as Northern Australia, have rates of more than 50 cases per 100 000 persons (level II evidence).16 According to W. De Groote, MD, among aboriginal children presenting to a Winnipeg, Man, referral centre serving northwest Ontario and Manitoba, ARF remains the most common underlying cause of cardiac murmurs (level III evidence) (written communication, March 2008). Despite this, no association with ethnicity has been identified in the literature (level III evidence).3

Changing epidemiology of GAS infections

Only rheumatogenic strains of GAS result in ARF (level II evidence).17 The endemic strains seen during World War II had particularly high ARF rates (3%).17 By the 1970s those streptococcus M antigen serotypes (which confer resistance to phagocytosis) had virtually disappeared in North America, although there was little change in the rate of endemic streptococcal sore throat (level II evidence).4 Lower incidence of ARF had more to do with changes in virulence and improvement in socioeconomic conditions than with use of antibiotics (level II evidence).4

By the 1980s a resurgence of the streptococcus M antigen subtype 5M occurred, along with outbreak reports in Utah and Colorado identifying more than 30 cases annually (level II evidence).5 The epidemiology of ARF in aboriginal communities in Australia challenges the historical belief that streptococcal skin infections cannot cause the disease. In these communities pyoderma is the most common manifestation of GAS, and typical rheumatic strains are not present (level II evidence).18 Researchers postulate that even non-GAS infections (eg, group C and G streptococci) might play a role in these high-incidence settings and note that the high rates seen among Australian aboriginal children parallel overcrowding and poor living conditions.19

Changing presentation of ARF

Acute rheumatic fever is a clinical diagnosis for which presentations are highly variable (level III evidence).20 Rheumatic carditis is associated with the murmurs of valvulitis and is more common in children; arthritis predominates in adults (level II evidence).21 In the Utah outbreak of the 1980s, carditis was seen in 80% of patients, arthritis in 70% of patients, and chorea in 28% of patients (level II evidence).22 The arthritis is migratory, with pain often worse than physical findings would suggest; any one site might resolve within 2 to 5 days (level II evidence).23 Chorea occurring alone meets the Jones criteria and typically presents 2 to 6 months after the initial infection as purposeless movements of the extremities, dysphonia, and possible emotional lability.20

Disease-related murmurs are most often caused by mitral insufficiency; aortic insufficiency is the second most common disease-related cause (level II evidence).15 Silent carditis is sometimes found during echocardiography (7% to 47% of cases).15 Mitral stenosis is a delayed complication with increasing age (level II evidence).7 There is a natural improvement of the carditis over several years, in the absence of ARF recurrences, with more than 65% of patients demonstrating resolution or improvement (level II evidence).5 Disease recurrence, however, increases long-term risk and the degree of valvular damage.5 Erythema marginatum and subcutaneous nodules are both infrequent, occurring in less than 5% of presentations (level II evidence).7

Investigations

Investigations are useful mainly to confirm the existence of prior streptococcal infection by elevated ASOT, to check for ongoing pharyngeal GAS infection, and to assess erythrocyte sedimentation rate. Echocardiography is not part of the diagnostic criteria, but is generally done as part of the initial workup to clarify cardiac involvement.

Prevention, treatment, and prophylaxis

Most cases of ARF can be prevented by antibiotic treatment received within 9 days of GAS pharyngitis (level II evidence).24 After onset of ARF there is no effective treatment for the immune reaction. Salicylates and antibiotics for any current GAS infection remain the cornerstone for treating most cases (level II evidence).25 Salicylates give relief from fever and arthritis. They have no role in the treatment of carditis, which is addressed with bed rest (level II evidence).1 Prednisone is not useful for arthritis but it is the drug of choice for those patients who experience chorea (level II evidence)1 (Box 2).

Secondary prophylaxis for prevention of new GAS pharyngitis is warranted for all patients, as there is an 8% to 10% recurrence rate of ARF within 5 years—with an attendant increase in cardiac involvement (level II evidence).15 Oral regimens can be used, but the treatment of choice remains intramuscular benzathine penicillin injections given monthly for 5 years, until adulthood, or for longer, depending on the severity of RHD or the existence of frequent ARF in the community.26

No vaccine currently exists, although development of a multivalent streptococcal M antigen vaccine is under way.4

Disease in northwest Ontario

Our observation of 5 unrelated ARF cases out of 60 000 patient-years gives an incidence rate of 8.33 cases per 100 000 persons. This is a much higher rate than generally reported for developed countries (0.1 to 2 cases per 100 000), but lower than rates seen in some indigenous communities of Australia. Our rate of cultures positive for streptococcus was in keeping with other general population studies (about 25%), and there was a correlation between the peak months of streptococcal throat infections and the timing of ARF presentations (level I evidence).27

Presentations were generally similar to those reported in the literature. We saw arthritis in 40% of cases and joint symptoms in 60% overall (literature shows up to 80%). Chorea was prominent as a presenting symptom, seen in 2 of 5 of cases (40%); both patients were female. The literature documents the increasing presence of chorea and a nonclassic oligoarthritis that does not respond to acetylsalicylic acid (level II evidence).7,24,26 Each of the rare manifestations (erythema marginatum and subcutaneous nodules) was found in at least 1 patient.

Our cases differed in that cardiac presentations were seen in 4 of 5 cases (literature suggests 40%). Echocardiography documented valvular cardiac involvement in 3 cases and an innocent murmur in a fourth case. The ASOT was positive in all cases, and the erythrocyte sedimentation rate was elevated in 4 of 5 cases, consistent with literature rates of 80% or higher.