Post-streptococcal Arthritis

Post-streptococcal arthritis (PSA) is a poorly understood clinical syndrome in which arthritis of one or more joints occurs following a Group A streptococcal infection of the pharynx. Clear guidelines do not exist for diagnosis and management of patients with PSA, nor have criteria for differentiating it from acute rheumatic fever been established. The purpose of this paper is to provide a summary of published information on PSA and to outline diagnostic and therapeutic recommendations.

Introduction

Post-streptococcal arthritis (PSA) is a poorly understood clinical syndrome that has generated much controversy and for which clear diagnostic criteria and therapeutic recommendations are lacking. Like acute rheumatic fever (ARF), PSA is a reactive arthritis characterized by a pharyngeal streptococcal infection, a symptom-free interval, and subsequent aseptic inflammation of one or more joints. However, unlike in the case of ARF, the risk of other post-infection complications is not clear. Interest in various aspects of group A streptococcal infection has been rekindled recently as a result of changing trends in both its suppurative and post-infection complications. With an apparent increase in the incidence of ARF in certain parts of North America, the diagnostic criteria have been revisited and refined (Table 1). Furthermore, a change in streptococcal virulence has been suggested by clusters of cases of ARF and by the emergence of cases of severe invasive disease. In this climate of enhanced interest in streptococcal infections and their complications, it is timely to examine some of the diagnostic and therapeutic controversies surrounding PSA.

Post-streptococcal arthritis

The term post-streptococcal arthritis was introduced in 1959 to denote patients who had arthritis following pharyngeal infection with beta hemolytic streptococcus, but in whom other major criteria of ARF were absent (Table 1). The term PSA is currently used inconsistently in the literature to indicate various constellations of signs and symptoms. In its original sense, PSA is used to designate the condition of patients in whom prolonged polyarthritis that does not respond to aspirin occurs approximately ten days following upper-respiratory-tract infection with group A beta hemolytic streptococcus when other signs of ARF are absent. In most reported series of PSA, however, cutaneous or cardiac disease was present in a significant proportion of patients. None of the 12 patients, as reported by Goldsmith and Long, had carditis. However, one had a pericardial effusion, and four had urticarial or maculopapular rashes; an additional four had cutaneous hyperaesthesia. Of the 16 children with PSA reported by Gibbas and Broussard, seven had pericarditis and at least three had valvular disease. Six of the seven patients, as reported by Emery et al, had evidence of carditis. Neither chorea nor nodules were reported in any of these series. It is likely that some patients identified in the literature as having PSA actually suffered from ARF.

It is not certain whether PSA represents a mild or early form of ARF, or whether it is an entirely separate entity. In support of the first possibility is the fact that outbreaks of PSA and ARF occur at the same time, and at least some children with PSA have been shown to go on to develop full-blown ARF. In support of the possibility that PSA and ARF are distinct disorders are the differences in interval between infection and disease, the differences in response to salicylate, and the differences in pattern of affected joints (Table 2). In children with ARF but not in those with streptococcal pharyngitis or post-streptococcal glomerulonephritis, there is a very strong association with the B-cell alloantigen D8/17. The occurrence of other manifestations of ARF, such as isolated chorea, may be associated with this antigen, and the same association may be found in children with isolated PSA; such an association would support the view that PSA is a limited form of ARF. It may be most accurate to regard PSA as a kind of ARF (with or without cardiac involvement) in which the arthritis differs significantly from the arthritis of typical ARF in the following ways:

• PSA begins sooner (approximately 10 days) after streptococcal infection than does ARF (approximately 21 days);
• it is cumulative, rather than flitting, and persistent, rather than fleeting;
• it responds poorly to salicylate, in contrast to the rapid, dramatic response that is usually seen in the case of ARF.

Substantial controversy exists regarding the optimal therapy for patients with PSA; this stems from concerns about the risk of other nonsuppurative complications upon recurrence. In particular, some authorities believe that the risk of carditis mandates the use of prophylactic antimicrobial therapy, like that recommended for patients with a history of ARF, to prevent group A streptococcal reinfection. In one study of twelve patients with PSA who did not receive antimicrobial prophylaxis, five experienced recurrent disease. Of the five, two had arthritis, two had arthralgia, and one experienced classic ARF with carditis, migratory arthritis, and subcutaneous nodules. In a classic description of scarlatinal arthritis, Crea and Mortimer suggested that patients may experience an illness that does not satisfy the diagnostic criteria for ARF but puts them at the same risk of acquiring cardiac valvular disease. If a patient thought to have PSA develops evidence of carditis, whereby the Jones criteria are satisfied, the diagnosis established is clearly ARF, and appropriate therapy should be initiated. Since PSA is so rare, the risk of nonsuppurative complications associated with first or recurrent attacks remains to be clearly established; expert advice should be sought in the diagnosis and therapy of patients thought to have either condition. Furthermore, patients initially identified as having PSA might, after prospective evaluation, prove to have ARF.

In the absence of a clear approach to prophylaxis and treatment of children with PSA and because the condition of patients initially thought to suffer from PSA is frequently associated with cardiac disease, it seems prudent to treat such patients as if they have ARF until more precise information is available. One reasonable approach is to initiate antibiotic prophylaxis as recommended for ARF. If, after three months, there is no evidence of either carditis or chorea, antibiotic prophylaxis may be discontinued.

Pitfalls in the diagnosis of streptococcal infection

Other diseases that should be considered in the differential diagnosis of patients thought to have post-streptococcal arthritis include juvenile rheumatoid arthritis, spondyloarthropathies, acute streptococcal arthritis, and reactive arthritis secondary to enteric (Yersinia, Campylobacter, Salmonella) or genitourinary-tract (Chlamydia) infections. Although clinical criteria help differentiate these disorders, the diagnosis of PSA hinges on the correct identification of patients with preceding group A streptococcal infection. The accurate diagnosis of streptococcal infection is difficult for a number of reasons, including the following:

• A substantial number of healthy children are chronically colonized with group A streptococci; therefore, recovery of the bacterium from the throat does not necessarily indicate acute infection.
• Group A streptococci may no longer be present in the throat at the time that nonsuppurative complications such as PSA appear, as a result of either effective antimicrobial therapy or spontaneous clearance.
• Serological evidence of recent streptococcal infection (when documented by a dilution indicating a twofold, or greater, rise in titers of ASO or anti-DNase B, or by means of other streptococcal antibody tests) is not present immediately after an acute streptococcal infection and may require weeks or months to appear.

Questions that need to be answered

Clinical criteria for establishing the diagnosis of PSA have been suggested, but a number of questions regarding the optimum diagnostic and therapeutic strategies remain unanswered. Issues that need to be addressed include the following:

• Is PSA a clinical entity distinct from ARF, or are the two manifestations of the same disease?
• To what extent are patients with PSA at increased risk of developing other manifestations of ARF, particularly carditis?
• Do patients with PSA require antibiotic prophylaxis to prevent recurrences (analogous to patients with ARF)?
• What is the present incidence of PSA and ARF?

Recommendations

1.        1.      In order to establish the diagnosis of PSA, antecedent infection with group A streptococcus must be established, either by culture of the bacterium from the throat or by demonstration of an elevation or a fourfold rise in ASO or anti-DNase B titer.

2.      2.    If group A streptococcus is recovered from the throat, specific antibacterial therapy should be given.

3.      3.    The diagnosis of PSA should be made only after careful historical and clinical evaluation for nonsuppurative streptococcal complications or other causes of polyarthritis.

4.      4.    In view of reports according to which patients thought to have PSA experience recurrences and may be at increased risk of developing carditis, antibiotic prophylaxis, as recommended for patients with ARF, may be considered on a short-term basis. If, after further evaluation, there is no evidence of carditis or choreal, prophylaxis may be discontinued.

5.      5.    A Canadian registry of patients with PSA should be set up in order to assess the prevalence, risks, and outcome of this rare disease and to establish an optimum strategy.